Kyle Evans

The Artist as Scientist: The Sonification of DNA

2009

Increased online accessibility to information in mathematics, genetics,

acoustics and myriad other scientific fields has resulted in the blurring of lines

between art and science. An example of this evolution can be seen in artists and

composers creating music utilizing quantified DNA and protein information

through the process of data sonification. According to the International

Community for Auditory Display, “Sonification is the use of non-speech audio to

convey information.”1 The ideas and practices of DNA sonification are not only

fundamentally based in biotechnology and genetics, but were initially conceived

by scientists in the field of research. This initiation point creates a platform from

which ideas and concepts can be transposed from the world of science into that

of art and music. As the field of biotechnology develops, and the information it

produces becomes increasingly more attainable, it is understandable that

composers and sonic artists would adopt this information into their own

practices and utilize the data for artistic means. This displacement of data from

one field of study to another spawns a dichotomy between the artist/composer

and the scientist/biologist. This dichotomy weighs heavily in defining the role of

the composer in the artistic process. Analyzing the merger of quantified

scientific data and musical composition is an important concept in the practice

of DNA sonification. Does a composer choose to fill the role of scientist and

utilize the process for research and scientific development, or does the

composer take on the role of artist and utilize the data produced by the scientist

to generate work with a focus on musical aesthetics? How accurately can data

derived from DNA sequences be reproduced?

The sonification of extra-musical measurements is a practice that has its

roots firmly grounded in the recent history of avant-garde composition. Though

not an extremely popular form of composition, sonification has been

successfully practiced through many techniques and approaches. In 1970,

Charles Dodge composed Earth’s Magnetic Field, which explores musical

representation of collected Kp index data; an average measurement of the

subtle changes in the earths magnetic field. In 1962, John Cage composed Atlas

Eclipticaias by using star charts superimposed with sheet music. In this process,

Cage not only exploits the random organization of stars to generate aleatoric

sequences of pitch and rhythm, but also utilized each star’s relative brightness

to signify the amplitude of each note. In pieces such as these, there exists a

fascinating juxtaposition of two fields of studies which, when merged, can

create unexpected alternative outcomes. Though the two pieces mentioned

above were created by composers and not by scientists, many scientists in the

area of genetics have attempted to solve the complexities concerning DNA

representation by taking on the help of composers and musicians to sonify the

complicated data structures into a more retainable and interesting medium. In

the opening statement of an article written by Andi Schoon and Florian

Dombois, this separation of scientific and artistic approach is explained and

applied to all forms of exta-musical data sonification:

“Seen from the vantage point of cultural history, contemporary sonification is

essentially characterized by two aspects that to date have seldom been considered in combination: the first aspect is sonification as the transformation of the inaudible into the sphere of the audible, and the second its use as an

instrument for gaining knowledge via the concrete listening experience.” 2 (Andi Schoon, Florian Dombois, 2009)

This statement introduces the two greatly differing approaches that scientists

and artists take, and the proportionally differing results of the sonified DNA data.

The connection between DNA sequence data and musical notation are

not immediately recognizable, but through study and examination it is possible

to find many different ways that the two can combine. The most immediate

realization of musical connection is the serialization of data that is used in

representing DNA sequences. For example, DNA is encoded using four base

types; adenine, cytosine, thymine, and guanine (A,C,T and G). These bases are

then combined into sets of three generating what is called a codon (AGT, TTC,

AGA, etc). This massive collection of potential combinations can portray many

parallels to melody and rhythm when an artist uses the data to represent

musical properties such as pitch, loudness, duration, timber, etc. Beyond this

“language” that represents basic DNA structures, composers have also utilized

other genetic measurements to incorporate in the sonification process including,

water solubility, Pk(a) and pH measurements, molecular weight, folding patterns

and even light absorption levels. Due to a continued public interest in

biotechnology and increases in data accessibility of modern times, there now

exists a plethora of representational genetic data that composers and musicians

have readily available at their disposal. With this wide-open access to genetic

data, many technical and conceptual challenges arise. Where is an artist to

begin?

