introduction-by j.j.p. the manuscript, its publication, and its

JOURNAL OF APPLIED BEHAVIOR ANALYSIS

SELF-CONTROL TECHNIQUES OF FAMOUS NOVELISTS'

IRVING WALLACE

AUTHORS GUILD AND AUTHORS LEAGUE OF AMERICA

(WITH INTRODUCTION AND DISCUSSION BY J. J. PEAR)(UNIVERSITY OF MANITOBA)

A statement by a famous contemporary novelist is presented that indicates how he andothers, independently of formal behaviorism, used behavioristic methods-specifically,self-recording charts and regularly scheduled daily work hours-to accelerate and main-tain their writing outputs. On the basis of his statement and an analysis of his self-recorded data, it is argued that a meaningful and useful analogy can be drawn betweenwriting a novel and emitting a simple operant response on a fixed-ratio schedule.DESCRIPTORS: self-control, self-recording, self-reinforcement, fixed-ratio schedules,

novel writing, famous novelists

INTRODUCTION-by J.J.P.

The persistence of the successful novelistseems, superficially at least, to defy prevalentnotions of reinforcement. Novel writing in-volves a supreme effort whose ultimate primaryreinforcement, by any conventional standard,must be described as very long deferred. Whatreinforces the behavior before the completion ofthe manuscript, its publication, and its publicacclaim? And if the manuscript is rejected by allpotential publishers, what reinforces the behav-ior that produces subsequent manuscripts untilfinally one is published? The behaviorist can, ofcourse, point to a number of possible intermedi-ate sources of reinforcement, perhaps eventuallyto be subjected to an experimental analysis(Skinner, 1957, pp. 396-402). Nevertheless, itis clear that these sources are often insufficient.Novel writing is a rather uncommon endeavorand, when it is undertaken, appears quite sus-ceptible to extinction (Skinner, 1953, pp. 71-72).

If consulted by a client who wished to writea novel, who appeared to have an adequate be-

tPrepration of this article was supported by GrantNo. MA-5647 from the Medical Research Council ofCanada. Reprints may be obtained from J. J. Pear,Department of Psychology, University of Manitoba,Winnipeg, Manitoba, Canada R3T 2N2.

havioral repertoire for completing such a task,but who nevertheless seemed unable to carry itout, an informed behaviorist would initiatetreatment with techniques that are based on cur-rent laboratory findings and that have provensuccessful in presumably similar cases. A suit-able locale for performing the writing taskwould be specified with the objective of obtain-ing stimulus control over the desired behavior,regular and frequent (e.g., daily) periods for en-gaging in the behavior would be scheduled,measurable units of the behavior would be de-fined, and supplementary reinforcement wouldbe made contingent on completion of an appro-priate number of units within a given time pe-riod (Goldiamond, 1965). As to locale, the be-haviorist would naturally suggest a well-lightedprivate work area equipped with desk and neces-sary writing materials. For units, either words orpages written might prove to be both adequateand convenient. Concerning reinforcers, thereare many possibilities. But, from a practicalpoint of view, a simple chart showing the num-ber of words or pages written each day mightprovide sufficient supplementary reinforcementas well as data for monitoring the desired be-havior (Kazdin, 1974).

Thus, in some such manner, the behavioristwould probably treat the problem as one of effi-ciently maintaining ratio-schedule performance

515

1977, 10. 515-525 NUMBER 3 (FALL) 1977

IRVING WALLACE and J. J. PEAR

.e., behavior whose reinforcement dependsdirectly on the quantity of behavior emitted(Ferster and Skinner, 1957). An analogy can bedrawn from Findley and Brady's (1965) study,in which a chimpanzee's button-pushing per-formance was maintained more efficiently on ahighly demanding fixed-ratio schedule whenconditioned reinforcement (a light flash) waspresented on a less demanding fixed-ratio sched-ule, than it was with primary reinforcement(food) alone. Moreover, a study by Zeiler (1970)indicated that efficiency might have been in-creased further by imposing a time limit on theavailability of conditioned reinforcement. Simi-larly, writing output might be enhanced by as-signing conditioned reinforcement value to thecompletion of a certain number of words orpages if and only if it is achieved within a giventime period.

