The Rosgen Wars and the Shifting Political Economy of Expertise

by

Rebecca Lave

A dissertation submitted in partial satisfaction of the

requirements for the degree of

Doctor of Philosophy

in

Geography

in the

Graduate Division

of the

University of California, Berkeley

Committee in charge: Professor Michael Watts, Chair

Professor Michael Burawoy Professor Kurt Cuffey Professor Nathan Sayre

Spring 2008

The Rosgen Wars and the Political Economy of Expertise

© 2008

Rebecca Lave

ii

Preface

In August 2003, 35 of the most respected academics, agency staff and consultants

involved in stream restoration in the U.S. met in Minneapolis. Their goal was to set the

research agenda for the first update to the National Research Council’s work on stream

restoration in over a decade. They were a disciplinarily diverse but otherwise fairly

homogenous group of mid-career scientists and professionals, relaxed in business casual

clothing, the shared language of science, and a high degree of personal success.

Then there was Dave Rosgen, by any conventional measure the most successful

person in the stream restoration world, not to mention the room. Rosgen was dressed like a

cowboy in a white hat, blue jeans, and large rodeo belt buckle. He talked like a cowboy, with

folksy turns of phrase that mixed oddly with the scientific jargon flying around the room.

And if it looks like a duck and it quacks like a duck….

In the stream restoration field, Rosgen holds the maverick position his self-

presentation suggests. Despite the fact that he has little formal training in restoration

science, Rosgen is the primary educator of restoration practitioners in the U.S., and training

in his approach is in many parts of the country considered preferable to a Ph.D. in

geomorphology or hydraulic engineering. His Natural Channel Design approach has been

adopted by federal agencies including the Environmental Protection Agency, US Fish &

Wildlife Service, the US Forest Service, and the Natural Resources Conservation Service, to

the exclusion of other approaches. And it is Natural Channel Design, not a more accepted

scientific approach, that forms the primary basis of the burgeoning restoration consulting

industry.

iii

Since the mid-1990s, Rosgen’s work has been vehemently opposed by many of the

most prominent university- and agency-based restoration scientists in the U.S., including

some of those present at the NRC conference. Thus it was surprising that Rosgen was

invited to the conference in the first place,1 and it was shocking that he accepted the

invitation given the tenor of the background document prepared by the NRC. This position

paper was sent to all of the participants in advance, and included an implicit but

unmistakable indictment of Natural Channel Design to anyone even passingly familiar with

the Rosgen Wars, which by 2003 included most people in the stream restoration community

and certainly everyone invited to the NRC conference.

The opening reception avoided any overt conflict, but within a few hours of the

conference’s formal start the following day the situation degenerated into one of the most

excrutiatingly intimate battles of the Rosgen Wars. The first session allotted each participant

a few minutes to formally introduce their work and their thoughts about the state of the

stream restoration field. A few people used this opportunity to take mild potshots at the

NCD approach and its application in practice, which was uncomfortable since Rosgen was

right there in front of them, but still within the bounds of propriety.

Things didn’t really heat up until the second session of the morning with the late

arrival of Matt Kondolf, a professor of geomorphology at UC Berkeley and one of the most

1 Gary Parker, a Rosgen Wars moderate, invited Rosgen over the objections of some participants. As Parker described it,

GP: There was a strong vote to make sure that Rosgen was not invited, but I made damn sure that he was.

RL: Why did you do that?

GP: Academics tend to be somewhat dismissive of Rosgen's approach. Rosgen, however, can accurately be described as one of the founders, and certainly the major popularizer of the field of river restoration. To my mind, then, it would be unthinkable not to invite him… to a major US workshop on the subject.

iv

vehement Rosgen critics. Having missed the icebreakers of the morning and previous

evening, Kondolf walked in without the restraints of sociality that held other attendees at

least somewhat in check, and proceeded to let loose a shotgun blast of critique which

sounded very loud in such a small room. Kondolf presented his Uvas Creek paper

(Kondolf et al. 2001), a scathing denunciation of the Natural Channel Design approach. The

Uvas Creek case study is a powerful analysis of a spectacular project failure in Gilroy, CA.

The before and after photographs, in particular, show a jaw-dropping contrast between a

perfectly manicured, suspiciously symmetrical single-thread meandering channel, and the

shaggy expanse of multi-thread gravel bed channel that replaced it only a few months later

after a medium-sized storm. As aptly demonstrated by Kondolf, the project design was an

obvious disaster in the making. His conclusion – that the Natural Channel Design approach

is a short-sighted, opportunistic piece of bad science – sounded like a call to the barricades.

Uncomfortably, however, the target of revolutionary wrath was located not a respectable

distance away at Versailles, but no more than 25 feet from the podium.

It was, to put it mildly, awkward.

But here’s the thing about the Uvas Creek case study: whatever the project’s failings,

its relationship to Rosgen’s work is tenuous; the designers did not employ anything

approaching the full Natural Channel Design approach. Uvas Creek is a telling example of

the dangers of allowing poorly trained, inexperienced designers to create new stream

channels, but as a sweeping indictment of Natural Channel Design it has little traction.

Rosgen pointed this out. Kondolf disagreed, and the conversation degenerated into barely-

veiled hostilities until the organizers cut it off. Both sides declared victory in private

conversations, and nobody in the room changed their minds.

v

This skirmish was unusual for the sheer level of social discomfort it created, but in all

other respects it was typical of the Rosgen Wars in both its substance and seeming

unresolvability. A vocal university- and agency-scientist opposition has been denouncing

Rosgen as a charlatan and snake oil salesman for more than a decade. These guardians of

scientific legitimacy – bearing academic sanctification in the form of prestigious degrees, jobs

and publications – have argued against the Natural Channel Design approach in print, at

conferences, and at short courses to remarkably little effect. Rosgen’s classification system,

design approach, and short course series are increasingly seen in the restoration field not just

as scientifically legitimate, but as a more legitimate basis for restoration practice than

academically-produced science and training.

This raises some very important questions about how scientific expertise is

constructed. First, what (and who) makes someone an expert: knowledge, a degree from a

top-ranked institution, market demand, the state? Second, how does the construction of

scientific expertise relate to the status of its science, its relative newness, level of uncertainty,

and public profile? Finally, how does the construction of expertise relate to the shifting

boundary between public research and private profit? Is Rosgen’s ability to supplant the

university as educator, researcher and developer of applied techniques a fluke, or a portent

of things to come? These are the questions at the heart of my research.

vi

Acknowledgments

Qualitative research is impossible unless a great number of busy people take time away from

their main activities and concerns to talk. I am thus very grateful both to the short course

students who filled out my surveys and put up with having one of their fellow participants

observe them, and to the people who sat down with me to talk about their restoration work

during my various rounds of interviews (interview subjects are listed by name in Appendix

A). Of the latter group, there are a few who were particularly generous, talking with me

multiple times to help me get a better grasp on stream restoration history, policy, practice

and science. I am especially grateful to Martin Doyle, Craig Fischenich, Matt Kondolf, Greg

Koonce, Jim MacBroom, Dale Miller, Dave Rosgen, Doug Shields, and Jim Wilcox.

Academic labor may be largely solitary, but the end product is collaborative. My work builds

on the conversations, arguments, and suggestions of my broader academic community:

Jason Delborne, Mike Dwyer, Ben Gardner, Julie Guthman, Charles Lave, Jean Lave, Sara

Shostak, Jason Strange, and my graduate school cohort: Andy Bliss, Jennifer Casolo, Wendy

Cheng, Rita Gaber, Shiloh Krupar, Jason Moore, and Madeline Solomon.

I was blessed with the kind of committee that makes other graduate students whistle, and

office hours appointments something to be prepared for with Great Care. Michael Watts,

my chair, is one of the most incisive readers I have ever had the good fortune to encounter,

and consistently pushed me to improve what I thought was already good work. He also

deserves deep thanks for guiding me through what was unquestionably the most difficult

part of my graduate career: learning to write successful grant applications. Michael Burawoy

vii

intervened at a number of key points in this dissertation, leading me to think more carefully

about the mechanisms of Rosgen’s success, to read Bourdieu’s work on fields, and to

consider more critically what Bourdieu says about conflict. Kurt Cuffey is remarkable

among natural scientists both in his willingness to engage seriously with social science

research, and his openness to teaching social scientists geomorphology. He was very patient

about helping me understand what to him must have seemed very basic aspects of natural

science practice. Nathan Sayre joined my committee after I finished my rough draft; I

drafted him because he actually volunteered to read it. Through this action and many

conversations over the years since he arrived at UC Berkeley, he has given me the gift of

consistent and enthusiastic support for my work. Anyone who has been through the ups

and downs of the dissertation process can recognize the enormous value in that.

Life is what happens when you are trying to get your research done; so is death. This project

has been punctuated by the loss of people very dear to me: Elizabeth Carter, my maternal

grandmother, in July 2005; Herbert Carter, my maternal grandfather, in March 2007; Henry

Fienning, my brother-in-law, the day I turned in my rough draft in July 2007; and my father,

Charles Lave, the week that I filed. My daughter Nell, the delight of my life, was born in

August 2005. This dissertation is dedicated with love to each of you, and to my husband,

Sam.

i

Table of Contents Preface ii

Acknowledgments vi

Chapter 1: Introduction 1

Chapter 2: The Rosgen Classification System and Design Approach 25

Chapter 3: Adoption and Use of Rosgen’s Approach 50

Chapter 4: The Structure of the Stream Restoration Field 71

Chapter 5: The Intellectual Content of the Rosgen Wars 111

Chapter 6: The Mobilization of Expertise in the Rosgen Wars 148

Chapter 7: The Rosgen Wars and Bourdieu’s Agonistic Fields 182

Conclusions 207

Bibliography 217

Appendix A: Interview and Survey Meta-Data 225

Appendix B: University-Based Stream Restoration Certificate Programs 230

Appendix C: Additional Critiques of Rosgen’s Classification System 238

Appendix D: The ASCE River Restoration Committee Calls for Design Standards 241

1

Chapter 1: Introduction

Proponents of ecological restoration claim that we can transcend the defensive

stance of traditional environmentalism and repair the damage already done (Jordan 2000).

The tremendous appeal of this image of humans as a positive contributor to environmental

health has rapidly made restoration, particularly stream restoration, a driving force of the

environmental movement, an institutionalized commitment at all levels of American

government, and a lucrative market (Bernhardt et. al. 2005). While there is no federal

legislation requiring stream restoration, the restoration field and market have been to a large

extent catalyzed by three classic pieces of environmental legislation: the National

Environmental Protection Act, the Clean Water Act, and the Endangered Species Act.

Together, they removed the cavalier ease with which public agencies and private landowners

formerly moved, straightened or buried streams that were inconveniently located. These

laws created conditions in which the water in American streams and rivers became clean

enough to support life again, some forms of that life had legal right to be there, and any

impacts on streams or their inhabitants had to be disclosed and, most importantly for the

stream restoration industry, mitigated. Restoration, on- or off-site, is the primary type of

mitigation.

Unfortunately, the demand for stream restoration projects thus created far outstrips

knowledge of how to implement them. The science called upon in stream restoration

practice is characterized by deep uncertainty. This is not only because streams are complex

systems, but also because human goals for stream restoration are notably conflicted. More

than perhaps any other type of ecological restoration, stream restoration melds the desire to

restore natural processes with demands to achieve human ends – bank stabilization and

2

flood protection in particular – that are in direct conflict with those processes.

Compounding the problem, the nascent field of stream restoration has yet to gain a strong

foothold in the academy; there was until recently no way to get even a certificate in stream

restoration from an accredited college or university, and there is still no such thing as an

academic department of stream restoration.1 This leaves everyone involved in stream

restoration – developers, community groups, practitioners, scientists, and regulators – in an

uncomfortable situation. There is tremendous public demand for restoration, a legislative

framework in place that promotes it, and a rapidly expanding market, but no established

academic discipline that provides the basic science, applications, and training needed to

guide the new field.

Dave Rosgen, a consultant in Colorado, stepped into the breach American

universities have so far failed to fill by developing a system for classifying and reconstructing

stream channels. Rosgen uses classic 1950s and 60s work on hydraulic geometry relations

and bankful discharge as the basis of a purportedly universally-applicable alphanumeric

classification system that divides channels into seven categories (A-G, each with multiple

numbered sub-divisions) based on a relatively brief evaluation of their geomorphology

(Rosgen 1994). According to Rosgen, practitioners can use this classification system as a

basis for designing stable natural channels (Rosgen 1996a). Such claims of stability and

naturalness are deeply appealing for regulators stuck with implementing Americans’

conflicting goals for restoration. Despite strenuous opposition from university- and agency-

1 As will be discussed in detail in Chapter 4 and Appendix B, at present there are two American universities that grant certificates in stream restoration: the University of Minnesota, which offers a one-year post-baccalaureate certificate program, and Portland State University, which offers a short course-based professional certificate in restoration through its extension program. Both programs began in 2006. West Virginia University offered a certificate program modeled loosely on Rosgen’s short course series from 2002 to 2004, but no longer does so.

3

based scientists, 2 Rosgen’s approach has been adopted and implemented by local, state and

federal agencies, including the Environmental Protection Agency, the U.S. Fish and Wildlife

Service, the U.S. Forest Service, and the National Resources Conservation Service.

While for the most part excluded from university curricula, Rosgen’s approach is

disseminated via a series of short courses. Roughly a dozen are held each year, heavily

attended by agency staff and private consultants despite a fairly steep fee ($1,500-3,000 per

person, depending on length, not including transportation, lodging or meals). After

completing the four-course series attendees are considered to be Rosgen-certified.

Attendance at the short courses is spurred by the fact that a growing number of agencies

have made completion of the course series a requirement for consultants bidding on

restoration projects.

For Rosgen’s opponents, it would be bad enough if Rosgen training were only

treated as necessary, but it is unacceptable that it is increasingly treated as sufficient.

Students who have completed Ph.D.s in fluvial geomorphology or hydraulic engineering are

being turned away from restoration jobs because they are considered unqualified. Professors

and full-time consultants with decades of experience cannot bid on projects because they

have not studied their own subject as taught by Rosgen. Rosgen’s classification system,

design approach, and short course series are increasingly seen not only as scientifically

legitimate, but as more legitimate than academically-produced science and training.

2 For journal articles critical of Rosgen’s work, see Doyle and Harbor 2000, Gillilan 1996, Juracek and Fitzpatrick 2003, Kondolf 1995b, Kondolf 1998, Kondolf et al. 2001, Miller and Ritter 1996, Roper et al. 2008, Sear 1994, Shields et al. 1999, Simon et al. 2007, and Smith and Prestegaard 2005. For conference presentations, see Ashmore 1999, Doyle et al. 1999, Simon et al. 2005, Shields and Copeland 2006, Simon 2006, Simon and Langendoen 2006, and Kondolf 2007. For national guidelines efforts where authors have argued against the broad application of Rosgen’s work, see Federal Interagency Stream Restoration Working Group 1998, Shields et al. 2003, Shields et al. in press, and Slate et al. 2007. For white papers on this topic, see Miller et al. 2001.

4

In response a cross-cutting coalition of academics, agency scientists, and university-

trained practitioners have thrown the considerable weight of their scientific legitimacy into a

concerted attack on Rosgen’s work. While the list of common criticisms is as exhausting as

it is exhaustive, the major themes are all focused on drawing the boundaries of the stream

restoration field with Rosgen firmly outside them. Critics argue that Rosgen’s knowledge

claims have no scientific basis, that he does not follow the norms of scientific practice, and

that his approach leads to inadequate restoration practice. Despite making these arguments

in a wide variety of venues – from peer-reviewed journals to short courses to national design

guidelines – Rosgen’s critics have been unable to stem the rising tide of support for his

work.

The methodological and political battle over Rosgen’s restoration approach has

defined the field of stream restoration over the last decade. The Rosgen Wars raise

questions about the role of uncertainty in environmental science, the importance of tensions

between basic and applied scientists, and the avenues of communication between scientific

and lay communities. Perhaps the most important questions raised by the Rosgen Wars,

however, are fundamental to their existence. How has a consultant with little formal

academic training been able to effectively supplant the role of universities as the consecrated

providers of scientific knowledge and education? Conversely, why have Rosgen’s critics

been so startling ineffective at using their scientific authority to dismiss his work?

I argue in this dissertation that Rosgen has been so successful, and his critics so

ineffectual, because he has provided structure that the stream restoration field desperately

needs: a common language for communication, a set of shared methods and standards of

practice, and the primary means of disciplinary reproduction and training. The Rosgen Wars

are thus not simply a conflict over truth and practice, but also a struggle for control over the

5

field of stream restoration, and the authority to produce new knowledge. To address these

issues, I draw on work in Political Ecology and Social Studies of Science, and on the work of

Pierre Bourdieu.

A. Theoretical Framework

In the chapters that follow, I work in two areas of the overlapping territory between

Social Studies of Science and Political Ecology. First, I explore issues of deep interest to

both fields: the construction of scientific expertise, and the changing character of the

university with the rise of neoliberalism. Second, I draw heavily on the work of French

sociologist Pierre Bourdieu who, through his focus on the character and political economy

of science, formed a potential one-man bridge between Social Studies of Science and

Political Ecology that is worth testing.

1. Political Ecology and Social Studies of Science

Over the last decade there have been a number of calls to bring Political Ecology and

Social Studies of Science (STS)3 analyses together (Watts and McCarthy 1997, Taylor 1997,

Demeritt 1998, Braun and Castree 1998, Forsyth 2003). Doing so has proved complicated

in practice for two main reasons. First, the fields have very different research sites: STS

typically focuses on sites of scientific production, most often in the 1st world, while Political

Ecology typically focuses on sites of conflict over access to natural resources, most often in

3 STS is the commonly used acronym for social studies of science, which is conspicuously lacking in a word to abbreviate with ‘T’. There is some conflict about what STS actually abbreviates. The majority of STS scholars choose the less political ‘Science and Technology Studies’. More politically-motivated scholars argue that STS abbreviates ‘Science and Technology in Society’.

6

the 3rd world. Second, there is the incompatibility between STS’ theoretical approaches,

often drawn from phenomenological traditions, and Political Ecology’s common sources of

theory, usually structuralist and post-structuralist. These incompatibilities are not small, but

neither are they insurmountable.

STS is built around the claim that science is not simply an objective mirror of reality

but a negotiated product of social and political forces (Latour & Woolgar 1986). STS

scholars study the ways in which science is, as Stephen Jay Gould memorably described it,

“practiced within a constraining and potentiating set of social, cultural, and historical

circumstances.” (Gould 2000) It is thus very helpful in providing a more nuanced

understanding of science as not only a physical, but also a profoundly social and political

enterprise. But while post-Mertonian STS got off to a political start with the Edinburgh

School’s interest-based analysis, more recent STS work is notably silent about the influences

of political-economy in scientific practice.4

Instead of pursuing the influences of political-economic relations, in the last three

decades STS has famously followed the scientist, examining science as a social practice in

labs (e.g., Latour 1987, Knorr-Cetina 1981, Shapin and Schaeffer 1985, Traweek 1988),

medical facilities (Star 1989, Epstein 1996), science museums (Haraway 1989, Star and

Griesemer 1989), and the regulatory and judicial process (Jasanoff 1990 and 2004, Hilgartner

2000, Ong and Glantz 2001). A few intrepid souls have even analyzed the initial

construction of environmental science in the field (Latour 1999, Kohler 2002, Raffles 2002).

But STS thus far has little to say about how science is practiced and deployed when it is no

longer under the control, however partial, of the scientists and institutions that produced it.

4 Epstein’s Impure Science and the work of Haraway, Mirowski, and Restivo are important exceptions.

7

How is scientific expertise constructed and sustained far from lab benches and statistical

modeling tools?

This is a question with which Political Ecologists are all too familiar. The interested

deployment of science to operationalize imperial and state exploitation of local natural

resources is a common theme in this body of work (Blaikie 1985, Hecht 1985, Fairhead and

Leach 1996). Despite a consistent engagement with the use of science and scientific

expertise as blunt instruments, Political Ecologists have until recently focused relatively little

attention on the initial production of that science. From its inception in the mid-1980s,

Political Ecology has criticized the interested nature of government- (particularly colonial

government-) produced science. Unfortunately, while criticizing the biased character of

dominant paradigms in environmental science, many of these authors were less reflective

about the truth status of the counter claims they produced in collaboration with local people

and natural scientists (Bell and Roberts 1991, Sullivan 2000), or through their own

environmental research (Hecht 1985, Fairhead and Leach 1996, Robbins 1998, Turner 1999).

In the last five years, this picture has shifted dramatically as Political Ecologists began to

more fully embrace STS. Political Ecologists, particularly but not exclusively in England,

have recently begun to produce far more sophisticated analyses of environmental science

(Fairhead and Leach 2003, Forsyth 2003, Goldman 2004, and Robertson 2006).

Clearly, these two fields have much to say to each other. I have tried to bring them

into conversation around their shared concern with the issues at the heart of this

dissertation: the construction of scientific expertise and its relationship to the changing

character of the university and public science more generally.

8

a. Expertise

The study of scientific expertise has a long history in STS, but has recently come very

much to the fore in debates over the appropriate level of inclusion of lay people in decisions

about science and technology policy. The STS debate over public engagement in science

and technology (or PEST, as it is sometimes referred to) has centered around issues of

democracy. How can the restriction of decisions about science policy to those with

expertise in that science be justified politically in a democracy? Conversely, how can the

extension of participation to (by definition) scientifically-uninformed lay people be justified

on technical grounds? This is typically referred to as the problem of extension.

In 2002, Collins and Evans first published a proposed solution to the problem of

extension that has remained at the center of STS debates on expertise for the last five years,

boosted once again by their 2007 publication of a monograph fleshing out their initial

arguments. Collins and Evans take what they describe as a realist position: experts must

“know what they’re talking about.” (Collins and Evans 2007, p.2). “Know” is the key word

in that phrase: the authors are explicitly trying to produce an epistemological answer to the

question of how to discern who is an expert and who is not.5 The distinctive move they

make is to expand the pool of those who could be considered to know what they were

talking about from merely those with scientific credentials to those who have relevant

expertise gained through experience (this is an interesting parallel to Political Ecology

arguments for indigenous technical knowledge). As Collins and Evans point out, their

approach is both conservative and liberal:

On the one hand, it is a conservative approach in that it restricts participation in the technical aspects of technical debate. On the other hand, it is a liberal approach in that it admits to the company of those who know what they are

5 For example, the heading of their section on how one becomes an expert is, “Acquiring Expertise: Five Kinds of Face-to-Face Knowledge Transfer.” (Collins and Evans 2007, p.40)

9

talking about many experience-based experts whose contribution would not have been countenanced in earlier times. (Collins and Evans 2007, p.113)

What their approach is not, however, is an exploration of the process by which

someone can gain recognition as an expert. Collins and Evans assume that the tacit

knowledge ubiquitous in a society should be sufficient to distinguish between those

deserving of expert status and those that are not. Their argument is that non-experts who

accord scientific expertise to someone who does not hold it are simply making a social error,

missing the social cues set up to recognize consecrated science:

[T]hose with little scientific knowledge can sometimes make what amounts to a technical judgment on the basis of their social understanding. The judgment turns on whether the author of a scientific claim appears to have the appropriate scientific demeanor and/or the appropriate location within the social networks of scientists and/or not too much in the way of a political and financial interest in the claim…. Poor social judgments are the problem with those who believe, in say, newspaper astrology as a scientific theory. They are making a social mistake – they do not yet know the locations in our society in which trustworthy expertise in respect of the influence of the stars and planets on our lives is to be found. (ibid., p. 45 and fn 1, p.46)

As the analysis in the body of this dissertation will demonstrate, this is not an adequate

description of the dynamics at work in the Rosgen Wars. The thousands of supporters who

accord Rosgen expert status despite the fact that he has neither an appropriate scientific

demeanor nor an appropriate scientific location cannot be simply described as making a

social mistake. While Collins and Evans’ arguments that expertise should be grounded in

experience clearly resonate with Rosgen’s supporters, these authors’ deliberate decision to

focus on expert knowledge and thus avoid consideration of the social and political-economic

forces involved in establishing expertise limits their applicability to the questions at the heart

of this dissertation.

Unsurprisingly, the Political Ecology literature on expertise is a mirror image of the

STS literature: instead of an epistemological focus on the grounds of expertise, authors such

10

as Goldman (2005), Mitchell (2002) and Mosse (2005) take a deliberately political-economic

view of how expertise is constructed, and a far more cynical view of the uses to which it is

put.6 For these scholars, expertise is not of interest as a possible justification for expanding

democratic participation in technical decisions, but as a means of political and economic

exclusion and expropriation. Mitchell, for example, describes the development of new

forms of expertise as an explicit part of the colonial project, whose outcome was not

improved efficiency but the stripping of control from locals:

The new statistical methods did not generate a more accurate knowledge, or even a greater amount of knowledge…. [What they] accomplished was a series of removals, of transferences, that tried to shift the site where calculation could occur. In place of the village surveyor walking with his rod along the boundaries of each plot, the employee in the survey office would walk his dividers across the feddan comb, stepping backwards and forwards from the thread to the scale…. [It] represents a reformatted knowledge, information that has been translated, moved, shrunk, simplified, redrawn. What is new is the site, and the forms of calculation and decision that can take place at this new site. (Mitchell 2002, pp. 115-116)

In his work on the World Bank and its key role in the production of development

expertise and science, Goldman describes mechanisms for the construction of that removed

expertise. He argues that to expand its reach, the World Bank effectively created

development science:

[N]o internal institutional support system existed to make McNamara’s audacious claims ring true that ‘investing in the poor’ was the most efficient route to growth with equity in the third world. Nor was anyone in the larger community of constituents, development officials, or economists prepared to accept the financial soundness of such investments. That was a perception that McNamara’s team had to create. The World Bank thus became central headquarters for research, economic modeling, data collection, report writing, and dissemination of information on the so-called less developed world…. (p.77)

6 It is worth noting that of these authors, Goldman and Mosse might consider themselves part of the larger Political Ecology field, or at least near neighbors; while Mitchell is widely read among Political Ecologists, his relationship to the field is more distant.

11

To ensure acceptance of the data and analysis thus produced, the Bank set up a

gigantic interlinked web of transnational policy networks. By training most of the key

players in development, and becoming the sole producers of the datasets used to justify

development policies, the WB created a situation in which it could ensure the rapid take-up

of its development “science” and claims (Goldman 2005, pp.85-86).7 Goldman writes that

the Bank’s network of training centers has,

trained approximately 50,000 scientists, many of whom subsequently took up prominent positions as ministers of state, agriculture, and finance, as well as CEOs and research directors for major multinational firms…. As the Bank reinvented the professional landscape in which the international agricultural scientist worked into one flush with financial and political rewards, this science-industry-government network enabled the Bank to overcome the historical skepticism of capital markets to invest in rural production.” (ibid., p.86)

At this point, Goldman argues, the weight of the Bank’s claims to expertise and the

hegemonic strength of its vision of development science have shifted the center of gravity in

the academic disciplines that study development:

The problem here lies at the crossroads of knowledge production and use…. [A]s this becomes the big business of development, it draws in hundreds of applied and development anthropologists as its hired practitioners. Thus the type of anthropology being produced for the World Bank becomes a strong influence on the trajectory of anthropology scholarship, writ large. (ibid., p.139)

This Political Ecology approach to the construction of expertise and its political-

economic motivations is central to the analysis of the Rosgen Wars that follows. The recent

STS focus on the epistemological grounds of expertise is also present, however. It is not

enough to analyze the ways in which Rosgen’s expert status was constructed and promoted,

7 The parallels between my and Goldman’s arguments about the creation of scientific expertise outside the academy are startling. Both Rosgen and the World Bank have built their legitimacy by creating broad networks of support produced by staking out a position as the primary trainer of practitioners in their field. Both create vast quantities of data which are considered proprietary and therefore shared only in small doses. Both have sufficient weight that they are shifting the center of gravity in the academic fields that share their substantive focus. Neither are subject to much monitoring.

12

it is also critical to analyze the substantive reasons why participants in the stream restoration

field are convinced by his claims of scientific legitimacy.

b. The Neoliberal University

Political Ecology and STS are on much firmer shared ground when it comes to

neoliberalism and the commercialization of science.8 STS scholars such as Kloppenberg (2004),

Krimsky (2003), Sismondo (2004), and Mirowski (2005) have written scathing works on the

increasing privatization of scientific practice and the commercialization of the university.

Political Ecology scholars, while typically not focused on the university per se, have made

neoliberalism and its discontents one of their primary strands of research over the last five

years (Mansfield 2004, Robertson 2004, Bakker 2005, Prudham 2005, and McCarthy 2006).

A combination of STS scholars’ focus on the commercialization of science and Political

Ecology’s careful analysis of the larger context of neoliberalism is a productive lens with

which to consider some the larger implication of the Rosgen Wars

The best example of such a combined approach is Mirowski and Van Horn’s

landmark 2005 article, “The Contract Research Organization and the Commercialization of

Scientific Research.” The authors, both scholars of the social and political history of

economics, argue that the STS literature to date is missing the boat because the university is

not at the forefront of the commercialization of science; private sector science is.9 Drawing

on their deep understanding of the roots and forms of neoliberalism, Mirowski and Van

8 The earliest use of the term ‘neoliberal university’ that I have found is from a 1997 article on a Chilean student uprising (Godoy-Anativa 1997). In the last few years, it has become an increasingly popular topic in critical studies of education and labor. See for example Ciafone 2005, Davies et al. 2006, and Canaan and Shumar 2008. 9 Studies of neoliberalism’s impacts on the university tend to focus on four factors: increasing dependence on private sources for research funding; the rise of patenting and material transfer agreements (MTAs) as a form of enclosure of the commons; the casualization of the workforce, as universities increase their dependence on non-tenure track faculty; and the rise of the “marketplace of ideas” paradigm, which emphasizes the economic rather than the scholarly value of research.

13

Horn argue that both positive and negative perspectives on the commercialization of the

university (the Economic Whigs vs. the Mertonian Whigs, in their framing) are,

unduly restricted by the unexamined presumption that the university is the primary field upon which the privatization of research has played out. We contend that the re-engineering of the structures of scientific research since the 1980s has occurred on many different fronts, that it has been a subset of larger political and economic trends, and that universities have been relative latecomers to the thorough-going privatization of the conduct of science. In many instances, the transformation was nurtured by the creation of new social structures of research, which act as prototypes outside the university: new forms of intellectual property, new communication technologies, new research protocols, new career paths, and new institutions of command and control…. We believe that the best way to encourage debate over the possible consequences of the commercialization of science since the 1980s is to pay more attention to functional innovations in the organization of scientific research within the corporate sphere. (Mirowski and Van Horn 2005, pp.504-505)

This insistence on the private sector as harbinger of changes to come in public sector proved

crucial in my analysis of the Rosgen Wars.

2. Bourdieu’s Field Concept and the Inherent Nature of Conflict

In addition to exploring relations between concepts of expertise and the university in

Political Ecology and STS, the chapters that follow draw on the work of Pierre Bourdieu.

Bourdieu was one of the major sociologists of the 20th century, and STS was one of the

many sociological subfields in which he participated. Bourdieu’s writings on the

construction of, and dynamics within, scientific and other fields provide a useful bridge

between Political Ecology and STS because unlike the vast majority of STS theorists,

Bourdieu gives pride of place to the political economic factors so central to Political

Ecologists. It is important to note, however, that despite his enormous influence on

Sociology and Anthropology, Bourdieu is not a central figure in either STS or Political

Ecology. For its part, the STS world is profoundly ambivalent about Bourdieu. On the one

14

hand, his classic 1975 article on the structure of the scientific field is included in the major

STS anthologies. On the other hand, his later critiques of STS researchers for taking

positions that he considered too strongly constructivist alienated many who might have

made use of his powerful analytical frameworks. Within Political Ecology, Bourdieu is not

much discussed (c.f. Sayre 2002), though as the larger Critical Geography community has

started to engage more with his work over the last five years, this may shift (see for example

Sayer 2003, Jackson et al. 2007, Pain 2007).

Bourdieu’s key intervention in STS was also the first formulation of his field concept

which, along with capital and habitus, form the heart of Bourdieu’s analytical framework

(Bourdieu 1975). According to Bourdieu, a field is a bounded, structured social arena which

provides a particular set of opportunities and constraints to those who participate in it.

Struggle – to delimit the boundaries of the field, to determine conditions of entry, and most

especially to define the types of capital of most value – is a defining feature of Bourdieu’s

profoundly agonistic field concept. This struggle takes place within the hierarchical structure

provided by fields, each of which is organized around an axis whose poles Bourdieu defines

as autonomous and heteronomous. At the autonomous end of any field are those actors

whose production is controlled most thoroughly by the forms of power and prestige (capital)

specific to that field; at the heteronomous end are those whose production is shaped primarily

by other forces, typically economic. According to Bourdieu, tensions between agents who

hold positions at the autonomous and heteronomous poles of a field provide one of the two

primary motors of struggle and change in a field; the other is the intrinsic conflict between

newcomers and incumbents.

Bourdieu’s understanding of fields and the structured struggles that define and

transform them prove to be powerful tools for analyzing the Rosgen Wars, as I will

15

demonstrate in the chapters that follow. At the same time, analysis of the Rosgen Wars

highlights the overly schematic character of Bourdieu’s theorization of the sources of

struggle, and I suggest ways in which the field concept could be fleshed out in Chapter 7.

B. The Structure of this Dissertation

The first three chapters of this dissertation introduce Rosgen’s work and the field of

stream restoration more generally. Chapter 2 begins with a brief history of restoration

practice in the U.S., and then introduces the major components of Rosgen’s Natural Channel

Design (NCD) approach: the classification system, standards of practice for design, and

specific reach-scale restoration techniques. Chapter 3 describes how Rosgen’s NCD

approach has been treated in practice at the national scale through consideration of

nationwide design guidelines efforts, and at the state scale through examination of

restoration practices in North Carolina, a center of Rosgen-oriented stream restoration.

Chapter 4 provides context for the earlier chapters by examining the stream restoration field

as a whole through the lens of Bourdieu’s field concept. I argue here that despite its short

time scale, Rosgen’s short course series should be seen as producing a distinctive habitus

among his students that has come to control the direction of the stream restoration field.

The second half of the dissertation focuses on the fight over Rosgen’s work.

Chapter 5 lays out in detail many of the main critiques of Rosgen’s work, and to the extent

possible assesses their fairness. Chapter 6 describes the ways in which these arguments and

various species of capital have been mobilized in practice in the particular theaters of the

Rosgen Wars, such as articles, conferences and short courses. In Chapter 7 I argue that the

16

two most puzzling aspects of the Rosgen Wars – the interlinked questions of why Rosgen’s

work has been so widely adopted, and conversely why the consecrated institutions of science

have been so ineffective in their efforts to block that adoption – have the same explanation.

Rosgen has provided a structure for the burgeoning stream restoration field that includes a

lingua franca to unify the field, standards of practice that enable regulatory staff to carry out

our fundamentally contradictory mandates for stream restoration, and the primary venue for

training field participants. He has thus usurped the traditional role of public science, locating

the core of the new field firmly in the private sector. I then use the analysis of the Rosgen

Wars presented over the previous chapters as a lens for reconsidering Bourdieu’s work on

fields, and ways in which the concept should be expanded and deepened. Finally, in the

Conclusions I address the implications of the Rosgen Wars for the construction of scientific

expertise and the role of public science in this age of neoliberalism, arguing that there are

very good reasons to view Rosgen not as a fluke, but as a portent of things to come.

C. Research Methods

I employed a variety of methods in conducting this research. One primary

component was semi-structured interviews investigating the history, usage in practice,

justifications for, and critiques of, Rosgen’s approach. More than 60 participants in the

Rosgen Wars – academics, practitioners, and agency and NGO staff – took the time to speak

with me at length. Many are quoted in the text that follows; because of the high level of

controversy around Rosgen’s work, some of my interviewees chose to speak off the record,

particularly consultants. A complete list of the people I interviewed at least partially on the

17

record can be found in Appendix A, along with more information about the interview

process.

To see how Rosgen’s work was presented and taught by opponents and proponents,

I conducted participant observation at restoration short courses and conferences. I attended

three of the restoration short courses that are so central to the Rosgen Wars: the

introductory short course led by Matt Kondolf and Peter Wilcock in Bishop, CA in October

2004; the Level I Rosgen course held in Santa Cruz, CA in January 2005; and the Level II

Rosgen course held in Fayetteville, AR in November 2006. I also attended several

conferences with stream restoration strands, including the first National Conference on

Ecosystem Restoration in Orlando, Florida in December 2004. I was privileged to be a

graduate student observer at the small 2003 conference, jointly-sponsored by the National

Academy of Sciences and the National Center for Earth-Surface Dynamics, to plan for an

updated version of the 1992 NRC report on stream restoration. I also had the rather odd

experience of conducting fieldwork at sessions on restoration at the annual meeting of my

own discipline, the Association of American Geographers! This part of my fieldwork

allowed me to observe how the arguments and species of capital at the heart of the Rosgen

Wars were mobilized in collective practice.

To delve more deeply into the motivations of short course students and their

understanding and utilization of the material presented, I conducted a mail survey of

participants in the Level I Rosgen course and the alternative short course taught by Kondolf

and Wilcock. Selected results of this survey can be found in Chapter 4. The survey form

can be found in Appendix A.

To explore how and why Rosgen’s approach is used in practice by the agencies that

fund, regulate and carry out restoration projects, I conducted a detailed case study of

18

adoption patterns among local, state and federal agencies in North Carolina, a hot spot of

restoration activity. This material is presented in Chapter 3 and the Conclusions.

Finally, I conducted mini-case studies of several restoration projects in California

that have been mobilized in the Rosgen Wars, which allowed me to explore how the politics

of the Rosgen Wars play out in practice and compare factional opinions with the physical

reality of restoration projects. In each case, I conducted interviews with project participants,

reviewed project documents, and collected and/or reviewed geomorphological survey data.

This data has deeply informed my understanding of the Rosgen Wars.

D. Why Restoration Matters

In conducting my dissertation research, one of the most common questions asked of

me was: why stream restoration? Given the incredibly compelling issues around human

access to fresh water worldwide, why focus on the seemingly far less critical effort to restore

fresh water habitat? Despite the low level of attention to ecological restoration in the social

science literature (c.f. Baldwin et al. 1994, Katz 1998, Smith 1998, Eden et al. 2000, Gobster

and Hull 2000, Throop 2000, Gross 2003, Robertson 2000, 2006, and 2007 ) I believe that

there are at least three reasons why stream restoration is worthy of study.

First, when Western scholars list the big dualisms we inherit from our philosophical

traditions, nature/culture – the distinction between the human, created sphere, and the natural,

given sphere – is invariably among them. Restoration dismantles this divide by insisting that

humans are part of nature, with responsibilities to undo harm to the physical environment

the same way we are responsible for undoing harm to other people. In stark contrast to the

19

idealization of wilderness as a pure nature removed from human impact, restorationists see a

world in which human intervention is unavoidable and, most importantly, can be a force for

a good. The study of restoration is thus of philosophical interest because it grapples with

questions fundamental to Western intellectual traditions.

Second, ecological restoration is interesting historically as a third wave of the

environmental movement. The first wave of environmentalism in the U.S. could be loosely

described as conservationism: the idea that we should carefully husband our renewable

natural resources in order to preserve their ongoing economic utility. The second wave of

environmentalism in the U.S., which certainly overlapped the first, was preservationism or

the wilderness ethic: the idea of nature as something physically and spiritually separate from

humans, and the ideal of preserving its purity from the evils of anthropogenic influence.

This is the most common model of environmentalism in the U.S. today.

Ecological restoration is the standard bearer of a third wave of the environmental

movement, one that could be dated with hesitation to the later 1800s, and with more

confidence to Aldo Leopold’s The Sand County Almanac, though it has really only taken off in

the last twenty years or so. As described above, this third wave envisions human beings as

part of the natural community: not necessarily good, not necessarily bad, just one (very

powerful) component with responsibilities concomitant with the scope of our influence.

This far more positive view of the appropriate relation of humans to their natural

environment is very appealing and has made restoration a force to be reckoned with in the

environmental movement, and a growing impact on the landscape itself. Thus restoration is

also worthy of study because of its transformational role in the environmental movement.

Third, ecologically speaking the physical practice of restoration is worthy of study. It

is a truism that negative anthropogenic change has now touched every nook and cranny of

20

our world. Restoration points towards one way to approach those changes: with positive

human intervention and management. Stream restoration attempts to address a wide range

of anthropogenic impacts, from the mercury and enormous sediment pulses that serve as

toxic souvenirs of mining booms, to the rigid sterile landscapes created by a century of

hydraulic engineering. If successful, restoration could become a powerful force for positive

anthropogenic change. Thus restoration is worthy of study as a potential beacon of hope.

Within the realm of restoration ecology, stream projects are unusual in two ways.

First, stream restoration is probably the most common, and certainly the most popular, form

of ecological restoration; it is the flagship of the restoration movement. This status holds a

certain irony because of the second way in which stream restoration projects are unusual:

more than any other type of restoration, they embody the conflicting human goals for nature

of both healing and control. As referred to above in the introduction to this chapter, stream

restoration projects almost always include goals that have nothing to do with improving

ecological health and may even be antithetical to it, such as controlling flooding and

preventing bank migration; there are very real conflicts between human desires and

hydrological process. For example, research over the last decade has determined that large

woody debris provides crucial habitat for fish and invertebrates (Montgomery et. al. 1996).

But getting large woody debris into rivers requires erosion to undercut banks and topple

trees into the water, while one of the major human goals in river restoration is bank

stabilization.10 Similarly, to restore a river to hydrologic stability, it needs room to move.

But while natural channels typically migrate across the landscape, private property does not,

and owners of shrinking lots on the outside of meander bends are often unreceptive to the

idea of letting the river wander. Thus in the practice of stream restoration, the balance

10 It also requires leaving large trees in place, which is in direct conflict with logging practices that often cherry pick the biggest trees.

21

between human desires and natural processes is under active and constant negotiation,

providing an evolving test case for how we can respond to critical environmental issues

where human interests are too vital to be set aside.

At the same time, it could be argued that stream restoration is worthy of study for

the ways in which it is typical not just of ecological restoration, but of many the more recent

environmental sciences. These relatively new environmental sciences, focused on topics

such as climate change, environmental mutagens, and biodiversity, study highly complex

systems. Scholars in these fields are forced to accept high levels of uncertainty; they must be

content with error bars and estimates of probability rather than the certainty that

characterized earlier scientific claims. Also, like stream restoration these fields tend to be

newly developed, driven by the environmental commitments of the public and scientists

themselves, rather than catalyzed by scientific breakthroughs. Finally, the tension mentioned

above between the human desires to respect nature while at the same time shaping it to

human purposes, is typical of these newer environmental sciences as well. Thus study of the

changing structure of the stream restoration field can shed light on the dynamics of other

environmental sciences.

E. A Note on Terminology

Some of the terminology I use in this dissertation requires explanation. The controversy

over Rosgen’s work spills over into disagreements about what to call the approach itself and

those who use it. In the academic literature, authors typically refer to the combination of

Rosgen’s classification system and design approach as the “Rosgen Method.” In part to

22

disassociate the method and the man, and in part out of recognition that many key elements

of his design approach spring from previous work by Luna Leopold and others, Rosgen and

his advocates prefer the name “Natural Channel Design.” There is, however, widespread

disagreement over whether Rosgen’s approach is synonymous with Natural Channel Design,

or merely a subset of a larger Natural Channel Design movement. The term is appealing,

and many practitioners wish it to be a big tent under which all restoration approaches

alternative to traditional hydraulic engineering can shelter. I have thus chosen to use the

more neutral phrase “Rosgen approach” in most cases. When I use the term Natural

Channel Design (NCD) it is in reference to Rosgen and his supporters unless I specify

otherwise.

Choosing how to refer to those who use Rosgen’s design approach is also

complicated. The level of personal devotion Rosgen inspires among many of his students

has led to some widely-used nicknames, such as Rosgenites, Rosgenauts, and Rosgefarians.

While these labels are certainly catchy, they are not respectful. A more appropriate label

could be “Rosgen students,” as a) the vast majority of Rosgen’s supporters have taken one

or more of his short courses, and b) course attendance provides a strong common bond;

but not all Rosgen supporters have taken a short course, and there are many who have who

consider Rosgen their colleague, not their teacher. In the interest of accuracy and good

manners, I have chosen to refer them as “Rosgen supporters.”

Similar difficulties arise in trying to label those who actively oppose Rosgen and his

work. Despite a level of fervor that many will admit reaches religious levels, the opposition

has not yet been graced with any pithy nicknames. Characterizing them by sector as

“academics” collapses immediately, as the opposition includes not only professors but also

consultants, state and federal agency staff and scientists, and non-profit staff; and, too, there

23

are a few Rosgen supporters in the hallowed halls of academe. Characterizing the

opposition by discipline as “fluvial geomorphologists” has the advantage of reaching outside

the university to describe all of those with a specific academic training wherever they may be

employed, but is inaccurate. First, the discipline of fluvial geomorphology, like all sciences,

can be divided between basic and applied researchers, and the former are for the most part

blissfully unconcerned with the Rosgen Wars. Second, Rosgen opponents can be found not

only among fluvial geomorphologists, but also among engineers, hydrologists and biologists;

fluvial geomorphologists have no monopoly. Third, Rosgen teaches fluvial geomorphology,

so many of his students call themselves fluvial geomorphologists. Clearly allowing either

side in the debate to claim the disciplinary title could only result in confusion.

Still searching for a descriptive label that would clarify rather than confuse, I

canvassed opposition members for suggestions. Several suggested I refer to them as the

“research community;” but what is it that Rosgen and his supporters who test hypotheses

and develop new knowledge are doing if not research? In the end, the diversity of both

camps has forced me to adopt the dull but uncontroversial label the “Rosgen opposition,”

and this is how I shall refer to the anti-Rosgen camp for the remainder of this dissertation.

One final note: it will strike some readers that I have made a sudden and unexplained

leap by equating what Rosgen teaches with the field of stream restoration in its entirety. As

Rosgen’s critics (and many of his supporters) have persuasively argued, stream restoration

must address a broad range of factors from the physical to the ecological to the social in

order to be successful. Rosgen teaches only fluvial geomorphology and hydrology; equating

that with stream restoration as a whole is clearly incorrect. And yet that is exactly what is

happening in many areas of the U.S. right now. For a growing number of people, stream

restoration is synonymous with Rosgen. That is the issue at the heart of the Rosgen Wars.

24

happening in many areas of the U.S. right now. For a growing number of people, stream

restoration is synonymous with Rosgen. That is the issue at the heart of the Rosgen Wars.

25

Chapter 2: The Rosgen Classification System and Design Approach

“I’d like to make this sound simple, but it’s not simple. It is doable, but you’re going to have spend a lot of time in the field, in measurement and observation.”

Dr. David Rosgen, Wildland Hydrology

Depending on which camp you occupy, the most widely-used method for classifying

and restoring streams in the U.S. today was developed by a consultant with more than 40

years of field experience and an unparalleled ability to read riparian landscapes, or it was

cobbled together from out-of-date science by a cowboy with a formidable talent for self-

promotion. Chapters 5, 6 and 7 will address the debate over Rosgen’s work; this chapter

takes a step back to describe its content and origins. Of what does Rosgen’s approach

consist? How was it developed and disseminated? How do people in the restoration field

account for Rosgen’s success? Before I answer these questions, however, it is worth taking

time for a brief history of the stream restoration field to provide some context for Rosgen

and his work.

A. A Brief History of Stream Restoration

Until the early 1980s, stream restoration1 in the United States consisted largely of the

placement of in-stream structures, such as weirs or vanes, with the goal of improving fish

1 Although less prominent today, there has in the past been a great deal of debate in the stream restoration field over the proper definition of “restoration.” This debate is not relevant to the fight over Rosgen’s work,

26

habitat. Although in the early part of the century a substantial number of habitat

improvement projects were designed and implemented by wealthy private landowners,

habitat improvement from the 1930s through the 1970s was for the most part a public sector

project, conducted by staff at state resource agencies, the Civilian Conservation Corps

(CCC), the U.S. Forest Service (USFS), and the U.S. Fish & Wildlife Service (USFWS)

(Thompson and Stull 2002). 2

The 1930s were a key decade for such public sector efforts in a number of ways.

The Institute of Fisheries Research at the University of Michigan produced influential

guidance documents promoting the use of in-stream structures, and the first USFS

handbook on habitat structures was released (Arthur 1936). The 1930s were also a

particularly productive decade in terms of sheer numbers of structures constructed. The

CCC, for example, ‘improved’ more than 7,950 km of streams and built more than 31,000

in-stream structure between 1933 and 1936 alone (Thompson 2005, p.38). Legislatively, the

1936 Flood Control Act expanded the USACE’s jurisdiction to all navigable waterways in

the U.S., setting the stage for what Riley refers to as the “Golden Age of Channelization.”

According to Riley, by the 1970s there were 20,724 mi of rivers and streams channelized and

5,897 mi of levees constructed by the USACE and the NRCS. (Riley 1998, p.220)

because both sides agree that the goal of restoration is to restore ecosystem function, not to recreate the pre-contact landscape. 2 Thompson and Stull wrote a very useful history of early stream restoration work in the U.S. (Thompson and Stull 2002), but no one has yet written about the history of stream restoration over the last four decades as it coalesced into a field, became a significant component of American environmentalism, and developed into a lucrative market. Thus the remainder of this section depends on two main sources: 1. limited information on stream restoration history from sources focused on other topics (Brookes 1988, Riley 1998, and O’Neill 2006) 2. interviews focused on the history of the field with founding principals of the earliest restoration consulting firms in the U.S. and senior federal agency staff: Dr. Philip Williams of Philip Williams Associates in California, Greg Koonce and Dale Miller of Inter-Fluve, Inc. in the Pacific Northwest, James MacBroom of Milone & MacBroom in the Northeast, Robbin Sottir of Sottir & Associates in the Southeast, Dr. Craig Fischenich of the USACE Waterways Experiment Station, and Dr. F. Douglas Shields of the USDA-ARS National Sedimentation Laboratory.

27

This rampant channelization and the rise of environmentalism in the U.S. sparked a

counter-movement. Opposition to traditional hydraulic engineering solutions to flooding

problems began in the U.S. in the late 1960s and spread globally (Brookes 1988, p.20).

Support for a more natural approach came from a number of sources. In 1967 the

Wisconsin Department of Natural Resources published new holistic guidelines for managing

trout habitat, the first alternative to in-stream habitat structures in almost a hundred years

(White and Brynildson 1967). Perhaps even more crucially, the late 60s and early 1970s set

the stage legislatively for the modern restoration movement. In 1968, the Wild and Scenic

Rivers Act put forward the first strong statement for the preservation of rivers, declaring

that,

the established national policy of dam and other construction at appropriate sections of the rivers of the United States needs to be complemented by a policy that would preserve other selected rivers or sections thereof in their free-flowing condition to protect the water quality of such rivers and to fulfill other vital national conservation purposes. (Wild and Scenic Rivers Act. P.L. 90-542.)

This startling departure from the traditional view of highest and best use of water

resources was followed in short order by the National Environmental Protection Act

(NEPA) in 1969, the Clean Water Act (CWA) in 1972, and the Endangered Species Act

(ESA) in 1973. Thus by the early 1980s, when this legislation began to be fully implemented,

most streams in the U.S. had water clean enough to once again support a range of aquatic

organisms, and some of those organisms now had the legal right to be there. Further, thanks

to NEPA any impacts to those streams (at least from federal projects) had to be disclosed to

a public that was increasingly opposed to channelization and other heavily-engineered

solutions to flooding problems.

This is the context in which the earliest stream restoration and bioengineering firms

were established, such as Inter-Fluve, Inc. (1982), Robbin Sotir and Associates, Inc. (1982),

28

and Philip Williams & Associates and Milone & MacBroom (both pre-existing firms whose

stream restoration practice began in late 1970s).3 With the exception of Robbin Sotir, who

had extensive training in soil bio-engineering in Europe, the pioneering group of stream

restoration consultants found themselves with little guidance about what might or might not

work. Several of them described going to conferences and eagerly canvassing other

attendees about whether they thought channel relocation projects were workable, and if so,

how. Much of the early work was done with an experimental, trial and error mindset.

These firms’ initial clients came in three main groups. Private landowners,

particularly the wealthy owners of trophy ranches in search of pristine trout streams, were a

big initial client base in the Northwest. Suburban community groups were a common client

group for all of the early restoration firms once implementation of the CWA cleaned up the

water sufficiently for streams to become potential amenities. Phil Williams’ first stream

restoration job, for example, was helping a suburban community group fend off a USACE

channelization project. A final common group of clients was private developers who, thanks

to enforcement of Section 404 of CWA, could no longer move or culvert any

inconveniently-located streams on property they wished to develop.4 These three sources,

combined with spending by federal agencies such as NRCS and USFS, got the stream

restoration market up and running. By 1986, even USACE got into the restoration game.

3 The backdrop to the growth of the stream restoration consulting industry is, of course, the development of the environmental consulting industry as a whole. Unfortunately, no has written a history of environmental consulting, nor are there any sources I could find on current or former numbers of environmental consultants. There are a number of sources with generic statements that NEPA catalyzed the growth of industry (see for example Hanson 1976, p. 627, Nelkin 1976 p.39 and 1977 p.31, and Todd 1980, p.61), but sadly these authors do not provide much in the way of supporting evidence. One indicator is that the National Association of Environmental Professionals started in 1975, and began a certification program in 1979. 4 Section 404 of the Clean Water Act regulates the discharge of dredged or fill materials into riparian systems. To do so legally requires a permit from the USACE. Any proposed disturbance of an existing stream must be proved to be 1) the only feasible alternative and 2) not a significant source of degradation to the stream; in practice this is very difficult to prove. Thus permits typically require developers to compensate for avoidable impacts, leading to a tremendous demand for on- and off-site restoration as compensatory mitigation.

29

The 1986 Flood Control Act started a restoration program within the USACE that today

includes massive restoration projects on rivers such as the Mississippi and the Kissimmee.

In the 1990s, the restoration market took an exponential leap upwards in numbers of

projects and dollars expended, hitting the $1 billion/year mark and continuing to grow

(Bernhardt 2005, and Figure 3.1 in Chapter 3). In terms of guidance documents, the

National Research Council Report on aquatic ecosystem restoration was released in 1992,

and the far more detailed Federal Interagency Stream Restoration Working Group’s manual

on stream corridor restoration was released in 1998. The client base for early restoration

firms expanded to include a broad range of local, state and federal agencies, as well as a

greatly increased number of community groups.

In the first decade of the 21st century, the trends from the 90s have continued:

greater numbers of projects, more dollars expended and a broadening range of clients of

which an increasingly prominent component is local Storm Water Management agencies

forced into restoration work by the implementation of water quality standards under CWA.

A powerful trend, particularly in the Southeast, is the growth of the stream mitigation

banking industry. Stream mitigation banking allows developers with streams on their

properties to offset any impacts by purchasing restoration credits from for-profit companies

that restore damaged streams on a speculative basis. In states such as North Carolina,

Georgia and Virginia, stream mitigation banking is a major private-sector source of

restoration funding. EPA is promoting this model aggressively, perhaps presaging a major

shift in what has been until now a predominantly publicly-funded market.

30

B. Rosgen and his Work5

After completing a B.S. in Forestry and Watershed Management at Humboldt State

University in 1965, Dave Rosgen joined the USFS. Over the next ten years, he worked as a

forest hydrologist on the Clearwater National Forest in Idaho, the Beaverhead and Lolo

National Forests in Montana, and then went back to Idaho to work on the Kaniksu National

Forest. In 1975 he took over the forest hydrologist job on the Arapaho & Roosevelt

National Forests in Colorado, and there he stayed until he left the USFS in 1985.

Having regularly fished in the Clearwater National Forest as a child, he was shocked

by the changes he saw when he returned there in 1965. As he described it in the Level I

course I attended in January 2005:

After I went to college, I went back to the very ranger district, right on those tributaries below the Bungalow Ranger Station. You didn’t have to pack in anymore, you could drive right up. The Clear River was wider, shallower, and sandy. I looked upstream and saw huge swathes of clearcuts and road failures…. I went and yelled at the ranger for allowing so much damage, but he said that anything we do is minor compared to Mother Nature. That’s when I realized I was part of a pecking order. I was a peckee, and you know what that made him. I was an ineffective, frantic booger until I realized what I needed: data. Without the data, it’s just an opinion.

There was no handbook for the kinds of questions Rosgen wanted to answer about

why some streams could experience severe disturbance without taking substantial damage,

while other streams in the same watershed were profoundly affected. So he began to teach

himself, reading, measuring, and talking to everyone he could get to sit still and listen, most

notably geomorphologist Luna Leopold, who he went to see in 1968 with data from his

work on the Beaverhead Forest in Montana. During these years with the USFS, Rosgen

5 Unless otherwise noted, the description of Rosgen’s biography is drawn from The material in this section is drawn from conversations with Rosgen in August 2003 at the NCER/NRC conference in Minneapolis; his Level I short course in Santa Cruz, CA in January 2005; his Level II short course in Fayetteville, AK in November 2006; and a formal interview on 4.22.07.

31

developed the core of his classification system and design approach, refining them in

consultation with Leopold, Lee Silvey (his supervisor on the Arapaho & Roosevelt National

Forest), and other forest hydrologists.

In 1984, while working as a forest hydrologist on the Arapaho & Roosevelt National

Forest near Ft. Collins, Colorado, Rosgen was asked to do work that he considered

unethical. He refused to do it, and in response the Regional Forester issued him a directed

reassignment (an involuntary transfer) to a forest to which he knew Rosgen would refuse to

go. Rosgen duly refused the reassignment. Technically, he resigned; effectively he was fired.

Regardless of what you call it, after 20 years with the USFS Rosgen was out of a job and at

loose ends.

This situation seems to have lasted for no more than five minutes before his phone

started ringing. By 1986 he was teaching short courses at the University of Nevada-Reno,

USFS had been forced to hire him back on a consulting basis for the pivotal Division I water

rights case, and he had embarked on some of his first big restoration projects on private land

in Colorado. As I described in the previous section, this was the time period when

restoration was just beginning to take off as a field. The earliest firms had started to develop

the market, and Rosgen dived right into it, providing a crucial early source of training for

restoration practitioners. Since then, Rosgen has completed more than 50 restoration

projects, taught more than 10,000 students, and written two textbooks and dozens of papers.

Rosgen’s approach to stream restoration consists of three basic pieces: a

classification system, a set of guidelines for designing new channels, and a suite of

recommended sub-reach scale structures for implementing that design. While the

classification system has not changed substantively since 1996, the recommended structures

and techniques continue to evolve, and thus any discussion of them is limited to a snapshot

32

of present practice. Although recently Rosgen has started to insist that in some cases non-

intervention is the best restoration approach (NRCS 2007), his structures remain iconic for

students and opponents alike, the visible imprint of his work.6

1. The Rosgen Classification System

The classification system is the least controversial and most broadly-used part of

Rosgen’s approach. Although Rosgen has tinkered with it over time, the core was in place

by 1996, when he self-published Applied Fluvial Geomorphology, the textbook used in his Level

I course. The goal of the classification system is to reveal the processes at work in a

particular reach through examination of its form: to “predict a river’s behavior from its

appearance.” (Rosgen 1996, p.3-3) Rosgen’s classification system divides channels into

alpha-numeric categories based on features such as slope, shape, level of entrenchment,

width/depth ratio and bed material. According to Rosgen, these morphological features are

relevant because they reflect the physical processes that created them, and thus allow easy

assessment of forces that would be difficult and time consuming to measure directly.

Rosgen started developing his classification system in 1969 while working for the

USFS in an attempt to understand why some channels failed during floods and others in the

same watershed did not. His first attempt stratified streams based on their Strahler order, an

assigned number that indicates a stream’s position in its drainage network,7 but Rosgen

quickly abandoned this attempt when he found that position in a drainage network bore no

relation to whether a channel was capable of surviving a flood. He organized his second

attempt at a classification system around riparian vegetation, which led to an unmanageable

6 Rosgen is almost invariably photographed on construction equipment for magazine profiles. 7 In determining Strahler order, a headwater stream is a 1; where two 1s merge they create a 2, which stays a 2 until it flows into another 2 (or higher number), regardless of how many 1s have joined it in the meantime, and so on.

33

number of categories that he was never able to connect firmly to channel resiliency. His

third attempt at a classification system, begun in the late 1970s, forms the core of his work

today. At that point, Rosgen already had a database of morphological information he had

collected, and so he started looking for breakpoints in the data. This time he found what he

was looking for.

The Rosgen classification system has four levels.8 Level I depends primarily on data

obtained from aerial photographs and topographic maps. The first step is to identify the

morphology of the surrounding landscape. Rosgen’s system divides river valleys into eleven

types, ranging from Valley Type I (steep, V-shaped canyons) to Valley Type XI (deltas).

These types serve as a broad brush morphological system for considering geology and

climatic regime, as well as the position of the channel within the drainage network.9 The

next step is to divide channels into one of nine lettered categories that broadly describe their

form, gradient, and relation to the valley type they traverse. An ‘A’ channel, for example, is a

steep gradient, relatively straight, step-pool system, while an ‘E’ channel is low gradient,

highly sinuous, deep and narrow. The primary seven categories are A through G, but Rosen

added two additional categories: Aa+ for particularly steep gradient channels and DA for

anastomosing10 channels. These lettered designations and valley types are the coarsest level

of classification.

Level II classification depends upon field measurements of slope, bed material, and

channel geometry to verify the Level I alphabetical designation and sort channels into one of

six numeric subcategories within the lettered categories. The primary factor in assigning the 8 Rosgen’s short course series, described below, also has four levels, which creates some confusion as they do not map onto each other. 9 While Rosgen spends a good deal of time on valley types in his Level I and II courses, first textbook, and field guide, none of the practitioners who use his approach mentioned valley types in my interviews with them. 10 Anastomozing rivers resemble braided rivers in that they are divided into multiple threads or channels, but in the case of anastomozing rivers the threads are separated by stable material, such as bedrock. The multiple thread form was created through erosion, rather than deposition.

34

numeric subcategory is dominant channel material, which decreases in size as the

classification number increases: 1 designates bedrock channels, 6 designates silt or clay.

When Level II is complete, the classified reach is described with a letter-number pair.11 A

further two levels of data collection are necessary to complete the analysis of existing

conditions for design purposes, but by the end of Level II any channel’s alphanumeric

designation should be identifiable. And that “any” is not a figure of speech. Rosgen claims

that his classification system is universally applicable, that the nearly 100 categories and

subcategories encompass the entire naturally occurring range of rivers and streams.

This is not a modest claim. But setting aside for the moment the question of the

demonstrability of Rosgen’s claims for his classification system, the last ten years have

proved that it has utility. Like any classification system, it allows the compression of a great

deal of data. But unlike many classification systems, people generally find it both easy to

apply and, just as crucially, repeatable. As Dr. Timothy Keane, a professor at Kansas State

University and passionate Rosgen supporter described it:

The thing that I liked about Rosgen’s classification system... [is that it] allows communication, because people know what you’re talking about when you’re talking about a C4 no matter where you’re at. And it’s quantitative and replicable, so that somebody in Tennessee can use the same method to measure and assess a stream that we use in Kansas or Colorado or Wyoming. (Author interview, Dr. Timothy Keane, Kansas State University, 7/15/06.)

To a trained practitioner, Rosgen’s simple designations allow visualization of

complex channels. Say “B3” to a person trained in the Rosgen classification system, and

they will know that you are talking about a not particularly sinuous, relatively wide and

shallow, predominantly cobble-bed, riffle-dominated system running through a narrow,

11 As a final, more detailed step, the classification can include a subletter to describe where in the allowable range of slope for a category a particular channel lies. This produces classifications such as B6a, or C2c-. These subletter classifications are not used frequently. In more than 60 interviews, they were mentioned only twice.

35

moderately steep valley. Or say “C4,” and the trained practitioner will visualize a

predominantly gravel-bed riffle/pool system, highly sensitive to disturbance with a well-

developed floodplain running through a wide, gently-sloped valley. And these longer verbal

descriptions are themselves compressions of pages of field data and days, perhaps even

weeks, of surveying and measurement.

To a research scientist or a consultant on a multi-year project trying to understand a

river system, there is no substitute for the raw data; too much of the specificity of a system is

lost as its description is pared down to a letter/number pair. But in a conversation between

practitioners about a stream system one of them hasn’t seen, the wealth of data contained in

that pair is a critical tool. This is particularly important when the speakers come from

different disciplines. Because of Rosgen’s emphasis on classification based on repeatable,

objective measurements,12 a C stream for biologists is a C stream for hydraulic engineers.

Everybody is talking about the same thing, and this is no small achievement given the range

of disciplinary training involved in a typical stream restoration project. All but one of the

restoration practitioners I interviewed can apply the Rosgen classification system, even if

they don’t agree with it. And almost everyone I talked to agreed that it has been very useful

in establishing a basis for communication among the widely disparate disciplines involved in

stream restoration, which range from aquatic ecology to hydraulic engineering to fluvial

geomorphology. 13

In today’s academic climate, interdisciplinary communication is an uncontroversial

good on par with motherhood and apple pie. However, as Bowker and Star remind us,

classification systems are not simply ideas – perceptions of the structure of nature that live in

12 I address critics arguments about the difficulties of determining bankful discharge, the heart of Rosgen’s classification system, in Chapter 5, Section E. 13 Those who deny any utility to the Rosgen classification system tend to be basic researchers who do not favor classification in any form, people with the time to work from the raw data.

36

the realm of the mind – they have material force (Bowker and Star 2003, p.39). The

categories in Rosgen’s classification system exert their material force in a number of ways,

but most powerfully through their impact on restoration design.

2. The Rosgen Design Approach

When he first started teaching and implementing restoration projects in the mid-

1980s, Rosgen’s description of his design process in short courses and papers was not

exactingly specified. In response to confused students and outside criticism, he developed a

precisely described 40-step design approach. The careful specification, in turn, led critics to

dismiss his restoration method as a “cookbook” approach, a one-size fits all, follow the same

steps everywhere recipe, a charge Rosgen vehemently denies.

Rosgen’s chapter in the recently released USDA-NRCS Stream Restoration Design

Handbook will be the first place that his design process has been made publicly-available. In

the chapter, Rosgen divides his approach into eight phases, of which the 40-step process

specifies all but the first. These eight phases are:

Phase 1. Clearly and concisely define the project’s objectives, be they related to

physical, chemical, biological or human goals.

Phase 2. Develop or verify regional curves for geomorphic, hydrologic, and hydraulic

data; determine valley types and stream classifications; obtain reference reach

data.

Phase 3. Assess the stability and sediment supply of the restoration reach in relation to

its watershed to determine the cause and direction of change or impairment.

Phase 4. Seek a passive solution, such as a change in land use management. If none is

available, move on to Phase 5.

37

Phase 5. Combine all of the data gathered in the previous steps. Based on this data,

complete a design and test its compatibility with the hydraulic and sediment

regimes in the watershed.

Phase 6. Select and design appropriate enhancement and stabilization structures, such

as cross-vanes, W-weirs, or j-hooks.

Phase 7. Implement the design, including daily construction supervision.

Phase 8. Develop and implement a plan for monitoring and maintenance. (NRCS

2007, pp.11-1 - 11-2.)

The most distinctive features of Rosgen’s approach appear in Phase 2: the reliance

on regional curves, reference reaches, and dimensionless ratios. Regional curves describe

empirically-derived relations within a given hydro-physiographic province between drainage

basin area and channel mean depth, width, cross-sectional area, and discharge.14 In Rosgen’s

approach, regional curves are used to determine the bankful elevation and discharge of a

given reach, and are thus critical to proper channel classification, particularly in areas where

there are no reliable physical indicators of bankful discharge. Without regional curves, it is

not possible to employ Rosgen’s approach.

Once the current form of the project reach is classified, and its potential form

determined via examination of the reach, historic aerial photographs, and conditions up and

down stream (Rosgen 1996a, p.6-5), Rosgen practitioners go in search of reference reach

data that corresponds to the project channel’s classification, valley type, and hydro-

physiographic province. The reference reach approach is not unique to Rosgen. In his

formulation, however, the goal is not a single pristine reach to serve as a natural template,

14 Regional curves are based on the hydraulic geometry relations described by Luna Leopold and other USGS staff in their seminal 1950s publications (see for example Leopold and Maddock 1953).

38

but as many examples as possible of comparable reaches that are stable in order to provide a

range of values to use in design.15

Once collected, the reference reach data ranges are converted into dimensionless

ratios, the final key feature of Rosgen’s approach. For some design parameters, such as

meander length and radius of curvature, dimensionless ratios are created by dividing

measured values by the bankful width of the channel (either measured or determined by

regional curves, depending on the form of the channel). For others key design parameters,

dimensionless ratios are created by dividing particular values by the mean; for example,

dimensionless ratios for slope are created by dividing the measured slope of riffles, runs and

pools by the average measured water surface slope. These dimensionless ratios allow

designers to correct for scale when using data from stable channels elsewhere to guide design

for the project reach. Once the overall form of the project reach is determined, Rosgen’s

approach turns to the specific sub-reach scale techniques needed to implement the design.

3. Rosgen Restoration Techniques

Although he has toned it down over the years, the most frequent adjectives with

which Rosgen modifies the noun engineer all mean idiot. He has built his career critiquing

the traditional practice of hydraulic engineering, as embodied in the work of the US Army

Corps of Engineers and the standards of practice codified by the American Society of Civil

Engineers, the primary licensing body in the US. Despite this, Rosgen’s approach shares a

key goal with traditional engineering: stabilize the channel to prevent it from moving

sideways or downcutting. To achieve the former, you must protect the outer edges of

meander bends, where the lateral erosive force of the channel concentrates; to achieve the

15 The possibility of finding, and also of constructing, stable stream channels is a cornerstone of Rosgen’s approach, and also a major target of criticism as discussed in Chapter 5.

39

latter, you need to slow the water flow enough to prevent incision without critically reducing

the stream’s ability to transport sediment.

Shared ends unsurprisingly lead to a commonality of means. There is a clear kinship

between the traditional tools of American hydraulic engineers – weirs and bank armoring

materials such as concrete, riprap (large rocks) or gabions (metal cages with smaller rocks

inside) – and the suite of techniques that Rosgen has developed, such as cross-vanes, j-hooks

and bank revetments built from woody debris. The family resemblance lies in function not

form: Rosgen’s structures are built from locally-available boulders and logs with roots and

branches still attached. They thus have a much more natural appearance than the typical

USACE project. They also exert less complete control than a straightened, concrete-lined

trapezoidal flood conveyance system, although they perform much of the same work:

preventing channel migration by dissipating the energy of the water flow directed at the bank

or armoring the outside of meander bends; and preventing incision by reducing shear stress.

Like the traditional engineering structures they replace, when placed correctly in appropriate

stream systems Rosgen’s structures do a reasonable job of locking in a channel: Natural

Channel Design does not necessarily mean letting a channel behave naturally.

Unlike traditional engineering structures, however, Rosgen’s have aesthetic and

biological goals as well – creating a more natural appearance and improving fish habitat16 –

and it is here that Rosgen’s approach departs most powerfully from the past practice of

hydraulic engineering in America. All of Rosgen’s recommended structures are designed

with natural materials intended to harmonize with their surroundings. Cross-vanes and j-

hooks are intended not simply to divert water flow away from the outside edge of meander

bends, but also to concentrate that flow to maintain scour pools, providing critical deep

16 In his chapter for the NRCS Stream Restoration Design Handbook, for example, Rosgen refers to the structures as “stabilization/fish habitat enhancement measures.” (NRCS 2007, p.11-53)

40

water fish habitat. The log revetments, designed with rootwads facing out into the stream,

not only armor the bank but also are intended to provide shade and cover. In fact, much of

Rosgen’s work comes from private clients who wish to restore or create trout fisheries. The

Three Forks Ranch project on the Little Snake River in Colorado, for example, used more

than 500 cross vanes and j-hooks to create pools crucial for trout habitat in support of a

commercial trout-fishing operation. The resultant step/pool system is probably not

representative of what would naturally occur in this upper-montane valley; most likely the

system historically was full of willows, beaver, and multi-thread channels.17 But there is far

more to the system, aesthetically and ecologically, than there would have been had the

channel been culverted, and that is a notable improvement over what used to pass as

business as usual in the engineering community.

B. Initial Explanations for Rosgen Success

Rosgen’s success is a topic of conversation among both friends and foes, and there is

a surprisingly high level of agreement among participants in the restoration field about why

his classification system and design approach have been so widely adopted. The first thing

everyone mentions is Rosgen himself. Rosgen is deeply knowledgeable about streams, with

more than 40 years of experience studying rivers. The depth of his field experience is

unusual in a field that is itself so young, winning him respect even from his critics. Rosgen is

also a very charismatic man; people light up when they talk about him. He is the antithesis

of a deskbound DC bureaucrat: an honest-to-god Westerner from ranching stock who trains

17 Author interview, Dr. Brian Bledsoe, Colorado State University, 9/21/06. Bledsoe has been monitoring the Three Forks Ranch project for the past five years.

41

cutting horses and wears rodeo belt buckles he earned, Western shirts, and a white cowboy

hat. He is energetic, opinionated and extremely self-confident, and speaks in practiced

folksy phrases: idiots have “a terminal case of the dumb shits,” well-meaning idiots have

“their heart to cranium ratio out of whack,” and the actions of either kind of idiot are like

“crapping your chaps and sitting in the saddle.”

The second factor pointed to by both supporters and opponents is Rosgen’s

consistent claim that while restoration is not easy, it is doable: if you go through all four

levels of his courses, and do thorough data collection and analysis, you can design a good

restoration project. Over the five days of the Level I course I took, he said things like, “You

have to appreciate the complexity of this business. But the answer is out there. That’s the

good news.” This message of doability is underlined by Rosgen’s self-confidence: he is

certain that he knows what he’s doing, and that certainty is contagious. By contrast, his

opponents continually emphasize the uncertainty involved in restoration practice, and many

of them argue for refraining from restoration projects in most cases until we know more

about why they fail or succeed. For agencies with a mandate to restore streams, Rosgen’s

message that the scientists are making it look too complicated and that restoration is in fact

quite doable holds far more appeal.

According to supporters and critics, a third critical factor in Rosgen’s success is his

anointment by Luna Leopold, who is widely regarded as the father of stream restoration in

America. There are many stories told by supporters and opponents about why Leopold

took Rosgen under his wing. We will never for sure as Dr. Leopold died before anyone

asked him. What is certain is that he gave Rosgen enormous legitimacy by supporting his

work, co-teaching his initial short courses, and writing the foreword to Rosgen’s first text

book. Further, because of his eminence in the field, Leopold was also able to serve as a

42

buffer between Rosgen and advocates of traditional hydraulic engineering until Rosgen

established a track record for his design approach. For example, a case study on one of

Rosgen’s early projects in the 1992 NRC report Restoration of Aquatic Ecosystems describes

Leopold serving as arbiter between the USACE and Rosgen on the design of the Blanco

River project:

The project almost failed to materialize when COE [the US Army Corps of Engineers] subjected the unique design to expert review and was told by its reviewers that the new system would not contain flood flows. The project design was then sent for review to Professor Luna Leopold at the University of California, Berkeley, Department of Geology and Geophysics; Leopold praised the project and expressed confidence that it would work. On the basis of his recommendation, COE withdrew its reservations, and the project was allowed to proceed. (NRC 1992, p.474)

Without Leopold’s support, both supporters and opponents agree that Rosgen’s work would

likely never have achieved its current prominence.

A fourth key source of appeal pointed to by both opponents and critics is Rosgen’s

claim that his classification system and design approach are both universally applicable and

objective because they are based on simple, repeatable measurements. As Theodore Porter

has discussed in his excellent study of the rise of statistical analysis, Trust in Numbers, agencies

depend on such claims to protect themselves, using standardization to counter outside

pressure and guard against charges of bias:

Faced with the contradictory expectations of the executive, a myriad of congressional committees, and the courts, it is little wonder that they [federal agencies] should seek to minimize responsibility by adhering whenever possible to rules. (Porter 1995, p.194)

Applied fields, at least those that bear on matters of policy, are almost always exposed to scrutiny and criticism by the interests they affect…. Such a situation encourages the greatest extremes of standardization and objectivity, a preoccupation with explicit, public forms of knowledge…. Strategies of impersonality must be understood partly as defenses against… suspicions. (ibid., p.229)

43

For example, a USFS staff person who was on the intra-agency panel that selected Rosgen’s

approach as the Forest Service’s in-house method reported that the claim of universal

applicability was central to its decision. USFS already had homegrown classification systems

tailored to specific regions of the country, and there were methods available, such as those

developed by Montgomery and Buffington and Peter Whiting, that were considered widely

applicable. But none of these methods shared Rosgen’s claims to applicability everywhere.

Further, the Rosgen classification system is unique in its claimed basis in simple, repeatable

measurements.18 Thus in keeping with Porter’s arguments, one strong source of appeal for

some agencies has clearly been Rosgen’s work’s ability to shield them from accusations of

subjectivity.

C. The Short Courses

Given the nearly ubiquitous use of Rosgen’s classification system in the U.S., and the

fact that his design approach is the most widely used for constructing new channels, it is

notable that with only a few exceptions you cannot learn his approach at a university. Most

academic fluvial geomorphology courses do a brief review of the Rosgen classification

system as part of an overview of a number of classification systems – Montgomery and

Buffington, Schumm, Whiting etc. – but you wouldn’t need to take off both shoes to count

the universities that teach students to apply the classification system, and you could count the

18 Rosgen and many of his students acknowledge the difficulty of determining the bankful flow, the measurement on which his method depends, but treat it as a problem that recedes with experience, rather than as an on-going core of subjectivity in an otherwise fairly stringently specified procedure. By contrast, the Montgomery/Buffington classification system is often criticized by Rosgen’s supporters for relying too heavily on the judgment and experience of those who implement it. This will be discussed further in Chapter 5.

44

programs that teach Rosgen’s design approach on one hand.19 And yet Rosgen’s is the most

widely learned restoration approach in this country because he has been extraordinarily

successful at disseminating his work through short courses.

Since he offered his first short course in 1986, he has taught more than 10,000

students. In 1995 he taught 534 students in 13 courses, lasting 65 days; in 2005 he taught

687 students in 15 courses, lasting 89 days; clearly the demand is not dwindling. It is

important to note that these numbers only include the courses that were open to the public;

they do not count the courses arranged by agencies or organizations exclusively for their

staff, which increase the student count by an average of 8% each year.20

Rosgen’s short courses have evolved dramatically over time. In the beginning, there

was only one course, which typically lasted only a few days. There are now six standard

courses, 21 each of which lasts from five to ten days. The core short courses are arranged

into four levels (which do not map onto the four levels of the classification system). Level I,

Applied Fluvial Geomorphology, is a five-day course that introduces students to the classification

system, field methods, and basic theory of fluvial geomorphology, sediment transport,

hydraulics, and restoration. Level II, River Morphology and Applications, is a five-day course

that focuses on application of Rosgen’s classification system and reference reach approach.

The ten-day Level III course, River Assessment and Monitoring, is the favorite of engineers who

have gone through Rosgen certification. It goes beyond the earlier courses’ focus on form

to delve into the processes that determine that form, studying erosion, sediment transport,

and the influence of riparian vegetation. Finally, the nine-day Level IV course, River 19 I know of three: North Carolina State University, the University of Louisville, and California Polytechnic San Louis Obispo. West Virginia University taught the design approach during summer short courses from 2002 until 2004. 20 Information provided by Wildland Hydrology, 10/16/06. 21 A recently added 5th course deals with the nuts and bolts of the construction process, and there is also an alternative first course for people with engineering degrees, who are presumed to need a different introduction to restoration.

45

Restoration and Natural Channel Design, provides the tools for assessing, designing and

implementing restoration projects. Each of these courses is the prerequisite for the one that

follows. Complete all four and you are considered to be Rosgen-certified, which requires a

minimum commitment of 29 days and $7,400, not counting either transportation time and

costs, or the cost of room and board during the courses themselves.

For the vast majority of his students, Rosgen’s short courses introduce them to the

field of fluvial geomorphology, which forms the core of his work. In the Level I course I

took, only 4 out of 46 students had previous training in fluvial geomorphology. The

composition of the student body in the short courses varies, but a fairly typical pattern is that

approximately 3/5 of the students are agency staff, a quarter are consultants, and the

remaining 15% are divided between non-profit staff and university students or professors.22

According to Wildland Hydrology, Rosgen’s consulting firm, approximately 61% of students

stop at Level I, 17% stop at Level II, 8% stop at Level III, and 13% go all the way through

Level IV. This steep tapering off is not too surprising given that many of the people who

take Rosgen’s course are regulators or managers; they need to get the overview, but will

never need to classify a stream themselves, much less design a restoration project.23

22 In the academic short courses, which will be discussed below, the typical course composition in 2007 was fairly similar with just over half agency staff, 1/3 consultants, and the remaining 13% divided among, academics, non-profit staff, and other. (data provided by Lael Gilbert for Utah State courses, Laura Craig for the Maryland course, and Shannah Anderson for the California course.) 23 Estimating the proportion of participants in the stream restoration field with Rosgen training is an imprecise business. We know from figures kept by Wildland Hydrology, Rosgen’s consulting firm, that as of 2006 more than 10,000 students had attended his Level I, II, III or IV courses. This number does not include students who took special courses contracted for by specific agencies. EPA, for example, used to organize Rosgen courses just for their staff. Wildland Hydrology does not track these numbers, however, so for the purposes of estimation I will stick with 10,000 students even though we know it to be low.

Estimating the total number of restoration practitioners in the US is even less precise. There is no professional organization of stream restoration practitioners to ask about membership numbers. Nor is there a national conference devoted solely to restoration to query about attendance, or a single journal read by the majority of restoration participants to ask about circulation figures. As a proxy, three of the major conferences that restoration researchers and practitioners attend are 1) River Restoration NorthWest (RRNW), 2) the North Carolina Stream Restoration Institute Southeast Regional Conference (NC SRI), and 3) the American Society of Civil Engineers-Environmental & Water Resources Institute (EWRI) conference, which includes a

46

Why have Rosgen’s short courses been so successful? There are a number of

contributing factors, several mentioned above. Rosgen’s charisma is clearly of central

importance: a big part of the draw of taking a course in the Rosgen design approach is taking

it from Dave, as all of his students call him. There is also his emphasis on doability.

Perhaps the key factor in the popularity of Rosgen’s short courses, however, has been the

near total lack of alternatives: throughout the 1990s and well into this decade, Rosgen was

the only game in town. Educating stream restoration practitioners is an enormous challenge,

as I will discuss in depth in Chapter 4. For now, it is sufficient to point out that even today,

when stream restoration is a billion dollar per year industry (Bernhardt et al. 2005), no

university in this country has a stream restoration department or grants a BA, MS or Ph.D.

in stream restoration. In 2006 the University of Minnesota began a one-year certificate

program in stream restoration that primarily targets graduate students, drawing mostly on

courses already being offered; scientists at the University of Washington have tried to start a

similar program but have so far failed. Also in 2006, Portland State University began a

certificate program targeting both students and mid-career professionals based on a short-

course format, and West Virginia University had a similar summer certificate program from

2002 through 2004. At every other campus in the US, students who are interested in stream

restoration have to cobble together whatever relevant classes happen to be taught while

they’re there, and by their own admission and the general consensus in the field do not come

substantial restoration-related strand of presentations. In 2006, approximately 263 people attended RRNW, approximately 500 people attended the NC SRI biennial conference, and 778 people attended the annual EWRI conference, for a total of 1,541 conference attendees. Ignoring a) the fact that many people attending the EWRI conference are not involved in stream restoration, and b) the possibility of overlapping attendance among these conferences in order to maximize the potential number of participants, and assuming that only 1 in 10 participants in the restoration community attend such conferences in any given year, there would be somewhere on the order of 15,000 restoration field participants in the US. Thus a fair, though very rough estimate would be that more than half of the stream restoration community in the US has been through at least one of Rosgen’s courses.

47

out well-prepared.24 Short courses are the only avenue for supplementing their education

other than apprenticeship. For mid-career would-be restoration practitioners, often with

families, mortgages and no ability to drop out of the workforce for two years to get a

Masters degree (which doesn’t focus entirely on stream restoration anyway), short courses

are the only option. And in the short course market Rosgen is dominant.

There are increasing numbers of other restoration short courses available, some

taught by USACE staff, some taught by professors or consultants who wish to provide an

alternative to Rosgen’s courses. Only one of these, a series of two courses taught by faculty

at UC Berkeley, Johns Hopkins, University of Maryland, and Utah State, has anything like

the scope of Rosgen’s courses. Because this two-course series is crammed into the already

over full teaching and research schedules of the academics who staff it, however, it is

structured on a far smaller scale. The introductory course is offered three times a year (once

each in California, Maryland and Utah), and the second course is offered just once a year in

Utah. 108 students took these four courses last year, approximately 15% of the number of

students that attended Rosgen’s courses.25 As of yet, Rosgen does not have any serious

competitors.

D. Conclusion

This chapter introduced both Rosgen’s work – his classification system, design

approach, and the short course series he developed and continues to teach. In the next

24 Possible exceptions to this would be students at Colorado State University and University of Washington, where there is a sufficient mass of restoration-related courses that students could put together an adequate program. 25 According to course instructors, in 2007 30 students attended the academic level I course in CA, 25 students in MD, and 27 students in Utah. An additional 28 students attended the academic level II course in Utah.

48

chapter, I explore how Rosgen’s work has been used at the national and state levels, and

address the question of whether it has actually gained the prominence that Rosgen and his

critics ascribe to it.

50

Chapter 3. Adoption and Use of Rosgen’s Approach

“Rosgen's work and Natural Channel Design are pretty much the standard here. If you talk about stream restoration, that’s the rule.”

David Phlegar, Water Quality Supervisor, Greensboro, North Carolina

In the previous chapter, I asserted the success of Rosgen’s design approach. In this

chapter I assess the extent to which Rosgen’s work actually dominates restoration in the US.

Sweeping claims about the extent of its use have been put forward by Rosgen’s supporters,

and even some of his critics describe his success in totalizing terms in an attempt to rally

opposition. But to what extent has Rosgen actually captured the restoration field and

market? Given that decisions about what approach to use are made by thousands of project

managers, regulators and funders in offices all over the U.S., it would be difficult to answer

this question comprehensively.1 Instead, I offer here two partial answers to stand as proxies.

First, I examine efforts to develop stream restoration handbooks or guidelines for practice at

the national level, tracking their reliance on Rosgen’s work. Then I focus on patterns of

training and adoption in North Carolina, a state that is considered by critics to be “captured”

1 For example, the mix of agencies in any state is complex. At the federal level, there are the local district/region offices of the USACE and EPA; any national forests, national parks, or BLM land within state boundaries; and local offices of USFWS and NRCS. In coastal areas, the National Oceanic and Atmospheric Administration may have some say in restoration funding as well. At the state level there is typically a department concentrating on protection of natural resources, such as endangered species, and in some cases there is also a department focused on fresh water. In many states one of the biggest clients for stream restoration is the state Department of Transportation, as mitigation for highway projects. At the local level, there are the public works departments of every county and city, as well as locally-organized resource conservation, water and stormwater management, and irrigation districts. Within these myriad agencies the picture is further complicated by the fact that in some agencies, project managers are granted considerable discretion about the methods they require for stream restoration projects, so an agency’s stance on Rosgen’s work may not be unified, and it may vary over short periods of time depending on staff turnover. This level of complexity made a comprehensive national study of use of Rosgen’s design approach infeasible.

51

by Rosgen, for a more concrete look at whether and why people choose to use the Natural

Channel Design approach.

A. National Stream Restoration Handbooks and Guideline Efforts

One way to track the prevalence of the Natural Channel Design approach is to trace

its treatment in national handbooks, manuals and guidelines for restoration practice. Over

the last fifteen years, there have been four major efforts to develop national guidelines for

stream restoration. Chronologically, they are the 1992 National Research Council Report

Restoration of Aquatic Ecosystems, the 1998 Federal Interagency Task Force Stream Corridor

Restoration Manual, the recently released design handbook from the Natural Resource

Conservation Service, and on-going American Society of Civil Engineers’ efforts to develop

standards of practice. These four national initiatives embody a range of reactions to the

NCD approach, from promoting it, to drawing tight boundaries around its potential utility,

to active critique.

1. National Research Council 1992: Restoration of Aquatic Ecosystems

The National Research Council (NRC) 1992 report, Restoration of Aquatic Ecosystems:

Science, Technology and Public Policy, is generally cited as the first attempt to compile information

on the science and practice of aquatic ecosystem restoration in the U.S., and also as a key

early boost to Rosgen’s legitimacy. The committee that prepared the report leaned heavily

towards the biological sciences. Roughly categorized, the 15 members included nine

biologists, one chemist, two engineers, two trained in law and land use management, and one

52

agricultural economist; there were no geomorphologists included (NRC 1992, Appendix C).

The committee’s composition is a source of some bitterness among many fluvial

geomorphologists, who widely assume that the report’s favorable response to Rosgen is a

result of the committee’s ignorance about geomorphology.

The 1992 NRC report includes only one substantial chapter on rivers. This chapter

combines a review of major anthropogenic impacts on riparian ecosystems, a biology/

ecology-centered discussion of river science, a review of current management and policy

issues, and recommendations to improve the science and practice of stream restoration.

More than 15% of the chapter focuses on Rosgen’s work (16 out of 96 pages), which in

addition to general discussion is also the subject of two sidebars, two tables (one of which is

8-pages long), and a case study. Only one other geomorphological approach is mentioned,

that of George Palmiter, a consultant with no formal in the field who has largely disappeared

from view in the intervening years. It is difficult to imagine this committee promoting the

work of ecological practitioners without formal training in that field, much less exclusively

promoting such work, and yet this is precisely what the NRC report did in relation to

geomorphology.

Interestingly, despite the very positive references to Rosgen’s restoration work

throughout the chapter on rivers there is no discussion of his classification system, even

though the need for a classification system that describes the relationship between rivers and

floodplains is pulled out as a key recommendation. Classification appears as a step in the

restoration process in Table 5.8, which is taken from one of Rosgen’s papers, and there is a

one-paragraph discussion of Rosgen’s classification of the reach he worked on in the Blanco

River case study, but that is in an appendix. The only place the classification system’s alpha-

numeric pairs appear in the chapter on rivers is in Table 5.9, which rates the utility of

53

different structural interventions depending upon their Rosgen stream type. The odd thing

is that no context is given to explain the classifications. A footnote at the end of the table

explains the parameters on which the classifications are based, but does not explain what the

particular alpha-numeric pairs mean. Thus this 8-page table is unintelligible if you do not

already know Rosgen’s classification system (in fact, it may still be unintelligible if you know

the current form of the classification system because the table is based on its pre-1994 form.)

Overall, however, the 1992 NRC report is clearly an early promotion of Rosgen’s approach.

2. Federal Interagency Stream Restoration Working Group Manual

The Federal Interagency Stream Restoration Working Group (FISRWG) Manual was

the creation of federal agency scientists and their hired consultants rather than of academics,

as was the 1992 NRC report. Produced by an interdisciplinary team from 15 federal

agencies, it is described in its introduction as, “an unprecedented cooperative effort by the

participating federal agencies to produce a common technical reference on stream corridor

restoration.” (FISRWG 1998, p.4). Its goal was not to set policy for the agencies, but to

provide an overview of stream restoration and the types of things an agency staff person

needed to know to assess the need for, design, and implement a restoration project. The

FISRWG Manual’s overall goal was to convey, “the interdisciplinary and technical nature of

stream restoration.” (ibid., p.8) This is not the discouraging emphasis on uncertainty

common to many Rosgen opponents, but it is a far cry from Rosgen’s emphasis on doability.

By the time the FISRWG Manual was released, the Rosgen Wars had begun.

Rosgen’s 1994 Catena paper had been out for four years, and Miller and Ritter’s scathing

response to it for two years (Rosgen 1994, Miller and Ritter 1996). It was thus impossible

for the staff scientists and consultants preparing the FISRWG Manual to view the NCD

54

aproach with the naively hopeful eyes of the NRC panel; by that time they were all

participants in the debate.

Nowhere in the FISRWG Manual can you find an explicit critique of Rosgen’s

design approach or classification system. On the other hand, unless you read with great

thoroughness you might miss its presence entirely. The section on reference reaches in

Chapter 4 is minimal and doesn’t cite his work. There are a bare handful of Rosgen citations

unaccompanied by discussion in Chapters 4, 8 and 9, and there are several figures taken

from Rosgen’s publications, including his classification chart, a standard data collection

sheet, and an updated version of the table on appropriate structures by stream type. In

Appendix A: Techniques, some Rosgen techniques are addressed implicitly. It is only in

Chapter 7: Analysis of Corridor Conditions that we get any substantive discussion of

Rosgen’s work: a one-page write-up on the classification system in a section that also describes

Montgomery and Buffington’s and Schumm’s systems (ibid., pp.7.29-7.30).

There is, however, implicit criticism throughout the FISRWG Manual in the form

of a consistent emphasis on things that Rosgen did not then emphasize: interdisciplinary

teams, study of process, and non-intervention. As stated in the Introduction:

As a general goal, this document promotes the use of ecological processes (physical, chemical, and biological) and minimally intrusive solutions to restore self-sustaining stream corridor functions…. In addition, the document recognizes the complexity of most stream restoration work and promotes an integrated approach to restoration. Note that there are several things that this document is not intended to be.

It is not a cookbook containing prescribed “recipes” or step-by-step instructions on how to restore a stream corridor…. (ibid., pp.5-6)

This a far cry from Rosgen’s treatment in the 1992 NRC Report. Part of the reason

for this is surely the broad and shallow nature of the FISRWG Manual itself. As described

by Doug Shields of the USDA’s Agricultural Research Service, one of the document’s

55

authors, the Manual, “has a lot of back-up information about stream biology, stream

geology, and water quality; it’s kind of the 101 course for engineers to learn biology and

biologists to learn hydrology. It’s a mile wide and an inch deep.” 2 But the Manual’s

minimization of Rosgen’s work is also a reflection of conflict within the team that wrote it,

which was divided between supporters of Rosgen and some of his most vehement critics.

There was a difficult process of negotiation that Shields good-humoredly compared to the

Middle East peace process in its length, rancor and seeming unresolvability. The final

compromise was that the pro-Rosgen camp was allowed to include the Rosgen classification

system, but only as part of a section comparing, contrasting, and critiquing various

classification systems so that it was not the only one presented. As quid pro quo, the

sections on channel design were written by the anti-Rosgen camp. While this solution may

have been seen as a compromise at the time, it certainly comes across as, if not a critique,

than at least a dismissal. Rosgen was no longer presented as the great white-hatted hope of

the emerging field of stream restoration.

3. National Engineering Handbook 654: The National Resource Conservation Service Stream Restoration Design Handbook

The NRCS released in 2007 a restoration design handbook that it describes as, “the

next logical step” after the 1998 FISRWG Manual. Jerry Bernard, one of the leaders of the

Production Team for the FISRWG Manual, headed up the effort, which was run solely by

the NRCS. To the surprise of many, Rosgen was invited to contribute not only an appendix

on his classification system, but also an entire chapter detailing his 40-step NCD approach.

As this is the only chapter in the NRCS Design Handbook devoted to a single method, the

document as a whole is perceived as having a highly positive attitude towards Rosgen’s work; 2 Author interview, Dr. F. Douglas Shields, USDA-ARS National Sedimentation Laboratory, 6/13/06.

56

instead of supplementing the FISRWG Manual, the NRCS Design Handbook’s editors’

explicit goal, it appears to Rosgen’s critics to be working against it.3

Even in the NRCS, however, which is one of the most pro-Rosgen federal agencies,

there is sufficient awareness of the debate over Natural Channel Design that the Executive

Summary of the Design Handbook ends with the following disclaimer:

Please note that this handbook makes no endorsement of one particular approach over another and is not intended as a requirement document for purposes of funding or permitting. The guidance provided can be used to design and implement some of the techniques used in stream restorations. It is anticipated that as new methods are validated, they will be added to this guidance document or a supporting Web site. (NRCS 2007, p.ii)

4. American Society of Civil Engineers’ Efforts to Establish Standards of Practice

Standards of practice are critical for engineers because unlike most other restoration

practitioners they are legally liable for the drawings they sign.4 If a restoration project they

design fails they can be sued, and their only defense is if there are nationally-recognized

standards of practice that they can prove they followed. There have thus been three efforts

by the American Society of Civil Engineers (ASCE) to develop formal standards of practice

since 2000. The first and second of these efforts are tightly linked, and both were led by

Doug Shields, a senior researcher at the USDA-ARS National Sedimentation Laboratory

quoted above in relation to the FISRWG Manual; the third effort is just beginning and is

being led by Louise Slate, an engineer at a North Carolina consulting firm, and by Shields.5

3 There is also some concern because although NRCS sent Rosgen’s chapter out for peer review along with the rest of the manual, Rosgen was not required to respond to reviewer comments. Both Dr. Andrew Simon, USDA-ARS, and Dr. G. Matthias Kondolf, UC Berkeley, were asked to review the chapter, but refused when told that Rosgen would not have to respond to their comments. 4 The other exception is landscape architects. 5 I asked Shields about why he had been centrally involved in most of the major national guidelines efforts when so many of his fellow Rosgen critics avoided them. Shields said, “[T]he way that science advances is through a disciplined form of communication, usually written. Somebody has to write down where things are now, and somebody else says, ‘Well, if that’s the way things are, let me tell you how you oughta do it.’ So

57

In 2000, the ASCE began an effort to update Manual 54, which sets standards of

practice for engineers in regards to sediment engineering. Doug Shields was asked to lead

the writing of a new chapter on stream restoration (Shields et al., in press). By 2003, the

updated edition of Manual 54 had not been released so Shields and his co-authors published

a shorter version of the chapter in the Journal of Hydraulic Engineering (Shields et al. 2003),

which is discussed in more detail in Chapter 5.6 While these guidelines are in no sense a

frontal assault on Rosgen, putting forward an alternative approach is an implicitly critical

project, and multiple citations of papers critical of Natural Channel Design make it clear that

the authors do not support Rosgen’s classification system or design approach.

The third of the recent ASCE efforts to develop standards of practice for stream

restoration has only just gotten started, sparked by discussion about liability issues for

engineers working on river restoration at the 2005 annual meeting of the ASCE

Environmental and Water Resources Institute (EWRI) River Restoration Committee. A

team of authors led by Louise Slate and Doug Shields have a recent paper in the Journal of

Hydraulic Engineering which describes the state of engineering practice in river restoration and

why standards of practice are needed (Slate et. al. 2007). According to the authors, the

prevalence of agencies that insist on use of Rosgen’s NCD approach is one of the key

reasons why engineers need standards of practice for engineering. The basic argument is

that unless engineers have a set of standards that they can show clients, there will be no way

to educate them that a) there are a number of other restoration approaches available, and b)

NCD is applicable in some situations but not others. This is by far the most explicitly

critical of the national guidelines efforts to date, and thus sparked considerable debate even

there’s a dialogue that continues through time. But someone has to grasp the nettle and write something down and be definite and clear about it…. And so that’s what I was trying to do: further the dialogue.” 6 As of February 2009, the new edition of Manual 54 still has not been released, though it is in galleys.

58

before it was published (see Appendix D for an analysis of the debate and the resulting

changes to the paper).

What can we conclude from reviewing these four guidelines efforts? I would argue

that they suggest two things. At the national level the Rosgen Wars are in full swing. It is no

longer possible to view his work with the uncomplicated hope of the 1992 NRC Report.

Even for agencies such as the NRCS, many of whose employees are supportive of Rosgen’s

work, the backlash from Rosgen’s opponents is such that their support must be qualified.

And yet neither side can declare victory. There is considerable variation in how Natural

Channel Design is treated in the documents reviewed above. However, it is worth noting

that while some critics have started the process to develop alternative standards of practice,

only Rosgen has detailed standards that are already in print. So for the moment, Rosgen is

in a better position than his critics.

B. North Carolina

While at the level of national policy, the Rosgen Wars still rage, there are many states

where Rosgen’s opponents concede defeat. The language they use to describe these states is

laced with the imagery of invasion: Rosgen has “captured” South Carolina, or is “starting to

take over” in Florida. Local, state and federal agencies are granted little self-determination

in this process. Nor are less clear-cut situations, in which some agencies in a state adopt

NCD and others do not, much discussed. Which raises the question of what it actually

means to be captured. What does the process of selecting or rejecting Rosgen’s approach

actually look like on the ground, and how does NCD work for those who choose it?

59

To answer these question, I chose to concentrate on gathering a reasonably

comprehensive picture of restoration practice in one state: North Carolina. North Carolina

is widely considered to have one of the most extensive restoration programs in the U.S. It

is also considered by Rosgen’s critics to be a “captured” state. It thus provides rich grounds

for examining how and why Rosgen’s work is used in practice.

1. Data Sources

To analyze conditions in North Carolina, I have drawn primarily on data from

separate rounds of semi-structured and structured interviews. In November and December

2007, I conducted twenty structured interviews with agency and non-profit employees

involved in stream restoration in North Carolina (see Appendix A for a list of all the

interview questions and subjects).7 I interviewed nine staff members from local agencies and

non-profits, four from state agencies, and seven from federal agencies. My interview

subjects had a fairly broad range of roles in restoration projects. Although the majority

worked as project managers (80%), there were also designers (40%), regulatory reviewers

(25%), and funders (20%).8 As shown in Figure 3.1, there was a wide range in the average

number of projects that subjects in North Carolina worked on each year, from a bare

handful to more than a hundred. In terms of geographic spread, ten subjects worked in the

western, mountainous region of the state, nine in urban areas in the Piedmont or coastal

plan, and one out of state (at the regional EPA offices in Atlanta).