Is this a gross oversimplification of the prob-lem? Obviously there is more to producing anovel than merely putting words on paper. Asmentioned, however, we are assuming (for ourpresent purpose) that the necessary precurrentbehaviors are already in the prospective writer'srepertoire. Given this assumption, it still re-mains to be shown that the approach is valid.One might easily be tempted to predict that sucha blatantly mechanical treatment, with its em-phasis on regular periods for writing and rein-forcement of a steady output, would stifle im-portant processes suggested by terms such as"creativity" and "inspiration". Perhaps study be-havior (Broden, Hall, and Mitts, 1971; Fox,1962), grade-school composition (Ballard andGlynn, 1975; Brigham, Graubard, and Stans,1972; Maloney and Hopkins, 1973), and evendoctoral dissertation writing (Harris, 1974;Nurnberger and Zimmerman, 1970) can flour-ish under behavioristic regimens-but creatinga successful novel would be thought by manyto belong to an entirely different category.

Since applied behavioral analysis is a very newtechnology, still focusing mainly on relativelysimple behavior problems and those foundwithin institutional settings, it might seem that

as yet there could be no data on the use of arti-ficially constructed ratio schedules to accelerateand maintain the writing of successful novels.But such data do exist. At least several well-known novelists have kept detailed records oftheir writing behavior precisely in the mannerand, apparently, precisely for the reason that amodern behaviorist-had it been possible ornecessary to consult one-could have suggestedto them. One of these novelists of contemporaryfame is Irving Wallace. In the next section, hesets forth ways in which he and others, inde-pendently of formal behaviorism, have con-trolled their own literary output through theuse of behavior-modification techniques. In thefinal section, samples of the data he has takenon his own behavior are presented in the formof cumulative records illustrating similaritieswith data generated by fixed-ratio schedules.

SELF-CONTROL TECHNIQUES OFFAMOUS NOVELISTS2

by Irving Wallace

I kept a work chart when I wrote my firstbook-which remains unpublished-at the ageof nineteen. I maintained work charts while writ-ing my first four published books. These chartsshowed the date I started each chapter, the dateI finished it, and the number of pages written inthat period. With my fifth book, I started keep-ing a more detailed chart, which also showedhow many pages I had written by the end ofevery working day [e.g., Figure 1). I am notsure why I started keeping such records. I sus-pect that it was because, as a free-lance writer,entirely on my own, without employer or dead-line, I wanted to create disciplines for myself,ones that were guilt-making when ignored. Achart on the wall served as such a discipline, itsfigures scolding me or encouraging me.

I had never told anyone about these charts,because I always feared that their existence

2From Wallace (1971, pp. 65-69) and reprintedwith the author's permission.

516

NOVEL-WRITING BEHAVIOR

would be considered eccentric or unliterary. Butthrough the years, I have learned that their usagehas not been uncommon among well-knownnovelists of the past. Anthony Trollope, author

D4r6. . _

PS a 11D?

FAc BOB

P4&Etv.>Ar

of more than fifty popular novels includingBarchester Towers, was perhaps the greater rec-

ord-keeper known to literature. In his Autobi-

ography, published in 1883, Trollope wrote:

I7 -

wg~~~~~~~_wk4s~~

.I_

* t LE___1_ =3 0 ___- v. --_____ .....