In addition, as part of my 2006 fieldwork on the Rosgen Wars as a national

phenomenon, I conducted detailed semi-structured interviews with four key players in the

7 While these interviews were based on a set list of questions, taking approximately 10 – 15 minutes to answer in trial runs, most of these interviews lasted from 45 minutes to an hour as people volunteered additional very helpful information about their restoration practices. 8 These percentages do not sum to 1 because most interview subjects held multiple roles in restoration projects.

60

North Carolina restoration community: Brian Bledsoe, now a professor of civil and

environmental engineering at Colorado State University, but formerly the Non-point Source

Program Coordinator for the state of North Carolina; Will Harman, a former North

Carolina State University Cooperative Extension staff member and founding principal of

Buck Engineering, one of the main restoration firms in the Southeast; Angela Greene, an

NRCS staff member who now works at the Canaan Valley Institute in West Virginia, the

main restoration non-profit on the west side of the Appalachian Mountains; and Greg

Jennings, a North Carolina State University (NCSU) professor who runs the NCSU Stream

Institute. All four were integral to the rise of the restoration movement in North Carolina.

Bledsoe attended the first Rosgen course in North Carolina, and then worked through his

agency to fund additional courses. Harman and Greene (along with Dick Everhart, with

whom I conducted a structured interview) were some of the first and most fervent users of

Rosgen’s work in the state. Together with Jennings, they worked to develop a form of NCD

tailored to local conditions in North Carolina and to start the Stream Restoration Institute at

NCSU.9

In the sections that follow, I draw on both of these sets of interview data to answer

three main questions:

1. How did Rosgen’s approach become so central to the restoration community in

North Carolina?

2. Which agencies use it, and when and why did they adopt it?

3. How well does it work for them in practice?

9 The NCSU Stream Restoration Institute is one of the primary institutional supports for restoration in North Carolina, offering a program of short courses quite similar to Rosgen’s as well as running a bi-annual restoration conference that is the primary gathering for the restoration community in the Southeast.

61

0

1

2

3

4

5

6

7

8

1 to 5 6 to 10 11 to 20 21 to 100 100+

Figure 3.1: Average number of restoration projects per year completed by NC interview subjects

2. Rosgen’s NCD Approach Takes Root in North Carolina

In North Carolina, the entities that have been involved with stream restoration the

longest are not state and federal agencies but local groups in the western part of the state

responding to community demand. A prime example of this is the Mitchell River Coalition,

a citizen group in northwestern North Carolina that started organizing to promote

restoration efforts in the early 1990s with the help of Dick Everhart, an NRCS employee

working with the Surry County Soil & Water Conservation District and an early Rosgen

supporter.

In 1995, while these local groups were still searching for more natural approaches to

fixing damaged streams, Rosgen taught his first class in North Carolina. Rosgen’s work thus

had tremendous and immediate appeal. As Angela Greene, a federal employee and strong

62

Rosgen advocate, described it, she had been working on streams from a more conventional

engineering approach for years, and then,

I went on a detail on the Mississippi after the 1993 flood and I was just astounded by how much devastation was caused by the controls that had been put in place because it extended not only on the Mississippi but all the way back up the tributaries. And so I started asking questions then, you know. How could we do a better job? And at that point and time, I was really good friends and had worked a lot with a fisheries biologist that worked in North Carolina, and he kept saying “You know, there is this guy named Dave Rosgen and you really have to take his training.”… [When I finally was able to take a course from Rosgen in 1996] I was sitting there listening to him, and I was thinking, “This is the answer! These are the clues that I needed.” Because like I said my background was not in fluvial geomorphology or anything like that, it was in straight civil engineering, but we started talking about the things that Leopold and Wolman had put together about the pattern, dimension and profile of a channel…. I sat there in that class and it was sort of like a light bulb coming on, you know: these are the sorts of things that we need to consider to be able to do a good job.10

Brian Bledsoe, who was at that point on staff with the state Coastal Management

Program, attended Rosgen’s 1995 course. Like Greene, Bledsoe had been looking for an

alternative approach, so Rosgen’s emphasis on geomorphology really caught his attention.

When Bledsoe moved to the state Division of Water Quality soon afterwards he got his new

agency to sponsor a second Rosgen course in North Carolina, with the deliberate intention

of bringing the various state and federal agencies starting to get involved in the new field

together in one course. It was this 1996 course that Greene, Everhart and other key figures

in the North Carolina stream restoration community attended:

When I was the NPS [Non-Point Source pollution] coordinator, we funded what I believe was the second course he taught in NC. We funded that through a Section 319 grant, and paid for several agencies to send staff…. The idea was to get a broad cross-section of the state government agencies, and maybe a few federal, dealing with restoration of streams and get them all together in the same class.11

10 Author interview, Angela Greene, NRCS/Canaan Valley Institute, 11/17/06. 11 Author interview, Dr. Brian Bledsoe, Colorado State University, 8/23/06.

63

The 1996 course created a core group of restoration project managers and permitting

staff who were in remarkable unison about how best to approach projects. As Allan Walker,

a veteran NRCS staff member in the state described it:

In North Carolina, there were two or three people I know that were very interested in stream-type work that brought a lot of those ideas here: Dick Everhart, Angela Greene/Jessup, and Joe Mickey that worked with NC Wildlife, and I became very interested right after they did….Those were the folks that have attended a lot of the classes. But at the same time in North Carolina you had folks with USFWS, …with USACE, … folks with DWQ [the North Carolina Department of Water Quality]. Everybody kind of ended up being in there together [in the courses]. It made for a real good relationship with everybody doing it together and keeping on board.12

This well-integrated group, who already had ties from their work on North

Carolina’s water quality programs, carried the torch of Natural Channel Design in the state.

Bledsoe said that he thought the centrality of Rosgen and Natural Channel Design

approaches in North Carolina were due to,

a combination of a lot of factors, one of which is personality. It can’t be taken out of this. There were a few agency folk – Dick Everhart and Angela Jessup [now Greene], and a couple of old friends at NCSU who really embraced Dave and his view of the world. And I think there was a relatively strong and advanced water quality program [in North Carolina], and it’s one of those timing things, perhaps.… I think we were kind of at a tipping point. We’d been working on wetlands a lot, but there was this group of us there who got out walking streams and thinking about how we managed water quality, and concluded that we really needed to integrate a geomorphic perspective. Maybe that 319 grant funded course had something to do with it. I don’t know…. [Rosgen] had some real champions there…. Angela and Dick are faithful followers, and they really spread the good news.13

Will Harman, for example, who has been one of the key figures in stream restoration

in North Carolina for more than a decade, learned about Rosgen’s work from Angela

Jessup/Greene and Dick Everhart in what sounds like a remarkably collaborative process.

12 Author interview, Allan Walker, NRCS, 12/18/07. 13 Author interview, Dr. Brian Bledsoe, Colorado State University, 8/23/06.

64

In 1996 Harman was working on the national 319 monitoring project on Long Creek, and

had reached the stage of applying agricultural BMPs: 14

Naturally, if you’re doing BMPs on a farm, you’re going to go to NRCS for some advice, and it turns out that they had just taken one of Dave’s workshops, and so were talking about things like root wads and these other concepts, and that kind of got us started…. [Angela Jessup] and Dick Everhart, who is the other NRCS person, they had already started [taking Rosgen’s courses], and so Angela would go and take… Level I, and we would kind of go out in the field together and play with all this stuff that she learned in Level I. And we developed this page of questions, and then I would go take Level I and ask all those questions…. [W]e did that all throughout… because Dave works in such a different setting than we work in, so we would find all these things, these issues and problems, and I could take those questions back.15

Like Bledsoe, Harman thinks the timing was critical. He pointed out that 1996 was

the first year that the state DOT was required to mitigate impacts to streams (previously,

mitigation had only been required for wetlands), which introduced a major source of

demand for stream restoration projects. The state’s Clean Water Management Trust Fund

was created the same year. So just as NCD was starting to take off in North Carolina, two

major new funding sources for restoration projects appeared.

3. Use of Rosgen’s Work in North Carolina Today

The concurrence of Rosgen’s courses with growing interest in stream restoration, a

stream community already well-integrated across agencies, and the appearance of new stream

restoration funding sources kicked off an emphasis on Rosgen’s approach that continues in

North Carolina to this day. The use of NCD is one of the defining characteristics of stream

restoration in North Carolina. As Greg Jennings, professor at NCSU and founder of the

Stream Restoration Institute said,

14 BMPs are the widely-used acronym for Best Management Practices, typically field-tested methods for reducing environmental impacts on air quality, water quality, etc. 15 Author interview, Will Harman, Buck Engineering, 8.14.06.

65

I think everybody in the practice now is trying to create naturally stable channels… [using NCD approaches. Whereas] in the past, where people were focused on creating rigid, hard channels, at least in North Carolina now we don’t have any regulatory flexibility about what we can create. If we’re doing a stream restoration it has to be a naturally stable functioning channel because we can’t armor a channel with concrete or rock and get a permit at this point except under very extreme circumstances.16

a. Overview of Funding, Permitting, and Mitigation

Like many states, North Carolina has multiple agencies that fund stream restoration.

At the federal level, there are grants available from agencies such as NRCS and EPA. At the

state level, the main sources of funding are the Department of Water Resources, the Clean

Water Management Trust Fund, the Soil and Water Conservation Division, and the

Ecosystem Enhancement Program. Of these six funding sources, the NRCS and state Soil

and Water Conservation agricultural cost share programs as well as the Ecosystem

Enhancement Program all require use of an NCD approach.17 EPA and the Clean Water

Management Trust Fund do not explicitly require a Rosgen-approach, but that is the

approach in which their staff is trained. The only funding program that does not have a

clear pro-NCD leaning is the state Department of Water Resources.18

On the permitting front, regulatory review in North Carolina is handled by the

USFWS, USACE, and EPA at the federal level, and by the North Carolina Wildlife

Commission and the Department of Environment and Natural Resources Divisions of

Water Quality and of Land Resources at the state level. Of these regulators, the Wilmington,

NC District of the USACE requires an NCD approach for mitigation projects, as does EPA

16 Author interview, Dr. Greg Jennings, North Carolina State University, 7.21.06. 17 As Rosgen approach advocate Jeff Parker from the Transylvania County S&WCD pointed out, “NRCS, they chose that method and started teaching it to us, and it caught on from there. In the [state] agricultural cost share manual, it requires us to use Rosgen Natural Stream Design method.” (Author interview, Jeff Parker, Transylvania County S&WCD, 12.18.07) 18 Foundations and corporations also fund some restoration work in North Carolina.

66

Region 4. USFWS has a somewhat divided presence. Its Asheville field office, which covers

the western part of the state, is an NCD stronghold, while in its Raleigh field office, covering

the Piedmont and coastal plain, staff hold more mixed views. The state Division of Water

Quality uses an NCD approach, as does the North Carolina Wildlife Commission.

One of the most distinctive features of restoration in North Carolina is the heavy

emphasis on mitigation: offsetting impacts to a stream in one place by restoring a damaged

stream somewhere else. A key issue for any state attempting to develop a regulatory

framework for stream mitigation banking is how to establish equivalence between streams to

be destroyed and streams to be restored in compensation. In North Carolina, the Rosgen

classification system is used to establish that equivalence (Lave et al. 2008).

b. Patterns of Use

The agencies for which my structured interview subjects worked became involved in

stream restoration for a number of reasons. As shown in Table 3.1, the most common was

compliance with the Clean Water Act, followed closely by community demand; this was

particularly applicable to community-based organizations such as non-profits, Resource

Conservation & Development Districts, and Soil & Water Conservation Districts. The

timing of these agencies’ involvement in stream restoration shows a clear peak in the mid-

1990s; 60% of subjects’ agencies had become involved in stream restoration by 1997 (see

Figure 3.2), and all but one had become involved by 2000.19 This corresponds with the

interview data in the previous section.

19 The sole exception is the City of Chapel Hill, which only embarked on planning for a stream restoration program in 2004.

67

State requirements 5%

Other 10%

ESA 15%

Agency Mission 25%

Community demand 35%

CWA 45% Table 3.1 Reasons why agency became involved in stream restoration (respondents could list

multiple reasons)

Only game in town 6%Communication tool 6%Funding requirement 19%Natural aesthetic 19%Permitting requirement 31%Positive environmental effects 31%Everybody uses it 38%

Table 3.2 Reasons for selecting Rosgen’s approach (respondents could list multiple reasons)

0

1

2

3

4

5

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Figure 3.2 Approximate year when agency became involved in stream restoration

68

Communication tool 13%

No answer 19%

Natural Aesthetic 25%

Quantifiable/predictive 38%

Positive environmental impact 50%

Stabilizes banks/It works 56% Table 3.3 Strengths of Rosgen’s approach

Every single one of my survey subjects had heard of Dave Rosgen, and all but two of

them employed his approach in their restoration work. As David Phlegar, the Water Quality

Supervisor for the city of Greensboro, North Carolina put it, “Rosgen's work and NCD are

pretty much the standard here. If you talk about stream restoration that’s the rule.”20 When

asked why they or their agency had chosen to use Rosgen’s work, the most common answers

were because it was already in wide use, because it had positive environmental impacts on

water quality and aquatic habitat, or because of permitting requirements (see Table 3.2).

When asked about the strengths of Rosgen’s work for their day-to-day needs, respondents’

most common answer was that they used the NCD approach because it works (see Table

3.3). Only three people declined to state strengths. Thus critics’ claims that North Carolina

is a bastion of pro-Rosgen sentiment seem to be borne out by my data.

C. Conclusion

This chapter assessed how the NCD approach is promulgated and used at the

national level – in design manuals and guidelines of practice – and at the state level in North

20 Author interview, David Phlegar, City of Greensboro, NC, 11.19.07.

69

Carolina, one of the national centers of Rosgen’s work. I have argued that, as measured by

usage in design guidelines, the national picture is complicated. With the release of the 2007

NRCS design manual Rosgen’s NCD approach seems to have the upper hand, but his critics

in the engineering field are rallying to create a credible alternative; whether they can do so

remains to be seen.

The picture at the state level is even more complex, but my data demonstrate that

there are indeed states where stream restoration is synonymous with Rosgen. Thanks to

early efforts to coordinate the training of agency staff, it is now the case that staff at all but

one of the state and federal agencies funding and permitting restoration in North Carolina

use NCD. This seems a clear illustration of Rosgen’s critics’ claims about agency capture.

To fully understand Rosgen’s success and the on-going conflict it has sparked, however, it is

necessary to take a step outwards and examine Rosgen’s place in the larger context of the

stream restoration field as a whole.

71

Chapter 4: The Structure of the Stream Restoration Field

“The practice of river restoration far outpaced the science of river restoration…. Only in the past five years [have basic scientists gotten involved].”

Dr. Martin Doyle, University of North Carolina-Chapel Hill

Chapters 2 and 3 introduced Rosgen’s classification, design and educational systems

and their use in practice, but gave little sense of their context in the stream restoration field

as a whole. This chapter provides the context – so critical for understanding both Rosgen’s

success and the ferocity of the opposition to his work – through a field analysis of stream

restoration. I first introduce Bourdieu’s field concept in general, and his analysis of the

dynamics of scientific fields in particular. I then apply that framework to the stream

restoration field, analyzing the autonomy and objective and subjective structures of the field,

and the positions that Rosgen, his students, and his critics take within it.

A. The Field Concept

Field is one of Bourdieu’s key analytical concepts, and one that he wielded frequently

in contexts as diverse as religion, science, literature, and the state (Bourdieu 1975, 1983,

1991, 1994, 1996a, 1996b, 1998; Bourdieu and Wacquant 1992).1 A field is a bounded,

structured social arena which provides a particular set of opportunities and constraints to

1 Like Marx, Bourdieu defines his central analytical concepts relationally. Thus it is impossible to sort out what Bourdieu is trying to accomplish with the idea of the field without also addressing the ways in which habitus and capital, the other two members of his conceptual triumvirate, shape it. For example, fields are defined by the types of capital that are valued within them, while conversely, there is no capital without a field in which to deploy it. Thus the discussion that follows, while holding field at its center, shuttles among all three concepts.

72

those who participate in it. To understand the specificity of any given field, Bourdieu argues,

focus on the forms of capital that are valued within it, and thus its logic, organization, and

limits:

In empirical work, it is one and the same thing to determine what the field is, where its limits lie, etc., and to determine what species of capital are active within it. At each moment, it is the state of the relations of forces between players that determines the structure of the field. (Bourdieu and Wacquant 1992, pp.98-99)

Struggle – to delimit the boundaries of the field, to determine conditions of entry,

and most especially to define the types of capital of most value – is a defining feature of

Bourdieu’s profoundly agonistic field concept. This struggle takes place within the

hierarchical structure provided by fields, each of which is organized around an axis whose

poles Bourdieu defines as autonomous and heteronomous. At the autonomous end of any

field are those actors whose production is controlled most thoroughly by the forms of power

and prestige (capital) specific to that field; at the heteronomous end are those whose production

is shaped primarily by outside forces. The relative autonomy of a field can be measured by,

“the extent to which it manages to impose its own norms and sanctions on the whole set of

producers,” including those closest to the heteronomous pole who are, “therefore the most

responsive to external demands….” (Bourdieu 1983, p.321)

According to Bourdieu, tensions between agents who hold positions at the

autonomous and heteronomous poles of a field provide one of the two primary motors of

struggle and change in a field; the other is the intrinsic conflict between newcomers and

incumbents. The following description comes from The Rules of Art, Bourdieu’s book on the

emergence of the literary field, but he explicitly states that it holds true for all fields of

cultural production including science:

[T]he unified literary field tends to organize itself according to two independent and hierarchized principles of differentiation: the principal

73

opposition, between pure production, destined for a market restricted to producers, and large-scale production, oriented towards the satisfaction of the demands of a wide audience…. This principle of differentiation is intersected by a secondary opposition that is established, within the subfield of pure production, between the avante-garde and the consecrated avante-garde. (1996b, p.121)

This rather sparsely theorized understanding of conflict within fields will be addressed in

Chapter 7.

To understand the field concept, it is crucial to grasp that for Bourdieu the positions

that make up a field – heternonomous vs. autonomous, old-guard vs. newcomer – are

analytically distinct from the agents who occupy them.2 There is thus a tension in field

analysis between the objective structure of a field (the hierarchical and structured relations

among positions within the field), and its subjective structure (the habitus that agents within the

field acquire through participation in it). Bourdieu argues that these two modes of analysis

– the objective positions of the field and the subjective dispositions, or position-takings, that

agents bring to it – must be analyzed together “as two translations of the same sentence”

(op. cit.). Despite this intimate analytic and actual intertwining, Bourdieu consistently argues

that once a field is established, “the space of positions tends to command the space of position-takings.”

(Bourdieu and Wacquant 1992, p.105) That is, the structure of an established field is more

determining than the habitus of the individuals occupying positions within it. The habitus

that agents bring with them into a field will guide them to positions that suit them,3 but

except in extraordinary circumstances individual agency does not change the structure of the

field itself.

2 “The notion of field reminds us that the true object of social science is not the individual, even though one cannot construct a field if not through individuals.” (Bourdieu and Wacquant 1992, p.107) 3 “[T]he different systems of dispositions they have acquired by internalizing a determinate type of social and economic condition … find in a definite trajectory within the field under consideration a more or less favorable opportunity to become actualized.” (Bourdieu and Wacquant 1992, p.105)

74

B. The Specificity of the Scientific Field

In one of his earliest writings on fields, “The Specificity of the Scientific Field and

the Social Conditions of the Progress of Reason,” (1975) Bourdieu analyzed science and its

distinctive struggles and structure. Because scientists are intellectuals and thus cultural

producers they have a particular position in Bourdieu’s hierarchy as the dominated fraction

of the dominant class, which allows science a relatively high level of autonomy because of

the useful applications it occasionally produces (ibid. p.36). More often, basic science results

only in more basic science, thus Bourdieu argues that one of the key characteristics of

science is that “producers tend to have no possible clients other than their competitors”

(ibid. p.23).4

This high level of autonomy means that science is fundamentally characterized by the

struggle to define its particular form of authority and legitimacy (newcomers vs. old guard),

rather than struggle over political and economic concerns (autonomy vs. heteronomy). 5 As

Bourdieu describes it,

As a system of objective relations between positions already won (in previous struggles), the scientific field is the locus of a competitive struggle, in which the specific issue at stake is the monopoly of scientific authority, defined inseparably as technical capacity and social power, or, to put in another way, the monopoly of scientific competence, in the sense of a particular agent’s socially recognized capacity to speak and act legitimately (i.e. in authorized and authoritative way) in scientific matters. (ibid. p.19)

For the individual scientist, what is at stake in scientific struggles is, at the most

fundamental level, defining science in order to valorize the capital s/he brings to the field

and increase her/his scientific authority:

4 These limits on the client base for science seem overly constrictive, given that sciences such as astronomy and paleontology have broad public audiences. 5 Some of the limitations of Bourdieu’s concept of the autonomy, both generally and in relation to the scientific field in particular, are discussed in Chapter 7 and the Conclusion.

75

[W]hat is at stake is in fact the power to impose the definition of science (i.e. the delimitation of the field of the problems, methods and theories that may be regarded as scientific) best suited to his specific interests, i.e. the definition most likely to enable him to occupy the dominant position in full legitimacy, by attributing the highest position in the hierarchy of scientific values to the scientific capacities which he personally or institutionally possesses (e.g. being highly trained in mathematics, having studied at a particular education institution, being a member of a particular scientific institution, etc.) (ibid. p.23)

Summarized more succinctly: “[T]he dominant are those who manage to impose the

definition of science which says that the most accomplished realisation of science consists in

having, being and doing what they have, are, or do.” (ibid. p.24)

It is critical to realize, however, that for Bourdieu the definition of science is not only

political:

An analysis which tried to isolate a purely ‘political’ dimension in struggles for domination of the scientific field would be as radically wrong as the (more frequent) opposite course of only attending to the ‘pure’, purely intellectual, determinations involved in scientific controversies. (Ibid., p.21)

Thus in contrast to both the Mertonian paradigm of a disinterested science disrupted by

external influences, or the social constructivist program of the then newly emerging post-

Mertonian wave of STS, Bourdieu described science as an agonistic practice in which

scientific methods and the political struggle for scientific authority are impossible to

separate.

Bourdieu never defines a specific set of scientific capitals, though he does refer to

things such as competence at mathematics, and working at particularly prestigious

institutions. The type of capital on which Bourdieu focuses most frequently, however, is the

educational capital with which someone enters the scientific field. This educational capital is

critical for inculcating the habitus of a scientist, but, Bourdieu argues, it goes beyond

producing competence to profoundly color perceptions of competence:

76

Competence [is] a social authority which legitimates itself by presenting itself as pure technical reason…. In reality, the august array of insignia adorning persons of ‘capacity’ and ‘competence’ – … the academic distinctions and scientific qualifications of modern researchers...– modifies social perception of strictly technical capacity. (ibid. p.20)

Given this focus on educational capital as a crucial instrument of scientific

legitimacy, it is unsurprising that Bourdieu argues that the scientific field cannot be reduced

to any conventional definition of ‘official science’ – the inherited instruments, methods and

institutions which define and enable modern scientific practice:

the aggregate of the scientific resources inherited from the past which exist in the state of objectification, in the form of instruments, texts, institutions, etc., and in the state of incorporation, in the form of scientific habitus, systems of generative schemes of perception, appreciation and action, produced by a specific form of educative action, which make possible the choice of objects, the solution of problems, and the evaluation of solutions (ibid. p.30).

While these are all critical to the function of any scientific field, Bourdieu argues that

science cannot be understood without also including the institutions involved in consecrating

scientific authority and expertise (the National Academy of Science, the National Science

Foundation, the Nobel committee, etc.), as well as the institutions responsible for circulating

scientific products, such as journals, “which, by selecting their articles in terms of the

dominant criteria, consecrate productions faithful to the principles of official science” (op.

cit.) Last, but certainly not least, science includes the educational system in charge of

inculcating scientific habitus, and thus ensuring the continuity of the field. As Bourdieu

describes it,

[T]he problems, methods and solutions immediately regarded as scientific… [are themselves] based on the whole set of institutional mechanisms which ensure the social and academic selection of researchers…, the training of the selected agents, control over access to the instruments of research and publication, etc. (ibid. p.34)

To study a field, Bourdieu directs us to make three analytical moves: 1) analyze the

position of the field in relation to the field of power (its relative level of autonomy); 2) map

77

the objective structure of the field, i.e. the hierarchical and structured relations between the

positions that those competing for the forms of capital and authority specific to the field

occupy; and 3) analyze the subjective structure of the field, the habitus that agents in the field

acquire from their participation in it (Bourdieu and Wacquant 1992, p.105) The field

analysis in the remainder of this chapter follows this three-pronged strategy. I use both the

field concept in general and Bourdieu’s description of the particularities of the scientific field

to structure a description of the characteristics of the field of stream restoration, and thus of

the context within which the Rosgen Wars are taking place.

C. The Autonomy of the Stream Restoration Field

According to Bourdieu, fields are structured along a dominant axis from an

autonomous to a heteronomous pole; that is, from a set of positions in which the types of

capital governing success are specific to the field itself, to a set of positions in which types of

capital specific to other fields, such as the field of power, set the criteria for success.6 It is

not simple to assess the relative autonomy of the stream restoration field because there is

very little data available about its basic characteristics, much less its relationship to the field

of power. As I mentioned in Chapter 2, no one has written a history of stream restoration

or attempted to document its exponential growth over the last forty years. One reason for

this is surely that the data are so widely scattered. Even the most basic data – such as

number, type, cost, and funding source of projects – are distributed across thousands of

6 Bourdieu states that only in the field of power does the axis run from cultural capital to economic capital, while in all other fields the axis runs from autonomy to heteronomy. In works such as The Rules of Art, however, Bourdieu consistently implies a conflation between heteronomy and economic capital (see for example pp.83, 114, 115, 120-121, 141, and 321 Bourdieu 1996b). The implications of this are discussed in detail in Chapter 7.

78

sources. When I worked as a research assistant on the National River Restoration Science

Synthesis (NRRSS) project, the first (and so far the only) nationwide attempt to compile data

about restoration practice in the U.S., we spent two years combing the paper and electronic

files of an overwhelming variety of local, state and federal agencies. Eventually, we compiled

data on more than 37,000 projects, but because we obtained records primarily from public

sources, we likely missed many privately-funded projects.

Despite the incomplete nature of the NRRSS database, it still has some light to shed

on the question of the stream restoration field’s level of autonomy. It is clear from the data

on number of projects completed per year (Figure 4.1, below) that there was an initial jump

in restoration projects in the early 1980s, followed by approximately fifteen years of slow but

steady growth, and then a much steeper growth curve from the mid-1990s on. Similarly,

NRRSS data demonstrate that the market for stream restoration reached more than a billion

dollars per year in the early 1990s, and has continued to grow since then (Bernhardt et al.

2005):7 according to a recent study, in 2006 the federal government alone spent more than

$4 billion dollars on restoration-related activities (ELI 2007).8 To determine the level of

autonomy of the field, the key question is whether this surge in the number of restoration

projects and the value of the restoration market pushed agency and university scientists to

shift their research focus to stream restoration. Several indirect indices, as well as data from

my interview subjects, suggest that it did.

7 For an anecdotal report on the growth of the restoration business, see also Lavendel 2002. 8 Although the report does not make clear how much of this was spent on streams, most likely, the majority of the funds were for wetlands restoration. But even if only ¼ of the $4 billion were spent on streams, this would suggest a large rise in spending on restoration in the U.S., since many more restoration projects are financed by private developers, non-profits, and state and local agencies.

79

0

500

1000

1500

2000

2500

3000

1970 1980 1990 2000

Number of Projects

Figure 4.1 Number of stream restoration projects in the U.S. per year based on the NRRSS database9

Figure 4.2 Number of peer-review journal articles on stream restoration published per year (Source: ISI Web of Science)

9 Because the National Oceanic and Atmospheric Administration did not allow NRRSS to make their data we public, this graph does not include approximately 19,000 projects from the NRRSS database.

80

0

5

10

15

20

25

1990

1992

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

number of awards

Figure 4.3 Number of NSF projects on stream restoration funded per year

$0

$1

$2

$3

$4

$5

$6

$7

$8

$9

$10

1990

1992

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Mil

lio

ns

award values

Figure 4.4 Total value of NSF grants on stream restoration per year10

10 The value of NSF grants awarded in 2002 is actually almost $25 million, $23.2 million of which went to fund the National Center for Earth-Surface Dynamics (NCED) at the University of Minnesota.

81

In terms of indirect indices, data on projects funded by the National Science

Foundation and on articles published in peer review journals show parallel trajectories. As

illustrated in Figure 4.2, above, there were relatively low levels of academic publications on

stream restoration through the 1990s, followed by a notable increase that began around 2000

and continues into the present day. Similarly, in Figures 4.3 and 4.4 the number and total

value of NSF grants for research on stream restoration remained relatively low through the

late 90s and then began a marked upwards trend.11 Suggestively, these increases in academic

research and publication on stream restoration shadow the increase in number of restoration

projects (shown in Figure 4.1). Can the increasingly applied focus of researchers in

hydrology, fluvial geomorphology and aquatic ecology be directly linked to expansions in the

restoration market, and to Rosgen’s success in capturing so much of that market?

My interview data speaks to this question; although I did not ask about it, many of

my research subjects brought it up on their own. Confirming the initial lack of interest in

stream restoration by agency and university researchers, consultant James MacBroom said

that when he started researching more natural approaches to channel design in the late 70s

and early 80s, he,

found that there really wasn’t what we think of as design guides or manuals on how to design a channel for something other than a rigid boundary, prismatic type of geometry. On my own I discovered writings by Luna Leopold, other USGS people, and Gordon Wollman dealing with fluvial geomorphology. But it was never a design manual, it was talking about the subject as a pure geologist or hydrologist.12

A number of my interview subjects confirmed the trajectory of limited academic and

agency researcher interest starting in the mid-90s, with a notable ramping up in 2000-2001,

from personal experience. Peter Wilcock, one of the most prominent sediment transport

11 NSF lists no grants for stream or river restoration before 1990. For Figures 3.2 – 3.4, I searched using the phrases “stream restoration” and “river restoration.” 12 Author interview, Dr. James MacBroom, Milone & MacBroom, 7/24/06.

82

researchers in the U.S., reported that his level of interest in restoration had changed levels at

a couple of key points:

I initially paid attention, back about … eleven or twelve years ago [1994/ 1995], when I first heard people remarking on this Rosgen phenomena… and complaining at scientific meetings about the dumb things that were being done in the name of stream restoration… And then, four years ago [2002]… [I started co-teaching with Matt Kondolf, and] that was the first time that I was actively involved in teaching a restoration short course. I certainly came into that as a sediment transport person. Then my level of activity hit the current maxed out level two years ago [2004] when I joined NCED [the National Center for Earth-Surface Dynamics]. They decided that stream restoration would be one of three major areas of emphasis, and asked me to lead that effort.13

ARS National Sedimentation Laboratory researcher Dr. Andrew Simon, a prominent

process researcher and Rosgen critic, described a similar trajectory in his career:

up until recently I haven’t been involved in stream restoration at all. I’ve been involved in the geomorphic, that is the quantitative analysis of unstable systems…. My experience with restoration has come from the other side of trying to work with people who are involved in stream restoration to educate them into what the important processes are… [because] I became uncomfortable with some of the applications I was seeing.14

Martin Doyle, a geomorphologist at the University of North Carolina, described the

same timeline from the perspective of someone who has been doing applied work on stream

restoration since the mid-90s when,

the premier geomorphologists in the country were condescending towards river restoration…. I distinctly remember telling people I was doing my Masters… [on stream restoration] and there was definitely a sense of scorning, oh, he’s in applied fluvial geomorph…. The academics were completely out, they were not touching it at all…. The practice of river restoration far outpaced the science of river restoration…. Only in the past five years [have basic scientists gotten involved in restoration].15

13 Author interview, Dr. Peter Wilcock, Johns Hopkins University, 7/24/06. 14 Author interview, Dr. Andrew Simon, USDA-ARS National Sedimentation Laboratory, 12/13/06. 15 Author interview, Dr. Martin Doyle, University of North Carolina-Chapel Hill, 6/15/05.

83

Steve Kite, a geomorphologist at West Virginia University and Rosgen Wars

moderate, mentioned the positive effects of both Rosgen and funding availability on the

timing of his decision to become involved in stream restoration:

Can’t say as I spent much time thinking about it [stream restoration] until 2001 when I was on sabbatical…. [S]ome colleagues in the Civil and Environmental Engineering department had taken his [Rosgen’s] courses. They were trying to put together a stream restoration team, and found out through networking with a fish biologist that I had worked with that there was a fluvial geomorphologist on campus. That’s where my active involvement in restoration started. I had shied away from those sorts of issues for a long time…. I realized this was an incredible opportunity to be involved in this [stream restoration]. I said to myself, not only is this something that’s important, it’s fundable, and at the same time because of the unique setting here there are problems and perspectives that are unique to people who are familiar with this area [the West Virginia Coalfields] that need to be integrated into the stream restoration workshops and literature.16

Gary Parker, a prominent sediment transport researcher and, like Kite, a Rosgen

Wars moderate, described the relation between the rise of Rosgen and the shift towards

applied research among agency and university researchers in both direct and approving

terms:

[U]ntil Rosgen started actually doing things, most academics had not the slightest intention of getting involved in an applied project and saying how things ought to be done. They spent most of their time telling people what they couldn’t do…. Rosgen has had the effect of moving the entire field of river geomorphology more in the direction of thinking about how to solve practical problems.17

A final example comes from an informal conversation with Jack Schmidt, a

geomorphologist from Utah State University. Schmidt said that he started the restoration

short courses at Utah State because the state Department of Natural Resources came to him

and said that they were tired of sending all these people out of state to Rosgen classes; since

Utah State was a landgrant school, would it provide an alternative? Schmidt now runs a

16 Author interview, Dr. Steven Kite, West Virginia University, 7/26/06. 17 Author interview, Dr. Gary Parker, University of Illinois, 6/26/06.

84

two-level short course series at Utah State (a core part of the alternative short course series

mentioned in Chapter 2) as a direct response to that government request.18

Taken together, my qualitative and quantitative data paint a picture of agency and

academic researchers gradually pulled into the study of stream restoration and away from

more basic research. Further, these data suggest that the autonomy of stream restoration as

a scientific field has been decreasing over time as research has been increasingly tailored to

meet the demands of the growing restoration market and to respond to Rosgen’s role within

it. As I will discuss in Chapter 7 and the Conclusion, this shift in stream restoration science

is happening at the same time as a more general shift in the political economy of the

American university system writ large. The work of Mirowski, Krimsky and others suggests

that in American scientific institutions the influence of heteronomous forces is becoming

increasingly significant.

D. The Objective Structure of the Stream Restoration Field

According to Bourdieu, the objective structure of a field is a system of necessities

and constraints that exist independently of the will and recognition of the individual actors

participating in the field. It consists of the objective relationships among, “the positions

occupied by the agents or institutions who compete for the legitimate form of specific

authority of which the field is the site.” (Bourdieu and Wacquant 1992, p.105). In An

Invitation to Reflexive Sociology, Bourdieu writes that:

These positions are objectively defined, in their existence and in the determinations they impose upon their occupants, agents or institutions, by

18 Personal communication with Dr. Jack Schmidt, Utah State University, 10/22/04.

85

their present and potential situation in the structure of the distribution of species of power (or capital) whose possession commands access to the specific profits that are at stake in the field, as well as by their objective relation to other positions (domination, subordination, homology, etc.).” (ibid., p.97)

The organizing principle of these positions is the axis from autonomous to heteronomous

production, described in the section above.

For reasons explained above and in the previous chapter, there is very little

quantitative data available about stream restoration funding, practice and practitioners. This

makes laying out the objective structure of the stream restoration field difficult. However,

having attended multiple restoration courses and conferences, and interviewed more than 60

of the major players in the stream restoration word, I do have substantial amounts of

qualitative data about the relations among positions in the stream restoration field. Thus

while the mapping and analysis below has relatively little quantitative support, it is still well-

substantiated by my research.

The objective structure of the field of stream restoration (at least until recently) is

shown below in Figure 4.5, reflecting two of the primary tensions within the field: research

vs. practice, and environmentalism vs. traditional hydraulic design.19

19 I will argue in Chapter 7 that Rosgen is successfully reorienting the axes of the field, and present an updated version of this diagram at that point.

86

Figure 4.5: The Objective Structure of the Stream Restoration Field

Towards the bottom of the hierarchy are the practitioners, primarily employees at

private consulting firms. There are still some positions that involve design and

implementation of stream restoration projects in state and federal resource agencies, most

notably USACE and NRCS. In the non-profit sector, there are two national NGO’s – The

Nature Conservancy (TNC) and Trout Unlimited – which conduct significant numbers of

restoration projects. There are also a small number of local non-profits, such as the Plumas

County CDC in California, the Hiawassee River Watershed Coalition in North Carolina, and

the Canaan Valley Institute in West Virginia, which design and carry out stream restoration

work.

Among practitioners, the larger science/practice divide that structures the field as a

whole repeats itself. In this initial view of the restoration field, he handful of consulting

firms and non-profits – such as Stillwater Sciences and TNC – which are staffed almost

entirely by Ph.D. scientists and function effectively as auxiliary research units hold more

87

prestige and capital than the far larger group of consulting firms whose work neither stems

from scientific training nor aspires to scientific practice.

2. Environmentalism vs. Traditional Hydraulic Engineering

The second, subsidiary structuring axis in this initial view of the stream restoration

field runs between those whose work is based primarily on the traditional principles of

hydraulic engineering, and those whose work is based on environmentalist ideals embodied

in the larger Natural Channel Design movement. As environmental regulations and

environmentally- and aesthetically-motivated citizens’ groups have gotten more powerful,

the power and influence of the traditional engineers have waned. There is now such

concerted opposition to channelizing rivers into straight, concrete-lined, ecological deserts

that even the USACE has begun a restoration program.20 The consulting firms and public

agencies dominated by engineers – such as the federal and state Departments of

Transportation, water and stormwater management districts, and even the Bureau of

Reclamation – have started to shift towards more environmentally-friendly positions. There

is still a great deal of friction between the groups. Disparaging comments about engineers

are common among Rosgen and his supporters, and many of my interview subjects from the

larger NCD community said that their initial commitment to more natural design came from

deep distress about the ravages of traditional hydraulic engineering practice.

20 The 1986 Flood Control Act, Section 1135, started the USACE restoration program (although it wasn’t until four years later that the 1990 Flood Control Act made it a permanent program). In 1989 Lt. Gen Henry J. Hatch, then commander of the USACE, went so far as to hold a press conference announcing that the Corps had to embrace environmental concerns (O’Neill 2006).

88

3. The (Relatively) Uncapitalized Masses: Funders, Clients, and Regulators

As in any field, the mass of participants in the stream restoration field do not hold

substantial amounts of individual capital (collectively, they hold a great deal of capital, as I

will argue later). These are the agencies, non-profits, and private landowners that fund,

manage, and regulate stream restoration projects.

Stream restoration funders are primarily employees of a wide range of federal and state

agencies. At the federal level, funding agencies (many of whom run multiple grant

programs) include EPA, BLM, the National Oceanic and Atmospheric Administration

(NOAA), NPS, DOT, USFS, USFWS, NRCS, and USACE. There is also some stream

restoration work funded by tribal governments. Funding at the state level has 50 different

variations, but in most cases grants are available from the state DOT and whichever agencies

are charged with protecting environmental and/or water resources.

In the non-profit realm, TNC and Trout Unlimited both fund substantial amounts of

restoration work. There are also private funding sources, such as wealthy landowners and

foundations. In Montana, for example, which is poor in state resources and rich in trophy

ranches, wealthy landowners are a primary driver of the restoration market. Up to this point,

however, public sources have been the primary funders of stream restoration in the U.S. For

example, in a California follow-up survey with a randomly-selected subset of 44 projects

from the NRRSS database, 84.1% of projects were publicly-funded, and 4.5% were funded

by non-profits, leaving a maximum of 11.4% that could have been funded primarily by

private sources (projects in the “I don’t know” and “other” categories).21

Managers are those who propose and oversee restoration projects. They are a broad

group that includes all of the types of federal, state, non-profit and private entities listed

21 This seem likely to change dramatically in the near future because of the explosive growth of the stream mitigation banking industry.

89

above, plus the cities, counties, community groups, irrigation districts, water and storm water

management districts, and soil and water conservation districts that apply for grants to fund

restoration projects. Because they hold the purse-strings (if not the prestige) in the stream

restoration field, funders and managers are the people who effectively set stream restoration

policy in the U.S., albeit in an extremely piecemeal fashion. Each local and state agency,

each regional office of a federal agency and even, in agencies such as the USFS and NRCS,

each project manager within that agency, has his or her own little fiefdom. Within the confines

of regulatory requirements, they have the autonomy to propose projects, write requests for

proposals, and select consultants. These are the people charged with implementing our

profoundly contradictory directives for restoration: control nature to meet human ends while

setting it free to pursue its own ends. It is this large group of agency staff at the local, state

and federal level who, on a day to day basis, decide the outcome of the thousands of tiny

battles that make up the Rosgen Wars. 22

The regulators at local, state, and federal agencies are a smaller but toothier group.

The primary regulatory review is at the federal level, and the main players are those agencies

charged with implementing the Endangered Species and Clean Water Acts: USACE, NOAA,

and USFWS. At the state level, agencies charged with protecting natural resources may have

some permitting involvement as well if there are state regulations that exceed federal

requirements. States typically also have state-wide water quality and flooding permits, which

they will enforce in relation to stream restoration projects. Lastly, municipalities review

restoration plans in accordance with local flood and water quality permits and land-use

regulations. Staff charged with regulatory oversight, particularly at the federal level, can

demand changes in projects, and in the case of Section 404 permitting, set the conditions for

22 The exception to this is in states such as North Carolina, where some permitting agencies require use of Rosgen’s approach, restricting managers’ freedom to choose a restoration approach.

90

requiring restoration projects in the first place. Thus regulatory staff, too, play decisive roles

in the Rosgen Wars despite their relatively low level of capital in the stream restoration field.

4. Species of Capital

The forms of capital in play in the stream restoration field constitute the primary

stakes of the Rosgen Wars, so they are currently in flux. Figure 4.5, above, reflects the initial

structure of the stream restoration field. I will talk in Chapter 7 about the ways in which

Rosgen is successfully changing the types of capital valued in stream restoration, and thus

the form of the field itself. Up until recently, however, the norms and practices of basic

science set the types of capital valued at the dominant autonomous pole of the stream

restoration field, while the demands of regulatory compliance and the market determined

those at the heteronomous, practice-oriented pole. Along the subsidiary axis, the

environmental capital associated with respect for nature and non-destructive outdoor

recreation are valued by the larger NCD community, while the quantitative rigor and

exactingly specified standards of traditional hydraulic engineering are valued at the opposite

end of that axis.

What is perhaps most interesting about these types of capital and the relations

among them is that rather than being specific to the stream restoration field, the species of

capital valued within it are drawn from its relations to other fields. This calls into question

the bounded model of fields that Bourdieu describes, as I will discuss in more detail in

Chapter 7.

91

E. The Subjective Structure of the Stream Restoration Field

The third part of a field analysis, according to Bourdieu, is to lay out the subjective

structure of the field, the habitus that participants in the field share. Habitus consists of a set

of embodied dispositions, learned through some form of educational institution, which do

not govern behavior but do make particular actions likely. In his article on the specificity of

scientific field, Bourdieu says that scientific habitus consists of the,

systems of generative schemes of perception, appreciation and action, produced by a specific form of educative action, which make possible the choice of objects, the solution of problems, and the evaluation of solutions. (Bourdieu 1975, p.30).

Attempting to analyze the habitus-producing subjective structure of the stream

restoration field raises an immediate question: does stream restoration actually have a unified

subjective structure? Ordinarily, this would be supplied by university training, but as I will

discuss in the first part of this section, the orthodox pathway into a profession has not

proven adequate to meet the growing stream restoration field’s demand for practitioners.

Instead many people have turned to an unorthodox pathway – short courses – to

supplement their university science education, or to train them in restoration from scratch.

Participants bring with them a variety of habiti stemming from different kinds of training

(university vs. short course, physical sciences vs. life sciences vs. engineering), as well as

different types of practice (in consulting, academia, public agencies, and NGOs).

I will argue here that to the extent that there is a discernible trend, a congealing in

the subjective structure of the stream restoration world, it is that provided by Rosgen’s short

course series, which has become the de facto industry standard. It is critical to ask, though,

what sort of habitus could be created by a short course, however deeply-focused and intense.

As many observers on both sides of the debate have wryly noted, Rosgen’s courses have

92

certainly produced a large group of people who want to be Dave Rosgen23 and a strong sense

of community among participants, but can this be described as something at the partially

unconscious level of a habitus?

To explore this, I analyze the educational experience created by Rosgen’s short

courses in the second part of this section. Because similarities in organization, a common

student pool, and overlapping content create notable similarities in the educational

experience provided by a number of restoration short courses, I compare Rosgen’s courses

and their potential for habitus creation to the closest competition: the two-level short course

series organized by a loose group of academics including G. Matthias Kondolf, a

geomorphologist at UC Berkeley; Peter Wilcock, a sediment transport researcher at Johns

Hopkins University; Jack Schmidt, a geomorphologist at Utah State; and Margaret Palmer,

an aquatic ecologist at the University of Maryland. For simplicity’s sake, I will refer to the

latter set of short courses as the academic short courses.

I turn first to the reasons why university training – the normal path into a profession

in this country – has so far been unable to provide the primary path into the stream

restoration field.

1. Explaining Academic Inertia

The growth of the stream restoration market has produced a significant demand for

trained practitioners, but academia has been slow to respond with programs focused on

restoration science or practice. As mentioned in Chapter 2, no university in this country has

a department of stream restoration. Students attempt to cobble together a suite of

23 My favorite of these stories comes from an agency staff person who told of watching with mounting disbelief as employees of a consulting firm that relied on Rosgen’s approach stopped to put on white cowboy hats before photographing each of the cross-sections they surveyed.

93

reasonably relevant courses in ecology, geomorphology, hydrology and engineering from the

existing course offerings at their university. There are some state and polytechnic

universities that offer concentrations in stream restoration as part of degrees in

environmental or watershed management (in California, for example, both California

Polytechnic Institute San Luis Obispo and California State University-Humboldt offer such

concentrations). There are also two universities (the University of Minnesota and Portland

State University) that have started restoration certificate programs, but it is not yet possible

to get an academic degree in stream restoration in the US.

The first question to consider is why existing arrangements at universities have not

proven sufficient to the task of educating the needed restoration workforce. Based on my

interviews data with academics who have tried to start such programs, I believe that there are

three basic reasons for this. First, there are the technical difficulties presented by the

disciplinary structure typical of American universities. Although interdisciplinary programs

seem to be very much in vogue at present, the vast majority of degree programs are still

organized along disciplinary lines. Fulfilling the requirements of her home department

typically makes it quite difficult for a student seeking a BA or MA degree to take the range of

geomorphology, hydrology, engineering and biology courses necessary to prepare her for

work in the very interdisciplinary field of stream restoration. While MA students often have

more flexibility, they are more limited in terms of available courses because they are only

enrolled for one or two years. Steve Gough, a private consultant, tried to put together an

interdisciplinary course of study for his Ph.D. at the University of Illinois, and was rebuffed:

I started fishing around and the walls between academic departments are just still very strong. I had a professor tell me that at the U of I you’re either going to be a geomorphologist, or you’re going to be a fish biologist, or you’re going to be a civil engineer, but you’re not going to be even two of

94

those things let alone three; you’re going to be one, that’s all you can be in today’s academic world.24

Second, there are not many universities where the range of course work necessary

for training in stream restoration is already in place. There are a few centers of restoration

science, most notably Colorado State University, the University of Minnesota, and the

University of Washington, where the broad range of necessary knowledge is covered by

existing course offerings. Most universities do not have this luxury.

The third reason why existing university-based education has not been sufficient to

meet the training needs of the new discipline is that what many would-be practitioners

require is re-training: they already have Bachelors degrees, and have neither the need nor the

desire to return to college. Ph.D. programs are both too narrowly-focused and too time

intensive for mid-career professionals with mortgages and families. Even Masters programs

are a stretch for this core body of would-be practitioners, particularly given existing Masters

programs’ limited ability to provide a targeted restoration education. As Laura Wildman, a

senior staff member at American Rivers, the most prominent river-focused NGO in the

U.S., described the situation, there are very limited options:

I have a strong interest in fluvial geomorphology, and I’d like to continue to educate myself. Who has a class in fluvial geomorphology? Very few universities in this country. How do I take it? I live in Connecticut. And yet I really want to take a bunch of high level courses from a professor. So I go to Yale to the geology department where there’s one fluvial geomorphology post-doc, who doesn’t teach. The post-doc’s supervising professor is more focused on his one particular topic of research and seems unconcerned with how I learn about this in the future. Do I want to be a Ph.D. student and work on his very specific project that has nothing to do with restoration? No. And guess what, everywhere I go it would be like that. Now there a few universities with fluvial geomorphologists interested in restoration, but I can’t move out there, and do I really want to be a Ph.D. student anyway? I’m applied. I’ve always been applied. I love applied. So… I’ve reached some bizarre roadblock where no one cares about teaching me. I want to be

24 Author interview, Steven Gough, Little River Consulting, 6/19/06.

95

better, I really want to train myself in this. My one door available is to give up my career for 6-7 years, and do a Ph.D. long distance.25

Given that the American university system does not currently meet the training

needs of the stream restoration field, why have universities been so slow off the mark in

developing restoration science programs? The demand has been there since at least the mid-

1990s (Bernhardt et al. 2005), and geomorphologists who work on restoration have been

aware that Rosgen training was supplanting academic training for at least that long.

A primary reason is the difficulty of establishing new academic programs.

Universities are like oil tankers; one should never underestimate their inertia. New programs

require resources of administrative time, space, and curricula, all of which are held extremely

dear. And too, new programs require a considerable degree of administrative scrutiny,

particularly at public universities.

Second, there is broad agreement, even among the opposing camps in the Rosgen

Wars, that becoming an adequate restoration practitioner requires a major component of

learning by doing. Although this is less true of many state schools, research universities do

not typically encourage apprenticeship as part of degree requirements. Nor do faculty at

research universities typically have the extensive hands-on project experience to be able to

provide such field training. So even if a university was willing to include an apprenticeship

component in a degree program, it would likely have to turn outside its tenure-track faculty

to implement it.

The only attempt I know of to start a hands-on restoration practice course for

students at major research universities was spearheaded by Ann Riley, one of the most

prominent restoration practitioners in the country. She tried to start an apprenticeship

program through her non-profit to serve students at UC Berkeley and UC Davis, but it only

25 Author interview, Laura Wildman, American Rivers, 6/4/04.

96

lasted one semester because of lack of support from the universities. When asked about the

(in)adequacy of current academic training, Riley had this to say:

Right now the universities are lightyears behind what the practitioners are doing. There’s no place for anyone to go to school to learn this stuff. Students are asked to assess a restoration project even though they have never received training on how to design restoration projects. This is akin to hiring a restaurant critic who has never cooked a meal or been in a kitchen to publish a critique of a restaurant. It is my personal experience that most students do not have the training to perceive what they are looking at when they go into the field.

So people like Dave [Rosgen] and I, we’re reduced to holding these workshops and trying to pull in as many people as possible. The answer is to have apprenticeships. People should get training at the universities, but the professors are not practitioners and don’t know how to do that kind of training. You need to have non-profits or professional schools or something like that where people can go to learn on the job.26

Developing a new degree program is not made any easier by the fact that the primary

proponents of restoration programs have been applied scientists, who are usually lower

status than basic researchers. Among geomorphologists, for example, most basic researchers

initially regarded river restoration as simply another form of human disturbance, and thus

had no motivation to start or support programs to educate restoration practitioners, as

described in Section C, above. Without the support of their more prestigious colleagues,

applied scientists did not have the capital to start new programs from scratch.

Despite the difficulties of starting new programs, in the last five years three

universities (West Virginia University, the University of Minnesota, and Portland State

University) started restoration certification programs, and a group of faculty at the University

of Washington made a serious attempt to develop one (see Appendix B for detailed

information about these programs). While West Virginia University’s program no longer

exists, the new University of Minnesota and Portland State University programs – both

26 Author interview, Dr. Ann Riley, State Water Quality Control Board, 6/19/06.

97

launched in 2006 – appear to be going strong. While these programs are both too small and

too new to have yet had a significant impact on the restoration field, they are one indication

that public science is finally attempting to assert its traditional role in the development of a

new scientific field. However, university-based restoration education is not yet developed

enough to create a subjective structure for the stream restoration field.

2. Short Courses

Since most would-be restoration practitioners have not been able to find the

programs they need via the traditional university pathway, many have turned to restoration

short courses. More than 10,000 of them, for example, have taken one or more of Dave

Rosgen’s short courses. Is a short course series, however intensive, sufficient to inculcate a

shared habitus among its students? To answer this question, I will compare the

organization, student pools, methods of data collection, and design processes taught in

Rosgen’s short courses with their more serious competition: a two-level short course series

which I will refer to as the academic short courses.

a. Organization

Short courses differ from the coursework typically required for a professional

qualification such as a Masters degree in a number of key ways that shape the educational

experience of their students. First, the amount of material that can be covered in a short

course, even the five-day or more short courses that Rosgen and his competitors teach, is

substantially smaller than the material in a university course. Thus the range of information

presented in short courses is necessarily much narrower than that included in a traditional

98

university education, although that narrow focus means that the material covered is far more

precisely targeted for relevance to stream restoration.

Second, because short courses compress a great deal of material into a very short

period of time, students have little time to reflect on the material presented, ask questions,

and attempt to put it to use. Trying to absorb the content presented in short courses has

more in common with attempting to drink from a fire hose than from the more measured

cup of university courses. Thus students often report that they come out of a short course

thinking they generally understood the material, and then discover substantial holes in their

understanding when they try to apply what they thought they had learned. These students

can’t sign up for office hours or ask questions in lecture or lab the following week. Their

only recourse is going to additional short courses. Craig Fischenich of the USACE

Waterways Institute, a highly-respected restoration practitioner and Rosgen Wars moderate,

describes the process in relation to Rosgen’s student like this:

[W]hat I’ve witnessed is that people who have taken Dave’s classes and gotten interested and engaged … ultimately start asking questions that go beyond what Dave taught. I’ve watched this so many times. People go to that class and they come out of there and they feel good about it, and they go, ‘you know, I didn’t really understand everything he said, or some things aren’t quite clear but I got the general gist. The general gist is that it’s pretty easy to just go out and do it.’ So then they get out and start trying to do it and start running into some of these problems that others have identified and made a big deal out of. Most of those people will then look outward and try to find ways to solve those problems. I’ve taught a lot of people whose primary background was Dave’s classes 1-3, and who said, “look, we like this approach. It’s what we’re comfortable with. It’s the way we think about rivers because that’s what we’ve been taught. But on the other hand we recognize there’s some problems with it, too. And what we’d like to do is have a week-long workshop that focuses on the things we need to know that we didn’t learn in that class.”27

Finally, short courses differ from the standard university structure in creating a total

immersion experience, far from family, work, or other instructors that might question the

27 Author interview, Dr. Craig Fischenich, USACE Waterways Experiment Station, 6/27/06.

99

premises being taught. They thus have the capacity to create intense bonds among students,

and even faith-like conversion experiences.

All three of these characteristics are common to most short courses, but there are

also aspects distinctive to Rosgen’s courses and students that do much to create a habitus

and a community of practitioners to reinforce and support it. Rosgen’s courses are

distinguished from the academic short course series by an intense focus on teamwork.

Based on a field experience survey all students must complete before the course begins,

Rosgen divides course participants into four teams that form the basis for all of the field

work exercises during the week. Not only do these teams work together in the field (for

one day in the Level I course and two days in the Level II course), they spend evenings

working up their data together. Bonds among teammates are even more firmly cemented by

the communal presentation and critique of their data, an educational hazing experience that

is the source of considerable bonding even among those who took Rosgen’s courses at

different times. By contrast, in the academic short courses, the time set aside for socializing

was minimal and the atmosphere was, well, academic. Team-oriented work played almost no

role despite the emphasis on restoration as a team endeavor. While students were divided

into groups for field exercises (three students to do the long-profile, three students to do

pebble counts, etc.), these groupings were spontaneous and not built into the organization of

the course. Other assignments were completed individually.

There are also notable differences between the take home materials distributed in

Rosgen’s courses and those distributed in other short courses. As the take-home materials

become the primary resource for students once the course is over, they are worthy of

attention. In Rosgen’s courses, the take home materials underline the legitimacy of the

information presented: a bound, tabbed course notebook with background material and

100

detailed instructions for the field exercises and, most importantly, a copy of one of Rosgen’s

text books.28 This is a far cry from the ad hoc handouts typical of courses organized by

academics and agency researchers, which tend to provide students with a far greater range of

information, but leave it up to the individual student to explore it.29 There are also take-

home gifts in Rosgen courses – in Level I a Wildland Hydrology ceramic mug, in Level II a

Wildland hydrology commuter mug and tape measure, and in both a signed certificate of

completion – which allow Rosgen students to advertise the fact that they’ve been through

the courses and are members of Rosgen’s camp.

b. Student Pools

Rosgen and his opponents are drawing from the same pool of potential students:

primarily mid-career agency staff and consulting firm employees moving into the stream

restoration field. Based on data from a survey I conducted of students in the introductory

Rosgen and academic short courses, they are attracting somewhat different sets of students

from that shared pool. Most notably, the educational levels and exposure to science and

engineering curriculum seem to be lower among Rosgen’s students (although these data

must be considered at least somewhat provisional as the response rate to my surveys was

28 In the Level I course, students receive Applied River Morphology, Rosgen’s 1996 text book on classifying streams and valley types. In the Level II course, students are given Rosgen’s Field Guide for Stream Classification, a light weight waterproof Cliff Notes version of Applied River Morphology intended to be carried into the field. Both of these texts are self-published. In the Level I course, students are also given a second bound manual with supporting materials for the lectures as well as Rosgen’s Catena paper, the Blanco River case study from the 1992 NRC Report, and a number of Rosgen’s conference papers. 29 In the first course, students are given copies of a textbook that Kondolf edited, Tools in Geomorphology; rather than being a focused discussion of the material presented in course lectures, as Rosgen’s textbook is, this text is a compendium of chapters about different measurement techniques, some of which were covered in course lectures but most of which were not. At the end of the first course, the instructors also distributed a compact disc with key scientific papers referred to in lectures, but this was clearly an afterthought as the cds were produced in the classroom on the last day of the course while students were presenting their work.

101

only 50% for the Rosgen course and just 44% for the non-Rosgen course, 23 and 12

respondents respectively).30

As shown in Table 4.1, the respondents from the academic short course had on

average attained higher levels of formal education than those in the Rosgen course: while

just under 17% of the respondents in the academic short course had stopped with Bachelors

degrees, 26% of Rosgen’s students had, and one had only a high school diploma (I know

from talking to students in the Rosgen Level I course that there were at least four people

with only high school diplomas in attendance, but the other three did not respond to the

survey). Further, while 58% of the respondents from the academic short course had a

background in the physical sciences or engineering, only 43% of Rosgen’s Level I students

did (none in engineering). Fully 17% of the respondents from Rosgen’s course had no

science background at all. Rosgen’s students thus bring a relatively lower level of familiarity

with accepted scientific material and practice, and are that much more open to developing a

restoration habitus based on Rosgen and his teachings.

The idea that Rosgen’s students are substantially more open to taking on his stream

restoration habitus is supported by other survey data as well. As shown in Table 4.2,

students in the academic short courses bring with them more experience designing and

managing restoration projects, suggesting that the initial academic course may be used to fill

in specific aspects of restoration – geomorphology and sediment transport – for people that

are already familiar with the field, whereas the Rosgen Level I course is used as an

introduction to stream restoration in general. Supporting this only one respondent from the

academic short course said that they took the course in order to learn more about stream

restoration in general, whereas almost half of the Rosgen course respondents did (Table 4.3).

30 I initially sent the surveys via U.S. mail with self-addressed stamped envelopes included, and also via email. After a month, I sent a follow up email with another electronic copy of the survey attached to those who had not yet responded to the initial mailing/emailing. The full text of the survey is included in Appendix A.

102

Educational Level Rosgen I Academic I HS 4.3% 0.0% BS life sciences* 13.0% 16.7% physical sciences** 13.0% 0.0% engineering 0.0% 0.0% other*** 0.0% 0.0% MS life sciences 17.4% 16.7% physical sciences 26.1% 25.0% engineering 0.0% 33.3% other 17.4% 0.0% Ph.D. life sciences 4.3% 8.3% physical sciences 4.3% 0.0% engineering 0.0% 0.0% other 0.0% 0.0% Total 100.0% 100.0% * ecology, chemistry, wildlife management ** hydrology, geology, forestry *** humanities, social sciences, public administration, law

Table 4.1 Highest level of educational attainment by short course participants

DesignerFunder/

RegulatorProject

Manager Other*

Rosgen Level I 30% 26% 43% 48%

Academic I 42% 25% 58% 33% *operator, habitat assessor, surveyor, monitor, educator (Respondents were allowed to check as many boxes as were relevant.)

Table 4.2. Current Roles in Restoration

The data in Table 4.3 also demonstrate that the NCD approach is firmly embedded

in the practice of many agencies and consulting firms. More than a third of respondents

from the Rosgen course were there because their employer or workplace required it; no one

attended the academic course for this reason. Further, 26% of Rosgen course respondents

103

said that they were there because NCD was so widely accepted and/or well-known; none of

the respondents from the academic course mentioned the instructors’ design approaches.

Reason/Goal Academic I Rosgen I

Required by supervisor/workplace 0% 35%

Recommended by co-worker or supervisor 33% 13%

Learn about geomorphology/hydrology/sediment transport 92% 22%

Learn more about restoration in general 8% 48%

Alternative to Rosgen 25% 0%

Learn Rosgen's classification and/or approach 0% 48%

Reputation of instructor(s) 42% 13%

Widely accepted/well-known method 0% 26% Table 4.3 Reasons for selecting/goals for particular short courses

c. Methods of Data Collection and Design Approach

Because both the initial academic course and the Rosgen Level I course share a

critique of traditional hydraulic engineering, and because they attempt to introduce the basics

of hydrology and fluvial geomorphology (upon which they agree up through the early

1960s), there is a substantial amount of overlap between the materials presented in these

courses. Further, the basic survey techniques taught during the field days on the two courses

are nearly identical. In both courses, students are taught to do long-profiles, cross-sections

and Wolman pebble counts. (Since the instructors disagree about the breakpoint between

the active channel and the banks, they teach different variations on Wolman’s technique, but

the core method for selecting particles and the system for binning by particle size are

identical).

104

Despite these substantial overlaps in basic science and methods of data collection,

the two types of short courses ask profoundly different sets of questions, and thus focus on

different measurements to answer them. For example, as described in Chapter 2, NCD

starts by considering regional curves: data on regionally-typical relations between the size of

a drainage basin and the discharge, cross-sectional area, and bankful channel width of

streams within it. Most of the non-NCD restoration consultants I interviewed said that far

from being a starting point for analysis and design, hydraulic geometry relations only figure

late in the process as one of multiple checks on the reasonability of proposed designs.31 For

NCD practitioners the regional curves are foundational, the non-negotiable starting point of

analysis. Non-NCD practitioners, by contrast, are more likely to start with questions about

flows of water and sediment, and the basin-wide history of channel movement and land-use.

A second key step in the NCD process is channel classification, which provides

another example of the differences in questions asked and methods used by NCD and non-

NCD practitioners. While sharing some core data collection with other classification

systems, Rosgen’s classification process includes morphological features that are not

considered in academic practice, most notably bankful and flood prone width. The

differences are more fundamental, however. It is not just that non-NCD practitioners avoid

the Rosgen classification system unless forced to use it by clients; most non-NCD

practitioners don’t bother to use any type of classification system as part of their evaluation

and design process. Academically-trained practitioners tend to start with a system’s unique

qualities, whereas NCD practitioners tend to start from its similarities to other systems.

31 As Scott McBain, an academically-trained restoration consultant in Northern California, said when asked about whether he uses hydraulic geometry: “I use it just for checking what we’re doing, but I don’t design from it.” Author interview, Scott McBain, Principal McBain & Trush, 7/17/06.

105

The next step in the NCD approach provides a third example. Channel design for

NCD practitioners is deeply dependent on the availability of reference reach data. Finding

stable stream reaches of the same classification and within the same valley type is crucial to

set ranges for sinuosity, depth and other key design parameters in the NCD approach. For

non-NCD practitioners, the comparability of reference reaches depends far more heavily on

assessment of geology and flow patterns for water and sediment; they do not reference

channel or valley type.32

d. The Relative Certainty of Practice

Perhaps the most telling contrast between Rosgen’s approach and his critics’ is not

these stark differences in the types of questions asked and design methods utilized, but in

their treatment of the relative uncertainty of restoration science and practice. One of the

most common criticisms put forward by Rosgen’s opponents is that he continually

emphasizes the doability of restoration, thus giving his students a false sense of restoration

as a relatively easy and simple practice. As I will describe in more detail in the next chapter,

this criticism is not entirely fair; while Rosgen certainly does emphasize doability, at no

point in either the Level I or Level II Rosgen courses I attended did he describe restoration

practice as simple or easy. Not simple, however, does not equal not certain. Rosgen’s clear

take home message is that river systems are predictable and, therefore, if done correctly

restoration practice is reliable. When asked about the level of uncertainty associated with

restoration practice, Rosgen had this to say:

I have a great degree of certainty in the designs IF you get them built right. That’s the biggest weak link that I see in this business. And that’s where most people fall apart…. When I see that done right I have great confidence…. I

32 This divide is only likely to widen given Dr. Richard Hey’s recent work claiming that hydro-physiographic province is immaterial to comparability when using Rosgen’s approach (Hey 2006).

106

haven’t been doing this for 47 years and thinking we’re taking a shot in the dark. That’s one of the strengths that I can stand up on, because I can take you to hundreds of rivers that have responded the way we predicted.33

By contrast, the fundamental uncertainty of current restoration science and practice was the

main take home message of the academic short course, reiterated multiple times a day and

made concrete in the most memorable model of the course, Wilcock’s Monte Carlo

spreadsheet.

In the survey of students from the Level I Rosgen and the first academic short

courses, I asked the following question: “Based on the information presented in the course,

how predictable or unpredictable do you think restoration work is?” and asked respondents

to circle a number on a scale from one to ten, where one was totally unpredictable and ten

was totally predictable. Despite the fact that a substantial portion of the students in each

course came in with some hands-on experience with restoration projects, and thus could be

expected to hold views on the relative certainty of restoration practice independent of what

the instructors taught, there are still clearly visible differences between the two groups of

respondents (fig. 4.6). The average level of certainty posited by Rosgen Level I course

respondents was 6.5 out of ten, almost two points higher than the 4.8 posited by academic

course respondents. Even more tellingly, while 83% of Rosgen course respondents believed

that restoration practice was more predictable than not, only 43% of academic short course

students did. At a very basic level, the Rosgen short courses encourage participants to view

stream restoration as an already viable practice. By contrast, the academic short courses

encourage participants to view the current practice of stream restoration as a crapshoot,

overstepping the limitations of the science underlying it.

33 Author interview, Dr. David Rosgen, Wildland Hydrology, 4/22/07.

107

0

1

2

3

4

5

6

7

8

9

1 2 3 4 5 6 7 8 9 10

1 = totally unpredicatable, 10 = totally predictable

Rosgen I

Academic I

Figure 4.6. Level of Certainty of Restoration Work According to Short Course Participants

e. Habitus?

I would thus argue that the answer to the question posed above – can Rosgen’s short

courses inculcate habitus in a Bourdieu-ian sense – is probably yes. The intensity and lack of

alternatives characteristic of the short course format in general means that they can have an

educative impact far out of proportion with their length, and Rosgen’s courses are organized

to concentrate that impact through intensive teamwork. My survey data suggest that

Rosgen’s students are less likely to come in with developed scientific or restoration-practice

habitus, thus making them more open to adopting Rosgen’s. Further, nearly half of the

students were there because they came from a workplace where NCD is standard practice,

suggesting that for many students, dispositions acquired during the short courses would be

reinforced after leaving them. Finally, it seems clear that as an educative act, Rosgen’s short

108

courses are inculcating an approach to restoration based on methods, questions studied, and

overall philosophy that is distinctively different from non-Rosgen courses. Rosgen’s short

courses have thus been steadily populating consulting firms and regulatory agencies within

the stream restoration field with people whose restoration habitus was shaped by Rosgen

and his design approach, creating a powerful epistemic community.

This has serious consequences for restoration practitioners with conventional

academic training. Even if a Request for Proposals does not specify that consultants must

use NCD, if that is the only method with which the client agency is familiar, consultants

proposing any other approach are fighting an uphill battle. Steve Gough, the restoration

consultant quoted above about the difficulty of designing interdisciplinary Ph.D. programs,

described a project which had been in litigation for more than a year because ascending

levels of the client’s supervisors, all of whom had been to Rosgen’s Level I course, objected

to the fact that Gough did not use NCD:

The case has been settled twice, but each time someone showed up and said, ‘Wait a minute, this guy didn’t use the Rosgen System? This can’t be right!’ It’s reached the level of absurdity. Another guy comes in the room and says, “What? You didn’t use the Rosgen system!” And everybody goes, “Oh shit, did you have to say that? Now we have to open it up again.”34

And the fight is on once more.

F. Conclusion

In this chapter, I have outlined the basic structure of the stream restoration field in

order to provide an overview of the context within which the Rosgen Wars take place.

34 Author interview, Steve Gough, Little River Consulting, 6/19/06.

109

Using Bourdieu’s field concept and the three-part framework he recommends for

conducting a field analysis, I have argued that:

the relative autonomy of the stream restoration field is decreasing, as even basic

scientists who study streams are pushed into applied work by Rosgen’s success and the

rapidly expanding restoration market;

the objective structure of the stream restoration field is characterized by axes of tension

between science and practice, and between environmentalists and proponents of

traditional hydraulic engineering approaches; and

that Rosgen’s short courses do inculcate a habitus of sufficient strength that it

increasingly provides a subjective structure for the stream restoration field.

The remainder of this dissertation focuses on the fight over Rosgen’s work.

Bourdieu is quite clear that in studying any scientific conflict it is crucial to analyze both its

intellectual and political content as the two are deeply interrelated (Bourdieu 1975, p.21).

Thus Chapter 5 reviews the substantive content of the Rosgen Wars, while Chapter 6

examines how those substantive arguments are mobilized for political ends.

111

Chapter 5: The Intellectual Content of the Rosgen Wars

“An analysis which tried to isolate a purely ‘political’ dimension in struggles for domination of the scientific field would be as radically wrong as the (more frequent) opposite course of only attending to the ‘pure’, purely intellectual, determinations involved in scientific controversies.”

Pierre Bourdieu, “The Specificity of the Scientific Field,” p.21.

The Rosgen Wars have been one of the organizing principles of the stream

restoration world for more than a decade. While no one denies the role of personality in

fanning the flames of the debate, critics have for the most part focused on substantive

critiques of the NCD approach. Rosgen’s opponents raise an exhaustive list of concerns

about the content and use of his classification system and design approach. In the spirit of

classification, these concerns can be loosely divided into five groups: critiques of 1) the

classification system, 2) the design approach, 3) the short courses, 4) Rosgen’s scientific

practice, and 5) the scientific content of his work. This chapter attempts to explain and

evaluate the most common and substantive critiques.1

There are clear themes in these critiques that one might expect to find in any conflict

about environmental practice: the rarified realm of basic science vs. applied science’s struggle

for accessibility; the intellectual rigor of science vs. the practical requirements of consulting.

Another theme revolves around questions of methodology and the struggle to develop

predictable practices, which seems typical of any relatively new field.

What is notable about the Rosgen Wars, however, is the seeming impossibility of

closure. Despite the fact that they have preoccupied the stream restoration community for

1 This chapter address 21 of the most common critiques. Additional critiques of the classification system, a less common focus for critics, are covered in Appendix C.

112

more than a decade there is remarkably little data available to resolve the substantive issues I

will describe below. In the 15 years since Rosgen’s Catena article was published, no one has

mounted the kind of broad-based study of project outcomes that would be necessary to

resolve the debate. The complexity of riparian systems and the high level of uncertainty of

restoration science make the production of such conclusive data a considerable challenge.

At the most basic level, it is not clear how researchers would establish equivalency between

projects in different watersheds and hydro-physiographic provinces. Establishing

comparability between practitioners is another very thorny issue, given the wide variety of

paths people travel to become successful restoration practitioners. Similarly, it would be

difficult to reliably categorize methods used into NCD and non-NCD approaches, given the

variety of approaches even practitioners trained in the same method employ. Simply

determining criteria for success is also complicated, given that for Rosgen and his supporters

a successful project does not move, while for his opponents the goal is a dynamic channel.

For all of these reasons a definitive, geographically-broad, comparative study of restoration

approaches would be difficult to carry out, but it is surprising that neither Rosgen nor his

critics have attempted such a study, and only one paper has even suggested it (Juracek and

Fitzpatrick 2003).2

Good case studies could provide useful, if not definitive, data. But here, too, we

come up short. While critics have conducted excellent case studies of particular restoration

projects (most notably Soar 2000, Kondolf et al. 2001, and Smith and Prestegaard 2005)

these studies have at best limited relevance as the designers did not follow anything

approaching Rosgen’s complete approach in any of the projects.

2 In an interview on 4.22.07, Rosgen showed me a sheet of notes speculating about how to do a broad study, but to my knowledge he has not tried to start such a project.

113

Rosgen’s supporters, for their part, have not conducted the type of detailed

published case studies necessary to meet scientific standards of evidence. Instead, they

refute their critics with nearly twenty years of anecdotal evidence about NCD projects,

typically claiming success rates of 80% and above.3 Given that Rosgen’s supporters are just

as committed to healing riparian systems as their critics, their claims for project success

should not be dismissed. Indeed some of them are very convincing. For example, Buck

Engineering, a North Carolina firm highly experienced in implementing NCD projects,

offers clients warranties on their projects that cover the costs of repairing or replacing any

structure that moves for a set number of years after project completion. Those warranties

are a net source of revenue for the company: once in place, the Rosgen structures Buck

Engineering installs rarely move.4 Still the informal presentation of these and other claims

limits their plausibility with Rosgen’s critics.

Without data from broad comparative projects or even case studies, we are stuck

with claim vs. counter claim. Thus neither I nor anyone else can adjudicate the disputes

described in this chapter.5 Where there is sufficient evidence to demonstrate that critics’

claims are fair or unfair I have done so, but many of the core questions remain

unsubstantiated.

3 The lowest claim I have seen comes from Mondry et al. 2006, who report a stability rate of 74% for Rosgen structures installed by the North Carolina Ecosystem Enhancement Project in restoration projects 2 to 5 years old. 4 Author interview, Will Harman, Buck Engineering, 8.14.06. 5 To be clear, I am not trying to claim that I am the most appropriate person to adjudicate these critiques. I do hope, however, that having listened carefully to both sides, attended their short courses and meetings, and walked their projects, I can shed some light on areas where the two sides are talking past each other.

114

A. Critiques of the Classification System

Rosgen’s classification system is by far the most widely used aspect of his work.

Even people who disagree with Rosgen intensely have learned to use it if they have to deal

with practitioners on a regular basis; all but one of the anti-Rosgen consultants I spoke with

were able to classify streams based on Rosgen’s system. The classification system is also the

least controversial aspect of his work. Even the most vehement Rosgen critics will

grudgingly acknowledge its utility, if only as a sort of consolation-prize preface to

concentrated attacks on the NCD approach.6

Despite a greater level of tolerance, or at least resignation, about the classification

system, Rosgen’s opponents do raise several objections to it. A common criticism is that

there are areas of the country or types of landscape where it does not work well. Some

critics have asserted specific limitations on the applicability of Rosgen’s classification system,

developing detailed lists of particular landscapes where the classification system does not

apply. Rosgen vehemently opposes these attempts to limit the applicability of his

classification system. As he said in response to a question about this at the Level II course I

attended:

There are maybe three states where I haven’t taught a course. And in every place I go I visit a gauge station and stream sites. I’m always looking for a new stream type, but I’ve never found one. It’s just more of a, a social thing. “Not invented here” syndrome…. [The classification system is universally applicable because] it’s physics. A 1% slope is the same expression of energy no matter where it occurs.

6 For example, the last paragraph of Miller and Ritter’s 1996 rebuttal of Rosgen’s Catena article begins: “We wish to stress that we have no major problem with using the Rosgen classification system as a communications tool. But as we have said above, the process significance of the defined stream types and their use in a predictive sense is questionable.” (p.298) An exception to this is a 2008 article by Roper et al.

115

One common argument is that Rosgen’s classification system does not apply in arid

climates because bankful discharge, the keystone of his approach, does not apply there (see

discussion of the bankful concept in Section E, below). A second commonly-claimed

exception is karst landscapes, because the relations between surface and sub-surface flows

are so different. Rosgen explicitly states that his classification system applies in both of these

landscapes. For example, in the Level II course I attended, the stream we were supposed on

to classify on the first field day happened to be within a belt of karst. It was quite clear that

Rosgen did not see the karst as an impediment to applying his classification system; he

simply took the existing regional curve data and stratified it by geology to reflect reduced

flows in the karst belt. Using his geologically-stratified regional curves, all four field teams

came to the same conclusion about the stream’s classification. Without further data, it is

impossible to say whether the particular character of arid or karst landscapes invalidates the

inferences Rosgen’s draws from his classification system, but it was certainly possible to apply

his classification system in the latter landscape in a reasonably repeatable fashion.

The argument that Rosgen’s classification system does not apply in drainage basins

that are undergoing land use change is intuitively more convincing. If a channel is still

adjusting, then its current form cannot yield accurate inferences about the shifting processes

at work within it. However, as explained in more detail below in Section E, Rosgen’s

classification system is not as dependent on current channel form as critics assume. Thus

this criticism is not as potent as it appears.

A third critique of the Rosgen classification system is that because it defines certain

types of streams as unstable, some classifications are tantamount to declaring the channel in

need of intervention, creating a rush to ‘fix’ streams that aren’t broken. As Peter Whiting, a

geomorphologist at Case Western Reserve University, put it:

116

Where I become far more concerned is when an identifier, a descriptor, is the… basis for deciding there's something the matter, and therefore we potentially need to do something…. [J]ust because a channel has been classified, doesn't mean that automatically it is damaged, but there's often the presupposition… on the basis of the characterization that there's something wrong if the channel is of a certain type…. [S]tarting from this standpoint that if there's something wrong, we should probably try to fix it, … often such fixing is oriented towards adjustments we can make to the channel to make it work "better" as opposed to looking at the system with a… larger perspective as to why things are going on.7

Rosgen himself explicitly states in his short courses that some channels are naturally unstable

and can point to a number of examples where he has incorporated such channels into his

designs. As he said when I interviewed him in April 2007, “It bothers me that people say I

hate braided channels. I’ve designed a lot of braided channels because that’s Mother

Nature’s way of depositing sediment.”8 Based on projects such as the Uvas Creek fiasco

documented in Kondolf et al. 2001, however, at least some of his students fail to accept this

point. The stability issue is addressed in more detail in Section E, below.

A last objection to the classification system made by Rosgen critics relates to how it

is used. It is very common for critics to assert that Rosgen practitioners use classification of

the project reach as the primary basis for design. The reality is more complicated. As

discussed in Chapter 2, the primary design source in NCD is not the project reach, but the

reference reach, preferably reference reaches. It is from the reference reaches that key design

parameters such as the value ranges for slope, sinuousity, point bar slope, and pool spacing

are derived. Reference reach selection, however, does tie directly into the classification of

the project reach: their channel type, valley type, and hydro-physiographic province must

match. Thus critics are partially correct: the project reach’s classified channel type does

have an impact on design, but not in the simple one-to-one relation they describe.

7 Author interview, Dr. Peter Whiting, Case Western Reserve University, 7/13/06. 8 Author interview, Dr. David L. Rosgen, Wildland Hydrology, 4/22/07.

117

Further, classifying the project reach is not a simple matter of applying the

classification system to existing conditions. Rosgen emphasizes the importance of

considering a reach’s potential, not just its current impacted form:

Streams need to be considered often, not in terms of their current degraded state, but in terms of their future potential as conditioned by their watershed and valley features. Applications of the evolutionary stages of stream types presented earlier in this chapter assist in the determination of the appropriate stream type. Similarly, it is essential that restoration goals not be based on conditions – regardless of how desirable – that bear no relationship to the river system. (Rosgen 1996a, p.8-33)

To assess the project reach’s actual potential, Rosgen suggests three possible methods, all of

which should be used with due consideration of his channel evolution models: 1) evaluating

the reach in relation to its channel type to see if one or more variables are at the outside edge

or beyond the ranges specified for key parameters; 2) using historical photographs or other

available data to determine the project reach’s pre-disturbance form; 3) examining the

channel conditions up- and down-stream of the project reach (Rosgen 1996a, p.6-5). This

extra-reach data feeds into the classification of the project reach, which in turn is a factor in

selecting the reference reaches from which the design parameters eventually emerge. Critics’

assertions that the project reach classification is the basis of design in NCD thus seem too

simplified a picture of what Rosgen actually recommends and practices; whether his students

incorporate these methods into their work is another question.

B. Critiques of the Natural Channel Design Approach

For many of Rosgen’s critics, the most powerful condemnation of the NCD

approach stems from its effects on the ground. Every member of the Rosgen opposition

118

with whom I spoke alluded to the potential for project failure because of the “cookbook”

nature of NCD combined with Rosgen’s students’ overconfident application of the small

amount of material they’ve learned in the short courses. Part of this arises from general

tensions between science and practice, but some of the objections are specific to the NCD

approach. For example, some critics point to the prevalence of what they call “sine-curve”

or “radio-wave” projects with perfectly symmetrical meander bends as evidence of the

failures of NCD. Others point to what have become the iconic case studies of failed

Rosgen projects: Deep Run in Maryland, as documented in Smith and Prestegaard (2005),

and Uvas Creek in California, as documented in Kondolf et al., (2001).9 While moderate

members of the opposition acknowledge that many non-NCD approach projects fail as well,

the consensus among critics is that Rosgen’s students are more prone to design failures.

Further, although a substantial minority of Rosgen opponents feel that he should be held no

more responsible for his students’ failures than university professors should be for their

students’, the majority of the opposition argues that as their sole educator and enabler, his

students’ actions can be laid at Rosgen’s feet.

The design failures are seen by critics as particularly problematic for three main

reasons: they waste taxpayer money, they may in some cases cause irrevocable ecological

harm, and widely-publicized failures may undermine the entire project of stream restoration

by eroding public support for it. As Daniel Levish, U.S. Bureau of Reclamation staff

member and vehement Rosgen critic, put it:

What could be quantified would be the huge sums of federal dollars wasted on inappropriate ‘stream restoration’…. When you consider that each of the people Rosgen trains applies his method to several hundred streams over their career, the potential environmental degradation is stupendous! These

9 As I explained above, the designers of these projects did not follow anything close to Rosgen’s complete design approach, but critics continue to cite the studies of them, perhaps unaware of their limited relevance to the debate over Rosgen’s work.

119

are things that can’t be undone! So there’s a huge amount of money that’s been wasted, but there’s also tremendous amounts of damage to stream systems, irreparable sorts of damage…. It’s awful to watch people waste so much money and think they’re doing good when they’re doing harm.10

Syd Brown, a staff scientist with the California Department of Parks and Recreation who is

generally supportive of Rosgen’s work, voiced similar concerns:

[T]ruly what’s at stake are resources: public trust resources and future funding. It’s so difficult to get goverment support, which is usually where these things come from, for altruistic activities. If it’s done poorly or if it fails the fauwcet runs dry. So future support, future restoration, future enhancements to the natural environment by humans become more difficult to get support for.11

In the absence of a broad study comparing the success of NCD and non-NCD

restoration projects, it is difficult to determine the fairness of these claims. While at first

glance it does seem hypocritical to nail the failures of Rosgen’s students to his door without

attributing similar levels of responsibility to university professors, the reality is more

complicated. As should be clear from the discussion in Chapter 4 about lack of educational

alternatives, very few professors actually teach restoration courses. Even those that do teach

restoration-related university classes or short courses often address only particular issues,

such as sediment transport, rather than broader design training. Part of the enduring

popularity of Rosgen’s courses is that there is no comparable alternative: only Rosgen

provides such a focused, intensive education in restoration supported by such developed

materials. Thus I do think there is some rationale for holding him more responsible for his

students’ work than, say, a professor teaching introduction to fluvial geomorphology.

Rosgen is hauled up on charges of empowering undertrained students frequently

enough that he has a stock answer:

10 Author interview, Dr. Daniel Levish, U.S. Bureau of Reclamation, 6/28/06. 11 Author interview, Syd Brown, California Department of Parks & Recreation, 6/24/04.

120

A lot of people say I make people dangerous because I make it [restoration] sound so simple. I always ask how many people are already doing restoration in the first class [Level I], and over half the hands go up.12 Am I responsible for these people? If I can help them I will. Their bosses have already forced them to fix rivers. They’re taking this training because it’s what’s available. It’s a complex issue.13

Because of this complexity, perhaps the most fair answer would be that while it is

not appropriate to hold Rosgen responsible for what his students do, it is appropriate to ask

that he try to discourage restoration work by the unprepared, and to provide avenues for

students to improve their design practice once they leave his courses. What might this look

like? Perhaps he could offer targeted follow-up courses to pass on new information or

modifications to existing techniques. He could moderate his famed emphasis on doability

(more on this is Section 5, below) with a, “kids, do not try this at home!” lecture at the end

of the Level I – III courses. Perhaps he could even require some sort of test of competency

to complete the Level IV course. None of these things would stop the problem of people

who take one or two Rosgen courses and think that’s all they need, but separately or

together they might at least moderate the problem.

Taking a step back, it is important to note that this issue arises because unlike other

design professions that have major impacts on the landscape – architecture, landscape

architecture and engineering – there is no national licensing or certification program for

restoration practitioners. The issue of accountability and quality control is universal in the

stream restoration community, though it may loom a bit larger for Rosgen and his students

since, judging from my survey data, they have a lower percentage of licensed engineers.14

Lack of certification has very concrete results: unless they are also licensed engineers or

12 In the Level I course I attended, just under a third of people had previous restoration experience, but that may have been atypical. 13 Author interview, Dr. David L. Rosgen, principal, Wildland Hydrology, 4/22/07. 14 While 33% of the students who took the academic short course described in Chapter 4 had engineering degrees, none of the students from Rosgen’s Level I course did, nor does Rosgen himself.

121

landscape architects, restoration practitioners are not required to sign and stamp their

construction drawings, so they have no legal liability for project failures. If a restoration

practitioner, Rosgen- or academically-trained, builds a project upstream from your property

and it causes a meander bend cut-off that floods your home, you cannot sue him or her for

bad practice. To address this, the National Center for Earth-Surface Dynamics (NCED)

convened a panel of experts in restoration education to propose restoration curricula and a

national certification process (Wilcock et al., in press, discussed in depth in the next chapter).

If the NCED proposal is implemented, the issue of empowering bad practice will go away

for all restoration instructors.

A second common critique of the NCD approach as restoration practice is that it

goes too far in fulfilling clients’ desires. These days all but the diehards think it is

inappropriate to try to recreate pre-disturbance conditions within a system that is no longer

suited to them. Inserting trout streams into landscapes where they would not occur on their

own is viewed with skepticism, however. As Karin Boyd, a university-trained restoration

consultant describes it, clients bring a particular set of demands to the table, and some

practitioners meet them whether they’re characteristic of the system or not:

One thing that the Rosgen Method has created is a bit of a shopping mentality for streams, where … everybody wants a C4 channel, and they’re just totally inappropriate in many places. You see people trying to build them where it’s not a sustainable channel type.15

An example of this type of practice, the Three Forks Ranch project in Colorado, was

discussed in Chapter 2. Brian Bledsoe, professor of hydraulic engineering at Colorado State

University, monitored this extensive project (more than 500 structures were used over a

distance of approximately 4 miles), which was designed by Rosgen himself, for five years. In

15 Author interview, Karin Boyd, Principal, Applied Geomorphology, Inc., 7/18/06.

122

describing it, Bledsoe remarked that while the project has certainly been successful at

creating and sustaining trout habitat, it is not an ecosystem type the ranch property could

sustain without intervention.16 Thus this critique seems fair: a shopping mentality does seem

to affect the design of at least some projects designed by Rosgen and his students.

A third common critique comes from ecologically-minded practitioners who argue

that NCD is not bioengineering but traditional hydraulic engineering in a more attractive

package. These practitioners argue that true bioengineering is deformable, and thus

predominantly biological. As consultant Scott Gillilan put it,

As far as the restoration method that has always been my number one gripe: the technology is so heavy-handed. If you’ve nailed your geomorph, how come you have to nail the bank in place with rock vane weirs and riprapped outer banks? It’s no different from traditional engineering. They should call it what it is. If it can’t evolve with the channel, I don’t see it as a restorative practice, I see it as channel tweaking.17

Or as consultant Steve Gough described it,

As I look at projects … that were designed by Dave or his disciples, I see a very common element: outside bends are armored, sometimes very heavily. And at each crossover or "riffle" zone, there is a structure I'd call a grade control…. I use these in urban areas to focus energy on an armored spot, and when used properly, these things are very powerful medicine. Armoring outside bends beyond what is needed to allow vegetation a chance to regrow is not "natural." And the grade control structures are not "riffles," in the natural sense, i.e. they are immoveable. These two elements… do not add up to "natural channel design."18

Boyd made a similar point in regards to rootwad revetments: “In certain situations they

[rootwads] can be good habitat elements, but as major armoring pieces that’s what they are:

armor. To call it bio-engineering I think is kind of a stretch.”19

16 Author interview, Dr. Brian Bledsoe, Colorado State University, 8/23/06. 17 Author interview, Scott Gillilan, Gillilan Associates, Inc., 8/10/06. 18 Author interview, Steve Gough, Little River Consulting, 6/19/06. 19 Author interview, Karin Boyd, Applied Geomorphology, Inc., 7/18/06.

123

Rosgen himself seems to have some ambivalence on this front. On the one hand, he

consistently describes as successful projects where the structures stayed put; I do not recall

him ever showing pictures of a channel where the structures he put in place have been

replaced by vegetation, or where the stream had adjusted out of the design channel. On the

other hand, he has made several recent comments to the effect that the point of his

structures is to hold the channel in place until the bank vegetation grows in and the bed

adapts to its regime of incoming water and sediment. In the final lecture of the Level II

course I attended he put it this way:

[Restoration is] way more than structures. Sometimes I feel bad that I invented J-hooks and cross-weirs, ‘cause I see people thinking restoration is just structures. I never put them in unless the dimension, pattern and profile are right. All these things have a reason. Structures buy time for vegetation to take hold, and for the pavement and sub-pavement20 to set up.21

Given what seems to some observers a mismatch between the size of rocks and LWD

specified for use in Rosgen structures and the size of material a restored channel is capable

of moving, the current iteration of Rosgen’s structures does not embody this bioengineering

ideal. I would thus argue that the bioengineering practitioners are correct on this one:

Rosgen’s current suite of recommended structures do not promote channel adjustment. On

the other hand, NCD critics clearly have Rosgen’s attention, and he has already proved

himself capable of shifting his positions to accommodate new information. It will be

interesting to see to what extent his methods evolve in coming years to more closely follow

bioengineering ideals.

20 In a stream context, pavement and sub-pavement refer not to asphalt, but to the structure of sediment (which could range from silt all the way up to boulders) that make up the surface and subsurface layers of the stream bed. 21 In my interview with Dr. Brian Bledsoe, he reported that Rosgen had made a similar remark to him during a field visit to the Three Forks Ranch project.

124

Finally, the most commonly-leveled charge against NCD is that it is a ‘cookbook

approach’. For those more familiar with the cooking channel than a braided channel, this

critique may seem mysterious, but in brief, the issue is this: critics argue that by breaking his

design process down into a step-by-step, recipe-like process, NCD ignores both the

complexity and the regional-specificity of stream channels in favor of universalizing and

simple techniques that can be used anywhere by people without the judgment and

experience to understand when they do or do not fit. This is a classic science vs. practice

argument.

In an ideal world, this critique would be fair: we would leave all but the most

threatened stream systems alone to heal themselves, and what restoration we did attempt

would be done by local experts of great experience. What those who raise this objection

overlook is both tremendous public demand and a stream restoration market that hit the

billion-dollar a year mark a decade ago and continues to expand rapidly. Many in the stream

restoration community, regardless of their stance on NCD, argue that we need some kind of

guidelines or standards of restoration practice, even if they are ringed round with caveats and

insistence on the specificity of local conditions. This is the motivation behind on-going

efforts to develop national restoration handbooks, such as the new NRCS design handbook

(NRCS 2007) and the ASCE effort that is just getting underway, (Slate et al. 2007, discussed

in Chapter 2 and Appendix D).

If the form of the NCD guidelines is a necessary evil then the issue becomes the

content of the 40-step process, and here the jury is still out. Without a large scale study

evaluating the success of projects that actually use the full NCD approach, there is no good

way to adjudicate between claims from critics that NCD projects are inordinately prone to

failure, and claims from supporters that their current success rate is 80 % or higher.

125

C. Critiques of the Short Course Series

Rosgen’s series of four short courses, described in Chapter 2, form the heart of his

enterprise. In stark contrast to the publication-based culture of academic science, they are

the primary means through which he disseminates his work. He has taught more than

10,000 students in the last twenty years. While there is no good estimate of the number of

restoration practitioners in the U.S. (I make a rough attempt at this in Chapter 2), this is

surely a high proportion of them. Thus the short courses are a lightening rod for negative

comments.

To recap, Rosgen’s short-course series lasts for 29 days, some of which are spent in

the classroom and some in the field. Level I introduces Rosgen’s classification system and

fluvial geomorphology in general, Level II covers application of the classification system and

reference reach method, Level III focuses on sediment transport and hydraulics, and Level

IV teaches channel and structure design. As in North Carolina (described in Chapter 3),

many local, state and federal agencies that fund restoration projects require either full

Rosgen training or completion of Levels I and II to bid on their projects, causing many

consulting agencies to require NCD for their employees.

Criticisms of the short courses typically begin with a blanket condemnation of the

short course format as the sole means for educating practitioners. Setting aside objections to

the content that Rosgen teaches, opponents believe it is impossible for anyone to convey the

information needed to practice stream restoration in six weeks of courses, however

126

intensive; there is simply too much that people need to know.22 Given the complexity of

stream systems, this objection seems fair, if somewhat hollow in the face of the lack of

opportunities for more in-depth restoration education described in Chapter 4.

Moving deeper, opponents commonly raise three issues specific to the way in

which Rosgen’s courses are organized: his emphasis on the doability of restoration, his

privileging of geomorphology, and the lack of pre-requisites or exams in his courses. First,

the emphasis on doability is one of the key objections to Rosgen’s teaching approach and to

his work in general, and is an obvious example of the science vs. consulting split. Rosgen is

widely seen as promoting the view that scientists have made restoration seem far more

complicated than it is, and he is thus accused of empowering legions of under-trained

enthusiasts with the confidence to get out the construction equipment and destroy streams

in the name of helping them. As federal agency scientist Dan Levish describes it:

I think that was the biggest disservice that Rosgen did: he sold stream restoration as something that’s simple.… [P]eople who were honest and educated realized that this was not a simple undertaking, and it might take years and lots of money to do correctly. And Rosgen came along and said…, “Hey, it’s really simple. Don’t listen to them there scientists. Just do some stuff!”23

Karin Boyd described the stark contrast between the short courses that she co-

taught in the 1990s while employed at Inter-Fluve, a major NCD consulting firm, and the

Rosgen short courses:

Our whole thing was: this is hard. There are a lot of things to understand, and very typically you don’t have the amount of money you need. It would be nice if all of these could be dissertation-level projects, but they’re not. And so you have to make some decisions and you need to be informed. Rosgen’s course, on the other hand, really makes it look easy. And I think that was the fundamental response we got from people. When I asked

22 This criticism is expressed with particular vehemence by the considerable number of critics who are under the misapprehension that the complete Rosgen training still lasts only one or two weeks. 23 Author interview, Dr. Daniel Levish, U.S. Bureau of Reclamation, 6/28/06.

127

someone, and I’ll never forget this, I asked him, “This is great! You went to Rosgen’s course three weeks ago. What is the difference?” And he said, “You know I’m leaving your class with so many questions, and I left his with all the answers.”… So ours might have been a tad more humbling, and his more empowering. But we felt good about that, we thought that was appropriate.24

While Rosgen certainly does emphasize doability, at no point in either the Level I or

Level II courses I attended did he describe restoration practice as simple or easy; quite the

contrary. Rosgen consistently emphasized that restoration is difficult and requires

experience. For example, at different points on the first day of the Level I course Rosgen

had this to say:

I want to caution you, sometimes I make things look a little simple, maybe because I talk fast. I hope you appreciate the fact that this is not simple. It takes another 280 hours of study at a minimum, and even then people have some difficulties because of lack of experience.

I’d like to make this sound simple, but it is not simple. It is doable, but you’re going to have spend a lot of time in the field, in measurement and observation.

It is crucial to emphasize, however, that for Rosgen ‘not simple’ does not equal

‘uncertain’, and it is here that Rosgen’s critics have some traction. His overall message was

clearly that it should be possible, with sufficient knowledge and training, to be confident that

changes you make to a river system will perform the way you expect them to. As discussed

in Chapter 4, this leads to big differences in opinion between NCD and academic students

(while 83% of Rosgen course respondents believed that restoration practice was more

predictable than not, only 43% of academic short course students did). Given the high

degree of uncertainty about current restoration practice, Rosgen’s critics’ concerns on this

24 Author interview, Karin Boyd, Applied Geomorphology, Inc., 7/18/06.

128

front seem reasonable.25 As argued above, however, I think this issue could be addressed

fairly easily by shifts in Rosgen’s short courses.

A second objection to Rosgen’s organization of his short courses is that because he

is the only instructor and lectures only on fluvial geomorphology and hydrology, he conveys

to his students they do not need to draw on any of the other disciplines typically involved in

restoration in order to complete successful projects. Rosgen occasionally mentions

consulting with local “fish-squeezers” to get limiting factors analyses and other data relevant

to his design process, but he never mentions a biologist collaborator or partner in crime the

way he constantly references Luna Leopold or Lee Silvey, a geomorphologist and

hydrologist respectively. While Rosgen does say in his courses that it is important to consult

with other disciplines, the solo practices of many of his students suggest that this message is

getting lost along the way.

By contrast Rosgen critics tend to stress that restoration is an inherently multi-

disciplinary practice that depends on collaboration among geomorphologists, engineers and

biologists. As engineer and consultant James MacBroom put it,

I think people need to be cross-trained and realize there’s more to stream restoration than just the Rosgen technique. I think people need to be trained in the fundamentals of biology and stream ecology, because ultimately that should be the final product.26

Critics’ short courses reflect this, drawing in multiple instructors from a broad range of

disciplines. In the academic short course described in the previous chapter, for example,

there were two primary instructors and no fewer than seven guest lecturers. The Inter-Fluve

short course described above by Karin Boyd included instruction in hydrology, hydraulics,

25 At the 2008 Stream Restoration Institute biennial conference Martin Doyle, a geomorphologist at University of North Carolina and former consultant with Inter-Fluve argued that when ecological response is taken into account, there are no successful restoration projects documented in the Southeastern U.S. This speaks to the uncertainty of all restoration attempts, Rosgen or otherwise. 26 Author interview, James MacBroom, Malone & MacBroom, 7/24/06.

129

sediment transport, geomorphology, fish biology and plant ecology. Whether you support

the idea of a single cross-trained uber-restorationist or a collaborative team of specialists, it is

difficult to imagine restoration projects succeeding on multiple fronts without a concomitant

range of expertise. Thus this critique seems fair: even if Rosgen wishes his students to work

in multi-disciplinary collaborations, his solo teaching and restoration practice and the

content he covers in his courses do not support that message. There has been some

discussion of adding a fisheries biology segment to his short course series;27 that would

certainly go a long way towards addressing this issue.

A third major objection to the way in which Rosgen runs his short courses is that

there is no quality control. He has no pre-requisites, and gives no exams to assess whether

people are learning what he’s trying to teach. As Doug Shields, a researcher at the National

Sedimentation Laboratory, a branch of USDA’s Agricultural Research Service, put it:

We were at this meeting last year and I said “Dave, are there any prerequisites for your short course? Do you have to be able to read and write? I know that formal education and intelligence are not correlated. I’ve met a lot of people that didn’t have my level of education who are smarter than I am. But the paradigm we’re working with here, has anybody ever failed your short course? Has anybody ever failed?” “Well, no, nobody’s ever failed.” If we’re going to have people working out here on streams and ecosystems I think we need to have high standards.28

It is worth separating this issue into two pieces: the need for pre-requisites and the

need for assessment. There is a general assumption by Rosgen’s opponents that many of the

people taking his short courses are ill-prepared in the scientific disciplines underpinning

restoration practice. Based on my data from surveys of students in the Rosgen Level I and

first academic short courses (shown in Table 4.1, Chapter 4), there are indeed notable

differences in the educational background of the students. While 58% of the students in the

27 Author interview, Dr. Steven Kite, West Virginia University, 7/26/06. 28 Author interview, Dr. F. Douglas Shields, ARS National Sedimentation Laboratory, 6/13/06.

130

academic short course had a background in the physical sciences or engineering, only 43% of

Rosgen’s Level I students did, and 30% of the students in Rosgen’s course had no college-

level science background. While 33% of the students who took the academic short course

had engineering degrees, none of the students from Rosgen’s Level I course did. Further,

the students in the academic short course had on average attained higher levels of formal

education than those in the Rosgen course: while just under 17% of the students in the

academic short course had stopped with Bachelors degrees, 26% of Rosgen’s students had,

and one had only a high school diploma.