* 2 . .13._._.__~~~~a .L._L. _l Ir. -____ __Z &_

et~~_ 2 4 2__

__~~4 __L ___.......__ ._ ..__.___.__

.._ .............2.I L. __ ___ _ _...________.

~~~~~~---1--_-__________--J2.

** ; 30 - --__- ' -tL E -----1 -

lb Ih) -. J1

_ , __ t. 19l---0-- 0 -u--__1--Fig. 1. A sheet from one of Irving Wallace's work charts. The sheet, which is Page II of the chart, shows

the daily number of pages written for the first draft of The Fan Club from December 17, 1972, until the com-

pletion of that draft on January 25, 1973.

A.1

_

5

1. -T.

517

.-Nt

1 0 1

. .-,_-I.. ,


"When I have commenced a new book, Ihave always prepared a diary, divided into weeks,and carried on for the period which I have al-lowed myself for the completion of the work.In this I have entered, day by day, the numberof pages I have written, so that if at any time Ihave slipped into idleness for a day or two, therecord of that idleness has been there, staring mein the face, and demanding of me increased la-bour, so that the deficiency might be supplied.According to the circumstances of the time,-whether my other business might then be heavyor light, or whether the book I was writing wasor was not wanted with speed,-I have allottedmyself so many pages a week. The average num-ber has been about 40. It has been placed as lowas 20, and has risen to 112. And as a page is anambiguous term, my page has been made tocontain 250 words; and as words, if not watched,will have a tendency to straggle, I have hadevery word counted as I went.... There has everbeen the record before me, and a week passedwith an insufficient number of pages has been ablister to my eye and a month so disgracedwould have been a sorrow to my heart.

"I have been told that such appliances arebeneath the notice of a man of genius. I havenever fancied myself to be a man of genius, buthad I been so I think I might well have sub-jected myself to these trammels. Nothing surelyis so potent as a law that may not be disobeyed.It has the force of the water-drop that hollowsthe stone. A small daily task, if it be really daily,will beat the labours of a spasmodic Hercules[Trollope, 1946, pp. 116-1171."

This revelation, as well as other confessionsmade by Trollope, indicated that "he treatedliterature as a trade and wrote by the clock," andthis offended literary assessors and damaged hisreputation for years after. Yet numerous authorshave been just as meticulous about their writingoutput and about recording it, and they havefared better in the eyes of the literati. ArnoldBennett, for one, devoutedly charted in his Jour-nal his daily progress, by word count, for eachnew novel [Bennett, 1971).

Ernest Hemingway is an example of a wordor page counter in recent times. According tothe Paris Review:

"He keeps track of his daily progress-'so asnot to kid myself'-on a large chart made outof the side of a cardboard packing case and setup against the wall under the nose of a mountedgazelle head. The numbers on the chart showingthe daily output of words differ from 450, 575,462, 1250, back to 512, the higher figures ondays Hemingway puts in extra work so he won'tfeel guilty spending the following day fishing onthe gulf stream [Plimpton, 1965, p. 219)."

Once, long ago, deceived by the instructors,professors, by an old romantic tradition, I hadbelieved that a writer writes only when he feelslike it, only when he is touched by mystic inspira-tion. But then, after studying the work habits ofnovelists of the past, I realized that most suc-cessful writers invest their work with profession-alism. From Balzac, who worked six to twelvehours a day, and Flaubert, seven hours a day,and Conrad, eight hours a day, to Maugham,who worked four hours a day, and AldousHuxley, five hours a day, and Hemingway, sixhours a day, these authors were uniformly indus-trious, and when they were once launched upona book they wrote regularly, day in and day out.While the story may be apocryphal-I shouldlike to believe it is not-it is said that VictorHugo sometimes forced himself to work regu-larly by confining himself to his study. To dothis, he had his valet take away every stitch ofhis clothing, and ordered this servant not to re-turn his attire until the hour when he expectedto be through with his day's writing.

In short, no matter how they effected theirroutines, the vast majority of published authorshave kept, and do keep, some semblance of reg-ular daily hours.... Occasionally the hourkeep-ers were inspired when they went to their desks,but if they were not, they simply wrote as well

518


as they could, as craftsmen, and hoped for thebest.

A FURTHER DISCUSSION ON NOVELWRITING AS FIXED-RATIO

BEHAVIOR3by J.J.P.