It thus seems fair to me to argue that Rosgen’s students are more likely to come in

under-prepared, and thus be at a disadvantage throughout the short-course series. Given the

security of Rosgen’s position in the field, it seems reasonable that he could require students

to complete some college-level course work in geomorphology, hydrology or ecology before

attending his courses.

The second part of this criticism – the need for some kind of final examination –

seems more important, but is just as applicable to Rosgen’s opponents’ courses. The

necessity of some kind of assessment of knowledge uptake to ensure quality control was

stressed by a number of people, including Steve Kite, a geomorphology professor at West

Virginia University who teaches similar short courses:

He [Rosgen] needs to work in some assessment of learning outcomes for individuals who takes his classes. And this is something we didn’t do in our courses for the first three or four years that we offered them, but I think there needs to be, if you will, an accounting, a test, an assessment vehicle, so that you can show that you have learned something…. I think back to my classes and a lot of the clientele we’ve had over the years didn’t have their heart in it. I remember having one person who went through the program and got a little certificate saying he completed the program who was mainly interested in coffee and cigars and hanging out at the back of the room. I think he was out of the lecture room as much as he was in. Went out in the field, saw the first stream crossing, and said, “I’ll work from here.” That person shouldn’t have gotten that certificate. But we at that point didn’t

131

have anything other than force of personality in order to get that person to execute. I think Dave’s pretty good at getting people to work because of his magnetic personality, but … I do think that we across the board need some sort of certification test.29

This seems to me a critical lack in most available restoration training. Interestingly, the new

certificate programs offered by the University of Minnesota and Portland State University

(discussed in Chapter 4), include examinations or graded assignment for each individual

course taken toward the certificate, though in neither case is there an overall examination.

This is where the proposed NCED certification program described above would come in.

A final criticism of Rosgen’s short courses is that attending them is treated as

necessary, and also as sufficient by a growing number of agencies and municipalities.30

Students who have completed Ph.D.s in fluvial geomorphology or hydraulic engineering are

being turned away from restoration jobs because they are considered unqualified without

NCD training.31 Professors and full-time consultants with decades of experience cannot bid

on projects because they have not gone back and studied their own subject as taught by

Rosgen.32 Critics argue that this exclusivity limits restoration practice by barring alternative

29 Author interview, Dr. Steven Kite, West Virginia University, 7/26/06. 30 Direct evidence for this is presented in Chapter 3, in data from North Carolina. Indirect evidence comes from my interviews and survey data. As described in Chapter 3, just over a third of respondents from the Rosgen course were there because their boss or workplace required it; no one attended the academic course for this reason. 31 For example, Dr. Steven Kite, a fluvial geomorphologist at West Virginia University, said that one of his first exposures to Rosgen’s work was when, “I had a graduate student who got an MS with a project that was in fluvial geomorphology, and applied for a job and was told he wasn’t a geomorphologist because he wasn’t Rosgen trained.” 32 Every Rosgen opponent I interviewed addressed this issue. For example, Karin Boyd, a consultant in Montana with 20 years of experience, including six years working for Stan Schumm, one of the most respected figures in 20th century geomorphology, put it: “I definitely see RFPs [that require Rosgen training] and because I have not had Rosgen training [I don’t bother applying for them]. You know it’s kind of funny, those of us who’ve been doing this for some period of time, … Stan Schumm is a good example: he isn’t going to go do Rosgen training, and so he can’t go after those projects.” Scott Gillilan, a consultant with 19 years of experience, has simply abandoned public sector work because there are so many agencies that will, “write into the RFP that you’re going to do it this [Rosgen’s] way. I’ve stopped trying to fight that battle and I don’t respond to those RFPs anymore. I got burned out trying to educate them, and for the last five years, I’ve mostly been working for private clients.”

132

approaches and experienced practitioners. Other critics argue that it stifles innovation. As

Martin Doyle, a geomorphologist at the University of North Carolina, said,

[A]s soon as you prescribe a way that things have to be done you have all these people coming out of school that can’t get creative. I mean, river restoration is nothing more than creativity. Here’s the end that we want; we gotta figure out a way to do it…. [The prevalence of the Rosgen Method has] basically stagnated the innovation of river restoration for at least five years.33

There is complete consensus among Rosgen’s opponents and even moderates in the

debate: the privileging of one method and one type of training over all others cannot stand.

While agency-based decisions cannot be attributed directly to Rosgen, critics argue that he

should use his authority to argue against such limitations. Rosgen’s response to this criticism

is not conciliatory. When I pointed out that agencies requiring NCD training bar even some

practitioners he respects from restoration work, Rosgen had this to say:

That was a major surprise to me [when agencies began requiring my work]. My answer to that is that we have to listen to the message rather than shooting the messenger. The … reason that they’re saying that is that consultants in the past responded to RFPs with more of the traditional [hydraulic engineering] work. And the states and agencies are tired of traditional work. The message is: we wanna go to more NCD. What are universities teaching? Do you see universal teaching of NCD? No. They’re doing courses, but they don’t do implementation. … Here’s my recommendation [for qualified practitioners excluded because they have not taken my courses]: go back and say here’s my qualifications towards NCD…. Just because you’ve got a Ph.D. doesn’t mean you’re qualified. All you got taught was traditional standard methods, … [and] what they’re saying is they don’t want to see anymore of the traditional way we’ve been dealing with the rivers. I don’t tell people I’m sorry. It’s up to you to figure out how to give them what they want.34

Following Rosgen’s advice turns out to be a lot harder than he acknowledges,

however. Bioengineering consulting firms such as Inter-Fluve and Milone & MacBroom

and sole practitioners such as Steve Gough, Scott Gillilan, and Karin Boyd report great

33 Author interview, Dr. Martin Doyle, University of North Carolina – Chapel Hill, 6/15/06. 34 Author interview, Dr. David Rosgen, Wildland Hydrology, 4/22/07.

133

difficulties convincing regulators who treat Applied River Morphology as gospel to accept

alternative approaches. Many report that they have simply stopped trying, and no longer

respond to RFPs that require an NCD approach (see footnote 32). Thus I find Rosgen’s

response insufficient. Although it is clear that he did not start this trend, it is within his

power to soften it by encouraging agencies to be more open to non-NCD approaches that

meet NCD goals.

D. Critiques of Rosgen’s Scientific Practice

Rosgen’s critics in academia and the research branches of federal agencies care about

restoration practice, but restoration science is their life’s work, and many are outraged by what

they see as Rosgen’s departures from accepted scientific practice. Some Rosgen critics begin

at the most fundamental level, questioning his lack of academic credentials. They point out

that he has little formal training in the subject he teaches, and that his recently-granted Ph.D.

required no course work, which is not unusual for British Ph.D.s. “Why does he have to

teach in this workshop format if he’s so legit?” remarked a USFS employee with academic

training in geomorphology who was forced to attend a Rosgen course by her supervisor. Or

as Dan Levish put it, “What I don’t understand is without any… real training or background

or anything else, how does he get written into the regulations?”35

35 Only the most vehement critics put it this bluntly, but even moderate critics sometimes point to Rosgen’s lack of formal credentials obliquely, or even unconsciously. Dr. Gary Parker, for example, who seems to be the most Rosgen-supportive of the process research stars, made the following comment: “I hope you have a chance to talk with Eric [Larsen] because he did learn as an apprentice to Rosgen,... [and] then went on to get a very solid background in engineering and river geomorphology.” The implication being that an apprenticeship with Rosgen would not in and of itself provide such a background.

134

Perhaps the most important set of critiques of Rosgen’s approach as scientific

practice, however, focuses on the ways in which he does not comply with the norms of peer-

review: he often does not indicate the number of data points or the geographic locations

from which the data was collected on his graphs, he does not publish in peer-reviewed

journals, and he does not make the data sets on which his knowledge claims and proprietary

models are based available for review. To Rosgen’s critics, this is not simply a clash of

cultures, or a failure to pay respect to sacred cows. In their view, the way that science

progresses is through communication and critique. By failing to provide access to the data

supporting his work and to participate in the peer-review system, Rosgen opts out of the

project of scientific progress in the eyes of his critics. As consultant Steve Gough describes

it:

This method is kind of a product. It’s not something that came together with broad cooperation and consent from the scientific community. It was put together mostly by just Dave, and I’m still waiting for all this data that it’s based on… I remember one of the things that really struck me scientifically at the course was the particle size information. He had conglomerated it together, he’ll say this came from however many streams, but there’s no N anywhere. There’s not an N anywhere in that book that I know of.… Don’t tell me that this is based on a large amount of empirical data but you’re not going to tell me what the N is or where the rivers are! … [When] you’re saying this is scientifically-based and spending large amounts of public money on it, but it’s this [proprietary] product that was developed so that even though it’s scientifically-based we can’t see the data on which it’s based. I mean, you see the philosophical holes there?36

To their credit, some university- and agency-based scientists are trying to

communicate this cultural rift to him. One of Rosgen’s most vehement critics, Andrew

Simon of the National Sedimentation Laboratory, reports saying this to him:

You need to get yourself out there and not just in conference proceedings. You need to write yourself up in the journals. Not so that you can put something after your name, but because this is how science moves, this is

36 Author interview, Steve Gough, Little River Hydrology, 6/19/06.

135

how it works. This is how you get to look at my work and critique and review it. And this is how I get to look at your work and critique and review it. That’s how we learn and move forward.37

Taking this further, Rollin Hotchkiss, a hydraulic engineer on the faculty at Brigham Young

University and the current editor of the Journal of Hydraulic Engineering, has been actively

working with Rosgen to help him publish his work.

Unlike the sections above, the criticisms in this section were not interleaved with my

evaluations of their fairness. Although Rosgen follows some academic forms, he is

fundamentally a consultant operating under the constraints of the market, not a scientist paid

to produce peer-reviewed articles. Thus I think that all of these criticisms are reasonably

accurate; the question is how much Rosgen’s refusal to follow scientific norms matters. It

clearly slows evaluation of his work considerably, and discomfits many of his critics, but if

his design approach and models are proven to work consistently this departure from

conventional scientific practice may not be particularly important.

E. Critiques of the Scientific Basis of Rosgen’s Approach

The final broad grouping of critiques focus on what critics judge to be errors in

Rosgen’s science, places where even without access to the data supporting his claims they are

seen as clearly wrong in light of the current consensus in the field of geomorphology. The

most common and substantive of these critiques are that Rosgen’s approach:

infers processes and/or predicts channel evolution from form,

relies on models for calculating sediment transport that cannot be correct,

37 Author interview, Dr. Andrew Simon, USDA NRCS National Sedimentation Laboratory, 12/13/06.

136

focuses too narrowly on the reach to be restored,

overemphasizes channel stability,

is too interventionist, and

relies too heavily on bankful discharge.

Without broad follow-up studies of Rosgen-style projects, it is not possible to definitively

evaluate these claims. Instead, I have focused on two tasks: clarifying each of the issues

listed above, and assessing their applicability to what Rosgen actually teaches.

1. Inferring Process from Form

Perhaps the most sweeping critique of the scientific basis of Rosgen’s work attacks

the core premise that it is possible to diagnose processes or predict channel response from

an analysis of form. Critics argue that what a channel reach looks like now does not tell you

why it looks that way, and it certainly does not tell you where the channel is going next. As

Kris Vyverberg, a geologist with the California Department of Fish and Game, put it,

People want to tell me it’s a ‘B’ channel and then move on. I want people to describe it, tell me what’s going on around the bend, why is it that way?... There is just a failure to realize that streams are more complex that can be accounted for by “C3.”38

A first step in evaluating this critique is to reflect it back at Rosgen’s critics. When

asked about their channel evaluation and design processes, every Rosgen critic I interviewed

included key measurements of form. Long-profiles, cross-sections, and slope were universal

objects of study, and without exception Rosgen’s critics thought there was something to be

learned from them about the processes at work in a reach. As Andrew Simon, a passionate

critic of Rosgen’s work, wrote in a textbook on fluvial geomorphology methods:

38 Author interview, Kris Vyverberg, California Department of Fish and Game, 6/3/04.

137

Channel form, or morphology, has long been recognized as a diagnostic tool in evaluating fluvial landforms…. The key to using channel form in the analysis of fluvial landforms must be based on either (1) measurements of parameters that aid in quantifying channel processes such as flow hydraulics, sediment transport, and bank stability or (2) observations of diagnostic characteristics that provide information on active channel processes. (Simon and Castro 2003, my emphasis)

Clearly, Rosgen’s critics consider channel form an important source of information

about process. The issue thus becomes whether Rosgen’s focus on form is as exclusive as

his critics assert. Again, I think the picture is more complicated. As demonstrated above in

the Simon and Castro quote, the two key processes critics typically point to are flows of

water and sediment. Contrary to critics’ assertions, the NCD approach does include

measures of hydraulics and sediment transport and capacity, some of which overlap with his

critics’. In fact, the entire 9-day Level III course is devoted to study of these processes.39

Overlapping methods for measuring sediment transport include Wolman pebble counts,

which are critical to determining channel classification and also ubiquitous among Rosgen’s

critics (though disagreements about where the active bed ends and the bank begins mean

that their sampling methodologies vary). And while perhaps not as commonly used by

Rosgen’s critics as by his supporters, scour chains and bank pins are accepted methods of

measuring short term erosion in both camps.40

The claim that the NCD approach simply infers process from form is clearly

incorrect; Rosgen, like his critics, emphasizes both process and form. Instead, the crucial

question is whether the particular methods for assessing flows of water and sediment

39 I heard from many interview subjects that Rosgen did not pay sufficient attention to sediment transport. This claim is surprising given that in addition to the Level III short course Rosgen puts a great deal of emphasis on sediment supply, capacity and competence, providing both detailed instruction about methods for measuring sediment transport, and models for developing sediment rating curves based on empirical data. I do not know if this was ever a fair characterization of Rosgen’s level of attention to sediment transport; certainly it is not a plausible claim about his current teaching and practice. 40 Perhaps the only way in which one could mistake a Rosgen lecture for one by Peter Wilcock is in their shared insistence on the need for good empirical sediment transport data!

138

incorporated into NCD are sufficient, and this cannot be assessed without a broad study of

the outcomes of Rosgen-style projects.

2. Sediment Transport

A second major critique of the scientific basis of the NCD approach focuses on the

transport models FLOWSED and POWERSED, recently developed in collaboration with

former USFS hydrologist Charles Troendle, which purportedly enable their users to develop

sediment rating curves based on a single sediment measurement and data regarding bankful

discharge. These models have become the basis of EPA’s new regulatory regime for

sediment TMDLs, giving them an influence beyond even Rosgen’s restoration work, as

discussed in more detail in Chapter 6.

There is broad agreement among the sediment transport researchers I interviewed

that it is impossible for a single equation to describe the vast diversity of sediment transport

regimes. For example, Peter Wilcock’s initial reaction to the FLOWSED and POWERSED

models was disbelief:

[Rosgen and Troendle] basically took some data with which I happened to have been familiar and set a sediment rating curve, a relationship between transport and discharge, and essentially claimed that the exponent was always the same, no matter what kind of stream, no matter where…. [I attended a presentation by Rosgen where he] had a bunch of overheads with data from other places, mostly from Idaho. Just by chance … he was showing all these sites that I actually knew the data and had memorized the exponents and knew they were different by a factor of four.41

Given how improbable sediment transport researchers find Rosgen’s claims, one

would expect a barrage of criticism from the ranks of basic science similar to the barrage

from applied scientists that followed Rosgen’s first publication of his classification system.

41 Author interview, Dr. Peter Wilcock, Johns Hopkins University, 7/24/06.

139

However, the FLOWSED and POWERSED models are proprietary, and Rosgen and

Troendle have refused to make the data points or technical details available for peer review.

Thus to their frustration, Wilcock and others have no solid basis for evaluating the models.

Wilcock described multiple attempts to obtain access to the data, including reviewing

Rosgen’s proceedings papers, asking him questions at conferences, corresponding with

Rosgen and Troendle, reviewing data on the EPA website, and even attempting to meet with

Rosgen in person. All were unsuccessful. Thus not only is it impossible to evaluate the on-

the-ground impacts of Rosgen’s sediment transport work, it is also impossible to rigorously

evaluate their scientific credibility.

3. Narrow Focus on Project Reach

A third critique of the scientific basis of NCD is that it focuses on the current form

of a particular reach, ignoring both the basin-wide forces that set the flows of water and

sediment that shape that reach, and the ways that those forces have changed over time. As

Dave Montgomery, a professor of geomorphology at the University of Washington put it:

The issue with Rosgen training is that you never get around to asking the right questions. What happened to this river? How did it change? How was it disturbed historically?... The more I’ve studied rivers, the more I realize that while the physics that underlie river processes are really important and generalizable, the manifestations of those processes in a particular river system are context and history dependent. If you never get to the point of asking about context and history, you have no business restoring rivers.42

The importance of such attention to context is intuitively obvious: however

convenient it may be to focus solely on the project reach, it is impossible to accurately assess

present conditions and predict channel response without considering how the channel got to

be the way it is and what forces are currently acting upon it. One of the many people who

42 Author interview, Dr. David Montgomery, University of Washington, 6/28/06.

140

espouse this position, however, is Rosgen himself. Rosgen’s colleague and fellow Luna

Leopold student Ann Riley described the first two steps of appropriate restoration design

practice as follows:

To recap, A) as Luna directed, the first thing we do is look at historic records. Rosgen talks about that as part of a Level I assessment. Some professors have complained that the [Rosgen] system doesn’t take into account historic conditions, but in his book and teaching he’ll tell you the first thing to do is get aerial photos and maps. That’s a no-brainer for anybody. B) We try to get an assessment of the watershed, of what the reasons are for the instabilities in this watershed, because unless we identify them we can “restore” the streams until the millennium and not fix anything. Again, Rosgen will tell students to do that, and again he is widely criticized in academic circles for not doing that.43

Riley’s claims are born out by Rosgen’s teaching and writing. In the first lecture of

the Level I course I attended, for example, Rosgen argued that the first step in restoration is

understanding the cause of change:

Changes in flow regime: peak, duration, magnitude, and timing. Changes in sediment regime: load and particle size. You need to understand the disturbance regime, and whether it’s natural or anthropogenic. And you need to know the difference. You also need to know the history. The cause of the problem may have happened 20 years ago or 40 years ago. Assessment is crucial because it tells us about change over time.

Rosgen recommends use of time series aerial photograph analysis in a number of

places in his 1996 textbook (see for example pp.6-5 and 8-31). In the recently released

NRCS restoration design handbook (2007), Rosgen begins his section on Watershed and

River Assessment with the following paragraph:

Land use history is a critical part of watershed assessment to understand the nature and extent of potential impacts to the water resources. Past erosional/depositional processes related to changes in vegetative cover, direct disturbance, flow and sediment regime changes provide insight into the direction and detail for assessment procedures required for restoration. Time-series aerial photos are of particular value to understand the nature, direction, magnitude and rate of change. This is very helpful as it assists in

43 Author interview, Dr. Ann Riley, California State Water Quality Control Board, 6/19/06.

141

assessing both short-term, as well as long-term river problems. (NRCS 2007, p.11-20)

Carrying out such an analysis of change over time is an enormous undertaking, but Rosgen

clearly acknowledges the importance of trying. Thus I would argue that this critique is not

applicable to the intentions of Rosgen’s approach, and even in practice is fully applicable only to

very limited applications of his work, such as the Uvas Creek project described in Kondolf et

al. 2001.

4. Insistence on Stability

A key scientific sticking point for many of Rosgen’s critics is his claims about

channel stability. This concern has two parts: a) objections to Rosgen’s foundational

assumption that dynamic equilibrium is a common condition and an achievable restoration

goal, and b) a widespread perception that both Rosgen and his students go beyond dynamic

equilibrium to favor actual honest-to-God “stability stability;” channels that are not going to

move no matter what flows hit them.

The first part of this objection – that the NCD approach assumes the possibility of

dynamic equilibrium – is an accurate description of Rosgen’s position. In his lectures and

writings, Rosgen continually puts forward stability as both the required characteristic of

reference reaches and the desired outcome of restoration, and his definition of stability

clearly corresponds to an idea of dynamic equilibrium: “the ability of a river, over time, in

the present climate to transport the flows and sediment produced by its watershed in such a

manner that the stream maintains its dimension, pattern and profile without either aggrading

or degrading.” (Rosgen 1994, 1996a, 2007)

This flies in the face of current scientific consensus. As Doug Shields of the ARS

National Sedimentation Laboratory argues, since the 1950s and 60s when the research on

142

which Rosgen’s work is based was published, geomorphology has moved towards, “the idea

that fluvial systems are inherently unstable… [T]he idea of dynamic equilibrium is useful, but

like an ideal gas it probably doesn’t exist.”44 It also contradicts the field experience of

consultants such as James MacBroom, who argues that the NCD approach is,

based too much on the concept of the equilibrium channel, which is something we don’t see very often. Most of our watersheds are very dynamic. We live in an urban area so our watersheds are rapidly changing due to hydromodification of discharge, impervious cover, and storm drains, and also they have variable sediment loads.45

If Shields, MacBroom and their colleagues are correct that dynamic equilibrium is the

exception rather than the rule, then Rosgen-style design projects should fail as soon as they

experience a flow of a magnitude sufficient to trigger adjustment of the reengineered

channel. But despite the handful of widely publicized project failures cited by the Rosgen

opposition, anecdotal evidence from Rosgen’s students suggests that when implemented by

experienced practitioners most Rosgen-style projects stay in place, as evidenced by the Buck

Engineering warranty program described in the introduction to this chapter.

I think it’s likely that the stability of these projects could be attributed to the second

part of Rosgen opponents’ concern: although Rosgen defines stability as dynamic

equilibrium, he and his students favor channels that stay where they were put. This seems to

me a fair characterization of Rosgen and many of his students. As Kris Vyverberg from the

California Department of Fish and Game put it:

There’s a basic misunderstanding in the idea that here’s where you put your meanders, here’s where you put your bank revetments. Meanders meander, but you’re developing static control measures. People forget that the channel is dynamic…. Rosgen folks are talking about stability stability: they’re saying it’s not going to change. For example, I worked on a culvert removal in Ukiah. It was a C3/C4 channel that had been hugely degraded by the culvert.

44 Author interview, Dr. F. Douglas Shields, ARS National Sedimentation Laboratory, 6/13/06. 45 Author interview, James MacBroom, Principal, Malone & MacBroom, 7/24/06.

143

They had a whole meadow to work with, so why fix the meanders in place? I can’t speak to Dave’s perspective on stability, but as the Rosgen Method is applied in the field people seem to believe stability means the channel won’t move.46

A preference for “stability stability” could explain why Rosgen’s recommended

restoration structures bear more resemblance to those of traditional hydraulic engineering

than they do to stricter bio-engineering approaches: deformation is a no-no. Even more

than the insistence on dynamic equilibrium, the emphasis on big ‘S’ stability is in direct

contradiction to the current scientific consensus that emphasizes rivers as dynamic systems.

These days, channel migration is a hoped for natural process, not a nuisance. As Craig

Fischenich from the USACE Waterways Experiment Station described it,

We want streams and rivers that behave, that don’t erode or change over time, because we want them to be predictable, we want to preserve conditions that we think are either advantageous for us or good for the environment or whatever. But the reality is if you have an opportunity to go work in some of the more pristine areas of the world that haven’t had the kind of anthropogenic impacts that we have on most of our river systems here, these wildly dynamic systems are constantly changing and evolving and it’s reasonable to speculate that the organisms that live in those systems have evolved to take advantage of it that change. And in fact there’s plenty of good scientific evidence that those kinds of dynamics are really important.47

Thus Rosgen’s critics seem to me to be accurately describing the importance of stability in

NCD: the practice of hardening the channel to create stable channels backs up the

foundational concept of dynamic equilibrium.

5. Overly Interventionist

Viewing river systems as inherently dynamic leads the Rosgen opposition directly

into objections to NCD’s interventionist emphasis, yet another classic science vs. practice

46 Author interview, Kris Vyverberg, California Department of Fish and Game, 6/3/04. 47 Author interview, Dr. Craig Fischenich, USACE, 6/27/06.

144

argument. Many members of the Rosgen opposition believe that we would be better off

leaving the majority of damaged channels alone to let them heal themselves. As consultant

Steve Gough describes it,

To me it’s much better to leave a stream alone completely than to go in and manipulate it. And this [NCD] system is extremely biased towards getting out the yellow machines and rearranging the creek. And very very seldom in my experience is that what needs to be done. I tell so many of my clients to leave things alone.... I see the Rosgen Method sort of being like, it’s not a question of whether we’re going to rearrange the stream or not, it’s like, what size of meanders are we going to put in when we get the bulldozers in there.48

Gary Parker, a moderate in the debate, expressed this is less partisan terms: “My own

view is that what I’ve seen tends to be in many cases too interventionist, and there has not

been enough emphasis on how to encourage rivers to fix themselves.”49 Or as Dan Levish

put it, “Our primary recommendation in many cases was just set back the levees, widen the

bridges, and leave it alone and you’ll get it back eventually at lower cost, no maintenance,

and not much monitoring. In anthropomorphic terms, the stream knows what to do.”50 A

primary justification for the “leave it alone” philosophy has been the high level of

uncertainty that Rosgen’s critics’ attribute to restoration (discussed in Chapter 4).

Whether or not to intervene is an issue which splits NCD critics by discipline,

however, since biologists confronted with crashing fish populations often argue that waiting

is not feasible. While Rosgen actually agrees with his geomorphologist critics that it is in

many cases best to leave streams to heal themselves, he argues with the biologists that non-

action is often a luxury he does not have, though he more often points to human

infrastructure impacts than to fisheries:

48 Author interview, Steve Gough, Little River Consulting, 6/19/06. 49 Author interview, Dr. Gary Parker, University of Illinois, 6/26/06. 50 Author interview, Dr. Daniel Levish, Federal Bureau of Reclamation, 6/28/06.

145

I’ve got some guy standing on the bank with his barn about to fall in the river saying, ‘hey, what can you do to help me?’ I’m gonna have to make a decision. It may be the wrong decision, but you have to do something. You don’t have the luxury of saying, ‘let me study this for 10 years.’ You’ve gotta make some calls.51

Whether the guy on the bank, not to mention the river flowing by him and the organisms it

hosts, would have been better off if he’d just rebuilt his barn a good distance from the

channel is unanswerable without a broad study of the outcomes of NCD projects.

6. Over-reliance on Bankful Discharge

Perhaps the most common charge raised by critics is that the NCD approach is too

dependent on the concept of bankful discharge. Many of Rosgen’s critics see the centrality

of the bankful concept as an irreparable fatal flaw at the heart of NCD.

Bankful discharge has been variously defined over the years,52 but its original source

is Wolman and Miller’s (1960) classic insight that the effective discharge controlling channel

form is not the rare big flood, but the smaller more regular flow that fills the active channel

without spilling out onto the floodplain – the bankful flow. The relation between the

frequency of flows and the magnitude of sediment they transport demonstrated in the

Wolman and Miller paper is one of the most influential analyses in 20th century

geomorphology, and has thus drawn a number of follow-up studies over the last fifty years.

There is widespread agreement that finding physical indicators of bankful discharge in the

field can be very difficult (Williams 1978, Knighton 1984 and Federal Interagency Stream

Restoration Working Group 1998, pp.7-10 – 7.11). Beyond that, the current scientific

consensus is that it is inappropriate to equate bankful discharge – the maximum flow a channel

can sustain without spilling out onto the floodplain – with effective discharge – the flow that

51 Author interview, Dr. David Rosgen, Wildland Hydrology, 4/22/06. 52 See FISRWG 1998, p.7-11 for a concise and useful overview of different definitions of bankful discharge.

146

determines channel form – because the relationship does not hold in arid and semi-arid

environments (Wolman and Gerson 1978), in systems that have no floodplains, in incised

channels, or in channels that are currently adjusting to upstream influences (Doyle et al.

2007).

Critics are certainly correct that the bankful discharge concept is the keystone of

Rosgen’s approach. (In)famously, the graduation gifts at the end of Rosgen short courses

used to include a t-shirt reading: “if you don’t know bankful, you don’t know shit!” And

teaching students to find physical indicators of bankful discharge in the field was one of the

primary foci of both the Level I and II courses I attended. It is critical to note, however,

that Rosgen’s definition of bankful discharge is not tied directly to physical features of the channel itself, as in

the original Wolman and Miller paper and most of the work that followed it. Instead,

following Dunne and Leopold (1978) Rosgen defines the bankful discharge as the effective

discharge:

The bankful stage corresponds to the discharge at which channel maintenance is the most effective, that is the discharge at which moving sediment, forming or removing bars, forming or changing bend and meanders, and generally doing work that results in the average morphologic characteristics of channels. (Rosgen 1996a, p.2-3)

Rosgen then ties this definition of bankful discharge into a specified return interval from 1.2

to 2 years, most typically 1.4 to 1.6 years, arguing that this is the frequency of bankful

discharge. By depending on recurrence intervals and, most importantly, regional curves,

Rosgen can derive the bankful discharges on which his approach depends even in systems

where physical indicators are misleading or absent.53 Critics’ arrows thus fly wide of the

53 I learned this the hard way. The fourth day of the Level II course I attended was spent surveying a stream reach outside of Fayetteville, AK. I was on my group’s long-profile team, so I spend most of a day working my way downstream looking for, among other things, the slope breaks that provide physical indicators of bankful flow. It was not simply up to me as rod-holder to find these breaks; the rest of the long-profile team and anyone else who was handy would come confer. When we did a quick sketch over lunch, we were pleased to

147

mark when they critique the NCD approach based on conventional definitions of bankful

flow.

F. Conclusion

Together, these criticisms of the NCD approach make a formidable and substantive

list. To those outside the stream restoration world and to critics, it is startling that a list of

such gravity has not dispelled the perception of Rosgen as scientifically legitimate, even

without the backing of data from a broad and comprehensive evaluation of completed

projects. Yet for Rosgen’s supporters these critiques do not hold much water, and so the

stream restoration community has been mired in more than a decade of unproductive

debate: recent critiques of the NCD approach include many of the issues already flagged in

publications from the mid-1990s.

Why haven’t these substantive critiques generated more traction? One factor is

critical ignorance. None of Rosgen’s major critics have attended his short courses or

carefully reviewed his published materials, so some of their critiques are based on

misapprehensions of what he actually teaches or practices. Notable examples of this include

misunderstanding how Rosgen defines bankful discharge, the keystone of his approach, and

failing to realize that Rosgen does emphasize analysis of hydraulic and sediment transport

see a very consistent profile, and confident we’d succeeded in determining the bankful flow. But when we worked up our data that evening back at the hotel, we discovered that what we thought was bankful did not correspond to what the regional curves said it should be. And in the NCD approach, the regional curves trump field data on bankful. So to our shock and frustration, we were instructed to disregard our field data because physical indicators may be misleading.

148

processes. Thus what seem to critics to be fatal blows appear to Rosgen’s supporters as

irrelevant.

Other critiques seem to be ineffective because they reflect the disconnect between

the preoccupations of the scientific community and those of NCD practitioners. For the

latter, issues of peer review and correspondence with current scientific consensus hold far

less weight then the practical examples of projects that stay in place and prevent the ravages

of traditional hydraulic engineering solutions to bank stabilization and erosion issues. For

scientists, quackery is the enemy to be fought off at all costs; for Rosgen’s supporters the

ultimate enemy is channelization.

I would argue, however, that the primary reason why the substantive issues central to

the Rosgen Wars have remained unresolved for so long is that fights over substance are the

form of the debate, not its primary content. If scientific evidence alone could explain or

settle the Rosgen Wars, peace would have broken out long ago. At its deepest level, the

fight over the NCD approach is over who is viewed as scientifically legitimate, and who

controls the emerging field of stream restoration: Rosgen or his opponents, private or

public sector science. As pointed to by Bourdieu in the epigraph of this chapter, the

conflicts endemic to science are inherently both substantive and political. In the following

chapter, I turn to the ways in which these substantive arguments are mobilized in political

practice.

149

Chapter 6. The Mobilization of Expertise in the Rosgen Wars

“In reading many recent proceedings papers, it is apparent to me that the recent authors are not at all familiar with my methods, many have little to no experience in river restoration projects, have not attended any of my courses, and are not the least familiar with what the “Rosgen method” is all about!”

Dr. David Rosgen, Wildland Hydrology

In the previous chapter, I described the substantive arguments over Rosgen’s work.

In this chapter I extend the analysis from the realm of ideas to the realm of practice in two

overlapping ways. I describe how the content of the debate is produced in, and shaped by,

the particular theaters of the Rosgen Wars, such as conferences, journal articles, and short

courses. At the same time I analyze the claims of expertise mobilized within these theaters.

The grounds on which participants in the Rosgen Wars base their assertions of scientific

expertise and authority are diverse, and most often defined in implicit opposition to those

they are fighting: the ability to produce practical tools vs. the capacity to generate theoretical

understanding; the possession of scientific credentials vs. the mastery of hands-on

experience; the ability to control vs. the acknowledgment of complexity; the shelter of the

scientific community vs. the exposed and lonely stance of the maverick. These are all claims

to authority, and thus types of capital, mobilized in the Rosgen Wars.

What do I mean by capital in this non-monetary sense? According to Bourdieu,

capital should be seen not as a fixed object that keep its character regardless of the field in

which it is mobilized, but instead as an active social power relation specific to a particular field

(Bourdieu 1996a, p.264). The species of capital that are valued in a given field both enable

those who hold them to succeed, and define the character of the field itself. In An Invitation

to Reflexive Sociology, Bourdieu states that:

150

a species of capital is what is efficacious in a given field, both as a weapon and as a stake of struggle, that which allows its possessors to wield a power, an influence, and thus to exist, in the field under consideration, instead of being considered a negligible quantity. In empirical work, it is one and the same thing to determine what the field is, where its limits lie, etc., and to determine what species of capital are active within it. (Bourdieu and Wacquant 1992, pp.98-99)

The Rosgen opposition’s basic tactic has been to insist that the types of capital that

matter in stream restoration should be the same as in any other scientific field: credentials

from prestigious institutions, compliance with the norms of scientific practice, ability to

abstract theoretical principles from empirical data (as in the development of models), and

correspondence with the current scientific consensus. Rosgen, his work, and his supporters

do not possess these types of capital, according to opponents, and thus should not be

considered authoritative or even legitimate parts of the stream restoration field. To make

this point, Rosgen’s critics mobilize some of the institutions responsible for consecrating

scientific authority, circulating scientific products, and inculcating scientific habitus: peer-

review journal articles, papers at scientific/technical conferences, short courses, national

design guidelines, and university curricula.

Rosgen’s supporters, by contrast, insist that the key types of capital defining success

in the stream restoration field should be hands-on experience, ability to provide practical

solutions to human problems such as flooding and bank erosion, accessibility to a broad

audience, and correspondence with nature (as opposed to traditional hydraulic engineering).

Rosgen and his supporters try to win support for the types of capital they hold in an

overlapping but not identical set of arenas, primarily self-published textbooks,

technical/practical conferences, short courses, and national design guidelines. In addition to

the advantage of reaching Rosgen’s core constituencies of consultants, community groups

151

and agency staff, these arenas have the advantage of not requiring passage through peer-

review gatekeeping from the academic community.

The discussion that follows analyzes attempts by Rosgen’s opponents and supporters

to win support for the types of capital they hold, and how the specific arenas in which these

attempts are made shapes the ways in which claims to scientific expertise and authority can

be mobilized.

A. Engage! Opposition Attempts to Mobilize Their Capital

The previous chapter laid out the substantive issues that Rosgen’s critics emphasize.

Though these critiques sometimes arise from ignorance of what Rosgen actually teaches,

they are most notably characterized by the difficulty of resolving them: the sheer complexity

of stream systems make the empirical questions at stake in the Rosgen Wars very difficult to

settle.

Rosgen’s opponents thus have to put a great deal of weight on the forms of capital

they hold to back up their arguments: credentials from prestigious institutions, ability to

abstract from empirical data, compliance with the norms of scientific practice, and

correspondence with the current scientific consensus. The substantive arguments discussed

in Chapter 5 are underlain by the more fundamental claim that because Rosgen does not

possess these kinds of science-generated capital, he and his work should not be included

within the legitimate boundaries of the stream restoration field. It is striking that even

casual critics seem comfortable exiling Rosgen to the far side of that boundary line.

Making the excommunication stick, however, requires the consecrated to effectively

broadcast the news to the laity, a difficult task in the stream restoration world. There is no

single national conference that all types of people involved in the stream restoration world

152

routinely attend. Nor is there a single journal that serves as the clearinghouse for papers on

stream restoration, or even a handful of journals that one could count on to reach the

diverse restoration community. Dave Montgomery, professor of geomorphology at

University of Washington, described the problem like this:

say someone in my position decided they wanted to engage in a big debate about Rosgen: where would you go, how would you engage? Would you send Dave a nasty letter? That’s not my style. Comment on one of his papers? Others are doing that. Most of the people that really need to hear what you could cast as a debate are not equipped to adjudicate it. They’re wanting knowledge. There’s no outlet for communicating directly with them except through a short course. So it’s not clear what the choices are for framing that debate. And if there’s one thing academics don’t have it’s spare time, so this is a significant entry barrier.1

In response to these complications, members of the Rosgen opposition are using a

variety of arenas to try to make their case. As Andrew Simon, a prominent Rosgen critic,

describes it, he and his colleagues:

decided we need to keep publishing in journals, but a lot of these people [Rosgen’s supporters] don’t read journals, they just go home with their handbooks. So we also have to teach short courses at national technical meetings, and teach classes at federal agencies to reeducate, or re-indoctrinate them. And we have to engage all the time at these meetings, and continually pump papers in. The problem is that we have to educate not just the people on the ground but their bosses and their bosses’ bosses.2

To the three arenas to which Simon points – peer-review journal articles, presentations at

technical conferences, and alternative short courses – I would add two: attempts to establish

national guidelines for restoration practice and to develop national level restoration curricula.

The entrance requirements for each of these arenas and the forms of communication they

enable are quite different; the types of arguments that can be made (and thus the species of

capital that can be mobilized) in the loosely-structured format of a five-day short course are

1 Author interview, Dr. David Montgomery, University of Washington, 6.28.06. 2 Author interview, Dr. Andrew Simon, USDA-ARS National Sedimentation Laboratory, 12.13.06.

153

quite different from those in the tight confines of a 20-page peer-review article, as are the

species of capital required to participate in them in the first place.

By way of illustration, I have chosen five examples: Smith and Prestegaard’s now

iconic journal article on the failed attempt to restore Deep Run in Maryland, Simon et al’s

2005 conference presentation attacking Rosgen’s classification system and design approach,

the treatment of NCD at an alternative short course taught by some of his critics, Shields et

al.’s efforts to develop restoration design standards for engineers, and the effort to develop a

national curriculum for stream restoration spearheaded by Peter Wilcock under the auspices

of the National Center for Earth-Surface Dynamics.

1. Preaching to the choir: Smith and Prestegaard 2005

As an arena, academic journal articles require a high level of scientific capital for

entrance. To get past the peer-review gatekeepers, authors typically need to demonstrate that

they hold particular types of capital valued in the scientific field writ large – possession of

prestigious credentials and compliance with the norms of scientific practice– as well as

mastery of the analytical skills and the currently accepted knowledge base particular to a

given field. Journal articles allow a high level detail and the careful working out of

arguments, but typically require a neutral tone and a viable dataset, encouraging the conduct

of case studies or particular experiments rather than the polemical expression of broad

critiques.

In 1995 - 1996, as research for his Masters thesis in geomorphology, Sean Smith

monitored a channel reconfiguration project on a stream in Maryland called Deep Run. The

project was justified and designed by a private firm based loosely on NCD. Five months

after construction the project was hit by a 10-year flood and failed spectacularly, giving

154

Smith the data to ground an unusually rich and well-analyzed case study. His exacting

collection and analysis of hydraulic data, and the article’s eventual publication with Smith’s

advisor (Karen Prestegaard, a senior and well-respected fluvial geomorphologist) in a

prestigious peer-reviewed journal has made the Deep Run case study one of the iconic

examples of the failure of NCD projects and a powerful weapon for asserting the superiority

of scientific practice and capital in the face of quackery. This despite the fact that the

project’s designer did not utilize the NCD approach, but instead based his design off the

generic description of a C4 channel.3

Smith and Prestegaard wield their considerable scientific capital with a precision that

gives the Deep Run case study both notable strengths and weaknesses as a weapon in the

Rosgen Wars. First, following the norm of scientific neutrality almost to a fault, the authors’

tone and content are so studiously dispassionate as to obscure the broad implications of

their argument. A casual reader (or one without advanced training in hydraulics) could be

forgiven for missing the fact that the article critiques the NCD approach. Rosgen is not

mentioned in the title or abstract, nor does the introduction inform us that the designers

relied on his work. It isn’t until the second page that the authors mention that the designers

used Rosgen’s classification system, and we don’t learn that they attempted to use the NCD

approach until the fourth page. For the most part, the article simply refers to “a

morphology-based” method, and most critiques are phrased as, “the field would benefit

from additional consideration of X,” whether that X be hydraulic resistance or land use

history, rather than, “it is a problem that NCD does not consider X.” The courtesy of this

approach greatly limits its rhetorical power.

3 For much of the period during which I conducted this research, I was surprised that Rosgen and his supporters never pointed this out or attempted to discredit Smith’s case study. Rosgen eventually got around to making this point in a recent conference presentation (2006a); to my knowledge it is the first and only time the issue has been raised.

155

Even more than the neutral language, the scientific credibility of, and thus the capital

generated by, the Deep Run case study stems from the data and analysis on channel

adjustment and feature-scale hydraulic resistance that make up the core of the paper. Firmly

couched in the concepts and equations of academic fluvial geomorphology, this section of

the paper is inaccessible to Rosgen opponents’ target audience. The data and analysis

presented, however, exemplify the norms of scientific practice that Rosgen is accused of

flouting, right down to the estimated ranges of error for measurements. The Deep Run case

study is thus an excellent tool for convincing those with advanced degrees in geomorphology

and hydraulic engineering that their deep suspicions about Rosgen’s claims are in fact

justified, but has little utility in luring Rosgen’s supporters back into the fold.

The neutral language and hydraulic analysis of the Deep Run case study, the

qualifications of its authors, and the prestige of the venue make it a substantial source of

scientific capital, but also set significant boundaries around its power to persuade by making

the details of the overall argument inaccessible to Rosgen’s supporters. As a weapon in the

Rosgen Wars, the case study’s value is effectively three words, four numbers, and a set of

parentheses: (Smith and Prestegaard 2005).

2. Blunt force instrument: Simon et al., ASCE Conference 2005

Conference presentations tend to have looser entrance requirements than peer-

review journals: particularly at technical (as opposed to academic) meetings, the scientific

capital produced by association with prestigious institutions is not necessary for

participation, so merely appearing as a conference presenter is not enough to ensure

credibility in the way that publishing in a peer-review journal does. Further, proceedings

papers do not require (and may actually inhibit) the kind of analytical capital on display in the

156

Deep Run case study. Conference presentations are typically less than half the length of

journal articles, so authors are not afforded the luxury of either detailed proof or subtlety of

argument. The Rosgen opposition is thus largely prevented from mobilizing another of its

primary sources of capital.

The good news for opponents is that both of these factors ensure that presentation

papers are usually far more accessible to their target audience than journal articles, a fine

feature given that at at least some of these conferences the target audience is actually present.

Better still, the types of arguments that can be made, and the partisan language wielded, in a

conference presentation are far less strictly regulated than in a peer-review journal: while

time constraints mean that only a small number of points can be made, there is far more

scope for expression of opinion and rhetorical gusto. Encouraging the departure from the

decorum of academic journal articles are those conference organizers who consider the

Rosgen Wars as something akin to gladiatorial battles in the Roman Colliseum: necessary

spectacle to keep the public entertained. As senior restoration scientist and practitioner

Craig Fischenich put it,

CF: I view a lot of it [the Rosgen debate] as being positive, but primarily for entertainment.

RL: The Rosgen debate as spectator sport?

CF: Well it is! I’ve put on two international conferences that I’ve chaired, and in both instances I set it up as one of the sideshows, like a circus. Watch everybody fight over whether Dave walks on water or has horns coming out of his head. Cause it’s either one or the other.4

Egged on by organizers, conference papers offer considerable scope for rhetorical

provocation.

4 Author interview, Dr. Craig Fischenich, USACE Waterways Experiment Station, 6/27/06.

157

Rosgen’s critics have been relying heavily on conference presentations to get their

message across since the late 1990s. One of the most recent and multiply-authored was a

paper entitled, “How Well Do the Rosgen Classification and Associated ‘Natural Channel

Design’ Methods Integrate and Quantify Fluvial Processes and Channel Response?”

presented at the 2005 World Environmental & Water Resources Institute of the American

Society of Engineers conference in Anchorage, Alaska. The ten authors include senior

federal agency scientists, academics and consultants, among them Andrew Simon, Doug

Shields and Matt Kondolf, perhaps the most visible Rosgen critics.

Rather than the neutral tone and careful data analysis of the Deep Run case study,

the Simon et al. conference paper is a frontal attack on the NCD approach. The bulk of the

paper is a whirlwind tour through a dozen common critiques of Rosgen’s work,5 and the

criticism is direct and at times biting. The last sentence of the introductory paragraph, for

example, reads:

[T]he para-professional training provided by some involved in ‘natural channel design’ empower[s] individuals and groups with limited backgrounds in stream and watershed sciences to engineer wholesale re-patterning of stream reaches using 50-year old technology that was never intended for engineering design. (Simon et al. 2005, p.1)

Throughout the paper the authors refer to NCD as “natural channel design,” maintaining

the quotation marks throughout the text as a way of constantly underlining their contestation

of that label.

5 Clients require use of Rosgen’s approach without regard to its relevance for the project at hand; well-trained geomorphologists and hydraulic engineers are considered unqualified; Rosgen’s short courses provide inadequate training; the classification system and design approach are based on 50 year-old work, and thus are not well-grounded in current peer-reviewed science; the difficulties in identifying bankful discharge make accurate classification difficult; Rosgen’s pebble count methodology mixes disparate populations and is thus misleading; the classification system and design approach are not universally-applicable as claimed; the breakpoints between categories in the classification system are artificial constructions; Rosgen’s channel evolution models are not useful for predicting or quantifying channel response; Rosgen’s work is too focused on bankful discharge, a concept of questionable utility; the data the Rosgen Method requires is only sufficient for qualitative analysis, and thus collection of it is a waste of taxpayer money; and Rosgen’s design approach is too focused on the reach at hand, ignoring needed data on historical and drainage basin context.

158

Simon et al. also mobilize their capital as part of the legitimate scientific

establishment. The last paragraph of the paper, for example, offers shelter for practitioners

within the established canon of science that Rosgen, as a lone gun, cannot match:

The foremost advantage of the process-based approach is that it is well established in the scientific and engineering literature. For decades, geomorphologists and hydraulic engineers have been quantifying river processes and developing models that have been tested and refined over time. Developing a design using this rich literature leverages off of a substantial scientific background, and thus provides a critical foundation from which to defend the design approach. Such literature and historical precedence is lacking for the classification approach. Practitioners concerned with professional liability and with the future of their professions would do well to provide design services based on peer-reviewed professional standards. (Ibid., p.10)

This combined play on the authors’ authority and their audience’s fear of liability is potent

stuff.

3. Re-education Camp: Fluvial Geomorphology Short Course

The capital required to teach a restoration short course is somewhat more restricted

than that required to present at a technical conference: you have to convince an audience of

potential students that you have not just 20 minutes, but five days of material to which it

would be worth their time to listen. Short course students typically enroll in order to learn

how to do stream restoration, so while the scientific capital held in such abundance by the

Rosgen opposition is not a hindrance in jumping into the short course market, it is not an

enormous help either. It is notable that the teaching rosters of short courses presented as

alternatives to Rosgen’s typically include both academics and consultants.6

6 For example, I have referred repeatedly to the two-course alternative series organized by a loose group of academics including Matt Kondolf from UC Berkeley, Peter Wilcock from Johns Hopkins, Margaret Palmer from University of Maryland, and Jack Schmidt from Utah State. Despite the fact that these courses have been spearheaded by academics, all three of the level 1 courses and the single level 2 course offered include consultants among the instructors.

159

Once these entrance requirements have been satisfied, however, the captive audience

at short courses allows the broadest canvas for mobilizing scientific capital while at the same

time presenting the fewest limitations on allowable rhetoric. The typical five-day plus length

of a restoration short course allows instructors to leverage their analytical skills, prestige as

part of the scientific establishment, and the shelter of the consecrated scientific

establishment. They have time to carefully explain and promote their position with no peer-

review filter to prevent sometimes scathing critiques of work with which they disagree. Thus

although they reach only a tiny fraction of Rosgen’s critics’ target audience (enrollment is

typically capped between 25 and 40 people), short courses are one of their most effective

means for discrediting the NCD approach and asserting the centrality of process-based

approaches.

Matt Kondolf was one of the first academic geomorphologists to identify Rosgen as

a threat worth addressing. After years of conference presentations and journal articles were

not visibly stemming the NCD tide, Kondolf started in 1995 to offer an annual short course

on fluvial geomorphology as an alternative to Rosgen’s.7 I attended the course in Fall 2004

as a teaching assistant.

For the first two days of the course, NCD was notable mostly for its absence; neither

Kondolf nor the other primary instructor, Peter Wilcock, mentioned Rosgen, nor did the

students bring up questions about his work. His only appearance was as an object of

humor. He was the punch-line of one of the guest lecturers’ jokes, who dismissed the

classification system by showing a slide of the selection panel of a vending machine as a

parody of Rosgen’s alpha-numeric format. And the second evening’s activity was to watch

the classic USFS video, “A Guide for Field Identification of Bankfull Stage in the Western

7 Personal communication, Dr. G. Matthias Kondolf, 4/30/08.

160

United States,” which stars Rosgen and his long-time colleague and collaborator Lee Silvey,

followed by a short recut entitled “Bankfull B-Boys” poking fun at the parts of the video to

which Rosgen opponents object most.8

While clearly having their intended impact of starting to make Rosgen’s work appear

laughable, the videos also opened the door for students to ask questions about the

relationship between the course material and the NCD approach. Wilcock and Kondolf had

very different responses to these questions, as demonstrated in the exchange below, with

Kondolf dismissing NCD as unworthy of discussion, and Wilcock attempting to refute its

core principles. The student was a senior fisheries biologist who had been through some

NCD training:

Student: I’ve been struggling with bankful stuff for years, and I have attended Rosgen’s classes. What are the strengths and weaknesses of his method? What should ordinary people do?

MK: On the bankful thing, clearly you haven’t been fasting enough and well-enough behaved to get the insights.

PW: This is not a good start. The essential conceptual model is that floodplains are built by lateral accretion rather than vertical accretion. Because if it’s aggrading, how do you know when to stop the stopwatch and say this is bankful? It’s a circular reasoning issue: people start looking for bankful at about the right height. Don’t confuse bankful with Rosgen. He didn’t invent the term and is just using it. As a way of comparing big and little rivers, bankful is essential; you have to use the frequency of flows, not actual discharges. But the final and biggest problem is that if your stream is truly disturbed, how do you find the correct bankful? You get people with little training and they’re basically going out there and divining. You need to understand the history. I don’t want to be overly critical of our geomorphological forefathers or any particular person doing restoration.

Student: I wasn’t asking for criticism of the Rosgen Method, I was asking for analysis of its strengths. Some of us on the lower end of the chain get pretty confused.

8 This video was not prepared by the course organizers and had not been reviewed by them before it was presented to students.

161

PW: It’s a template problem. The main problem with applying that kind of methodology is that classification is a great thing for certain purposes, if we use it to organize our knowledge. But classification is not prediction.

At this point, the student gave up and dropped the question.

Treatment of NCD continued in this piecemeal fashion until the evening of the

fourth day, when Kondolf presented a public lecture that the short course students attended.

The lecture was Kondolf at his most engaging, using humor and striking visual images to

make his argument about the dangers of NCD’s form-based approach, the superiority of

process-based approaches used in California on the Carmel and Tuolumne Rivers, and the

human psychology that engenders a preference for single-thread meandering streams.

Kondolf narrowed his arguments against Rosgen’s classification system and design approach

down to a few key points: 1) Rosgen’s followers are ‘fixing’ systems that aren’t broken; 2) it

is critical to consider larger scales historically and spatially; 3) regional differences resulting

from climate and geology cannot be ignored; and 4) only design approaches that focus on

natural processes rather than idealized forms will work. He then illustrated his points with

case studies showing how lack of attention to natural processes and temporal and spatial

context resulted in dramatic project failures. For each project Kondolf presented design

documents showing suspiciously symmetrical sine curve design channels, and striking before

and after photos showing devastated channels. He ended the negative case study section

with the table below, leaving the audience to wonder how anyone could promote Rosgen’s

classification system or design approach with a straight face (Figure 6.1). The subtext of the

presentation was that the NCD approach is fundamentally not-scientific; not, as Kondolf’s

final slide stated, “real geomorphology.”

162

Slide 22: C4/C3 Meandering Channels in California (Some dates below are appx, now being checked) ------------------------------------------------------------------------- 1990 Wolf Ck, Sierra Nevada Washed out, buried 1990 Cuneo Ck, Coast Ranges Washed out 1993 Mattole Cyn Ck, Coast Ranges Washed out 1994 Greenhorn Ck, Sierra Nevada Washed out 1995 Jamison Ck, Sierra Nevada Washed out 1995 Uvas Ck, Coast Ranges Washed out 1995 Cold Ck, Lake Tahoe Filled then scoured 1997 West Walker R, Sierra Rocked meanders 1999 Bear Ck, Cascades Channel moved to meadow (success), Many constructed riffles (undulating bed cut into clay) washed out 1999 Ackerman Ck, Coast Ranges Washed out -----------------------------------------------------------------------

Figure 6.1: slide from Kondolf lecture

This stirringly anti-NCD presentation was followed up by a morning lecture on the

fifth and final day entitled, “Geomorphic Channel Classification: tool or crutch?” Here

Kondolf took a more scientific, neutral tone, addressing not just Rosgen’s but all stream

classification systems, although the focus consistently returned to Rosgen. Kondolf

expanded the strengths beyond the traditional communication tool exemption to include the

ability to enable rapid inventories, to provide a framework for detailed observations, to

define project goals, and to provide an initial basis for selecting restoration strategies. But he

then went into far greater detail about why classification systems in general are problematic,

focusing on three key points. First, classification systems can start to serve as substitute for

scientific inquiry, as practitioners start to believe that they understand a system because they

know its classification. Second, classifications are at best only partial explanations because

they focus on particular features. Whatever partial or mis-understandings you begin with are

perpetuated rather than questioned. Third, the ability to predict channel behavior from

classification is illusory. This last point was a particular dig at the NCD approach. Overall,

163

Kondolf argued that all stream classification systems are artificial because there are no

natural clusters in geomorphic variables. Practitioners thus run the risk of forgetting that the

underlying variables are a continuum and starting to think that the differences between types

are meaningful.

By the time the structured portion of the course ended with discussion of student

projects on the afternoon of the fifth day, the equation Rosgen = joke was firmly entrenched

in students’ minds. During the second presentation, a consultant showed a design for a

channel reconfiguration project and asked for suggestions. “Why don’t you make a nice

meandering channel?” a student jokingly suggested to general laughter. And the presenters

of the third case study brought the house down with an initial image in which they showed a

design that reconfigured a channel into graceful meanders – that spelled out “Red Sox” in

cursive script (the Red Sox defeated the Yankees in Game Seven to win the 2004 American

League Pennant the third night of the short course).

4. Attempting to establish standards of practice: Shields et al. 2003

Standards of practice and design handbooks are typically collective efforts that go

through painstaking review and consensus-building processes. The major national efforts to

date, such as Stream Corridor Restoration: Principles, Processes, and Practices (Federal Interagency

Stream Restoration Working Group 1998) and the recently released USDA-NRCS Stream

Restoration Design Handbook (NRCS 2007), are the collective work of literally dozens of

authors invited to contribute because of their recognized expertise in a particular area.

Scientific capital is not a necessary entrance requirement, but some form of accepted

authority is. Like journal articles, design guidelines require neutral language, but they also

insist that the language be broadly accessible. The point is to straight-forwardly present the

164

approaches that are already agreed upon as standard practice. There is no room for technical

or rhetorical fireworks, nor is there a place for controversy in such a document. The scope

for expression and argument is largely in the editorial process: the decisions about what

material is anointed as ‘standard practice’, and what is not is the key to the significance of

design standards in the Rosgen Wars.

In 2003 the first attempt to provide design guidelines for channel reconfiguration

projects specifically targeted at engineers was published in the Journal of Hydraulic Engineering

(JHE). The JHE is a prestigious publication, and the authors are an impressive group with

both scientific and practical capital. All five authors hold doctorates, as well as Masters

degrees in engineering.9 This was a publication with clout.

The authors describe the rationale for the guidelines as follows: “Although the

number and scope of stream restoration projects are increasing, designs for these projects

are often weak in hydraulic engineering. This paper represents an attempt to outline

acceptable standards for hydraulic design for channel reconstructions.” (Shields et al. 2003,

p.582) Unlike the Simon et al. conference paper described above (which shares three of the

five authors), the language is studiously neutral. And unlike the Deep Run case study, these

design guidelines are accessible. One does not have to be trained in hydraulics to understand

what the authors are recommending. This is clear as early as the abstract where, instead of

using the opaque euphemism ‘a morphological approach’ the authors use the transparent

euphemism a ‘relatively simple… [design approach] based on stream classification and

regional hydraulic geometry relations’. (Ibid., p.575) While Rosgen and his supporters would

9 Doug Shields, the lead author, and Andrew Simon are both researchers at the National Sedimentation Laboratory, which is part of the Agricultural Research Service, the research branch of the USDA. Ron Copeland is a former senior researcher with the USACE, who at the time of publication had recently moved to the private sector. Peter Klingeman and Martin Doyle are both academics, the former at Oregon State University and the later at the University of North Carolina – Chapel Hill.

165

contest the ‘relatively simple’, the remainder of the description is recognizable enough to tell

them that this message is being delivered to their address.

The majority of the proposed design guidelines do not explicitly challenge the NCD

approach, but putting forward an alternative approach is an implicitly critical project. For

example, the authors argue that an important first step in assessing channel stability is to

determine the appropriate spatial domain, which they parenthetically define as follows:

“(certainly more than the project reach, but economic constraints may prevent inclusion of

the entire watershed)” (Ibid., p.576). Given that the NCD approach is frequently criticized

for its limited spatial scale of analysis, this is a rebuke.

At other times the critique comes closer to the surface, as when the authors argue

that the key flow to consider in designing a new channel is the channel-forming discharge,

the flow that, “given enough time would produce width, depth, and slope equivalent to

those produced by the natural hydrograph,”10 (op cit.), rather than bankful discharge,11 the

lynchpin of NCD. Perhaps the most explicit critique comes in Table 1, which reviews the

strengths and weaknesses of a number of tools for assessing channel stability. One category

is ‘channel classification’, which clearly covers more than Rosgen’s system. But the authors

single Rosgen out specifically, even if not by name, when they write that “Form-based

systems do not provide indication of future response,” and then cite two papers whose

primary focus is critiquing NCD, including the Kondolf et al. Uvas Creek case study (2001),

which they cite six times in the article (ibid., p.577). Thus while the language of the

guidelines article is studiously neutral, the content is not.

10 Hydrograph is a short way of saying the range of flows the channel experiences over a given period of time. 11 Bankful discharge is a concept with many definitions, as was discussed in Chapters 2 and 5. The definition used here by the authors is, “the maximum discharge that the channel can contain without overflow onto the floodplain.”

166

5. Proposed National Curriculum: Wilcock et al. in preparation

Proposals for national stream restoration curricula are similar to efforts to establish

standards of practice in having high entrance requirements, but because curricula are by their

nature academic products the primary qualification is the academic capital of association

with prestigious scientific institutions. Curriculum proposals tend to be far less massive

efforts than standards of practice, because rather than necessitating the compilation of an

encyclopedic range of information about a field, they typically focus on a list of key topics to

cover in coursework. They are similar, however, in that the Rosgen Wars play out for the

most part in decisions about what to include and exclude, rather than explicitly in the text.

The National Center for Earth-Surface Dynamics (NCED) attempt to establish a

national curriculum and start the process of developing a national exam-based certification

for stream restoration is an ongoing effort. It has been spearheaded by Peter Wilcock, a

sediment transport researcher at Johns Hopkins University who co-instructs the alternative

two-part short course series described above and in Chapter 4, and is the NCED principal

investigator in charge of stream restoration. NCED’s proposed curriculum and certification

procedure is a group effort that includes many of the most prominent stream restoration

educators in the country, but not Dave Rosgen or any of his students who teach restoration

short courses.12 The curriculum and certification proposal is being developed in the form of

a paper to be submitted to the Journal of the American Water Resources Association.

The proposed curriculum starts from the premise that in order for stream restoration

to become a ‘mature profession’, there need to be rigorously defined and broadly tested

12 The authors include Peter Wilcock, Johns Hopkins University; Derek Booth, University of Washington; Janine Castro, USFWS; Craig Fischenich, USACE; Karen Gran, NCED; Steve Kite, West Virginia University; Matt Kondolf, UC Berkeley; James MacBroom, Milone and MacBroom; Margaret Palmer, University of Maryland; Mary Powers, UC Berkeley; Jack Schmidt, Utah State University; Doug Shields, ARS National Sediment Laboratory; and Vaughn Voller, University of Minnesota.

167

methods as well as an educational structure that transmits them, “a fully effective symbiosis

between research, training, and practice….” (Wilcock et al. in preparation, p.2) The paper

lays out many of the current educational challenges described in Chapter 4:

there are very few existing university courses that focus directly on restoration, and developing new ones requires resources of time and funding that are in short supply in the university;

short courses, which take up the educational slack, are uneven in quality and

comprehensiveness, rarely incorporate any assessment of learning, and do not provide sufficient breadth or depth; and

educational needs vary widely within the restoration community. (ibid., p.3)

The authors argue that while short courses have a role to play in providing continuing

education, the core of restoration education should come from university degree programs.

Using a combination of university degrees and short courses, the authors

recommend four levels of training:

an overview level sufficient for those who merely supervise restoration projects;

a principles level sufficient for planners and regulators;

a design level sufficient for those who wish to actually design restoration projects;

a specialty courses level intended to fill out specific skills, such as sediment transport

analysis. (ibid., pp.4-5)

The authors then address the issue of recommended curriculum in two ways: 1. a detailed

table that takes all of the topics necessary for stream restoration and lists what should be

taught about that topic at each of the three main educational levels specified above, and 2. a

set of core principles that should underlie all restoration teaching at whatever level. The list

of curriculum topics is notably broad, ranging from watershed hydrology and channel form

through water quality, biogeochemistry and ecological dynamics to more social concerns

such as aesthetics, regulatory programs and construction management.

168

While NCED has deliberately avoided taking a stance in the Rosgen Wars, the

curriculum topics practice critique via exclusion, clearly indicating that the NCD approach

would be no more than a minor implement in the restoration toolbox envisioned by the

authors – a nail-set rather than a hammer. Channel classification and hydraulic geometry

relations are mentioned, but they are subtopics of subtopics, and there is no mention of

reference reach or analogue methods at all. The authors’ core principles for restoration

education illustrate this implicit critique through inclusion, as they incorporate some of the

most common critiques of the NCD approach. For any Rosgen Wars veteran, statements of

principle such as, “No two streams are alike,” “There is no simple or universally applicable

design solution,” “Problems must be assessed in the context of the entire watershed, and

sometimes multiple watersheds,” and the emphasis throughout on uncertainty are clear

partisan statements.

The authors close with a proposal for a national exam-based stream restoration

certification program with prerequisites based on educational qualifications (at minimum an

undergraduate degree in a relevant field); a minimum level of professional experience; and a

minimum amount of continuing education. Reflecting the authors’ commitment to the

applied character of the field, the exam would have not just a written component, but also a

field component and a portfolio component. They then propose the formation of a steering

committee to create the exam, consisting of, “carefully selected experts from relevant fields

of the stream restoration profession. This would include representatives from national

societies such as ASCE, ASLA, RRNW, AFS, and SRE, as well as regulatory

representatives.”13 (ibid., p.11) The national societies listed are not noted for their support

13 These five organizations include three whose members already have to be professionally licensed: the American Society of Civil Engineers, the American Society of Landscape Architects, and Society for Reliability Engineering. The other two are the American Fisheries Society, whose members include many fisheries

169

of the NCD approach (although ‘regulatory representatives’ could be a different story

depending on the agency). While the authors never come out and say it, it is abundantly

clear that Rosgen’s certification is not what they have in mind for the proposed national

standard.

Thus while the NCED curriculum has not yet been published, it represents a

substantive effort by Rosgen’s opponents to promote their definition of stream restoration

as the core of the field, and the species of capital they hold as the most valuable types of

capital with it. Tellingly, the intended implementers of this curriculum are universities, an

attempt to move the center of disciplinary reproduction back into academia and away from

the private sector. And the premise of the article is that the university should occupy the

central position it holds in other resource sciences not only as the primary trainer of

practitioners, but also as the main developer of new knowledge and applications.

B. Rosgen Strikes Back

While Rosgen is clearly frustrated by the constant barrage of critique, he

demonstrates considerable poise under fire, meeting his critics’ barbs with grace and good

cheer. He could not be described as taking it lying down, however, as he constantly pushes

biologists involved in stream restoration, and one of the primary regional restoration conferences: River Restoration Northwest (RRNW). It is not clear why RRNW is on this list, but given that it is, it is worth noting that the other major regional restoration conference, the bi-annual conference organized by the North Carolina State Stream Restoration Institute (a national center of support for Rosgen’s work), is not included. Again, this demonstrates critique via exclusion.

170

back at his critics. The following sections describe first the species of capital that Rosgen

and his supporters attempt to mobilize, and second the arenas in which they do so.

1. Tactics and Species of Capital

Interestingly, Rosgen and his supporters rarely launch counter-attacks on the

substantive content of their opponents’ claims (c.f. Rosgen 1996b and 2006a, the latter of

which is discussed in the next section). Instead, Rosgen and his supporters base their

arguments on four primary claims: that practical capital is more critical to stream restoration

than scientific capital; that Rosgen possesses more environmental capital than his critics

because he is taking on the real enemy: the USACE; that it is not Rosgen but Rosgen’s critics

who fail to follow the norms of scientific practice; and that Rosgen does indeed hold some

of the types of scientific capital his critics claim he lacks.

The first of these tactics is to assert the importance of practical capital over the

species of scientific capital held by Rosgen’s critics. The typical argument is that because

academics never actually try to implement projects, they don’t have anything useful to offer,

only criticism. John Potyondy, a moderate in the Rosgen Wars and the program manager for

the USFS Stream Systems Technology Center, put it this way:

Academics lament that people don’t come to them for restoration advice. Maybe they understand rivers, but they haven’t been able to communicate that well to people. And many don’t have practical experience. It’s easy to be critical of failed projects, but they can’t say, in most cases, ‘this is what I’ve done that worked better.’14

Rosgen often phrases it in ways less directly critical of academia, but it’s the same message.

During the Level I course I attended, Rosgen said:

14 Author interview, John Potyondy, USFS, 4/29/04.

171

There’s a rule you guys: if you criticize someone else’s restoration project, you better be able to explain what went wrong and give them some advice about how to fix it…. You can’t come along and just be the critic, you’ve got to say what they should have done.

Or as he put it in a paper he presented at the 2006 Omaha ASCE conference: “Surprisingly,

the most vocal critics also have the least experience in conducting river restoration.” (Rosgen

2006b, p.10)

A second tactic Rosgen uses is to assert that he holds more environmental capital

than his opponents because while they waste their time attacking him he is taking on the real

enemy: traditional hydraulic engineering. Given that many of the people in the stream

restoration field were originally drawn to it by outrage over the impact of traditional

hydraulic engineering, this argument has great power.

A third tactic involves the counter-accusation that it is the critics who are not

following the norms of scientific practice, and thus that they do not themselves possess one

of the species of scientific capital they attack Rosgen for lacking. Rosgen and his supporters

repeatedly point out that many, perhaps even most, of his critics don’t know enough about

what he teaches or practices to have a strong basis for critique. In more reflective moments,

many of his critics acknowledge this as well, and it has caused some to step back from the

Rosgen Wars.15 As Rosgen wrote in an email regarding a proposed session on his work at

the 2006 ASCE conference in Omaha:

15 For example, Peter Wilcock explained that he has not spoken publicly about his concerns with Rosgen’s sediment transport work for EPA (discussed in the next section) because he has not been able to find out enough about it to confirm or deny his concerns: “I have been careful not to publicly criticize his design methods because some of his more vocal defenders complain, I think correctly, that there are lots of people who complain about Dave’s methods without actually having learned what they are. Most of us who are professionals in this learn what other people’s methods are by reading about them in the scientific literature, not by taking courses on them. And so that information is not available [to us].” Other critics admit ignorance and barrel ahead anyway. For example, Dave Montgomery said during our interview when I asked him about particular strengths and weaknesses of Rosgen’s work, “The biggest problem in the restoration world is that there appears to be, not that I’m speaking with great authority never having taken one of his classes, it appears that if you can convince people that they only need to know a little bit to do restoration then you’ve done great harm.”

172

In reading many recent proceedings papers, it is apparent to me that the recent authors are not at all familiar with my methods, many have little to no experience in river restoration projects, have not attended any of my courses, and are not the least familiar with what the “Rosgen method” is all about! I have not had any E Mails, letters or phone inquiries from these authors seeking data, clarification or explanations on how stream classification is actually used in natural channel design. Too many written documents are filled with half-truths, miss-information, incorrect assumptions, supposition, suspicion, and personal bias. (Rosgen 2005)

Because it is so easy to demonstrate critics’ ignorance (as discussed in the previous chapter

with regards to claims about bankful discharge and form vs. process, for example) and

therefore the hypocrisy of their attacks on Rosgen for not following the norms of scientific

practice, this is a powerful set of arguments for Rosgen.

The last commonly-used tactic is to assert that Rosgen does hold scientific capital.

This is done in a number of ways. From the very beginning of his consulting career Rosgen

has drawn heavily on his association with prestigious scientific figures, first and foremost

Luna Leopold. Leopold is widely viewed as the father of stream restoration and remains

after his death a revered figure throughout the restoration community. The association

between Rosgen and Leopold was remarked on even in the 1992 NRC report (NRC 1992),

described in Chapter 2, and was proudly noted by almost every Rosgen supporter I

interviewed. There seems to be general consensus among participants in the Rosgen Wars

that without Leopold Rosgen would not be where he is today.

More recently, Rosgen and his supporters have made much of his association with

Dr. Richard Hey. Hey inspires far more ambivalence than Leopold, but is still widely

regarded as one of the foremost researchers on sediment transport, and thus is able to

burnish Rosgen’s legitimacy to some extent. As Karin Boyd, a consultant in Montana put it:

[O]ne of the biggest complaints about his courses… is the lack of sediment transport knowledge. And what he did was he brought Dick Hey from Britain to help with the sediment transport component of his class. And

173

Richard Hey is an incredibly well-respected sediment transport guy. So I was kind of impressed by that. I thought, “Well, good for him! He’s responded to that criticism.”16

In addition, Rosgen and his supporters point to his impressive field experience

(something few if any of his critics can match) and to his unmatched empirical dataset on

stream morphology. Kansas State University professor Dr. Timothy Keane said that:

When I decided to go this route, I looked at what was out there and what was available in terms of basis of teaching, basis of replicable research, and this is the technique and method that I was sold on just because it comes from so many years of field observation and data collection. And I haven’t changed my mind. He reads streams like nobody I’ve ever seen.17

A last way in which Rosgen and his supporters insist that he does indeed hold

scientific capital is by invoking the language and trappings of science. Rosgen consistently

talks about the importance of “peer-reviewing” designs and emphasizes the importance of

the data and “having the numbers.” Most notably, despite his long-term habit of poking fun

at the ivory tower and its denizens, he chose to pursue a Ph.D. from the University of East

Anglia under Richard Hey’s supervision, and since he received his doctorate in 2003 he has

not been shy about using his newly-earned title “Dr. Dave Rosgen.” I asked him about this

during one of our conversations:

RL: What motivated you to go back and get a Ph.D.? What did you feel like it got you at this point, given how successful you are?

DR: A lot of the controversy that I heard was from academics and some researchers, that Rosgen doesn’t have a Ph.D., so anything that he develops can’t be valid. Or he’s not one of us, he’s not in our camp, so it can’t be true. So I thought well, maybe it would give some validity to the method if people saw that I had the titles that a lot of people feel are necessary to develop, ah, methods or procedures or write a book or something. You have to have some academic excellence and a certain title to justify your position, to justify what you say. So I thought it was to gain a little more professional credibility. Second, it was to help a lot of people I’ve trained, because, and

16 Author interview, Karin Boyd, Applied Geomorphology, Inc., 7/18/06. 17 Author interview, Dr. Timothy Keane, Kansas State University, 7/15/06.

174

this is probably some of my own doing, people see me in the past as this cowboy from Colorado who’s teaching geomorphology. And it’s like, “What?!?” (laughs). There’s this natural reaction. So the people that I’ve trained, which are well over 13,000, 40 at a time, when they want to implement this then their boss says, “Well, does he have a Ph.D.? Then we’re not going to accept this just because you went to this class and this is what he said to do. If he doesn’t have a Ph.D. how could this be true?” So the other reason I got it is to help a lot of people who’ve been through the training by giving some credibility to the method. Whether that’s valid or not, it just seems to be more how things go.

Clearly, Rosgen is trying to assert his own claims to the scientific capital generated by formal

academic qualifications.

2. Arenas of Mobilization

Where do Rosgen and his supporters mobilize these arguments? Not in the peer-

review literature. Very few of Rosgen’s supporters have sufficient scientific capital and

knowledge to a) conduct the types of research accepted in peer review journals, b) analyze

and present their data in a typical academic format, and c) drum up the financial resources to

support research, analysis and writing. This presumes that they wish to publish in the peer

review literature in the first place; publications are only currency in the world of the

university and the federal research agencies. Rosgen and his supporters typically have

concerns they consider more pressing. Even if Rosgen’s supporters decided to engage in

academic journals, they would face peer-review boards that are typically quite hostile to an

NCD approach. The only person who manages to get pro-Rosgen articles published in top-

ranked journals consistently is Richard Hey, whose stock of academic capital has not yet

been depleted by his colleagues’ disdain for his association with Rosgen (most recently Hey

2006 and Bhuiyan et al. 2007). Thus Rosgen and his supporters have focused their efforts to

persuade their critics not on the peer-review academic literature, but on the range of arenas

175

that are more open to them: self-published text books, conference proceedings papers, and

short courses.

For illustrative purposes, I have chosen Rosgen’s latest textbook, Watershed Assessment

of River Stability and Sediment Supply; a presentation paper from the 2006 ASCE EWRI

conference in Omaha, NE; and the lectures in the Level I short course I attended in January

2005 in Santa Cruz, CA. Each demonstrates the ways in which the types of capital described

in the section above are mobilized in practice.

a. The Next Battle: Rosgen’s WARSSS

Obviously, the entrance requirements for a self-published text book are not so high.

Despite this, Rosgen’s latest textbook gains a great deal of legitimacy because it describes a

set of methods for assessing channel stability and sediment supply that were commissioned

by the EPA. Watershed Assessment of River Stability and Sediment Supply (WARSS) lays out

EPA’s recommended approach for developing sediment TMDL’s nationwide. This work

explodes Rosgen’s reach from just those streams considered for bank stabilization or

reconfiguration to all flowing bodies of water in the U.S. It also expands the base of

Rosgen’s critics into the realm of sediment transport researchers, who consider sediment

models their province of expertise and have been frustrated by their lack of access to the

data or technical specifications on which the WARSSS method is based.

The WARSSS textbook is in many ways typical of Rosgen’s written work: rather than

taking on or even acknowledging his critics, he accentuates the positive: his particular kind of

practical capital. The introduction, for example, mentions Rosgen’s decades of practical

restoration experience. Rosgen also asserts his scientific capital in the introduction by

mentioning his connection to Luna Leopold and his decades of fieldwork, and by describing

176

himself as a teacher of scientists: the intended audience for the WARSSS method is consistently

described as “scientists” by both Rosgen and the EPA. For example, the EPA placed this

explanation for why WARSSS was developed on its website:

The U.S. Environmental Protection Agency supported the development of WARSSS by Dr. David L. Rosgen because there is limited guidance on assessing sediment impairments. WARSSS can be used to analyze known or suspected sediment problems, develop sediment remediation and management components of watershed plans, develop sediment TMDLs (Total Maximum Daily Loads), and other uses. This web-based assessment tool was designed for scientists who need to assess sediment-impaired waters in planning for their restoration. (emphasis added, http://www.epa.gov/warsss/about.htm)

Perhaps the most notable way in which Rosgen’s WARSSS textbook boosts his scientific

legitimacy is simply by existing. By commissioning Rosgen to develop WARSSS, adopting it

as a preferred method, and promoting it heavily, the EPA has effectively asserted Rosgen’s

scientific legitimacy, increasing his stock of scientific capital and bolstering his claims to

expertise.

2. A Rare Rebuttal: Rosgen 2006

As discussed above, conference papers have fairly low entrance requirements.

Rather than the scientific capital required to publish a peer-review article, presentation

abstracts merely have to look plausible and follow the basic conventions for the conference

in question; nor do they require the kinds of analytical prowess necessary for peer review

papers. Thus conferences are an easy arena for Rosgen and his supporters to access, and the

relatively loose structure of conference papers mean that they are good places to assert the

big points (and particular claims to capital) even if one does not have the exactingly collected

data or technical analysis to back it up.

177

At the 2006 ASCE EWRI conference in Omaha, NE, conference organizers once

again set up Rosgen Wars sessions, but this time with a twist: under the guidance of

moderator (and Rosgen Wars moderate) Janine Castro of USFWS, Dave Rosgen and

Andrew Simon presented a series of papers intended as a conversation. Each man presented

twice, attempting to make their case to the large crowd in attendance. Neither Rosgen nor

Simon came away satisfied from the session, but in preparation for it Rosgen made one of

his rare attempts to systematically rebut his critics’ objections.

The primary tactic in Rosgen’s paper is to accuse his critics of failing to comply with

the norms of scientific practice they claim to champion because they critique his work

without knowing anything about it. He thus denies one of their claims to scientific capital:

However, as none of the afore-mentioned authors [who include Simon, Kondolf, Juracek and Fitzpatrick] have attended the formal training courses offered or contacted the author, their lack of familiarity with the method may result in misinformed conclusions. In science, one must test a hypothesis based on a robust program of data collection and analysis. (ibid., p.2)

The introduction alone contains multiple illustrations of critical ignorance, rebutting critiques

of:

the seriousness of the short course series (Rosgen describes it as involving, “over 520 hours of formal instruction, homework and field application.”18 (op cit.));

the use of stream classifications as the basis for predicting stability (they’re not; that analysis doesn’t begin until Level III);

the lack of consideration of historical or land use context (also featured in Level III, but referenced throughout Rosgen’s approach).

The main body of the paper continues this argument that critics are guilty of

scientific hypocrisy in arguing against an approach they have not bothered to study. For

example, Rosgen points out that in Juracek and Fitzpatrick 2003, an article often cited by 18 It is not clear to me how Rosgen arrived at this number. In the other paper he presented at the 2006 Omaha conference, he described the short course training as lasting a minimum of 400 hours (Rosgen 2006b, p.3). Given that the four short courses cover 29 days, and that in my experience the workload averages out somewhere in the ballpark of 12 or 13 hours per day, Rosgen’s second estimate sounds more realistic to me. However, it is important to note that I have not attended either the Level III or IV courses.

178

critics as a definitive debunking of the utility of Rosgen’s classification system, the authors

did not correctly classify the river they describe:

Many authors, including Juracek and Fitzpatrick (2003), have correctly stated that if one cannot identify bankful stage in the field on an unstable (actively incising) channel, one could misclassify the reach. This is precisely why USGS gauges must be used to develop regional curves for bankful discharge versus drainage area for given hydro-physiographic provinces. The streams in Wisconsin evaluated using the RSC [Rosgen stream classification] were incised, making it difficult to observe bankful (Juracek and Fitzpatrick 2003). The authors classified the streams as B types, though B stream types are not incised. Since the authors did not properly apply the continuum of morphological variables to adjust the delineative criteria (Rosgen 1996[a]), the streams were misclassified.19 (Rosgen 2006a, pp3-4)

A second tactic Rosgen uses in this conference paper is to assert that his critics lack

practical capital because they focus on criticizing other people’s work instead of developing

the practical tools that the stream restoration field needs. In relation to the critique that his

approach is too dependent on analysis of form, Rosgen argues that a) his approach uses both

process- and form-based analysis and design techniques, and b) part of why he continues to

rely on form-based techniques is because critics have yet to produce a process-based model

that can determine appropriate channel morphology:

To accomplish the NCD methodology, one must use both form and process-based approaches. For example, to the author’s best knowledge, no available analytical or process-based models predict the depth and slope for runs and glides, transverse bar features, point bar slope, and other features of riffle/pool stream types such as a C4. To design and construct such features, dimensionless ratios and morphological relations of these bed features of similar stream types are used. This is a form-based calculation using analog methods from reference reach data by stream type; however, the final design is checked for hydraulic and sedimentological response using analytical approaches. Using such a form-based calculation is appropriate, as the author is not aware of any analytical model or other option for design and construction of such features. Unfortunately, critics do not offer alternate

19 In the Level II course I attended, we worked in an incised stream and Rosgen instructed us to proceed exactly as he describes here: find the bankful discharge based on USGS gage data, not physical indicators in the channel. I note also that the slopes reported in Juracek and Fitzpatrick (2003) were substantially lower than the range of slopes for B channels under Rosgen’s classification system, another reason to suspect that the channels were misclassified.

179

design strategies that would improve the current state of the science of river restoration implementation. (Rosgen 2006a, pp.4-5)20

3. Level I Lectures, 2005

As noted above in Section A, a greater degree of acknowledged expertise (and thus

more capital) seems to be required to teach a restoration short course than to present at a

technical conference because you have to convince potential students that you have at least

five days of material to teach them. Because most students attend restoration short courses

in order to learn how to practice stream restoration, the practical capital Rosgen holds in

such abundance is a major advantage over his critics. He owns this market. Thus it is

unsurprising that Rosgen does not devote much time in his short courses to discussion of

his critics. In fact, many Level I students do not seem to realize that there is anything

20 A last notable feature of this paper is that it is the only place I know of where Rosgen addresses the two iconic studies of failed projects to which his critics constantly refer: the Kondolf et al. 2001 Uvas Creek case study, and the Smith and Prestegaard Deep Run case study (2005), described above. Rosgen has long argued in conversation that the Uvas Creek failure cannot be laid at his doorstep since the designer who completed it had never taken one of his courses. Here, he finally puts that argument in print, and explicitly contrasts the Uvas Creek and Deep Run studies.

The “rest of the story” [about Uvas Creek] that has been formally documented, but not reported, is that the river restoration designer was a landscape architect with no training or experience, and who chose to disregard the advice of trained peer-reviewers [Rosgen and his student Steve Zembsch, who did the initial analysis and proposed design for the project.]. The channel constructed has a width/depth ratio twice the value of the stable form of a C4 stream type, a flat gradient point bar and a uniform grade (no differential between riffles/pools). No sediment competence or transport capacity was calculated. This design violated the fundamental tenets of the NCD procedure, yet its failure was blamed on stream classification…. The author encourages reviewers to obtain and present all the facts in order to make in-depth technical suggestions for improvement rather than incorrectly assessing blame to a procedure that was not followed.

The Deep Run rehabilitation project in Maryland, as reported by Smith and Prestegaard (2005), also proved unstable. Generic classification values were used for the design morphology instead of values from stable reference reach data in the Piedmont/Coastal Plain transition province. This valuable and thorough review points out the importance of using the correct reference reach data and checking the design based on hydraulic resistance for various flow levels. It also illustrates the importance of maintaining floodplain vegetation for added flow resistance during flood events. (Rosgen 2006a, p.6)

The carefully neutral language of the Smith and Prestegaard article allows Rosgen to treat what is effectively the same condition – the failure of a design that did not utilize Rosgen’s approach being blamed on his approach – in a much more receptive way.

180

controversial about Rosgen’s work at all, much less a debate that has pre-occupied the river

restoration community for more than a decade. Instead of attempting to rebut objections to

his work, Rosgen focuses in his short courses on asserting his expertise by promoting

practical capital over scientific capital, and at the same time burnishing his scientific capital.

For example, in response to a question about how to improve the problem-solving

process, Rosgen pointed to the importance of his decades of field experience, getting the

numbers, and peer-review, all of which assert his scientific capital:

This is my 40th year. When I look at things I can just see them. The more you do the numbers, the more data you collect and analyze, the easier it will be. You start to be able to see it. Familiarity makes it easier to come up with solutions. Part of the problem-solving process is to think about how does this occur in nature, then do the calculations to verify. There’s a 40-step procedure of analysis, much more complex than doing a trapezoidal channel, to do Natural Channel Design…. It requires you to understand hydraulic, sedimentory, geomorphic, and biological processes. One of those 40 steps has 30 substeps. It’s not simple!

There’s always going to be people who will be your critics. What’s important is that you better have the numbers, and have done your homework, and gotten peer review of your design. Once you’ve got the experience go ahead, but don’t forget peer review …. When the critics, if they ever dare, come and talk to you you can show them the numbers. Take them out in the field and show them what worked and what didn’t.

Rosgen also promotes his environmental capital in taking on traditional hydraulic

engineering, and even goes so far as to identify his critics as old school engineers, rather than

the actual cross-cutting coalition of academics, engineers, and alternative NCD practitioners

that make up the Rosgen opposition. For example, Rosgen implied that his methods were

excluded from textbooks because they were not traditional engineering methods (rather than

the actual situation, which is that academia, the most common source of textbook authors, is

a hotbed of anti-NCD sentiment):

The geomorphic solution, the Natural Channel Design method, you’re not going to find in a model, you’re not going to find in a lot of textbooks…. We

181

have to get away from our traditional river engineering approach, which is still in the textbooks. Right now, the better option is not in the book.

C. Conclusion

In practice, the Rosgen Wars have been fought in a range of written and spoken

venues from papers in the peer-review literature to lectures in short courses. As I have

described above, the particularities of these arenas shape both the kinds of capital that can

be asserted and the way those assertions are made. Rosgen’s opponents have defended their

claims to scientific expertise by asserting four traditional kinds of scientific capital:

association with prestigious institutions, analytical and theoretical capacity, conformity with

the norms of scientific practice, and congruence with the current scientific consensus.

Rosgen has asserted his expertise via claims to practical, environmental, and scientific capital.

He has also, however, had his expertise asserted by the state: EPA is promoting methods

that they commissioned Rosgen, not scientific sediment transport researchers, to develop,

and the NRCS’ long-awaited design manual (NRCS 2007) ratifies in print what has long been

the case in practice: Rosgen is that agency’s method of choice. Clearly, the state has played a

key role in bolstering Rosgen’s claims to expert status.

The Rosgen Wars have been a central preoccupation of the stream restoration field

for more than a decade without either side managing to convince the other of the error of its

ways. What the substantive questions and dueling assertions of expertise beg, however, is

the question of why the fight over Rosgen’s work exists at all. Given that Rosgen’s

opponents hold far more scientific capital, why haven’t they simply ignored him, or

successfully dismissed him by consigning his work to a footnote on outmoded classification

182

systems? Tackling the same question from the opposite angle, why has Rosgen been so

phenomenally successful? I answer these questions in Chapter 7.

183

Chapter 7: The Rosgen Wars and Bourdieu’s Agonistic Fields

“I feel [about the Rosgen Wars] a little bit like we feel down here in the South about the Civil War: deep down in our heart of hearts we have to admit we lost that one; even though our cause was noble, the conflict was a terrible waste.”

An agency research scientist

Despite attempts to leverage their scientific capital and legitimacy, Rosgen’s

opponents have been startlingly ineffective at slowing the spread of his work; far from

nipping his success in the bud, they haven’t even been able to prune him to a standstill. The

consecrated defenders of science seem at this point to be losing the Rosgen Wars. Scholars

have put a great deal of time and ink into demonstrating the erosion of scientific authority in

Western culture (e.g., Beck 1992), but that authority is still quite powerful when it comes to

delineating the bounds of science itself, as dissident scientists such as Ignacio Chapela have

learned to their cost (Gieryn 1999, Delborne 2008). Why then have Rosgen’s opponents,

who include almost all of the most respected academic and agency researchers on stream

restoration, been unable to use their societally-granted monopoly on the definition and

production of science to exclude one consultant in a cowboy hat, no matter how

charismatic?

There are many reasons put forward by Rosgen’s opponents and supporters to

explain the success of his design approach and the intensity of the controversy that

surrounds it. Depending on your point of view, the design approach is successful because

Rosgen did a superb job of distilling the principles of geomorphological science and making

them accessible; because he emphasizes doability instead of a more realistic uncertainty;

because he is a highly charismatic individual and a skilled salesman; and/or because he

184

simply had perfect timing in launching his consulting business just as the stream restoration

market came into existence. Similarly, depending on the side you take in the Rosgen Wars,

the passion of the opposition is fueled by a potent combination of ignorance of Rosgen’s

approach and professional jealousy, or by the desire to protect streams from the well-

meaning interventions of the poorly-trained.

As far as I can tell, all of these assertions are to some extent true.

And yet in their focus on the individuals involved – their passions, knowledge, and

ignorance – these explanations ignore the broader social and political-economic forces that

shape the stream restoration field. I will argue below that however powerful Rosgen’s

charisma, it is the relations among regulatory, market, and grassroots demand for stream

restoration that ground Rosgen’s success; individual characteristics cannot explain the wide

and exclusive adoption of the NCD approach, or the ineffectiveness of his critics.

In An Invitation to Reflexive Sociology, Bourdieu cautions researchers against limiting

their focus to individual members of a field, arguing that a,

general property of fields is that they are systems of relations that are independent of the populations which these relations define…. The notion of field reminds us that the true object of social science is not the individual, even though one cannot construct a field if not through individuals. (Bourdieu and Wacquant1992, pp.105-107)

Instead, Bourdieu consistently focuses on the influence of both the market and the

state. The primary structure of a field, in Bourdieu’s conceptualization, is the axis running

from autonomous to heteronomous poles, and the differential between these poles is one of

the primary motors of conflict in his schema. To recap the discussion of autonomy and

heteronomy in Chapter 4, Bourdieu defines the autonomous pole as the part of a field where

production is controlled most thoroughly by the forms of power and prestige (capital)

185

specific to that field, while the heteronomous pole is where production is shaped primarily by

outside forces: political and economic relations.

While Bourdieu seems to argue in his article on “The Specificity of the Scientific Field”

that the autonomous/heteronomous relation is not a defining characteristic of the natural

sciences (Bourdieu 1975, p.36), I believe this is a naïve view of the scientific field in this age

of the neo-liberal university, and one that Bourdieu himself would not espouse if he were

writing on this topic today. The Rosgen Wars clearly demonstrate the importance of state

and market forces; I will argue below that the ability to satisfy the needs of resource and

regulatory agencies and the growing restoration market were both enormous factors

bolstering the success of the NCD approach. Thus the political-economic forces Bourdieu

emphasizes in relation to fields such as religion and culture should be acknowledged for their

power in scientific fields as well.

In the first part of this chapter, I argue that the success of the NCD approach and

the ferocity of the opposition to it stem from the same cause. Stream restoration is an

emerging field with a deep base of grassroots interest (pushing a powerfully contradictory

agenda of control and empowerment of nature, as described in the Introduction), a strong

legislative mandate, and a lucrative market. The public sector, in the form of academia and

the federal research agencies, has not yet been able to provide the framework that the new

field badly needs; Rosgen has. He has been so successful, and his critics so ineffectual,

because his work is fundamental to the flourishing restoration market, while at the same time

serving as an indispensable guide for regulators tasked with implementing our conflicting

demands for restoration projects.

In the second part of this chapter, I reflect back on Bourdieu’s field concept, asking

how the dynamics of the Rosgen Wars can illuminate areas in which the field concept needs

186

additional development. As a conceptual umbrella the field – with its inherent, structurally-

driven struggles and the strong emphasis Bourdieu places on the role of political and

economic power – has proved quite useful for analyzing the controversy over Rosgen’s

work. I will argue, however, that the types of conflict and strategies of combat that

Bourdieu describes are not exhaustive: Rosgen’s challenge is both far less and far more

radical than the strategies Bourdieu lays out. I will also argue that, contrary to Bourdieu’s

conceptualization of fields as bounded entities, the Rosgen Wars demonstrate that the

relations among fields are a crucial factor at stake in conflicts within them. Lastly, I argue that

Bourdieu’s assumption that the only players within a field that matter are those who hold

substantial amounts of the capital specific to that field is unnecessarily limiting. As Rosgen’s

rise to expert status demonstrates, the mass of participants who hold little or no capital can

still be crucial in setting the general direction of a field, and in accepting or rejecting claims

to expert status within it.

A. Explaining Rosgen’s Success

As described in Chapter 2, the stream restoration market was propelled by grassroots

demand and regulatory requirements into rapid expansion long before university- and

agency-based scientists paid much attention to it. While there were a few major figures in

the scientific community that made restoration a major focus of their work, such as Luna

Leopold, in general stream restoration was not considered central to the field of

geomorphology until the mid-90s or later. Martin Doyle, a geomorphology professor at the

University of North Carolina-Chapel Hill described it this way:

187

In the mid-90s the premier geomorphologists in the country were condescending towards river restoration…. There was very much a thing of ignoring Rosgen because they were pure geomorphologists…. So there was not an academic backlash to Rosgen [at that point] because there was more of an academic backlash to the existence of river restoration in the first place.1

As a result, the emerging field lacked the kind of institutional framework that public

sector science typically provides. Rosgen stepped into this breach by developing his own

structure for the stream restoration field: a common language for communication, standards

of practice, and the field’s primary means of disciplinary reproduction. All three of these

structuring elements have helped to build a powerful epistemic community in support of the

NCD approach. The NCD community has supported Rosgen’ss claims of expertise and

helped to fend off attacks from the consecrated institutions of public science.

This success in turn catalyzed opposition to his work. As Bourdieu reminds us, the

internal structure of even a relatively young field is powerfully relational. This

interdependence causes a particular position in the field to change, “whenever there is

change in the universe of options that are simultaneously offered for producers and

consumers to choose from.” (Bourdieu 1983, p.312) Clearly a crucial reason why academic

and agency-based scientists bothered to engage with Rosgen rather than ignoring him is that

he shifted their positions simply by entering the field, calling their authority into question

without doing, or even intending to do, anything to them. As Doyle tells it:

It’s only in the past five years, that [top geomorphologists] ... started hearing from their students who are now consulting that they weren’t being considered qualified to do river work…. I think that’s what got people personally involved: being told that they’re not qualified to do things.2

1 Author interview, Dr. Martin Doyle, University of North Carolina-Chapel Hill, 6/15/06. 2 Op cit.

188

1. The Rosgen Classification System as Communication Tool

As described in Chapter 5, even the staunchest Rosgen critics grant his classification

system’s utility as a communication aid.3 Why is such a communication tool so critical?

Stream restoration requires coordinated efforts among a number of different disciplines,

most notably biologists, ecologists, engineers, hydrologists, and geomorphologists. These

biology- and physics-based disciplines have very different training, which causes confusion

over even the most basic things, such as which bank of the river is left and which right!4

Communication is thus a significant hurdle: if differently-trained experts cannot agree on a

description of the current character of the system they wish to modify, how can they design

its future form? Vaughan Voller, a civil engineering professor at the University of Minnesota

and a principal investigator at NCED, described the problem as follows:

[T]he nomenclature is really critical. You’ve got to be talking about the same thing. You often have long discussion and debates with ecologists, which is an extremely descriptive science, talking to an engineer, which is extremely quantitative. And sometimes it’s very difficult to bridge that gap, because you think you’re talking about the same thing, but you’re not; you’re arguing about something that basically boils down to semantics.5

Rosgen’s alphanumeric classification system fills this gap, creating a shared

terminology that allows practitioners to quickly grasp the key morphological characteristics

of a stream system. As Syd Brown from the California Department of Parks and Recreation,

described it:

Hydrologists speak very well to hydrologists, engineers speak very well to engineers. What I think Dave did was transcend those … very real

3 Although a recent paper in the Journal of the American Water Resources Association may challenge even that (Roper et. al., 2008). 4 Geomorphologists reference right and left off of the flow of water and sediment downstream; fisheries biologists reference off upstream movement of anadromous fish. 5 Author Interview, Dr. Vaughan Voller, University of Minnesota, 9.19.06.

189

vocabulary barriers and provide a way for people to actually have a common understanding and a common language….6

Beyond disciplinary boundaries, the Rosgen classification system is also intended to transcend

geographic boundaries, providing a universally applicable framework for describing streams.

As Tim Keane, a professor at Kansas State University and staunch Rosgen supporter said,

“It allows communication, because people know what you’re talking about when you’re

talking about a C4 no matter where you’re at.”7

The Rosgen classification system is thus widely used, or at least understood, even by

people who would not dream of using the other components of Rosgen’s design approach.

As Dave Montgomery, professor of geomorphology at the University of Washington, pithily

put it: “The great strength is that it’s a convenient shorthand, and at that level it’s brilliant,

and I really mean that. It’s simple, and it’s elegant. In terms of shortcomings, you know it’s

kind of like where do you start?”8 By providing a lingua franca, Rosgen’s classification system

has become a central support of stream restoration practice in the U.S.

The common language also anchors the epistemic community built around the NCD

approach, setting it apart from other restoration approaches and allowing initiates to

recognize each other with ease. Fluency in Rosgen’s classification systems is thus an

important species of symbolic capital among Rosgen’s supporters.

2. Rosgen as Provider of Shared Methods and Standards of Practice

Rosgen has developed the only set of purportedly universally-applicable methods for

channel reconstruction projects. As described above in Chapter 2, his design approach

includes both guidance on the overall form a newly-constructed channel should take, and

6 Author Interview, Syd Brown, California Department of Parks and Recreation. 6.24.04. 7 Author interview, Dr. Timothy Keane, Kansas State University. 7.15.06. 8 Author interview. Dr. David Montgomery, University of Washington. 6.28.06.

190

specific restoration techniques to achieve that form. There is no competing methodological

framework being put forward by academics because their insistence on the complexity of

stream systems makes such a universally-applicable system seem ludicrous. The engineering

community, spearheaded by Doug Shields, has started to develop such guidelines. But the

design approach laid out in Shields’ 2003 paper and his forthcoming Manual 54 update is

only roughed in, as Shields himself acknowledges (Shields et al. 2003, Shields et al. in press).

He has thus been a driving force behind the movement to develop standards of practice for

engineers working on stream restoration projects (the ASCE effort described briefly in

Chapter 3 and in more depth in Appendix D), but this initiative is only just getting

underway, and is unlikely to produce solid results for years. Thus for the moment, the NCD

approach provides the only solid footing of specified, agreed upon practice for restoration

practitioners left without concrete guidance from the academy or federal researchers.

Even though critics argue that the cookbook nature of the NCD approach enables

poorly-trained practitioners to wreak havoc on streams, practitioners are not the only

constituency for whom Rosgen’s work is a central pillar. Rosgen’s critics either ignore or

perhaps misunderstand the fact that the step-wise, articulated form of his design approach is

perhaps even more critical to the functioning of the regulatory community. 9 There is a large

group of regulators at the local, state, and federal levels who have been confronted since at

least the early 1990s with the Sisyphysian task of choosing consultants, managing contracts,

issuing permits, and writing legislation to produce successful stream restoration projects.

These regulators are the ones charged with reconciling our deeply-contradictory goals for

stream restoration: the return of nature, and the simultaneous extension of human control.

Regulatory staff are thus central to the development of stream restoration as a field despite

9 None of the Rosgen critics I interviewed who raised the cookbook issue mentioned it in the context of agency adoption.

191

the fact that it is a subject about which many, perhaps even most, have little knowledge or

experience. How is someone with little or no relevant training to evaluate the differences

and decide between a proposal prepared by an academically-trained fluvial geomorphologist

and one prepared by an NCD practitioner? And perhaps even more importantly, on what

basis can they justify that decision to their superiors and the public at large?

Bill Heatherman, the Stormwater Engineer for the City of Overland Park, KS and

the only local government official who is a member of the ASCE’s River Restoration

Committee, described the quandry he faces as follows:

One reason why regulatory agencies may be so inclined to require Rosgen courses is because of the lack of credentialing (so far) or even consistent academic coursework sequence that one could use to screen out people who should have some skills at design-scale stream geomorphology and those that clearly don't. This is not to say Rosgen's course is an adequate, acceptable, or optimal training program - only that it's the only one readily advertised and easily recognizable as such.

As a City official without a strong background in the discipline 6 years ago, but a desire to see our streams managed better, I first gravitated to Rosgen because his work was so noticeable in the "Interagency Stream Manual" and because he offered a short course in our region. Before long, I was able to realize the limitations of his approach, and by the time we tried to begin putting things in practice, we had already widened our perspectives quite a bit. It was difficult, though, to justify why we accepted the work of some consultants in geomorphology, but not everyone who came knocking on the door. (Heatherman 2005)

How can this decision to accept some consultants’ work and not others be justified?

Content seems the obvious ground for distinguishing between the approaches, but there is

sufficient overlap between the two camps that it requires a substantial knowledge base to see

the differences, much less to understand why they matter.10 For example, several of the

biology-trained restoration practitioners I interviewed said that they could not differentiate

10 Bourdieu writes that, “The field of argument which orthodoxy and heterodoxy define by their struggles is demarcated against the background of the field of doxa, the aggregate of presuppositions which the antagonists regard as self-evident and outside the area of argument because they constitute the tacit condition of argument.” (Bourdieu 1975, p.34)

192

between the material presented in Rosgen’s and Kondolf’s initial short courses.11 Experience

could be another possible justification for accepting one camp of consultants over the other,

but there are very experienced practitioners in both.