It was argued in the Introduction that ameaningful and useful analogy can be drawnbetween writing a novel and emitting a simpleoperant response on a fixed-ratio schedule. Thereare, of course, important differences between thetwo types of behavior. One obvious difference isthat writing a novel necessarily follows a cer-tain sequential progression, so that the form ofthe behavior changes continuously, whereas theoperants typically studied in behavioral labora-tories are repetitive (for a notable exception,however, see Pryor, Haag, and O'Reilly, 1969).No doubt this is an important consideration inestablishing the complex behavioral repertoireprerequisite to novel writing; however, it maybe less important insofar as accelerating andmaintaining the product of that repertoire isconcerned. Another obvious difference is thatnovels vary in length, whereas the amount ofresponding required to produce reinforcementon a fixed-ratio schedule does not, of course, varyfrom one reinforcement to the next. An experi-enced novelist, however, is probably continuallyexposed to numerous stimuli indicating the ap-proximate amount of writing required to com-plete his or her current manuscript.One plausible empirical test of the analogy

would be to compare the temporal pattern ofnovel writing with that of simple fixed-ratio be-havior. The latter is characterized by a pauseimmediately after reinforcement, followed by arapid transition to a high rate of responding that

3Based on a paper presented at the Thirty-SeventhAnnual Meeting of the Canadian Psychological Asso-ciation, Toronto, Ontario, June 9 to 11, 1976. Theauthor expresses his gratitude to Irving Wallace forhis helpful comments and for generously making hisprivate records available to be used in preparing thissection of the article.

persists until the next reinforcement (Ferster andSkinner, 195 7). One might therefore expect tofind a relatively low rate of writing near thestart of a manuscript and a higher rate towardits completion. Moreover, fixed-ratio scheduleswhose units consist of smaller fixed-ratio sched-ules have been studied, and it has been foundthat shorter pauses tend to occur after comple-tion of each of these smaller units than aftercompletion of the larger unit that they consti-tute (Kelleher, 1958; Lee and Gollub, 1971).Since each chapter of a novel might be consid-ered to be a small fixed-ratio unit within thelarger fixed-ratio unit consisting of the entiremanuscript, one might further expect relativelyshort pauses to occur after completion of eachchapter and longer pauses to occur between suc-cessive drafts of the novel.

The detailed self-recording charts (e.g., Fig-ure 1) that Irving Wallace kept on his dailynovel-writing behavior permit a test of theabove predictions. It should first be pointed out,however, that the data in these charts appear torepresent more the reworking and elaborationof previously generated verbal behavior (aprocess Skinner, 1957, pp. 344 ff., has used theterm composition to describe) than the genera-tion of completely new verbal behavior. Beforebeginning each novel, Wallace prepared exten-sive notes and outlines, as indicated in thefollowing quote: "On each new novel, I havealways written many outlines for myself, de-veloping scenes and characters, underliningstory problems that need further thought. I worka novel out, in chronological sequence, overmany weeks, in my head and then roughly onpaper before beginning it.... But at the sametime I try to leave a broad area for spontaneityin my outlines [Wallace, 1971, pp. 51-523."Although Wallace generally made many revi-sions in his outlines after beginning to writeeach novel, most of this revising was done out-side of the daily hours that he scheduled for hiswriting. Referring specifically to the writing ofThe Prize and The Man, Wallace (personal com-munication) stated: "On both books I usually

519


woke up at ten o'clock in the morning, went towork at noon, and wrote until five-thirty-withperhaps two half hour breaks for light lunch andopening mail. Often I worked two hours beforemidnight, planning or roughing out scenes I in-tended to tackle in the next day or week."

Samples of Wallace's daily self-recordingcharts have been reproduced in Wallace (1971)and Leverence (1974). Additional data for thepurpose of this paper were supplied by Wal-lace from his personal records. Due to spacelimitations, only the data for The Prize, TheMan, and The Plot will be presented here. Thesedata, however, are generally representative of allthe available data.

Figure 2 shows the cumulative record of thefirst-draft pages written for The Prize. Note that

THE PRIZE Star'

the overall rate of writing tended to conform tothe expected fixed-ratio pattern in that it wasrelatively low near the beginning of the manu-script and accelerated to a high level toward theend. Describing the daily work chart on whichFigure 2 is based, Wallace (1971, p. 87) ap-pears also to have been impressed with this pat-tern when he wrote: "Reviewing this workchart now, I see that in November of 1960,there were five out of twenty-six working dayswhen I produced nothing, nary a page, whereasin the last five days of that month I wrote fifty-four first-draft pages.... As I came nearer andnearer to the climax and to the end, I wrotemore and more steadily, entirely absorbed, to-tally pulled, and I started passing up meals, lim-iting my time with my family, canceling social

ted: Oct. 19, 1960

Finished Feb. 24,1961

&nMu0

cL

0

pp/dayI'00'/1

0 IDAY30 DAYS

OCT "NOV DEC -JAN FE BFig. 2. Cumulative record of the daily number of pages written by Irving Wallace for the first draft of The

Prize. The oblique lines denote completion of the indicated chapters. The starred chapter number indicates a'run-through"--i.e., that the next chapter was started on the same day that the indicated chapter was com-

pleted. The letters correspond to the following comments: (d) "Sunday, day of rest [this comment is fromWallace, 1971, p. 87; the others are from the work chart]" (except toward the completion of the manuscript,Wallace did not work on Sundays); (b) "[made research] notes"; (c) "research"; (d) "holiday"; (e) "research";(f) "holiday"; (g) "research"; (h) "holiday"; (i) "research"; (j) "research". (Data from Wallace, 1971, pp. 88-91.)

520


engagements. On February 18, 1961, I startedChapter XII, the last chapter of The Prize, andI wrote with such intensity that I completed 127pages in six days."

The cumulative record in Figure 2 also pro-vides some evidence that rate of writing tendedto decrease immediately following chapter com-pletions (indicated by the oblique lines). Thisevidence supports the view that chapters func-tioned as small fixed-ratio units for the novelist.Stronger evidence on this point is seen in thefact that the completion of a chapter almostalways coincided with the termination of writingfor the day on which the chapter was completed.There was only one instance, out of 11 possibleinstances, in which a new chapter was startedon the same day that the previous chapter wascompleted. As can be seen in Figure 2, this "run-through" (indicated by an asterisk) occurred atthe end of Chapter VIII. It is interesting that thesole "run-through" occurred toward the end ofthe manuscript, when the rate of writing hadaccelerated to a high level.

The comments (indicated by the letters in thefigure) Wallace entered in his daily work chartfor The Prize provide a flavor of the events andcircumstances surrounding the writing of thenovel. As subsequent data are presented, it is in-teresting to note the consistencies in the writingpattern that were maintained despite wide vari-ations in many of the situational factors relatingto the behavior. Concerning The Prize, note thatmost of the comments refer to research. Thisreflects the fact that The Prize was the mostheavily researched of Wallace's novels (Lever-ence, 1974, p. 187).

Figure 3 shows the cumulative record of thefirst-draft pages written for The Man. Again,note that the overall rate of responding tendedto conform to the expected fixed-ratio pattern,in that it was relatively low near the beginningof the novel and accelerated to a high leveltoward the end. The similarity between the over-all response pattern shown in Figure 3 and thatshown in Figure 2 is striking, especially consid-ering the different background circumstances for

the writing of the two novels. The differentamounts of research required for these books isreflected in the comments Wallace recorded inhis daily work charts for the two novels (see thefigure captions). Another difference can be seenfrom the following statement by Wallace(quoted in Leverence, 1974, p. 410): "Normally,a book gestates inside me for some time.... ButThe Man just came to me, and when it came Iknew it was right. The characters were there.Every major ingredient of the book came early."The continuation of the quote is also interesting:"This was important, for when you have a strik-ingly unusual idea to superimpose upon yourcharacters, there's always a danger that you'llget into trouble in the last part of the book, be-cause the idea begins to dominate the characters,suffocate them, so that the characters can't evolvethrough the novel naturally, and you are leftwith an overwhelming idea that can't be resolvedin the end." This suggests that the last part of anovel is not necessarily the "easiest" to write.

In addition to the overall fixed-ratio patternin the writing of The Man, there were alsosmaller fixed-ratio patterns at the chapter level.As with The Prize, some of these can be seen inthe cumulative record. Stronger evidence for theexistence of these smaller fixed-ratio units, how-ever, is the fact that there were only two "run-throughs" (occurring after Chapters II and VII,as indicated by the asterisks in the figure) out ofeight opportunities for "run-throughs" to occur.