What remains, then, is the bureaucrat’s safe haven of justifiability: the application of

accepted standards (Bowker and Star 1999, Porter 1995). The only set of standards, spelled

out step by step, against which it is possible to check a channel reconfiguration design are

Rosgen’s; even if a regulator does not herself have the ability to run that check, she knows

that it is in theory possible to do so. Thus the standards of practice supplied by Rosgen’s

‘cookbook’ approach are critical to its utility for the new field

The state, in return, has played an enormous role in promoting Rosgen’s claims to

scientific expertise, as discussed at the end of the previous chapter. Agency staff at the local,

state and federal levels are the anchors of the NCD epistemic community, and federal

agencies such as EPA and NRCS have provided critical support for Rosgen’s claims to

scientific expertise by commissioning him to develop national standards and protocols. In

his article “Rethinking the State,” Bourdieu describes the state’s role as arbiter between

different types of capital, and thus between different claims to power or expertise. While the

quote below describes the state’s ability to settle competing claims to power among fields, it

seems to me that this same dynamic can apply within fields as well:

Concentration of the different species of capital (which proceeds hand in hand with the construction of corresponding fields) leads indeed to the emergence of a specific, properly statist capital which enables the state to exercise power over the different fields and over the different particular species of capital, and especially over the rates of conversion between them (and thereby over the relations of force between their respective holders). (Bourdieu 1998, pp.41-42)

11 These were people that took Kondolf’s short course before Peter Wilcock signed on as co-instructor. No one who has taken Kondolf’s course since then mistakes Wilcock’s sediment transport equations and Monte Carlo spread sheets for Rosgen’s far simpler “high puke factor” equations.

193

Resource and regulatory agencies powerfully support Rosgen’s claims to legitimacy and

expert status by asserting that the types of capital Rosgen holds are more valuable than the

more traditional scientific capital held by his critics.12

3. Disciplinary Reproduction and Training

Stream restoration has had such a sudden rise to environmental and economic

prominence that there is a serious disconnect between the jobs available and the pool of

qualified practitioners. The National River Restoration Science Synthesis project found that

by the late-90s there was approximately $1 billion per year spent on restoration in the U.S.

(Bernhardt et al. 2005), and that amount is still growing; a recent EPA study found over $4

billion in federal spending alone on compensatory mitigation of streams and wetlands in

2006 (ELI 2007).13 The issue of how to supply a trained restoration workforce is of deep

concern to everyone involved in the field.

Universities, for a variety of reasons discussed in Chapter 4, have been very

slow to respond with programs focused on restoration science or practice. Although

strongly-motivated students have been able to cobble together a selection of

reasonably relevant classes, no one believes they are graduating with sufficient

training. This poses a substantial challenge to the stream restoration field. Even in

scientific fields with a strong industry presence in knowledge production (such as

biotechnology or the petroleum industry), the university has been seen as the main

12 To be clear, federal research agencies and public universities are components of the state as well, but in this case the resource management and regulatory branches of the state have very different interests than the scientific branches of the state. 13 The report does not make clear what proportion of this was spent on streams. Most likely, the majority of the funds were for wetlands restoration, but even if only ¼ of the $4 billion were spent on streams, this would suggest a large rise in spending on restoration in the U.S., since many restoration projects are financed by private developers as part of the conditions for Section 404 permits.

194

training ground for practitioners. In stream restoration, by contrast, universities

have not stepped up, so the private sector has stepped in. As described in Chapter 4,

Rosgen has used his series of short courses to set up a powerful rival educational

system inculcating a habitus that is increasingly becoming the predominant subjective

structure of the stream restoration field. In the process, he created an epistemic

community that has been crucial in asserting the legitimacy of the NCD approach,

and at the same time revoking the claims to expertise of consecrated scientific

institutions when it comes to stream restoration. As NCD training is increasingly

adopted as the standard for practicing stream restoration, university-educated

restoration designers are being shut out of the field.

As a result, both the location and the relative porosity of the boundaries of

the stream restoration field are shifting. One of Bourdieu’s key assertions about

fields is that their boundaries are always a stake in the struggles for power within

them (Bourdieu and Wacquant 1992, p.100). Rosgen’s dramatic rise to the top of the

hierarchy in the restoration field, and his mass production of practitioners, many of

whom have few formal scientific credentials, have redefined what constitutes a

legitimate producer as well as the related issue of the conditions of entry to the

stream restoration field. Bourdieu argues that,

[A]t every moment, there is something like an ‘admission fee’ that each field imposes and which defines eligibility for participation, thereby selecting certain agents over others. People are at once founded and legitimized to enter the field by their possessing a definite configuration of properties. One of the goals of research is to identify these active properties,… these forms of specific capital. (ibid. pp.107-8)

Rosgen’s critics, almost of all whom possess substantial educational and institutional

capital, in effect argue that the boundaries to the restoration field had been fairly

195

closed, with the universities (and professional certification requirements in the case

of landscape architects and engineers) as gatekeepers. Rosgen’s short courses prized

the boundaries wide open, redefining the stream restoration field in the process.

B. Reflecting on Bourdieu through the Lens of the Rosgen Wars

Up until this point I have framed my analysis of the Rosgen Wars in terms of the

broad elements of Bourdieu’s field concept. I want now to focus specifically on Bourdieu’s

understanding of such field-defining conflicts, and ask how Bourdieu’s conceptualization of

struggle stands up to the analysis of the Rosgen Wars presented so far.

This is a fair question: conflict is one of the most central aspects of Bourdieu’s field

concept. He consistently describes fields as fundamentally agonistic, with struggle

programmed in by the disparities in capital created by fields’ hierarchical structure. It is thus

worth considering carefully what Bourdieu has to say about such structured struggles. Does

his understanding of the engines driving these conflicts, the types of participants involved,

and the strategies they employ adequately describe the dynamics of the Rosgen Wars?

I will argue below that they do not. Analysis of the Rosgen Wars demonstrates that

key elements of Bourdieu’s conceptualization of intra-field conflict require further

elaboration. I will focus on three of the most important: Bourdieu’s sparse theorization of

the motors of intra-field conflict, his characterization of the ways in which conflicts are

settled (in particular the factors that determine whether or not someone is a significant

player), and his understanding of fields as bounded entities. Before I begin laying out these

196

arguments, however, it is worth revisiting Bourdieu’s characterization of struggle within a

field.

Bourdieu distills the engines driving struggle within a field down to two types:

conflict caused by the differential between the autonomous and heteronomous poles of the

field, and conflict between the old guard and newcomers.14 Bourdieu argued in his article,

“The Specificities of the Scientific Field” (1975) that only the latter of these types matters in

the natural sciences. The struggle of autonomy vs. heteronomy – whether a field will be

dominated by its own logic and norms or by outside demands from the field of power – was

to his mind not an important factor in the scientific field because elites grant it a relatively

high level of autonomy (Bourdieu 1975, p.36). This assumption that outside forces are not

important to natural science fields seems to me a mistake, and will be discussed in more

detail below.

In his work on the scientific field, Bourdieu focuses exclusively on the second type

of conflict listed above: the struggle between those who are already dominant and those who

have just arrived. It is this conflict between the old and new guards that he posits as the

primary motor of struggle and change in the scientific field. Bourdieu argues that

incumbents and newcomers have only a few strategies available to them because the

positions they occupy, and thus their interests and the means they have to satisfy them, are

largely structured by the field itself (ibid., p.29). The old guard employs strategies of conservation

to preserve the existing scientific order (ibid. pp.29-30), while the new guard follows either

strategies of succession or strategies of subversion depending on the dispositions they bring into the

field. The majority toe the line and slot into the existing structure of the field, but a tiny

14 This second type of conflict should not to be confused with generational conflict because, as Bourdieu points out repeatedly (see Bourdieu 1996b, for multiple examples), age is not definitive of either incumbents or new arrivals in a field.

197

minority who enter with substantial capital try to change the existing structure of the field by

redefining the types of capital seen as crucial for scientific competence, and thus success in

the field. As Bourdieu describes it,

‘new entrants’ may find themselves orientated either towards the risk-free investments of succession strategies…, [realizing] the official ideal of scientific excellence through limited innovations within authorized limits; or towards subversion strategies, infinitely more costly and more hazardous investments which will not bring them the profits accruing to the holders of the monopoly of scientific legitimacy unless they can achieve a complete redefinition of the principles legitimating domination. (ibid. p.30)

According to Bourdieu, the crucial factor in whether newcomers choose the path of

meek compliance or radical opposition is the amount of capital they bring with them when

they enter the field:

the scientific equipment required to effect a scientific revolution can only be acquired in and by the citadel of the scientific establishment. As accumulated scientific resources increase, so the incorporated scientific capital needed in order to appropriate them and thereby gain access to scientific problems and tools, and thus to the scientific struggle, becomes greater and greater (the cost of entry). The consequence is that scientific revolution is the business not of the poorest but of the richest (in scientific capital) among the new entrants. (ibid. p.33)

As I will explain in more detail below, this characterization of the drivers, strategies,

and agents of conflict seems to me too schematic. Analysis of the Rosgen Wars suggests

that Bourdieu’s description of the conflicts at the heart of any field needs to be fleshed out

in several crucial ways.

1. Conflict Between Incumbents and New Arrivals

According to Bourdieu, the primary motor of struggle and change in the scientific

field is the old guard vs. new guard dynamic, which brings with it a highly circumscribed set

of possible strategies for participants. Unfortunately, neither of the two strategies that

198

Bourdieu ascribes to newcomers adequately describes Rosgen’s actions, which seem to me

both more and less radical than the highly-limited palette of options Bourdieu acknowledges.

It should be obvious by now that Rosgen is not following a strategy of succession by gamely

upholding the norms of science, but neither could he be described as following the only

other course of action open to the new guard in Bourdieu’s conceptualization – strategies of

subversion. Rosgen is not attempting to go beyond the current consensus in restoration

science, but is instead relying on science from the 1950s and 60s and largely ignoring

developments in the field since then. Further, although it seems fair to describe him as part

of the new guard given when he arrived in the field, he built his career in part through an

alliance with the oldest of the old guard: Luna Leopold, one of the three eminences grises of the

last few decades of fluvial geomorphology (along with Reds Wollman and Stan Schumm),

and the one least willing to ride the changing tides of geomorphology. Last, Rosgen is in

some ways following the structured expectations of a field in which science has until recently

stood at the top of the hierarchy. He took the time to earn a Ph.D. in 2003. He presents at

the major national technical conferences that include restoration threads, and he structures

his new knowledge claims into papers and text-books (though non-peer reviewed). These

strategies seem too tame to fit well under Bourdieu’s rubric of subversion.

Instead, Rosgen is engaged in a different kind of struggle. Instead of asserting the

value of a different type of scientific capital, as Bourdieu expects, Rosgen is attempting to

assert the value of a different species of capital altogether. Following something that could

be more accurately described as strategies of reorientation, he is shifting the axis of the entire

field, transforming stream restoration from its initial configuration as an applied science into

a field in which science is peripheral (see figure 7.2, below). By pulling the locus of

199

figure 7.1. the initial structure of the stream restoration field

figure 7.2: the emerging structure of the stream restoration field

200

disciplinary reproduction out of the university into the private sector, producing new

knowledge claims without formal educational training, and refusing to participate in

scientific institutions such as peer review, Rosgen is effectively dethroning scientific capital

from the apex of the stream restoration field, even as he continues to follow some of its

conventions. This seems to me far more radical than the strategies of subversion Bourdieu

describes, which try to redefine the peak of scientific achievement, not to depose science

entirely. Like Monty Python’s old men at the bank, Rosgen appears to be releasing the

stream restoration field from its academic moorings and sailing it off in a new direction.

2. The Relative Poverty of Participants

I have argued above for expanding the range of strategies possible in the structured

conflicts Bourdieu describes. What then of how those conflicts are resolved? Bourdieu

expends a great deal of effort describing dueling claims to authority and status, but he never

goes into much detail about what allows a particular claim to win. There are a few general

statements, such as:

In the scientific field as in the field of class relations, no arbitrating authority exists to legitimate legitimacy-giving authorities; claims of legitimacy draw their legitimacy from the relative strength of the groups whose interests they express…. (ibid., p.24)

This seems to me to beg the question, however. What factors make a group strong enough

to defend a particular species of capital? What are the mechanisms for winning such a fight?

The Rosgen Wars demonstrate that factors outside of a field, such as state interests, can play

a part in determining the outcome of an intra-field struggle, but what about the mass of

participants in the stream restoration field itself?

201

Bourdieu consistently argues that only those participants who hold substantial

amounts of capital can be considered to exist in the field in any substantive sense. The

Rosgen Wars problematize this limited view in two quite different ways. First, there is

Rosgen himself. Bourdieu argues that the habitus an agent brings into the field profoundly

determines the position and trajectory s/he is likely to take within it:

[I]n a given state of the field, researchers’ investments depend both in their amount… and their nature… on the amount of actual and potential recognition-capital which they possess, and on their actual and potential positions in the field… [Thus] researchers’ aspirations… rise as their capital of recognition rises…. [I]nsofar as the qualification [academic degree], as scholastic capital reconvertible into … scientific capital, contains a probable trajectory, it governs the agent’s whole relationship with his scientific career (the choice of more or less ‘ambitious’ projects, greater or lesser productivity, etc.) through the intermediary of the ‘reasonable aspirations’ which it authorizes. (ibid., pp.27-28)

According to this description of the powerful relationship between the dispositions that

people bring to a field and the positions they assume within them, Rosgen should not exist:

when he entered the field, he did not hold sufficient amounts of scientific habitus or

educational capital to be a viewed as a contender for greatness or a legitimate participant in

the constant struggle to define the field itself. As demonstrated in Bernhardt et. al. 2005, the

field of stream restoration began to expand rapidly in the mid-1980s when Rosgen was

making his start. Given the number of people joining the field at that time, many of whom

brought with them prestigious educational qualifications and a great deal of scientific capital,

it is hard to imagine an interpretation which would view Rosgen – fresh from being fired by

the USFS, with only a B.S. from Humboldt State University and a determinedly maverick

stance – as comparatively rich in scientific capital. Yet despite entering the field with a

relatively low stock of capital, Rosgen still managed to grasp the resources needed to mount

a powerful challenge to those at the top of hierarchy when he arrived. Clearly, we need to

202

more carefully think through Bourdieu’s assertions about the power of the dispositions an

agent brings into the field in structuring his trajectory within it.15

Second, the critical role played by both grassroots demand for stream restoration and

the NCD epistemic community suggests that we need to pay more attention to the masses,

those low level participants in a field who, according to Bourdieu, do not hold sufficient

amounts of capital to have any importance within it. As I described in the Introduction, the

most basic goals for stream restoration in this country are deeply contradictory. More than

any other type of ecological restoration, human goals are a critical component of stream

restoration, even though they often undermine or obstruct the ecological goals for a project.

By setting a Janus-faced foundation for stream restoration (the fundamentally conflicting

goals of both human control and natural self-determination), the grassroots movement at the

base of the stream restoration field set the conditions for Rosgen’s success. Similarly, the

power of the relatively powerless – as collected and focused through the lens of the NCD

epistemic community – has been a key factor in Rosgen’s success. Thus I believe that

Bourdieu’s field framework must be expanded to address the role of the uncapitalized

masses in determining the outcomes of battles of the capitalized elite, as I have done above

in Section A of this chapter.16

15 One possible way to salvage this argument while remaining within Bourdieu’s framework would be to point out that while Rosgen lacked the educational and scientific capital that Bourdieu privileges, Rosgen was comparatively rich in social capital because of his connections with Luna Leopold, and his broad ties to hydrologists throughout the USFS. These ties were fundamental for propelling his initial success in the consulting world as both practitioner and teacher. Even though social capital is not a factor Bourdieu includes when speaking about the species of capital relevant to science, it would fall within his overall assertion that those who are richest in some form of capital are licensed for revolution. 16 The only place in his work on fields I found where Bourdieu entertains this possibility even for a moment is in his book The State Nobility, where in the midst of an almost Gramscian take on the ways in which the dominated can benefit from conflict among the dominant, he says that, “the dominated can always take advantage of or benefit from conflicts among the powerful, who, quite often, need their cooperation in order to triumph in these conflicts….” (Bourdieu 1996a, p.389) Unfortunately, he does not elaborate on this potential.

203

3. Autonomy, Heteronomy and Boundedness of Fields

Perhaps most critically, analysis of the Rosgen Wars suggests that the fundamental

structure of the field – Bourdieu’s posited axis between autonomous and heteronomous

producers – requires further thought. According to Bourdieu, every field is structured by an

axis running between those whose production is shaped by the species of capital specific to

that field, and those whose production is shaped primarily by outside forces. This

immediately raises the question of what constitutes outside forces. Demand from the field

of power is clearly one of them in Bourdieu’s framework, but what about market forces?

Bourdieu asserts in one of his later works on fields that it is only in the field of power that

the structuring axis runs from cultural capital to economic capital (Bourdieu 1998); in other

fields the forces at work at the heteronomous end of production are assumed not to be

primarily economic. Despite this, in works such as The Rules of Art, his most developed

examination of the birth of a field, Bourdieu repeatedly implies a conflation between

heteronomy and economic capital (see Bourdieu 1996b, pp.83, 114, 115, 120-121, 141).

Heteronomy is consistently equated with influence by market forces. Thus I would argue

that Bourdieu has built political-economic tensions into the structure of all fields.17

How then, would the conflict between autonomous and heteronomous participants

describe the dynamics of the Rosgen Wars? The obvious read is the one presented in Figure

7.1, above: public sector science – in the form of university and federal research agency

scientists – occupies the autonomous pole and Rosgen and the NCD community hold down

the heteronomous pole. After all, Rosgen’s vision of what constitutes the core of stream

17 As I mentioned above, Bourdieu explicitly excludes the natural sciences from this autonomous/ heteronomous divide, arguing that the field of power grants them a high level of autonomy. As I will describe in more depth in the Conclusion, this view of the relationship between public science and private profit seems out of date, and I very much doubt that Bourdieu would take such a position if he were writing today.

204

restoration is based on practice, with all of the market and regulatory influences that implies.

And yet, as is evident in Figure 7.2 above, I do not believe it is reasonable to characterize

Rosgen as the heteronomous pole of the field and the Rosgen Wars as simply the fight over

whether stream restoration will be defined by its own pure goals or instead by the goals of

the market and state. Although I have argued that the demands of government agencies and

the restoration market were fundamental to Rosgen’s success, his work is not driven only by

the goal of meeting those demands. According to Rosgen and his supporters, the types of

capital that define the stream restoration field and grant their holders legitimacy and expert

status are fundamentally about field experience, success in hands-on work, outdoorsmanship

and protection of natural resources, most notably game fish. While these are indeed not the

species of capital typical of scientific fields, and while they clearly display a wide range of

influences from fields other than stream restoration, they are not mere measures of market-

based success or satisfaction of state goals. Success in the NCD community is not described

by the dollar value of projects or the number of publicly-awarded contracts, for example.

Instead, I think it is more reasonable to argue that the confusion over where Rosgen

and the Natural Channel Design community stand on the autonomy/heteronomy axis is a

symptom of a larger issue: the fact that Bourdieu defines production determined by the core

values of the field as autonomous, bounded off from other fields. In contrast to this contained

vision, a major part of what is at stake in the Rosgen Wars is the question of which outside

fields should serve as the pole stars orienting the field’s axis. For the applied scientists that

have until recently dominated the stream restoration field, its key relation is with the more

rarified field of basic science; the types of capital they privilege are all derived from orthodox

scientific practice. It is their world view that is evident in Figure 7.1, above. By contrast,

Rosgen and his supporters’ push to reorient the field seems to be based on the belief that

205

stream restoration should have stronger ties to the environmental and outdoor recreation

fields than to the field of science, as is evident from the types of capital they prioritize. The

Rosgen Wars seem to me conclusive evidence that the core values of a field can be, and

perhaps always are, determined in relation to other fields rather than in the isolation the term

autonomy implies.

C. Conclusion

I have argued in this chapter that the reason why Rosgen has been so successful, and

his opponents so ineffective, is that he developed a structure for the emerging stream

restoration field when public sector science failed to create one. Rosgen’s classification

system serves as the lingua franca of the field, enabling collaboration among the disparate

disciplines called on in restoration practice. The NCD standards of practice enable agency

staff to carry out stream restoration’s contradictory mandate of extending human control

while simultaneously setting nature free. Finally, the short course series continues to provide

the most detailed, targeted education in stream restoration available in the U.S. All three of

these elements served to create an epistemic community supporting Rosgen’s claims to

expertise and promoting the types of capital he holds. He has thus deeply influenced both

the objective and subjective structures of the stream restoration field.

Analysis of the fundamental causes of the Rosgen Wars, however, suggests that while

many aspects of Bourdieu’s field concept are quite useful, his characterizations of the

inherently agonistic nature of fields and of the ways in which those conflicts are resolved are

too simplistic. First, the Rosgen Wars owe their longevity and ferocity to the fact that they

206

are at core a struggle not over who occupies the top positions in the stream restoration field,

but over the orientation of the structuring axis of the field itself. At the moment Rosgen is

successfully tilting the axis of the field away from a science vs. practice orientation (7.1),

towards a Natural Channel Design vs Traditional Engineering orientation (7.2). If he

succeeds in making this reorientation stick, the scientists who have to date occupied the

most prestigious positions in the stream restoration field would move from the apex to the

periphery, a geographical difference of substantial consequence. By pulling the locus of

disciplinary reproduction out of the university into the private sector, Rosgen is challenging

the entire consecrated structure for bestowing scientific legitimacy and constructing

expertise.

Second, there is the question of how the conflicts so central to Bourdieu’s field

concept are settled in practice. While Bourdieu explicitly acknowledges the role of the state,

and implicitly acknowledges the role of market forces, he seems to regard the mass of

relatively powerless participants in a field as peripheral at best to the great struggles that

determine its trajectory. I have argued instead that the grassroots restoration movement was

profoundly implicated in Rosgen’s success because it set the contradictory agenda for

restoration, and thus the conditions for his success. Thus any field analysis should go

beyond analysis of the powerful to examine the potentially decisive role of the uncapitalized

masses in determining the outcomes of endemic intra-field struggle.

Finally, analysis of the Rosgen Wars calls into question a key component of the field

concept: Bourdieu’s assertion that fields are isolated entities. While Bourdieu concentrates

fiercely on relations within fields, he treats their relations with any field other than the field

207

of power as incidental.18 But it is the issue of which fields stream restoration should point to

– environmental practice vs. basic science – that is at the heart of the Rosgen Wars.

Applying the field concept to the Rosgen Wars highlights a major blind spot in Bourdieu’s

conceptual framework: his tendency to treat the arenas he studies as bounded systems.

18 The one place I have found a discussion of the relations between two fields is in The Rules of Art, where Bourdieu briefly addresses the relation between the literary and artistic fields, but only to explain how they split off from each other. It seems a true case of an exception proving a rule.

208

Chapter 8: Conclusions

I have argued in this dissertation that Rosgen transcends the traditional role of a

consultant as applier of knowledge developed elsewhere, and functions instead as an active

producer of knowledge, an arbiter of legitimacy, and a powerful rival to established scientific

institutions. Increasingly in danger of marginalization or even irrelevance in their own field,

university- and agency-based scientists have fought back in defense of their students, science,

and understanding of streams. The Rosgen Wars are at heart a struggle over the orientation

of the stream restoration field: towards practice or science, the private sector or the public.

I have also attempted to demonstrate in this dissertation that the work of Pierre

Bourdieu can serve an important role in expanding political ecological analysis to more

carefully consider the production of science, and STS analysis to address the role of political-

economy. Bourdieu’s field concept has proved to be a very useful heuristic device for

analyzing the political economy of a scientific field, focusing attention on the objective and

subjective structures of the field, the key institutions, the species of capital in play, and the

critical role that the state and markets play in determining the outcome of substantive

scientific debates.

At the same time, I have argued that analysis of the Rosgen Wars suggests a few key

ways in which Bourdieu’s conceptualization of the field must be expanded. First, Bourdieu’s

schematic understanding of the structured conflicts that define a field needs fleshing out.

The Rosgen Wars suggests that an important additional category would be strategies of

reorientation, as participants struggle not just to change the species of capital awarded the most

value in a field, but to reorient the field entirely. Related to this, and drawing on the

emphasis in Political Ecology and Critical Geography more generally on relationality,

209

Bourdieu’s understanding of fields as bounded entities defined by internal struggles must be

reworked. Analysis of the Rosgen Wars suggests that in fact the locus of struggle is over

relations with other fields, not simply the internal hierarchy of the field. Finally, I argued

that Bourdieu’s lack of attention to the power of the uncapitalized masses must be remedied.

In the Rosgen Wars, the lower level participants in the stream restoration field have been

critical to determining the outcome so far.

What then, of the future of stream restoration? There are a few clear trends that I

will discuss below. Beyond the stream restoration field, however, the Rosgen Wars have

important theoretical implications for the political economy of expertise and the neoliberal

university.

A. The Future of Stream Restoration

Throughout my research, I paid particular attention to discussions of the

likely trajectory of the NCD approach. Despite a number of defeatist comments,

such as the wry quip from the beginning of Chapter 7, many of Rosgen’s opponents

and supporters seemed to feel that NCD is destined to lose its current prominence.

Rosgen’s supporters and some moderates point to two factors that might cause

NCD’s phenomenal popularity to fade: Rosgen’s retirement, and the rise of new

methods. Rosgen is 67 years old, and he works at a pace most people would find

unsustainable in their twenties. It is not clear how long he can maintain it, and yet

doing so is considered critical by many of his supporters because they see his

knowledge and charisma as central to the appeal of his work. Many speculated that

210

attendance at short courses would drop dramatically if Rosgen no longer taught

them, with a resulting reduction in agencies and consultants employing his methods.

At that point, speculation goes, other promising NCD methods would gain

prominence in their turn.1 As Gary Parker, a moderate in the Rosgen Wars, put it:

you know what’s gonna happen:… Dave’s gonna retire, and that is going to be probably the biggest change…. He’s getting older and he’s teaching at an exhausting pace. I can’t even imagine how he does it. When he tapers off, he has been THE person doing this…. I know he has people working with him at Wildland Hydrology, but I very much doubt that they have the charisma that he does. So it might taper off then and we might see other things rising.2

Unsurprisingly, opponents have a different vision of how NCD’s prominence may

wane. They point to what is consistently described as the ‘lifecycle’ of adoption: first,

agencies enthusiastically embrace the NCD approach based on claims of universal

applicability, doability, stability, etc. and fund a swath of NCD projects; second, at some

point thereafter, a 5 – 10 year storm hits and there is a rash of dramatic project failures; at

that point, agencies come to their senses and start looking for other approaches. When I

asked critic Martin Doyle about his sense of the NCD approach’s trajectory, he said that,

It seems like there’s a life cycle: love Rosgen, get over enamoured with him, start to see some failures and shortcomings of the approach, and then start to do other things…. Tennessee has just gotten into river restoration in the last five years and they’re all about Rosgen. So I figure they’ve got another 3-5 years before they start to come down off the high.3

I would like to suggest a third possible trajectory based on my data from North

Carolina: adaptation and expansion. The data I presented in Chapter 3 demonstrated that

that state is indeed a bastion of support for the NCD approach. Upon closer examination,

1 This almost Weberian vision of charismatic authority seems to me to misread the deeper roots of Rosgen’s success. Clearly Rosgen’s charisma is not trivial, but I have argued in this dissertation that focusing on the man hides the far more important function his approach serves in enabling the market and regulatory practices of stream restoration. 2 Author interview, Dr. Gary Parker, University of Illinois – Champagne Urbana, 6/26/06. 3 Author interview, Dr. Martin Doyle, University of North Carolina – Chapel Hill, 6/15/06.

211

however, the picture is more complicated. The North Carolina local, state and federal

agency staff members I interviewed tended to have well thought out critiques of the theory

and practice of the NCD approach (Table 8.1). The most common critiques were that NCD

is often applied thoughtlessly, and that it does not apply well to the particular conditions

either in North Carolina as a whole, or in urban conditions in particular. Other criticisms

included the limited focus on the project reach, and the lack of attention to biology. All of

these should sound familiar: they are common objections raised by the Rosgen opposition,

as discussed in Chapter 5. Only two people said Rosgen’s approach had no weaknesses,

which contrasts sharply with the image of Rosgenites – mindless members of the Cult of

Dave – that opponents often promote.

new and untested 10%

no consideration of biology 10%

limits focus to project reach 15%

expensive 25%

often poorly implemented 25%

not applicable here without modification 30%

applied thoughtlessly 35%

no answer 10% Table 8.1 Weaknesses of Rosgen’s approach (percentages to not sum to 1 because respondents frequently mentioned more than one concern).

Further, interview subjects’ balanced view of the strengths and weaknesses of the

NCD approach has produced changes in their practice. The restoration community in

North Carolina may use NCD almost to a person, but many of them are adapting it in to

better address local conditions and broader concerns. For example, Will Harman, one of the

founding principles of Buck Engineering, a top NCD consulting firm, said that he’s

212

modified the reference reach aspect of Rosgen’s approach to fit local conditions because he

came to realize that:

all of our reference reaches, and there are not very many of them because the Southeast has been so modified, but the few that we found were in mature, bottomland hardwood forest, 100-year-old forest, 50-year-old at least, and that pattern was really dictated by that vegetation…. They are not ‘free to form channels’ like they have out in the arid West, where vegetation doesn’t play a key role because it’s so slow in establishing that the river sort of establishes its pattern first…. That’s not really the case here…. Those streams were probably channelized long ago, and then as they were recovering the veg was growing as well. At some point, that veg sort of locked the pattern in place…. [By contrast] we are working in a denuded flood plain, and when we would try to put in the radius of curvature that worked great in that reference setting where there’s a 50-year-old oak tree holding the bend in place, now we’ve got nothing but some root wads or a cross-vane or something [to hold things in place], and life taught us better. So… we’ve really modified that approach.4

Harman is not alone in modifying the NCD approach. A majority of the people

employing NCD in North Carolina said that they were supplementing it with additional

restoration approaches. Figure 8.1 below shows that while 35% of respondents were using

NCD as their sole restoration method, 40% were adding other approaches to a primary

focus on NCD, and 15% were using NCD as simply one of many tools in their toolkit.

Additional restoration techniques employed include careful attention to aquatic habitat,

sophisticated hydraulic modeling, broad watershed-scale analysis of land use factors causing

channel change, adaptation to regional climate and geology, and even a move towards less

structural treatments. In fact many of the changes that Rosgen’s critics say must be made to

correct his work seem to already be in process in North Carolina.

4 Author interview, William Harman, Buck Engineering/Michael Baker Corporation, 8/14/06. In a follow-up email, Harman noted that reference reaches are not their only, or even their primary, source for designing new channels: “we use the reference reach as one approach for developing design criteria. We also use monitoring results from past projects, computer models, and equations from the literature. We use different techniques depending on the type of project and look for ‘converging lines of evidence’. It's about applying the right solution to the right problem.” Clearly, this is not a case of mindless application of the NCD approach.

213

NCD alone35%

primary +40%

one tool of many15%

no NCD10%

Figure 8.1 Restoration approaches reported in North Carolina

The practice of Rosgen’s most skilled students is more sophisticated than his critics

assume. At the end of our interview Dick Everhart, one of primary supporters of NCD in

North Carolina, said he tells students in his workshops, “to keep your eyes open and

continue to learn. If it works, try to refine it; if it fails, try to figure out why…. If you don’t

keep learning you’re not doing your job!”5 Thus in contrast both Rosgen’s supporters and

opponents, I think that twenty years from now it may be that most practitioners in the U.S.

claim to be using Natural Channel Design, but that Rosgen’s contribution to that approach

is increasingly peripheral.

For the moment, however, evidence from the regulatory arena suggests that

Rosgen’s prominence is not likely to peak for some time. As discussed in Chapter 6, the

5 Author interview, Richard Everhart, NRCS/Surry County Soil & Water Conservation District, 11/30/07.

214

EPA is now promoting Rosgen’s work as its recommended method for setting TMDLs for

sediment nation-wide. This vastly expands the reach of Rosgen’s work from the large but

still limited pool of streams perceived to be in need of stabilization or reconfiguration, to all

flowing bodies of water in the US. Because of the high level of uncertainty involved in

estimating sediment flows, and the difficulties (some would say the impossibility) of getting

accurate sediment transport field data, Rosgen’s relatively simple WARSSS method is likely

to have tremendous appeal for state and local regulators trying to develop TMDLs. Given

that the sediment transport research community has not been granted access to the technical

details supporting Rosgen’s WARSSS method and models, there is no obvious way in which

it can be evaluated until TMDLs based on it have been in use long enough to evaluate their

effects on the ground. Thus Rosgen’s work is likely to be prominent for many years to

come, even if retirement, adoption lifecycles, or adaptation eventually reduce its usage in the

stream restoration field.

A second major growth area for Rosgen’s work is the emerging stream mitigation

banking industry referred to in Chapter 3 (see Lave et al. 2008 for an overview). Stream

mitigation banking is a practice which allows developers to destroy riparian resources on

their property as long as they purchase equivalent restoration credits from mitigation banks

that buy and restore streams on a speculative basis. In a growing number of states, the

Rosgen classification system is being used to establish the equivalence between destroyed

and restored streams, in large part because it meets regulatory agency and market needs for a

simple, repeatable measurement system. Rosgen’s work has thus become the basis for

recent neoliberal incursions into stream restoration, and the expanded commoditization of

streams. The stream mitigation banking industry is growing rapidly, thus it, too, seems likely

to ensure Rosgen’s prominence for some time to come.

215

B. Implications for the Political Economy of Expertise

Reaching beyond the arena of riparian systems in the U.S., however, the Rosgen

Wars have serious implications for the construction of scientific authority and the relations

between public and private sector science more broadly. There has been much concern

about the reach of private industry into the academy (at UC Berkeley, for example, we have

had both Novartis buying a five-year lease on the Plant and Microbial Biology Department,

and the still unfolding saga of BP-funded alternative energy research institute) and I believe

that the dynamics of the Rosgen Wars sheds some light on the questions raised by these

initiatives.

I have argued that regardless of your opinion about the substantive content of

Rosgen’s work, the Rosgen Wars demonstrate the possibility of private industry not merely

infiltrating, but supplanting, the academy. Rosgen has transcended the consultant’s

traditional role as applier of knowledge developed elsewhere, creating a classification system

for ordering new data, developing and promulgating a set of knowledge claims about how

best to understand riparian ecosystems, and building up an educational system that has

gradually superceded the consecrated academic institutions. He has ascended to the peak of

the restoration field (or more accurately, he arrived somewhere and is in the process of

successfully declaring it the peak) with little formal academic training, without conforming to

many of the established norms of scientific practice, and without support from the official

gatekeepers of scientific authority.

Is this a fluke, or are the Rosgen Wars a portent of larger trends in the environmental

sciences? There is no sure answer to this question, but my sense is that the Rosgen Wars do

indeed presage a larger shift in the political economy of expertise. I argued in the

216

Introduction that one of the reasons why the stream restoration field is worthy of study is

that it has much in common with other newly prominent environmental sciences, such as

the study of climate change and the impacts of environmental mutagens. First, like stream

restoration these fields are driven by the environmental commitments and perceptions of

crisis of the public and of scientists themselves rather than by scientific breakthroughs. They

thus are characterized by their relatively undeveloped content, often drawn from pieces of

existing fields and not yet fully integrated. Second, like stream restoration, these fields both

study highly complicated systems, and are based on new scientific paradigms that emphasize

complexity. This gives their findings a high level of uncertainty that is not commensurate

with the expected role of science as arbiter of truth in the policy process. The relative lack

of substantive development compared to older more established fields, and the insistence on

uncertainty, render these sciences far less able to police their boundaries and fend off claims

to legitimacy from outside the consecrated realms of public science.

Third, like stream restoration, the new environmental sciences deal with issues in

which the general population has powerful interests. Humans want to be able to control the

consequences of potential crises from flooding or climate change, thus they may intervene in

scientific debates to support combatants who propose solutions that suit their needs. In

many cases the new environmental sciences deal with issues that are similarly central to

developing or established markets, which adds to the demand for certainty. I am thinking

here of both the stream mitigation banking industry, and the growing weather derivatives

market.

Finally, the increasingly commercialized structure of the neoliberal university is

reducing the resources of public science to defend its territory against private incursion. The

217

growing emphasis on the market place of ideas and on evaluating research on its economic

merits make outsiders offering applied science solutions look increasingly credible.

All of these factors suggest that the implications of Rosgen’s success cannot be

limited to the stream restoration field; others of the new environmental sciences are similarly

vulnerable to outside challengers. The Rosgen Wars may thus presage a fundamental shift in

the political economy of scientific expertise.

218

Bibliography Arthur, M.B. "Fish Stream Improvement Handbook." 33. Washington, D.C.: U.S. Forest

Service, 1936. Ashmore, Peter. "What Would We Do without Rosgen?: Rational Regime Relations and

Natural Channels." Paper presented at the 2nd International Conference on Natural Systems Design, Niagara Falls, Canada 1999.

Bakker, Karen. "Neoliberalizing Nature? Market Environmentalism in Water Supply in England and Wales." Annals of the Association of American Geographers 95, no. 3: 542-65.

Baldwin, A. Dwight, Judith De Luce, and Carl Pletsch, eds. Beyond Preservation: Restoring and Inventing Landscapes. Minneapolis: University of Minnesota Press, 1994.

Beck, Ulrich. Risk Society: Towards a New Modernity. Thousand Oaks, CA: Sage Publications, 1992.

Bell, Morag, and N. Roberts. "The Political Ecology of Dambo Soil and Water Resources in Zimbabwe." Transactions of the Institute of British Geographers NS 16 (1991): 301-18.

Bernhardt, E.S., M.A. Palmer, J. D. Allan, G. Alexander, K. Barnas, S. Brooks, J. Carr, S. Clayton, C. Dahm, J. Follstad-Shah, D. Galat, S. Gloss, P. Goodwin, D. Hart, B. Hassett, R. Jenkinson, S. Katz, G. M. Kondolf, P. S. Lake, R. Lave, J. L. Meyer, T. K. O’Donnell, L. Pagano, B. Powell, and E. Sudduth. "Synthesizing U.S. River Restoration Efforts." Science 308 (2005): 636-37.

Bhuiyan, F., R. Hey, and P. Wormleaton. "Hydraulic Evaluation of W-Weir for River Restoration." Journal of Hydraulic Engineering 133, no. 6 (2007): 1-14.

Blaikie, Piers. The Political Economy of Soil Erosion in Developing Countries. Edited by Dr. D.J. Dwyer. Fourth ed, Longman Development Series. New York: John Wiley & Sons Inc., 1985.

Bowker, Geoffrey C., and Susan Leigh Star. Sorting Things Out: Classification and Its Consequences. Edited by Wiebe E. et. al. Bijker. fourth paperback ed, Inside Technology. Cambridge, MA: MIT Press, 1999. Reprint, 2002.

Bourdieu, Pierre. "The Specificity of the Scientific Field and the Social Conditions of the Progress of Reason." Social Science Information 14, no. 6 (1975): 19-47.

Bourdieu, Pierre. "The Field of Cultural Production, Or: The Economic World Reversed." Poetics 12 (1983): 311-56.

_____. "Genesis and Structure of the Religious Field." In Comparative Social Research: Religious Institutions, edited by Craig Calhoun, 1-44. Greenwich, Connecticut: Jai Press Inc., 1991.

_____. The State Nobility: Elite Schools in the Field of Power. Translated by Lauretta C. Clough. Stanford: Stanford University Press, 1996a.

_____. The Rules of Art: Genesis and Structure of the Literary Field. Translated by Susan Emanuel. Stanford, CA: Stanford University Press, 1996b.

_____. "Rethinking the State: Genesis and Structure of the Bureaucratic Field." In Practical Reason: On the Theory of Action, edited by Pierre Bourdieu, 35-63. Stanford, CA: Stanford University Press, 1998.

Bourdieu, Pierre, and Loic J.D. Wacquant. An Invitation to Reflexive Sociology. Chicago: University of Chicago Press, 1992.

Bourdieu, Pierre, Loic J.D. Wacquant, and Samar Farage. "Rethinking the State: Genesis and Structure of the Bureaucratic Field." Sociological Theory 12, no. 1 (1994): 1-18.

Brookes, A. Channelized Rivers: Perspectives for Environmental Management. Chichester: John Wiley & Sons, 1988.

219

Braun, Bruce, and N. Castree. "The Construction of Nature and the Nature of Construction: Analytical and Political Tools for Building Survivable Futures." In Remaking Reality: Nature at the Millenium, edited by Bruce Braun and Noel Castree, 3-42. New York: Routledge, 1998.

Canaan, Joyce E., and Wesley Shumar, eds. Structure and Agency in the Neoliberal University: Routledge Press, 2008.

Ciafone, Amanda. "Endowing the Neoliberal University." Work and Culture 2005, no. 7 (2005).

Collins, H.M., and Robert Evans. Rethinking Expertise. Chicago: University of Chicago Press, 2007.

Davies, Bronwyn, Michael Gottsche, and Peter Bansel. "The Rise and Fall of the Neo-Liberal University." European Journal of Education 41, no. 2 (2006): 305-19.

Delborne, Jason. "Transgenes and Transgressions: Scientific Dissent as Heterogeneous Practice." Social Studies of Science, 2008.

Demeritt, David. "Science, Social Constructivism and Nature." In Remaking Reality: Nature at the Millenium, edited by Bruce Braun and Noel Castree, 173-93. New York: Routledge, 1998.

Doyle, Martin W., Dale E. Miller, and Jon M. Harbor. "Should River Restoration Be Based on Classification Schemes or Process Models?: Insights from the History of Geomorphology." Paper presented at the ASCE International Conference on Water Resources Engineering, Seattle, Washington 1999.

Doyle, Martin W., and Jon M. Harbor. "Discussion of ‘Evaluation of Rosgen’s Streambank Erosion Potential Assessment in Northeast Oklahoma’ by Harmel." Journal of the American Water Resources Assocation 36, no. 5 (2000): 1191-92.

Doyle, Martin W., F. Douglas Shields, Karin F. Boyd, Peter B. Skidmore, and DeWitt Dominick. "Channel-Forming Discharge Selection in River Restoration Design." Journal of Hydraulic Engineering 133, no. 7 (2007): 831-37.

Dunne, Thomas, and Luna B. Leopold. Water in Environmental Planning. New York: W.H. Freeman and Company, 1978.

Eden, Sally, S. Tunstall, and S. Tapsell. "Translating Nature: River Restoration as Nature-Culture." Environment and Planning D: Society and Space 18, no. 2 (2000): 257-73.

Institute, Environmental Law. "Mitigation of Impacts to Fish and Wildlife Habitat: Estimating Costs and Identifying Opportunities." 126, 2007.

Epstein, Steven. Impure Science: Aids, Activism, and the Politics of Knowledge. Edited by Andrew Scull. first ed, Medicine and Society. Berkeley: University of California, 1996.

Fairhead, J., and Melissa Leach. Misreading the African Landscape: Society and Ecology in a Forest-Savana Mosaic. Cambridge: Cambridge University Press, 1996.

_____. Science, Society and Power: Environmental Knowledge and Policy in West Africa and the Carribean. Cambridge: Cambridge University Press, 2003.

Federal Interagency Stream Restoration Working Group. Stream Corridor Restoration: Principles, Processes, and Practices, 1998.

Forsyth, Tim. Critical Political Ecology: The Politics of Environmental Science. 1st ed. London: Routledge, 2003.

Gieryn, Thomas F. Cultural Boundaries of Science: Credibility on the Line. Chicago: University of Chicago Press, 1999.

Gillilan, Scott. "Use and Misuse of Channel Classification Schemes." Stream Notes, October 1996, 2-3.

220

Gobster, Paul H., and R. Bruce Hull, eds. Restoring Nature: Perspectives from the Social Sciences and Humanities. first ed. Washington, D.C.: Island Press, 2000.

Godoy-Anativia, Marcial. "Chilean Students Take on the 'Neoliberal University'." NACLA Report on the Americas 31, no. 2 (1997): 42-44.

Goldman, Michael. "Imperial Science, Imperial Nature: Environmental Knowledge for the World (Bank)." In Earthly Politics: Local and Global in Environmental Governance, edited by Sheila Jasanoff and Marybeth Long Martello, 55-80. Cambridge, MA: The MIT Press, 2004.

_____. Imperial Nature: The World Bank and Struggles for Social Justice in the Age of Globalization. Edited by James C. Scott, Yale Agrarian Studies Series. New Haven, CT: Yale University Press, 2005.

Gould, Stephen Jay. "Deconstructing the 'Science Wars' by Reconstructing an Old Mold." Science 287 (2000): 253-61.

Gross, Matthias. Inventing Nature: Ecological Restoration by Public Experiments. New York: Lexington Books, 2003.

Haas, Peter. Saving the Mediterranean: The Politics of International Environmental Cooperation. New York: Columbia University Press, 1990.

Hanson, Charles H. "Ethics in the Business of Science." Ecology 57, no. 4 (1976): 627-28. Haraway, Donna. Primate Visions: Gender, Race and Nature in the World of Modern Science. New

York: Routledge, 1989. Heatherman, William. email, September 2 2005. Hecht, Susanna. "Environment, Development and Politics: Capital Accumulation and the

Livestock Sector in Eastern Amazonia." World Development 13, no. 6 (1985): 663-84. Hey, Richard. "Fluvial Geomorphological Methodology for Natural Stable Channel Design."

Journal of the American Water Resources Assocation 42, no. 2 (2006): 357-72. Hilgartner, Stephen. Science on Stage - Expert Advice as Public Drama. Stanford, CA: Stanford

University Press, 2000. Jackson, Peter, Nicola Thomas, and Claire Dwyer. "Consuming Transnational Fashion in

London and Mumbai." Geoforum 38, no. 5 (2007): 908-24. Jasanoff, Sheila. The Fifth Branch: Science Advisors as Policymakers. first paperback ed.

Cambridge: Harvard University Press, 1990. Reprint, 1994. Jasanoff, Sheila, and Marybeth Long Martello, eds. Earthly Politics: Local and Global in

Environmental Governance. Edited by Peter Haas, Sheila Jasanoff and Gene Rochlin, Politics, Science & the Environment. Boston: MIT, 2004.

Johnson, P.A., M. Tessier, R.D. Hey, and D.L. Rosgen. "The Use of Vanes for Local Abutment Scour Reduction." Journal of Hydraulic Engineering 127, no. 9 (2001): 772-79.

Jordan III, William R. "Restoration, Community, and Wilderness." In Restoring Nature: Perspectives from the Social Sciences and Humanities, edited by Paul H. Gobster and R. Bruce Hull, 21-36. Washington, D.C.: Island Press, 2000.

Juracek, Kyle E., and Faith A. Fitzpatrick. "Limitations and Implications of Stream Classification." Journal of the American Water Resources Assocation 39, no. 3 (2003): 659-70.

Katz, Cindi. "Whose Nature, Whose Culture?: Private Productions of Space and the 'Preservation' of Nature." In Remaking Reality: Nature at the Millenium, edited by Bruce Braun and Noel Castree, 46-63. New York: Routledge, 1998.

Kloppenburg, Jack Ralph. First the Seed: The Political Economy of Plant Biotechnology. 2nd ed. Madison, WI: The University of Wisconsin Press, 2004.

221

Kondolf, G. Mathias. "Geomorphological Stream Channel Classification in Aquatic Habitat Restoration: Uses and Limitations." Aquatic Conservation: Marine and Freshwater Ecosystems 5 (1995b): 127-41.

_____. "Lessons Learned from River Restoration Projects in California." Aquatic Conservation: Marine and Freshwater Ecosystems 8 (1998): 39-52.

_____. "River Restoration in North America: Meandering Channels for All?" Paper presented at the Association of American Geographers, San Francisco 2007.

Kondolf, G. Mathias, M.W. Smeltzer, and S. Railsback. "Design and Performance of a Channel Reconstruction Project in a Coastal California Gravel-Bed Stream." Environmental Management 28, no. 6 (2001): 761-76.

Knorr-Cetina, Karin. The Manufacture of Knowledge: An Essay on the Constructivist and Contextual Nature of Science. Oxford: Pergamon Press, 1981.

Kohler, Robert E. Landscapes & Labscapes: Exploring the Lab-Field Border in Biology. Chicago: University of Chicago Press, 2002.

Krimsky, Sheldon. Science in the Private Interest. Lanham, MD: Rowman & Littlefield, 2003. Latour, Bruno. Science in Action: How to Follow Scientists and Engineers through Society. 10th ed.

Cambridge, MA: Harvard University Press, 1987. Reprint, 2002. _____. Pandora's Hope: Essays on the Reality of Science Studies. Cambridge, MA: Harvard

University Press, 1999. Latour, Bruno, and Steve Woolgar. Laboratory Life: The Social Construction of Scientific Facts.

Princeton: Princeton University Press, 1986. Lave, R., Morgan M. Robertson, and Martin W. Doyle. 2008. Why you should pay attention

to stream mitigation banking. Ecological Restoration 26 (4):287-289. Lavendel, Brian. "The Business of Ecological Restoration." Ecological Restoration 20, no. 3

(2002): 173-79. Leopold, Luna B., and T. Maddock. "The Hydraulic Geometry of Stream Channels and

Some Physiological Implications." 57. Washington, D.C.: U.S. Geological Survey, 1953.

Mansfield, B. "Rules of Privatization: Contradictions in Neoliberal Regulation of North Pacific Fisheries." Annals of the Association of American Geographers 94, no. 3 (2004): 565-84.

McCarthy, James. "Neoliberalism and the Politics of Alternatives: Community Forestry in British Columbia and the United States." Annals of the Association of American Geographers 96, no. 1 (2006): 84-104.

Miller, Dale E., Peter B. Skidmore, and Dale J. White. "Channel Design." 117: Washington Department of Fish and Wildlife, Washington Department of Ecology, Washington Department of Transportation, 2001.

Miller, Jerry R., and John B. Ritter. "An Examination of the Rosgen Classification of Natural Rivers." Catena 27, no. 295-299 (1996).

Mirowski, Philip, and Robert Van Horn. "The Contract Research Organization and the Commercialization of Scientific Research." Social Studies of Science 35, no. 4 (2005): 503-48.

Mitchell, Timothy. Rule of Experts: Egypt, Techno-Politics, Modernity. Berkeley: University of California Press, 2002.

Mondry, Zackary, Greg Melia, and Mac Haupt. Stability of Engineered Stream Structures in North Carolina Restoration Projects. http://www.bae.ncsu.edu/programs/ extension/wqg/sri/2006conference/ presentations.html. North Carolina Ecosystem Enhancement Program, 2006 [cited 4/28/08 2008].

222

Montgomery, David R., and John M. Buffington. "Channel Processes, Classification, and Response." 13-42, 1998.

Mosse, David. Cultivating Development: An Ethnography of Aid Policy and Practice. London: Pluto Press, 2005.

NRC. Restoration of Aquatic Ecosystems: Science, Technology, and Public Policy. Washington, D.C.: National Academy Press, 1992.

_____. "Planning Workshop for an Nrc Study on Stream Restoration." 4. Washington, D.C.: NRC Water Science and Techology Board, 2003a.