Figure 4 shows the cumulative record of thefirst-draft pages written for The Plot. This is thelargest of Wallace's novels. Nevertheless, theoverall rate is remarkably high and steady, al-though somewhat lower than the overall ratesfor the previous two manuscripts. However, thecharacteristic pattern of a relatively low rateearly in the writing of the manuscript and ahigh terminal rate, is clearly present. Note, also,the large number of pages rewritten during theearly part of the manuscript (see the points ata, b, f, i, and o).

As indicated by the Arabic and Roman nu-merals (and as explained in the figure caption),

521


THE MAN Started: Oct. 31,1963Finished: Mar. 8, 1964

e

ppp/dayab~~~~~~~c~

6|/ ffi 30DAYS

as~~~ Ig,1 1NOV DEC JAN FEB MAR

Fig. 3. Cumulative record of the daily number of pages written by Irving Wallace for the first draft of TheMan. The oblique lines denote completion of the indicated chapters. The starred chapter numbers indicate"run-throughs"-i.e., that the next chapter was started on the same day that each of the indicated chapterswas completed. The letters correspond to the following comments: (a) "President Kennedy assassinated"; (b)"JFK's funeral"; (c) "Thanksgiving"; (d) "Life mag[azinej interview" (2 days); (e) "ill" (2 days); (f) "rewrit-ing" (2 days). (Data from Leverence, 1974, pp. 408-409.)

two possible chapter breakdowns-one consist-ing of nine and one consisting of 12 chapters-were considered for this novel. No "run-through" occurred after completion of anyof the proposed chapters. The absence of "run-throughs" may have been related to the fact thatthe chapters averaged somewhat longer for thisnovel than for the previous novels. It is knownthat large fixed-ratio schedules tend to producelonger pauses than do small ones (Ferster andSkinner, 1957).

Wallace typed each first draft straightthrough, occasionally rewriting pages he con-sidered "poorly done" or "false starts". Then, hecarefully reread the entire manuscript, revisingit as he went along. Five or six such revisionswere usually done for each novel. Sometimes

Wallace recorded the daily number of first-draftpages he covered when rereading and revisinga manuscript. Figure 5 shows the cumulativerecord of the daily number of pages covered forthe first revision of The Plot. Note the longpause of 40 days that occurred between comple-tion of the first draft and the start of the seconddraft. Concerning this, Wallace (quoted in Lev-erence, 1974, p. 417) stated: "And when thefirst draft was finally finished, I took a brief restand then began to rewrite." Wallace began grad-ually, however, for he did not start workingfull-time until the point indicated at a. Respond-ing then increased over days to a high terminalrate, thus replicating the overall response pat-tern seen for the other first drafts in the previousfigures.

522


THE PLOT Started Nov.8 1965Finished: May 15, 1966

#4 pp/day0 / P/*

O~~~~~~~~ 130 DAYS

Fig. 4. Cumulative record of the daily number of pages written by Irving Wallace on the first draft of ThePlot. The oblique lines denote completion of the indicated chapters. The Arabic numerals represent one chap-ter breakdown that was considered, and the Roman numerals represent an alternative whereby the originalChapter One became Book One consisting of the four chapters indicated. The letters correspond to the fol-lowing comments: (a) "rewrote 5 pages into 12"; (b) "plus rewriting 4 of yesterday's pages"; (c) "5 pages ofresearch notes"; (d) "Thanksgiving holiday"; (e) "5 pages of research notes"; (f) "4 of yesterday's pages re-written"; (g) "research"; (h) "mother to hospital; research"; (i) "1 of yesterday's rewritten, also"; (j) "Xmasholiday" (3 days); (k) "New Years day"; (1) "mother to sanatorium"; (m) "Sonny [Wallace's cousin] died";(n) "Sonny's funeral"; (o) "plus rewriting yesterday's 8 pages completely"; (p) "ill"; (q) "wrote N[ew] Y[ork]Times review"; (r) "Abe Wallace [Wallace's uncle] died"; (s) "Uncle Abe's funeral".

As mentioned, the above data are typical ofall of the data available on Irving Wallace'snovel-writing behavior. The rate of respondingtended to be relatively low near the beginning ofeach first draft or revision, and then acceleratedto a high, steady level that was maintained untilthe end of the manuscript. (Of 11 cumulativerecords that were examined, there was only oneexception to this generalization; viz., the datafor the first revision of The Fan Club.) More-over, relatively short pauses tended to occur im-mediately following first-draft chapter comple-tions, and considerably longer pauses occurredbetween successive drafts of the novels. It cantherefore be concluded that Irving Wallace'snovel-writing behavior closely resembles behav-ior generated by fixed-ratio schedules.