_____. "Assessment of the Science and Practice of Stream Restoration." 8. Washington, D.C.: NRC Water Science and Technology Board, 2003b.

NRCS. "Stream Restoration Design. National Engineering Handbook Part 654." U.S. Department of Agriculture National Resources Conservation Service, 2007.

Nelkin, Dorothy. "Ecologists and the Public Interest." The Hastings Center Report 6, no. 1 (1976): 38-44.

_____. "Scientists and Professional Responsibility: The Experience of American Ecologists." Social Studies of Science 7, no. 1 (1977): 75-95.

O'Neill, Karen M. Rivers by Design: State Power and the Origins of U.S. Flood Control. Durham, NC: Duke University Press, 2006.

Ong, Elisa K., and Stanton A. Glantz. "Constructing "Sound Science" and "Good Epidemiology": Tobacco, Lawyers, and Public Relations Firms." American Journal of Public Health 91, no. 11 (2001): 1749-57.

Pain, Kathy. "Spaces of Practice in Advanced Business Services: Rethinking London – Frankfurt Relations." Environment and Planning D: Society and Space advance online publication (2007).

Porter, Theodore M. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life. first paperback ed. Princeton, NJ: Princeton University Press, 1995.

Prudham, W. Scott. Knock on Wood: Nature as Commodity in Douglas-Fir Country. New York: Routledge, 2005.

Raffles, Hugh. In Amazonia: A Natural History. 1st ed. Princeton, NJ: Princeton University Press, 2002.

Riley, Ann L. Restoring Streams in Cities: A Guide for Planners, Policymakers, and Citizens. Washington, D.C.: Island Press, 1998.

Robertson, Morgan M. "No Net Loss: Wetland Restoration and the Incomplete Capitalization of Nature." Antipode 32, no. 4 (2000): 463-93.

_____. "The Neoliberalization of Ecosystem Services: Wetland Mitigation Banking and Problems in Environmental Governance." Geoforum 35, no. 3 (2004): 361-73.

_____. "The Nature That Capital Can See: Science, State and Market in the Commodification of Ecosystem Services." Environment and Planning D: Society and Space 24, no. 3 (2006): 367-87.

_____. "Looking for Price in All the Wrong Places: The Work of Commodity Definition and Price under Neoliberal Environmental Policy." Antipode (in press).

Robbins, Paul. "Authority and Environment: Institutional Landscapes in Rajasthan, India." Annals of the Association of American Geographers 88, no. 3 (1998): 410-35.

Roper, Brett B., John M. Buffington, Eric Archer, Chris Moyer, and Mike Ward. "The Role of Observer Variation in Determining Rosgen Stream Types in Northeastern Oregeon Mountain Streams." Journal of the American Water Resources Assocation 44, no. 2 (2008): 417-27.

Rosgen, David L. "A Classification of Natural Rivers." Catena 22, no. 3 (1994): 169-99.

223

_____. Applied River Morphology. 2nd ed. Pagosa Springs, CO: Wildland Hydrology, 1996a. _____. email, July 30 2005. _____. "The Application of Stream Classification Using the Fluvial Geomorphology

Approach for Natural Channel Design: The Rest of the Story." Paper presented at the ASCE Environmental & Water Resources Institute, Omaha, NE 2006a.

_____. "The Natural Channel Design Method for River Restoration." Paper presented at the ASCE Environmental & Water Resources Institute, Omaha, NE 2006b.

_____. "The Rosgen Geomorphic Approach for Natural Channel Design." In USDA-NRCS Stream Restoration Design Handbook, edited by Jerry Bernard, Jon Fripp and Kerry Robinson, 2007.

Sayer, Andrew. "(De)Commodification, Consumer Culture, and Moral Economy." Environment and Planning D: Society and Space 21, no. 3 (2003): 341-57.

Sayre, Nathan F. Ranching, Endangered Species, and Urbanization in the Southwest: Species of Capital. Tucson, AZ: University of Arizona Press, 2002.

Sear, D. A. "River Restoration and Geomorphology." Aquatic Conservation: Marine and Freshwater Ecosystems 4 (1994): 169-77.

Shabman, L.A., P. Scodari, and D. King. "National Wetland Mitigation Banking Study: Expanding Opportunities for Successful Mitigation: The Private Credit Market Alternative." Alexandria, VA: US Army Corps of Engineers Institute for Water Resources, IWR 94-WMB-3, 2004.

Shapin, Steven, and Simon Schaeffer. Leviathan and the Air Pump: Hobbes, Boyle and the Experimental Life. 2nd paperback ed. Princeton, NJ: Princeton University Press, 1985. Reprint, 1989.

Shields, F. Douglas, A. Brookes, and Jeffrey P. Haltiner. "Geomorphological Approaches to Incised Stream Channel Restoration in the United States and Europe." In Incised River Channels, edited by Steven Darby and Andrew Simon, 371-94. New York: John Wiley & Sons, 1999.

Shields, F. Douglas, Ronald R. Copeland, Peter C. Klingeman, Martin W. Doyle, and Andrew Simon. "Design for Stream Restoration." Journal of Hydraulic Engineering 129, no. 3 (2003): 575-84.

Shields, F. Douglas, and Ronald Copeland. "Empirical and Analytical Approaches for Stream Channel Design." Paper presented at the Federal Interagency Sediment Conference, Reno, NV, April 2-6 2006.

Shields, F. Douglas, Ronald W. Copeland, Peter C. Klingeman, Martin W. Doyle, and Andrew Simon. "Chapter 9. Stream Restoration." In Manual 54 Update, in press.

Simon, Andrew. "Flow Energy, Time, and Evolution of Dynamic Fluvial Systems: Implications for Stabilization and Restoration of Unstable Systems." Paper presented at the 2006 World Environmental and Water Resources Congress, Omaha, Nebraska, May 21-25 2006.

Simon, Andrew, and Janine Castro. "Measurement and Analysis of Alluvial Channel Form." In Tools in Fluvial Geomorphology, edited by G. M. Kondolf and Hervé Piégay, 291-322. Chichester: John Wiley & Sons, 2003.

Simon, Andrew, Martin W. Doyle, G. M. Kondolf, F. Douglas Shields, B. Rhoads, Gordon Grant, Faith A. Fitzpatrick, Kyle E. Juracek, Munsell McPhillips, and James MacBroom. "How Well Do the Rosgen Classification and Associated "Natural Channel Design" Methods Integrate and Quantify Fluvial Processes and Channel Response?" Paper presented at the ASCE EWRI, Anchorage, AL 2005.

224

Simon, Andrew, and E.J. Langendoen. "A Deterministic Bank-Stability and Toe-Erosion Model for Stream Restoration." Paper presented at the Environmental and Water Resources Congress, Omaha, Nebraska, May 21-25 2006.

Simon, Andrew, Martin W. Doyle, G. M. Kondolf, F. Douglas Shields, B. Rhoads, and Munsell McPhillips. "Critical Evaluation of How the Rosgen Classification and Associated "Natural Channel Design" Methods Fail to Integrate and Quantify Fluvial Processes and Channel Response." Journal of the American Water Resources Assocation 43, no. 5 (2007): 1-15.

Sismondo, Sergio. "Intersections of Pharmaceutical Research and Marketing." Special Issue of Social Studies of Science 34, no. 2 (2004).

Slate, Louise O., F. Douglas Shields, John S. Schwartz, Donald D. Carpenter, and Gary E. Freeman. "Engineering Design Standards and Liability for Stream Channel Restoration." Journal of Hydraulic Engineering (2007): 1099-1102.

Smith, Neil. "Nature at the Millenium: Production and Re-Enchantment." In Remaking Reality: Nature at the Millenium, edited by Bruce Braun and N. Castree, 271-85. New York: Routledge, 1998.

Smith, Sean M., and Karen L. Prestegaard. "Hydraulic Performance of a Morphology-Based Stream Channel Design." Water Resources Research 41, no. W11413 (2005).

Soar, P.J. "Channel Restoration Design for Meandering Rivers." doctoral dissertation, University of Nottingham, 2000.

Star, Susan Leigh. Regions of the Mind: Brain Research and the Quest for Scientific Certainty. 1st ed. Stanford, CA: Stanford University Press, 1989.

Star, Susan Leigh, and James B. Griesemer. "Institutional Ecology, "Translation," and Boundary Objects: Amateurs and Professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39." Social Studies of Science 19, no. 3 (1989): 387-420.

Sullivan, Sian. "Getting the Science Right, or Introducing Science in the First Place? Local 'Facts', Global Discourse -- 'Dessertification' in North-West Namibia." In Political Ecology: Science, Myth and Power, edited by Philip Stott and Sian Sullivan, 15-44. New York: Arnold, 2000.

Taylor, Peter. "Appearances Notwithstanding, We Are All Doing Something Like Political Ecology." Social Epistemology 11, no. 1 (1997): 111-27.

Thompson, Douglas M., and Gregory N. Stull. "The Development and Historic Use of Habitat Structures in Channel Restoration in the United States: The Grand Experiment in Fisheries Management." Geographie physique et Quaternaire 56, no. 1 (2002): 45-60.

Thompson, Douglas M. "The History of the Use and Effectiveness of Instream Structures in the United States." In Humans as Geologic Agents, edited by J. Ehlen, W.C. Haneberg and R.A. Larson, 35-50. Boulder, CO: The Geological Society of America Reviews in Engineering Geology, 2005.

Throop, William, ed. Environmental Restoration: Ethics, Theory, and Practice. Edited by Roger S. Gottlieb. first ed, Flashpoints. Amherst, NY: Humanity Books, 2000.

Todd, Jeffrey W. "The Wildlife Biologist in the Minerals Industry." Wildlife Society Bulletin 8, no. 1 (1980): 61-63.

Traweek, Sharon. Beamtimes and Lifetimes. First Paperback ed. Cambridge: Harvard University Press, 1988. Reprint, 1992.

Turner, Matt. "Conflict, Environmental Change and Social Institutions in Dryland Africa." Society and Natural Resources 12 (1999).

225

Watts, Michael J., and James McCarthy. "Nature as Artifice, Nature as Artefact: Development, Environment and Modernity in the Late Twentieth Century." In Geographies of Economies, edited by Roger Lee and Jane Wills, 71-86. London: Arnold, 1997.

White, R.J., and O.M. Brynildson. "Guidelines for Management of Trout Stream Habitat." 66. Madison, WI: Department of Natural Resources, 1967.

Wild and Scenic Rivers Act. P.L. 90-542. Wilcock, Peter R., J.J. Clark, Derek B . Booth, Janine Castro, Craig Fischenich, P. Grams,

Karen Gran, J. Steven Kite, G. M. Kondolf, J.D.G. Marr, James MacBroom, J. Moore, R. Newbury, M.A. Palmer, M.E. Power, J.C. Schmidt, F. Douglas Shields, and V. Voller. "Stream Restoration Training: Status and Proprosal for a New Approach." Journal of the American Water Resources Assocation (in press).

Wolman, M. Gordon, and J. P. Miller. "Magnitude & Frequency of Forces in Geomorphic Processes." Journal of Geology 68 (1960): 57-74.

Wolman, M. Gordon, and Ran Gerson. "Relative Scales of Time and Effectiveness of Climate in Watershed Geomorphology." Earth Surface Processes 3 (1978): 189-208.

226

Appendix A: Interview and Survey Meta-Data A. Semi-Structured Interview Subjects I conducted semi-structured interviews with the following agency staff, consultants, professors, and non-profit staff. This list does not include subjects interviewed completely off-the record. Brian Bledsoe, Colorado State University Margaret Bowman, American Rivers Karin Boyd, Applied Geomorphology,

Inc. Syd Brown, California Department of

Parks & Recreation John Buffington, USFS Janine Castro, USFWS Jock Conyngham, USACE Brian Dietterick, California Polytechnic

Institute San Luis Obispo Peter Downs, Stillwater Sciences Martin Doyle, University of North

Carolina-Chapel Hill Craig Fishenich, USACE Ted Geier, USFS Scott Gillilan, Gillilan Associates, Inc. Steve Gough, Little River Consulting Brian Graber, MA Waterways Karen Gran, National Center for Earth-

Surface Dynamics Angela Greene, NRCS/Canaan Valley

Institute Jeffrey Haltiner, Philip Williams

Associates William Harman, Buck

Engineering/Michael Baker Corporation

Cheryl Harrelson, Steady Stream Hydrology

Richard Hey, University of Birmingham Rollin Hotchkiss, Brigham Young

University Greg Jennings, North Carolina State

University Timothy Keane, Kansas State University Steven Kite, West Virginia University Greg Koonce, Inter-Fluve, Inc.

Eric Larsen, University of California at Davis

Daniel Levish, Bureau of Reclamation James MacBroom, Milone & MacBroom Scott McBain, McBain & Trush Dale Miller, Mainstream Restoration, Inc. David Montgomery, University of

Washington Gary Parker, University of Illinois John Potyondy, USFS Rhonda Reed, CALFED Bay-Delta

Program/NOAA Ann Riley, San Francisco Bay Regional

Water Quality Control Board David Rosgen, Wildland Hydrology Conor Shea, USFWS Doug Shields, USDA- ARS National

Sedimentation Laboratory Andrew Simon, USDA-ARS National

Sedimentation Laboratory Peter Skidmore, The Nature

Conservancy/Skidmore Restoration Consulting LLC

Louise Slate, Stewart Engineering Sean Smith, Maryland Department of

Natural Resources Robbin Sotir, Robbin Sotir and Associates Tracy Sylte, USFS Vaughan Voller, University of Minnesota Kris Vyverberg, California Department of

Fish & Game Peter Whiting, Case Western Reserve

University Peter Wilcock, Johns Hopkins University Jim Wilcox, Plumas County CDC Laura Wildman, American Rivers Philip Williams, Philip Williams Associates Ellen Wohl, Colorado State University

227

Because these interviews were semi-structured, the questions I asked varied from person to

person. However, there were some questions I asked everyone I interviewed, and other

questions that I asked everyone in a particular group (pro-Rosgen consultants, anti-Rosgen

agency staff, etc.)

1. List of common questions asked of all subjects, time permitting: Please give me a brief overview of your education and any previous restoration-related

jobs. When and how did you get interested in restoration? When and in what context did you first hear of Dave Rosgen or the Rosgen

Method/Natural Channel Design? What has your experience with Rosgen and his work been since then? What do you see as the strengths and weaknesses of the classification system and design

approach? What do you believe is at stake in the debate over Rosgen’s work? What trajectory do you think the Rosgen Method/Natural Channel Design is on: rising,

holding steady, declining in influence? What do you see as the best way to educate restoration practitioners? What do you believe is the current state of restoration science in terms of the relative

certainty of practice? Who should I talk to next? 2. Questions for all staff at consulting firms, public agencies, and non-profits: What is your role at __________ and how does it relate to restoration?

3. Questions for all academics: How have you learned about the content of Rosgen’s work? In your classes or short courses do you say anything about Rosgen’s work? Have any of your students ever been considered unqualified for a job because they had

not been through Rosgen’s training? 4. Questions for all consultants: Can you apply the Rosgen classification system? Could you walk me through your design approach? 5. Questions for all anti-Rosgen consultants: Have you ever lost a job because of your refusal to use Rosgen’s approach?

228

B. Structured Interviews

I conducted structured interviews about restoration practice in North Carolina with the

following subjects:

Tony Able, EPA Charles Anderson, NRCS/Pilot View Resource Conservation & Development

District Jeff Bruton, Department of Water Resources, Water Resources Development Grant

Program Marella Buncick, USFWS Asheville Field Office John Cox, Durham Stormwater Engineering Trish D’arconte, Chapel Hill Stormwater Utility Brady Dodd, USFS Dick Everhart, NRCS/Surry County Soil & Water Conservation District Tim Garrett, Southwestern Resource Conservation & Development District Tom Gerow, North Carolina Division of Forest Resources Anita Goetz, USFWS Asheville Field Office Keith Huff, Winston-Salem Stormwater Division Gary Jordan, USFWS Raleigh Field Office Jennifer Krupowicz, Charlotte Mecklenberg Stormwater Utility Eric Kulz, Division of Water Quality Andrea Leslie, Ecosystem Enhancement Program Western Branch Callie Moore, Hiwassee River Watershed Coalition Jeff Parker, Transyvlania County Soil & Water Conservation District David Phlegar, Greesnboro Stormwater Management Alan Walker, NRCS

I asked all of these subjects the following set of questions:

1. In what ways are your agency/office involved in stream restoration or channel stabilization work: funding, managing, designing, etc. 2. Approximately when did your agency first become involved in stream restoration/channel stabilization projects? 3. Why did your agency become involved in stream restoration/channel stabilization projects? 4. Approximately how many projects are you involved in per year?

229

5. Have you heard of the Rosgen Method of Natural Channel Design, and if so, in what context? 6. Do you utilize Natural Channel Design in your work? (If yes, skip to Q.8, below) 7a. Did your agency use Natural Channel Design in the past? If so, when, and why did you abandon it? 7b. What approach do you use instead? 8. Why did you select that approach? 9. What are its strengths and weaknesses for your purposes, in your day-to-day work? 10. Have you or other staff in your agency attended Rosgen training courses? If so, who, and to what level? 11. At the local scale, which of the following agencies or groups in your area are involved in restoration work? (Please check all that apply. Contact suggestions would be greatly appreciated!)

City public works or planning departments. Suggested contacts: County public works or planning departments. Suggested contacts: Soil & Water Conservation Districts. Suggested contacts: Non-profit or watershed groups. Suggested contacts: Other.

C. Mail Surveys of Short Course Participants

I sent the following survey questions to all students in the Rosgen Level I course I attended

in Santa Cruz, California in January 2005, and to all of the students in the Academic Level I

short course I attended in Bishop, California in October 2004.

1. Name: 2. When and in what context did you first hear of fluvial geomorphology? 3. Why did you choose this short course? 4. What were your goals for the course?

230

5. Did the course meet those goals? How so? 6. Based on the information presented in the course, how predictable or unpredictable do you think restoration work is? (please circle the number on the scale below that best applies) Unpredictable Moderately

unpredictable Moderately predictable

Predictable

1 2 3 4 5 6 7 8 9 10 7. Have you used any of the knowledge or skills gained in this course? If yes, please describe. 8. Education: please list any degrees you have completed and the dates when they were earned. 9. Have you taken any restoration-related short courses before or since this course? If yes, please list the courses and their approximate dates. 10. Do you plan to take any additional restoration-related courses? If so, please list the courses you are considering. 11. What is your current role in restoration projects? Designer Funder Project Manager Other (please explain) 12. Do you work for a: Consulting firm Government agency Non-profit Other (please explain)

231

Appendix B: University-based Stream Restoration Certificate Programs

` Despite the difficulties of starting new academic programs outlined in Chapter 4,

since 2000 three universities (West Virginia University, the University of Minnesota, and

Portland State University) have started restoration certification programs. All of these

programs are new enough that it is not yet possible to draw conclusions about their success

or the career trajectories of their students. However, given the range of difficulties

enumerated above, the existence of these programs is interesting in and of itself. Why was

one able to get off the ground and another not? What are the trade-offs of a professional

program vs. an academic program?

A. The University of Minnesota: it’s easier if it’s free1

The stream restoration program at the University of Minnesota matriculated its first

class of students in the fall of 2006. It is a one-year program that awards a post-bac

certificate in stream restoration, and is open both to enrolled graduate students and to

people outside the university who already have BAs. Fifteen students enrolled its first year,

4/5 of whom were already graduate students, perhaps because advertising for the new

program was primarily via word of mouth as it was not formally approved until February

2006. The program consists of mandatory introductory and capstone courses. In between

students are required to take three to four courses selected from a large roster of relevant

1 The information in this section is drawn primarily from interviews with Dr. Vaughan Voller, University of Minnesota professor of civil engineering and NCED principal investigator who spearheaded the effort to develop the certificate program, and with Karen Gran, a post-doctoral student who was at the time serving as the certificate program’s primary staff person.

232

courses already offered at the University. The allowed courses are grouped into four

concentrations: river and floodplain science and engineering (all students must take at least

one course in this concentration), river and floodplain ecology, water quality, and water

policy and management. While students are required to pass the courses they take, there is

no comprehensive final exam associated with the University of Minnesota certificate

program.

There are a number of factors responsible for the successful development of this

certificate program, some of them specific to the University of Minnesota, others not. One

clear factor was that the program proposal was developed by Vaughan Voller, a professor

with substantial experience in both writing and reviewing such proposals. Voller was able to

get the program approved in only one year at least in part because of knowing his way

around the university bureaucracy and the types of things it looks for in such proposals. He

was also savvy enough to make the rounds and drum up broad faculty support before he

submitted the first draft.

I’m familiar [with the proposal process] from having been on committees for previous proposals. You try to clear as many hurdles as you can before you actually put the piece of work in. So I tried to contact, first of all, as many people on campus that I thought were interested or I knew were interested in stream restoration and made sure that they looked at the development of the idea and the development of courses. And I got lots of good ideas about courses that should be added, courses that should be dropped, and also the balance they would expect. At UM we have a Water Resources Science program, and I was very conscious not to be treading on their toes, because it’s clearly something that they could be or would be interested in in the future. Luckily I worked very closely with Ray Newman, the Dean of Graduate Studies for WRS, and he was very instrumental in helping me define and confine the courses that we chose for the program. So it was a kind of consensus-building thing. I was very conscious that I had to build consensus before I went to the graduate school. And even as it was, there were still things that had to be put right. But having had a good consensus base, things went much smoother.2

2 Author interview, Dr. Vaughan Voller, University of Minnesota St. Paul, 9/19/06.

233

A second primary driver, alluded to above, was the considerable intellectual

resources of the University of Minnesota system. Voller was able to compile a substantial

list of existing courses relevant to stream restoration that were already being taught. Thus

only two new courses had to be developed for the certification program: an introductory

seminar, and a hands-on capstone course at the end.

A third key piece seems to be not over-reaching. When I asked Voller why they had

tried for a certificate program instead of a full-fledged MA program, he argued that

appropriate limits on ambition were critical to program approval:

I’m of the philosophy that you want to walk before you can run…. If we did an MS degree, we’d probably have to create about four to seven new courses specifically directed at stream restoration-type activities. In order to get the thing going forward, the certificate program is seen in the first place as a kind of litmus test to see how much interest there is, and in the second place it allows a slower, more methodical pace for development of courses. Because the way in which it’s structured we do have two new courses: an introduction to stream restoration and a capstone experience that they’ll go out and do at the end. But the courses that lie in between those two endpoints are already on the books at UM in terms of courses in ecology, geology, civil engineering. There’s content in these courses that are clearly relevant to stream restoration, so that’s why we structured it in this way…. [I]t’s biting off more than we can chew maybe to go after the MS degree in one shot. I think down the line, especially if we keep those [admissions] numbers up for 3 or 4 years, then I would see it as becoming a full MS track, and then we might have more courses particularly developed for, specific courses in that area. (op. cit.)

While these factors of experience, existing resources, and limited ambition were

clearly very important to getting the certificate program approved, the primary driving force

behind the program was the National Center for Earth-Surface Dynamics (NCED), which is

housed at the University of Minnesota. NCED provides the organizational focus for the

program (though the Department of Civil Engineering is the official host); Voller wrote the

proposal in his capacity as an NCED PI. Perhaps even more importantly, NCED funds a

234

part time staff member to administer the program, as well as the introductory and capstone

courses.

Thus the University of Minnesota program seems to have had several key things

going for it: an administratively experienced proponent, a strong base of existing courses to

pull from, a moderate level of ambition, and the intellectual and financial support of a major

national research center.

B. Portland State University: location, location, location

In contrast to the University of Minnesota, the stream restoration program at

Portland State University (PSU) confers a professional certificate rather than a formal

academic degree, although it is also possible to take the courses for academic credit. Because

it is intended to improve multi-disciplinary collaboration rather than to train practitioners

from scratch, the program is targeted at professionals who already have degrees relevant to

stream restoration. As Janine Castro, a geomorphologist with the USFWS and one of the

four people running the program, described it, rather than teach the basic science underlying

the various disciplines involved in stream restoration,

the intent of the program is … to bring the engineer in and say: this is stream ecology, this is what biologists bring to the table. To show people all the different players, and take them through restoration all the way through design and monitoring. That’s the intent. We have a separate stream ecology class, but the ecologists still have to take it because they need to see the other folks in the class, how they interpret and what they know and don’t know. For engineers taking the design class, hopefully they’ll gain a better understanding of how people from different disciplines understand the science. That’s what we’re shooting for.3

Instead of drawing on existing courses at the university the PSU program is based on

a short course model, drawing on experienced local consultants and agency staff as faculty.

3 Author interview, Dr. Janine Castro, USFWS, 8/16/06.

235

According to its organizers, it took four years to develop the program proposal and get it

approved. The program has two tracks: a four-week intensive summer program, or a series

of five courses, each lasting between 2.5 and 4.5 days. The two tracks cover the same broad

range of material: stream geomorphology and hydrology, stream ecology, assessment tools

and techniques (such as classification, modeling, and stability assessment), restoration design,

and project management and team building. In addition, participants in either track of the

certificate program are required to take two elective short courses at PSU or another

accredited institution. Unusually, the instructors are all volunteers, which has allowed the

organizers to keep tuition costs quite low ($495 - $825 per course, or just $3,100 for the

summer institute), and accessible to non-profit staff.4

Unlike most professional certification programs, the PSU courses are all pass/fail

with graded in-class, field, and group exercises and presentations. The course instructors

seem to be taking the evaluations seriously: according to Castro, in 2006 – 2007, the first

year of the program, two or three students failed each class because of low grades or poor

attendance. And because course work focuses on actual restoration projects, with results

presented to real clients, the bar for course work is set quite high.

So far, the PSU program seems to be quite successful if measured in terms of course

attendance. Eleven students completed all the program requirements in the first year and

received the certificate, and many more were poised to receive the certificate in 2007.

Courses, which are capped at 30, have filled consistently and some even have waiting lists.

There are several factors that are likely responsible for the successful launch of the

PSU program in Fall 2006. First, Portland State University is the largest provider of

continuing education programs in Oregon. There was already a structure in place for

4 http://epp.esr.pdx.edu/

236

professional certificate programs in general, and environmentally-oriented certificate

programs in particular. Thus Castro and her co-founders were not trying to force a square

peg into a round hole. Second, the current Provost at PSU is a hydrologist, so the certificate

program organizers felt sure of a positive audience for their proposal. Third, the tuition

costs for the program cover all of PSU’s administrative costs, and none of the instructors are

tenure-track faculty, so the program does not cost the university in terms of time or money.

Fourth, the Pacific Northwest is a hotbed of restoration activity driven in large part by

Endangered Species Act compliance on behalf of salmonids. According to the National

River Restoration Science Synthesis, more than half of all restoration projects conducted in

the U.S. were located there (Bernhardt et al. 2005). Thus the program organizers were more

or less guaranteed a student base simply because of their location.

C. University of West Virginia5

The first university-based certificate program in the U.S. was based at West Virginia

University (WVU). In its original incarnation, it ran from 2002 to 2004 and consisted of a

series of five week-long short courses. They were targeted at both WVU students and

professionals in the restoration field, particularly staff from the state Department of

Highways, which effectively sponsored the short course series. While modeled on Rosgen’s

courses, WVU’s short course series was in some ways quite different, most notably in

focusing on the specificity of conditions in West Virginia and in devoting substantial

amounts of time to ecology, including an entire course devoted to aquatic habitat

5 The information in this section is drawn from an interview with Dr. Steven Kite, West Virginia University.

237

assessment. As in Rosgen’s courses, there were no exams or other assessment of students’

understanding of course material; certificates were awarded based on attendance, although

this is something the course directors had decided to change had the courses continued.

The certificate itself was not a formal academic degree or professional certificate. As such, it

did not require formal recognition from the university as a new degree program. However,

students were offered course credit at WVU.

The initial courses in the five course WVU series introduced the geomorphology and

ecology of streams, with the first concentrating in the lecture hall and the second in the field.

In each case, the course material combined general principles of geomorphology, geology,

and ecology with particular attention, “paid to geological history, on-going processes, land

uses (especially urbanization, mining, forestry, and agriculture), and restoration goals unique

to Mid-Atlantic Highland rivers and streams.” (Riverstone Academy 2007) The third course

in the series focused exclusively on techniques for assessing the habitat value of streams, and

the fourth and fifth courses focused on design of restoration projects.

The motivation to start the short course series came from Ron Fortney, a

charismatic professor in the Civil and Environmental Engineering Department, who

attended a Rosgen course and was inspired to address the intense impacts on West Virginia

streams from coal mining and the resulting flood control projects. Another faculty member,

Todd Petty in the Department of Fisheries Management, joined Fortney in getting the

program started, with substantial assistance from Steve Kite, a faculty member in the

Department of Geography and Geology. The major factor that enabled them to start the

short course series was sponsorship from the West Virginia Department of Highways, which

helped to fund the courses and sent many staff members to take them. When Department

238

of Highways funding dried up, courses had to be cancelled because of insufficient

enrollment.

In 2007, the series restarted under the auspices of Riverstone Academy, the

educational arm of the Canaan Valley Institute, a prominent restoration NGO that also

sponsors Rosgen’s short courses in West Virginia. Only the first three of the WVU courses

are included in the new Riverstone Academy program; the design courses were discontinued.

The relationship between the Rosgen courses offered by the Riverstone Academy and the

WVU courses is, unsurprisingly, somewhat complicated. The course catalogue offers the

following caveat for the first two courses:

WVU/CVI Natural Stream Restoration Cooperative workshops may serve as a short-term remedial step for individuals who require immediate understanding of natural stream process before they advance to more complex, applied stream assessment and design courses. These workshops will strive to provide a broad understanding of stream science for those who play a limited or indirect role in stream restoration, but need to understand how their efforts fit into a larger watershed management process. (Riverstone Academy 2007)

And while the WVU courses accept Rosgen’s Level I course as an alternative pre-requisite

for their second course, the relationship is not reciprocal.

239

Appendix C: Additional Critiques of Rosgen’s Classification System

Here are four additional common critiques of Rosgen’s classification system. First,

many critics object to claims that Rosgen’s classification system produces results that are

objective and repeatable. For example, Cheryl Harrelson, a restoration consultant and

Rosgen supporter, pointed to subjective elements of the NCD approach:

CH: [A]lthough it’s quantitative, it has a subjective element that requires a degree of experience. And that experience needs to be experience in observing and being an observer. And I think it can be misapplied if someone doesn’t take the time to acquire or generate good observation skills.

RL: What would you point to specifically as the subjective element? Would that be identifying the bankful channel?

CH: That’s one thing that is disputed for sure. I think the subjectivity is more related to [the fact that] all streams change, and you can have a lot of complex situations where you have types inside of types, or types that change in very short segments. So do you break it into reaches or do you write down every one of those sections? … [The subjectivity is in] how you use it.1

But is this subjectivity a weakness particular to Rosgen’s system? The few STS

studies of field science (Latour 1999, Kohler 2002, Raffles 2002) have convincingly

demonstrated the subjectivity inherent in even the most seemingly objective field measures.

Thus this objection seems to hold, but without substantially weakening Rosgen’s position.

Other critics argue that while Rosgen mentions impacts from differences in geology

in a few places in Applied River Morphology, the classification system largely ignores them. It is

certainly the case that there is no step in the classification process devoted to evaluating

geology. There is implicit consideration inasmuch as geology impacts channel form, but

there is no data as to whether this is sufficient, or instead reduces the utility of the

classification system. Because there has not been a large-scale study of the interrelation 1 Author interview, Cheryl Harrelson, Steady Stream Hydrology, 8/30/06.

240

between geology and Rosgen stream type, it is not possible to definitively assess this

critique.2

Other criticisms center on the format of the classification system itself: for many

geomorphologists, the alpha-numeric labels are a sticking point, seen as unintuitive and hard

to remember because they bear no relationship to the channel types described. These critics

argue that, for example, it is far more descriptive to classify a channel as a steep gradient

step/pool system than as an ‘A’ channel. As Peter Whiting, professor of Geomorphology at

Case Western University, put it:

I have to double check my little crib sheet because I don’t walk out to channels and mentally classify them in quite that way. So, if somebody is really into it and they say, “It’s a B3,” I go, “B3; what is that? Ok, I know what they’re saying.” Whereas if they simply said, “It’s a gravel bed channel with a low terrace and a planar bed,” I get what I believe is a more helpful process-based picture....3

Other Rosgen opponents, however, find the alpha-numerics catchy and easy to remember.

It thus seems to me that this issue is more a matter of personal style, and perhaps of

differences in training (alphabetical and alpha-numeric categorization systems are more

common in the USFS, where Rosgen spent his formative years, than in academia, where

most of his critics did).

Finally, some critics argue that the classification system’s categories do not reflect

significant natural breakpoints in sediment transport or energy dissipation regimes. As Matt

Kondolf, a vehement Rosgen critic, has argued repeatedly in lectures to restoration

practitioners, fellow academics, and the public, it is critical to distinguish between natural

2 At present, when selecting a reference reach, Rosgen directs his students to find a reach with a similar channel and valley type in the same hydro-physiographic province (a concept that explicitly includes geology). A 2006 article by Richard Hey in the Journal of the American Water Resources Association, however, argues that hydro-physiographic province is irrelevant. If Rosgen adopts his co-teacher and Ph.D. advisor’s position, geology may become even more peripheral to his work. 3 Author interview, Dr. Peter Whiting, Case Western Reserve University, 7/13/06.

241

classification systems (which reflect actual categorical differences in nature) and special

classification systems (which represent arbitrary breakpoints within data that is actually a

continuum). Kondolf claims that Rosgen’s system is a special system, and thus is a matter

of convenience rather than science. Far from disagreeing, in his 1996 textbook Applied River

Morphology Rosgen explicitly states that stream morphology is a continuum, although the

conclusions he draws from that fact are unsurprisingly different from Kondolf’s:

The stream morphology classification uses discrete classes for a suite of morphologic parameters to set parameters or prescribe intervals for categorizing stream types. Discrete categories are necessary to create consistent, reproducible rules for inclusion of a stream within a stream type. The cutoff values for the category intervals were developed from frequency histograms for each of the morphologic parameters. In reality, however, stream morphology displays a continuum of form. The physical processes that create continuous adjustments in river morphology do not permit rigid, arbitrary boundaries to exist in nature. Thus the stream morphology hierarchy also uses a companion set of subclasses to accommodate a continuous range of parameter values. (Rosgen 1996, pp.3-4 – 3-5)

In effect, Rosgen is arguing that his system is designed the way it is precisely because

of the necessarily special nature of any stream classification. Thus the issue is not whether

Rosgen’s system carves up a continuum, but whether the fact that it does matters, and that

can only be determined by how it is used in practice.

242

Appendix D. The ASCE River Restoration Committee Calls for Design Standards

As described in Chapter 2, the national manuals and handbooks on stream

restoration produced in the last fifteen years treat Rosgen’s classification system and design

approach in a range of different ways, from allotting NCD an entire chapter (NRCS 2007),

to implicit critique (Shields et al. 2003). The American Society of Civil Engineers (ASCE) is

the latest group to attempt to develop standards of practice for restoration, and the

passionate argument sparked by their initial effort illustrates how the tactics of Rosgen’s

opponents and supporters play out in practice.

ASCE’s River Restoration Committee has just over 90 members, comprised of

roughly equal groups of academics, consultants, and government employees (predominantly

federal, although there are a few state and local). The vast majority of the members are

based in the U.S., although there are academic members in Canada, the UK, and Australia.

Most of the River Restoration Committee members are registered engineers. The

Committee’s role within ASCE is, “To focus on broader hydraulic engineering and

morphological issues of river restoration for improved project planning and natural channel

design.” (http://www.ars.usda.gov/Research/docs.htm?docid=5525)

At the May 2005 ASCE EWRI conference in Anchorage, Alaska, discussion amongst

the Committee members revolved around the need for standards of practice for engineers

working on stream restoration, since they are in the unique position (shared only by

landscape architects) of being legally liable for the success or failure of their restoration

projects. To jump start the process of developing such standards, five members of the River

Restoration Committee drafted a paper (submitted to the Journal of Hydraulic Engineering)

243

intended to rally enthusiasm for what promises to be a long and challenging project, and to

explain why such standards are needed.1

The paper’s authors are an impressively experienced and well-educated group. The

lead author is Louise Slate, an engineer who runs the river restoration department for a

major consulting firm in North Carolina. She is Rosgen-certified (having completed all four

levels of Rosgen training), and has also taken the first course in the academic short course

series described in Chapter 4. Doug Shields has already been mentioned a number of times,

but to recap, he is a licensed engineer and also has a doctorate. He does applied research on

restoration at the National Sedimentation Laboratory, which is part of the Agricultural

Research Service, the research branch of the USDA. Shields has been heavily involved in

most of the major guidelines efforts to date. John Schwartz is a professor in the

Department of Civil and Environmental Engineering at the University of Tennessee. Like

Shields, he is a licensed engineer and holds a doctorate. Donald Carpenter is a professor in

the Civil Engineering Department at Lawrence Technological University in Michigan. The

fifth author, Gary Freeman, is a principal for the firm River Research & Design in Arizona.

He too is a registered engineer and holds a doctorate.

The draft paper was sent out to the entire River Restoration Committee for

comment on July 19, 2005. Over the following two months, in a debate consisting of nearly

90 email messages, Committee members critiqued the draft. More than a third of these

emails addressed the treatment of NCD in the draft. The authors and many of the

Committee members were startled by this, and a number of them mentioned it to me in

interviews. Yet the heated tone of some of the email messages is perhaps unsurprising given

1 As part of the email exchange that follows, the authors discovered that the process of developing officially-adopted ASCE standards is intensive, understandably so given the liability issues. Thus one of the issues currently under discussion is whether the process on which the River Restoration Committee is embarking will be to develop standards or the less stringent, but correspondingly less-seriously regarded, guidelines.

244

that the draft took the most explicitly critical stance towards NCD of any of the national

guidelines efforts to date. It is worth quoting at length from the initial version of the

introduction, as these two paragraphs seem to have triggered the strongest reactions:

Over the last ten to fifteen years a geomorphic-based alternative to the generally accepted engineering methodology has been developed. This geomorphic design methodology was originally developed from data sets on coarse bed, montane streams (Rosgen 1996). In general, the approach is empirically-based on concepts from fluvial geomorphology that emphasize form over process (Doyle 1999, Simon et al. 2005), and does not consider the dynamic geomorphic context of the project reach or the sediment budget (Kondolf et al. 2001). Form-based geomorphology assumes that channel form reflects current processes. However, at any given point in time, channel form reflects the chronological sequence of fluvial events layered upon each other with vegetative recovery and human-caused perturbations between flood events.

This geomorphic-based methodology for channel design has become very popular over a 10- to 15-year period, in large part because its simplicity has allowed non-engineers to practice stream restoration…. Problems arise when guidelines only require an empirically-based approach that assumes equilibrium conditions, which are clearly inappropriate in areas of changing flow regimes and sediment loads, such as urbanizing areas…. Project failures are of concern to engineers because some project sponsors have expressed interest in suing for damages and paying for repairs from liability insurance. Due to the current popularity of a form-based empirical methodology, engineers designing stream channels are often required by clients (who often lack engineering training) to use this approach even when the engineers question its applicability to a specific project and the risk-level associated with it. Furthermore, some guidelines promoted by state agencies advance the notion that an empirical approach is “standard” practice, often without mention of the underlying assumptions of constant flow and sediment supply. (Slate et al., 2006)

Even the ubiquitous communication aid consolation prize critics usually award Rosgen is

missing; this is unleavened critique.

The responses supportive of the draft cite a wide range of reasons why the NCD

approach is inadequate, including claims that it is out of step with accepted science; that it

does not adequately treat sediment transport; that it is not peer-reviewed; that there are no

pre-requisites or examinations for the short courses, and that therefore Rosgen certification

245

is meaningless; that study of form cannot explain process; and that it is not a universally-

applicable method. By far the most common critique, raised in 15 of the 23 messages

responding to the initial draft, was that clients require the use of the NCD approach rather

than allowing engineers to utilize the full range of tools accepted as standard practice. It was

this more than any other factor that spurred the call for guidelines that went beyond NCD.

David Williams, a member of the Committee who is a licensed engineer, holds a

Ph.D., and is the director of hydrology for a large international engineering company, tried

to explain the two sides to each other in the following words:

Much of the animosity on Rosgen's method is not necessarily the method itself, but rather how it is used. The engineer's side (and associates) recommend looking at various methods and pick the appropriate one(s) for use. The "other" side says use Rosgen or you don't get the job. The engineers see a forcing of their judgment and the "other" side can't see what the problem is because they "know" what is the best method. Rosgen does not help much in this manner because he doesn't speak out enough that there are other methods that should be considered - and does not discourage the disciples in their thought that you "don't know jack" unless you have gone through his courses. I have seen RFPs where they state that the Rosgen method must be used and you MUST have gone through the course (at least the basic one). I don't think that engineers and geomorphologists would mind it if it was suggested to use Rosgen's method, but [it] raises their hackles when told to. (Williams 2005)

While 23 messages explicitly or implicitly supported the authors’ argument, only

thirteen argued that the authors should revise their arguments, and of those a number were

from people who supported the authors’ position but found it politically untenable to use

the article as a frontal assault on Natural Channel Design. For example, Martin Matlock, a

professor at the University of Arkansas and a licensed engineer, wrote, “While I share the

group's concerns I would tone down the criticisms of this approach [Rosgen’s] and limit the

critique to dealing with dynamic complexity.” (Matlock 2005) Clearly, this is not a

fundamental challenge to the authors’ position.

246

When examined closely, only four messages came from people that seemed to be

whole-hearted Rosgen supporters, and most of their comments were general objections to

the tone of the draft rather than rebuttals of its primary claims. For example, Michael Clar,

an engineering consultant who employs Rosgen’s design approach extensively, wrote: “The

introduction comes off as a broadside against the geomorphic based design and Rosgen in

particular. I personally do not like to be involved with criticism of this nature, and I believe

that it takes away from the focus of the paper.” The one exception to this was the

comments submitted by Richard Hey. Hey intended his comments to fan the embers of

debate into a more spectacular conflagration, as he made quite clear in his email. He titled it,

“The Maverick Colorado Cowboy vs. The Rest,” and the text, which served as a cover letter

to an attached six page memo, said only this:

The discussion on the draft forum paper has identified my collaboration with Dave Rosgen.

Not one to stand idly by, attached are my thoughts on the wider issues raised.

Having lit the blue touch paper, I will retire to a safe distance. (Hey 2006)

In my interview with Hey, conducted after he sent this email, he was convinced that his

commentary did such a thorough demolition job on the draft paper that it could not be

submitted. In the end, all his match seems to have sparked is a flurry of email and limited

revisions, more whimper than bang as explosions go.

In his commentary, Hey begins by spending two pages attempting to discredit the

approach promoted by the authors and putting forward Rosgen’s NCD approach in its

place. Hey does so by describing his own intellectual journey from key developer of the

approach promoted by the authors, to discoverer of its inadequacies, to Rosgen supporter,

247

complete with citations.2 The alternative approach thus vanquished, he moves on to the

authors’ criticisms of Rosgen’s approach.

In the initial draft paper, the main critiques of NCD are that it:

1. was developed for coarse-bed montane streams (implying that it is not applicable elsewhere);

2. assumes that study of form allows adequate understanding of process; 3. neglects dynamic watershed-wide processes and sediment transport; 4. enables untrained people to practice design; 5. is inappropriate in areas where flows of water and sediment are changing (i.e. the

urbanizing areas where so many restoration projects are constructed); 6. is promoted as standard practice by many agencies despite these weaknesses, leading

to project failures for which engineers can be held legally-liable; and 7. is not peer-reviewed.

In his commentary, Hey argues that many of the critiques in the Slate et al. draft are

scientifically incorrect. This head-on rebuttal of the scientific content of critics’ assertions is

unusual; as discussed in Chapter 6, when Rosgen and his critics directly confront their critics

they typically limit themselves to pointing out their ignorance of Rosgen’s teachings. Most

strikingly, Hey argues against point 2, above – the claim that form-based approaches cannot

be used to diagnose process – by making the bold claim (very much contrary to current

scientific consensus) that identical channel forms can only be created by identical processes.

He addresses the argument that Rosgen neglects the watershed-scale forces that impact local

2 “In the 1960s it was considered that the morphology of alluvial rivers was indeterminate (Maddock, 1970). This view was challenged by Hey (1978, 1982) who showed that the 3D morphology of stable natural rivers could be determined given information on the stable controlling variables (boundary conditions) provided general, physically-based equations could be prescribed for each of the operative channel forming processes. This was referred to as the rational approach to channel design. As there are more unknowns than equations, a full solution is not currently possible. However the use of an extremal hypothesis or an empirical width equation will provide closure. The latter approach is being advocated as the design standard (Shields et al. 2003; Shields et al. in press). I am hopeful that this rational approach will eventually become the design standard. However, after evaluating its predictive capabilities, it is apparent that it cannot predict the morphology of natural rivers with sufficient accuracy for it to remain stable without heavy engineering support (Hey 2004). Different resistance and bed material transport equations produced very different results because they contain empirical calibration coefficients. None account for the effect of meandering on these processes and neither can they prescribe the morphology of riffles, runs, pools and glides. Empirical regime equations offer a possible alternative.” (Hey 2005)

248

reaches (point 3) by insisting that the most important of those forces will be those operating

in the immediate watershed, which are easily analyzable via the NCD approach.

In other cases, Hey uses some of the techniques described in Chapter 6. For

example, he addresses many of the critiques listed above by arguing that they are incorrect

characterizations of Rosgen’s work, spawned from the authors’ ignorance of it. He thus calls

their claims to scientific capital into question. Hey argues that the claim (part of point 3)

that Rosgen’s work does not adequately address the issue of sediment transport is out of

date, failing to reflect the focus on sediment produced by Rosgen’s collaboration with Hey.

Hey dismisses Point 4, the claim that Rosgen is enabling untrained people to practice design,

because Rosgen certification requires 400 hours of course work, which Hey claims is the

equivalent of an MS degree. Most tellingly, he ends his commentary by asserting the

authors’ ignorance in a stinging final paragraph:

To move the debate forward, it is imperative that it is based on fact rather than supposition. Those seeking to mount a robust offence against the fluvial geomorphological methodology should make the effort to fully familiarise themselves with the procedure and then compare its predictive capability against the rational approach they are advocating as an engineering standard of practice. This would establish whether they have a case to make. Prosecuting an argument without evidence in a robust manner is not only unethical, it is tantamount to a witch-hunt. The issue of engineer’s liability in this debate is a red herring as standard engineering procedures can, and should, be adopted to evaluate the design. Not surprisingly, I don’t recommend publication of the draft paper in its present state. (Hey 2005)

For Hey and other Rosgen supporters on the ASCE River Restoration Committee,

these arguments were decisive. The paper’s authors, however, merely toned down some of

the language in the first two paragraphs and soldiered onwards, so that:

the Committee would like to see regulatory agencies develop criteria for river restoration designs rather than guidelines or rules that force engineers to use a single design approach. By gearing designs to satisfy specified criteria, we will be able to use all of the relevant methods at our disposal. Objective-

249

based design approaches will be easier to justify than those resulting from trying to fit a prescribed design approach to all types of projects. (Slate et al. 2006, p.1. Strikethrough indicates language removed from the published version of the paper.)

became instead:

the Committee would like to see regulatory agencies adopt performance-based criteria for river restoration designs to meet specific objectives rather than guidelines or rules that constrain engineers to use a prescribed design approach. By gearing designs to satisfy specified, measurable criteria, engineers will be able to select the most appropriate design methods for a given project across a wide variety of boundary conditions and system processes. Tailoring a solution to satisfy quantifiable goals for the given stream and watershed conditions over the lifespan of the design makes more sense than trying to fit a prescribed design approach to all types of projects over a wide range of project and watershed scales. (Slate et al. 2007, p.1. Underlining indicates language added.)

These changes flesh out the authors’ point and tone down their language, while retaining the

main idea that clients specifying a single design approach is unacceptable.

Similarly, the paragraph that read:

This analog-empirical methodology, often referred to as the Rosgen or reference reach method, was developed from data sets on coarse-bed montane streams (Rosgen 1996). In general, this approach is based on concepts from fluvial geomorphology that emphasize form over process (Doyle 1999, Simon et al. 2005), and does not consider the dynamic geomorphic context of the project reach (Kondolf et al. 2001, Smith and Prestegaard 2005). Form-based geomorphology assumes that channel form reflects current processes. However, at any given point in time, channel form reflects the chronological sequence of fluvial events layered upon each other with vegetative recovery and human-caused perturbations between flood events (Ward et al. 2002). (Slate et al. 2006, p.2. Strikethrough indicates language removed from the published version of the paper.)

now reads:

This analog-empirical methodology is often referred to as the Rosgen or reference reach method (Rosgen 1996; Hey 2006). In general, this approach is based on concepts from fluvial geomorphology that emphasize form over process (Doyle 1999; Kondolf et al. 2001; Simon et al. 2005; Smith and Prestegaard 2005). Importantly, this essay is not a criticism of the reference reach approach per se; nor is it a debate of form versus process geomorphology. (Slate et al. 2007, p.3. Underlining indicates language added to the published version of the paper.)

250

The authors now address the Rosgen debate explicitly, and have removed almost all of the

points that critique it, but the list of critical references has been expanded!

Perhaps most interestingly, the discussion of why the current state of affairs is

unacceptable and standards of practice are needed has been expanded, and I believe it is

worth quoting at length:

The crucial concern to the civil engineering community involved with river restoration is increasing recognition of the reference reach approach as the standard of practice by state and other regulators. Professional engineers designing channels are often required by clients to use the Rosgen approach by name even when its applicability to a specific project and the associated risk level are in question. Moreover, current state guidelines for stream channel mitigation projects in North Carolina and Kentucky are based on Rosgen’s design approach coupled with USEPA Rapid Bioassessment Protocols (USACE Wilmington District et al. 2003; KY Division of Water 2002). If there are no references to engineering hydrology, hydraulics, sediment transport, open channel flow or explicit requirements to perform engineering analyses, then why is an engineer’s seal required on the drawings? What parts of these drawings are considered engineered and for what will the engineer be held liable? The issue is critical to the engineering community because project failures have occurred nationally, and the public, rightly or wrongly, often identifies liability with the engineer. (Slate et al. 2007, p.4)

For engineers, an agency’s insistence on NCD poses serious hazards.

The paper ends with a call for guidelines based on peer-reviewed science and general

scientific principles that may be a less explicit challenge to NCD than the initial draft, but is

still a critique that borders on dismissal.

This committee proposes that stream restoration channel design standards be based on measurable criteria supported by current peer-reviewed scientific publications. Design goals should be clearly defined and based on general physical principles, rather than referenced to an empirically defined equilibrium state…. By gearing designs to satisfy specified criteria, engineers will be able to use all of the relevant methods that have withstood the scientific rigors of peer-review, better manage risk, and improve project success. (Slate et al. 2007, pp.9-10)

251

If and when the ASCE River Restoration Committee manages to develop restoration

standards of practice for engineers, it seems likely that they will be by far the most critical of

Rosgen’s NCD approach of any of the national guidelines efforts to date.