It can, of course, be argued that these findingswere due, entirely or in part, to processes thatonly superficially resemble the processes operat-

ing in simple operants that are reinforced onfixed-ratio schedules. While alternative explana-tions are possible, extreme care must be taken intheir formulation. For example, it might be sug-gested that writing tends to accelerate towardthe latter part of a novel because the novelistbecomes more "familiar" with the characters andsituations he has invented, or more "engrossed"by them, and that writing about them thereforebecomes "easier". The words in quotes, and simi-lar terms that come to mind, seem to denoteprocesses that are quite different from those seenin simple fixed-ratio behavior. However, suchterms must be precisely defined before they canbe useful in furthering a scientific analysis ofbehavior. In their present context, they mightmean essentially that as the moment of rein-forcement on a fixed-ratio schedule approaches,certain variables exert increasingly stronger con-trol over behavior. Exactly what these variables

523

524 IRVING WALLACE and J. J. PEAR

THE PLOT Started: June 25, 1966First Rewrite" Finished: Aug. 8, 1966

I 1~44 pp/day

I 0 ~

b I 0 DAYS

a I

JUNE JULY AUGFig. 5. Cumulative record of the daily number of pages covered by Irving Wallace on the second draft of

The Plot. The letters correspond to the following comments: (a) "started full-time"; (b) "rewrote full pages,or wrote new ones".

are and exactly how they exert their control arequestions that still require extensive study-evenfor the simple operant.

REFERENCES

Ballard, K. D. and Glynn, T. Behavioral self-man-agement in story writing with elementary schoolchildren. Journal of Applied Behavior Analysis,1975, 8, 387-398.

Bennett, A. The journals. Baltimore: PenguinBooks, 1971.

Brigham, T. A., Graubard, P. S., and Stans, A.Analysis of the effects of sequential reinforcementcontingencies on aspects of composition. Journalof Applied Behavior Analysis, 1972, 5, 421-429.

Broden, M., Hall, R. V., and Mitts, B. The effect ofself-recording on the classroom behavior of twoeighth-grade students. journal of Applied Behav-ior Analysis, 1971, 4, 191-199.

Ferster, C. B. and Skinner, B. F. Schedules of 'ein-

forcement. New York: Appleton-Century-Crofts,1957.

Findley, J. D. and Brady, J. V. Facilitation of largeratio performance by use of conditioned reinforce-ment. Journal of the Experimental Analysis ofBehavior, 1965, 8, 125-129.

Fox, L. Effecting the use of efficient study habits.Journal of Mathetics, 1962, 1, 75-86.

Goldiamond, I. Self-control procedures in personalbehavior problems. Psychological Reports, 1965,17, 851-868.

Harris, M. B. Accelerating dissertation writing: casestudy. Psychological Reports, 1974, 34, 984-986.

Kazdin, A. E. Self-monitoring and behavior change.In M. J. Mahoney and C. E. Thorensen (Eds),Self-control: power to the person. Monterey, Cali-fornia: Brooks/Cole, 1974. Pp. 218-246.

Kelleher, R. T. Fixed-ratio schedules of condi-tioned reinforcement with chimpanzees. Journalof the Experimental Analysis of Behavior, 1958,1, 281-289.

Lee, J. K. and Gollub, L. R. Second-order scheduleswith fixed-ratio components: variation of compo-

NOVEL-WRITING BEHAVIOR 525

nent size. Journal of the Experimental Analysis ofBehavior, 1971, 15, 303-210.

Leverence, J. Irving Wallace: a writer's profile(with an introduction by J. Weidman, an inter-view by S. L. Grogg, Jr., and an afterword byR. B. Browne). Bowling Green, Ohio: The Popu-lar Press, 1974.

Maloney, K. B. and Hopkins, B. L. The modificationof sentence structure and its relationship to sub-jective judgements of creativity in writing. Jour-nal of Applied Behavior Analysis, 1973, 6, 425-433.

Nurnberger, J. I. and Zimmerman, J. Applied analy-sis of human behavior: an alternative to conven-tional motivation inferences and unconsciousdetermination in therapeutic programming. Be-havior Therapy, 1970, 1, 59-69.

Plimpton, G. Ernest Hemingway. In G. Plimpton(Ed), Writers at work: the Paris Review inter-views. Second series. New York: The VikingPress, 1965. Pp. 215-239.

Pryor, K. W., Haag, R., and O'Reilly, J. The cre-ative porpoise: training for novel behavior. Jour-nal of the Experimental Analysis of Behavior,1969, 12, 653-661.

Skinner, B. F. Science and human behavior. NewYork: The Free Press, 1953.

Skinner, B. F. Verbal behavior. New York: Apple-ton-Century-Crofts, 1957.

Trollope, A. An autobiography. London: Williams& Norgate, 1946.

Wallace, I. The writing of one novel. RichmondHill, Ontario: Simon & Schuster of Canada(Pocket Book Edition), 1971.

Zeiler, M. D. Time limits for completing fixed ra-tios. Journal of the Experimental Analysis of Be-havior, 1970, 14, 275-286.

Received 12 February 1976.(Final acceptance 12 October 1976.)

introduction-by j.j.p. the manuscript, its publication, and its

Documents