Different musicians and scientists have approached this issue with greatly

differing methodologies. The artist must make aesthetic and conceptual

decisions as how to translate incredibly large amounts of data. But how does an

artist represent sonically that which is meant to represent something else? DNA

sequences naturally exist outside the realm of numerical value, but science has

chosen to represent these sequences as data strings in order to help our minds

decipher the complexity of the biological information. These methods of DNA

representation were created by scientists for scientists, which in turn, forms an

obstacle the composer must first encounter before any aesthetic or artistic

decisions can be made; should the process be approached from the

perspective of an artist or that of a scientist. Of course, a combination of both

must exist in order to use scientific data for musical composition, but the weight

of scientific or artistic approach usually lies in the practice of the composer.

An example of the separations and interconnections between scientist

and artist can be seen in comparing the practices and works of John Dunn and

Marry Anne Clark. The two began collaborating in the 1980’s with investigations

into DNA sonification. John Dunn, as a musician and artist, required the help of

a scientist, Marry Anne Clark, to develop the techniques and methods of

translating DNA sequences into the acoustic realm. The two generated methods

involving the translation of amino acids into frequencies on a traditional western

musical scale, and eventually added more complexity by utilizing data collected

from water solubility measurements and folding patterns. Their relationship, as

documented in a co-written article by Dunn and Clark entitled Life Music: The

Sonification of Proteins, shows the dichotomy of science and art through a

verbal exchange between the two; Dunn asks Clark, “Where is the art in your

science?” in which Clark responds by asking, “Where is the science in your

art?” 3 (Dunn and Clark)

Interestingly enough, it was the work of scientists, solely for the reasons

of research and teaching that the first experiments in DNA sonification practices

surfaced. Scientists Kenshi Hayashi and Nobuo Munakata suggest the approach

in their writing Basically Musical in 1984:

“Recent progress in gene cloning and DNA sequencing techniques has produced an enormous amount of base sequence data. …We propose an acoustic method to minimize the distress of handling such information. …One certain advantage of this method is that the sequences are now more easily recognized and memorized. … This practice may help to bring back some of the pleasure of decoding the mysteries of life from computers.” 4 (Hayashi and

Mankau, 1984)

As well as Hayashi and Mankau, N. Munakata has also been exploring the

scientific frontier of DNA sonification. Munakata approached the concept with

the notion that musical tunes are easier and more pleasant to memorize than

alphanumeric strings of bases and amino acids. Through this process, he

attempts to “reveal the meaning of specific sequences of DNA bases.”5 His

technique involved assigning pitches to DNA bases according to their thermal

stability. In Munakata’s words,

“Genes and music are two heritable systems that underlie our life. Both of them are made of linear and quantized information. I try to explore the correspondence and metaphor between them by converting gene (DNA, RNA and protein) sequences to MIDI sequences. Hopefully, gene music can capture

and inspire appreciation of the diversity, mystery and beauty of life.” 6 (Munakata)

Though his methods are slightly more intricate than those of Hayashi and

Mankau, he still limits his results to common western musical scales, therefore

limiting any distinction from other musical work. The restriction to scale

structures in attempts to make the sonified patterns of DNA more “listenable” is

a common practice among many composers. This shows the strong role the

composer can play in a process whose data sources exist beyond the

composer’s control. If the method in which the DNA is sonified has the ability to

structure a score due completely to the composer’s personal aesthetic opinions,

then does the result any longer represent a DNA sequence? Or does it then only

represent the composer?

This role of the composer can be heard very clearly in the work of John

Dunn and Marry Anne Clark. Each separately composed a piece from base

sequences collected of the enzyme lysozyme C. When listening to each artist’s

personal sonification of the enzyme, it becomes very clear the large amount of

liberties a composer can take in the process. Dunn’s thunderous sounds and

sharply pronounced rhythms of his rendition stand in stark contrast to the

cheerful, light hearted and floating melodies of Clark’s version. Dunn and Clark

describe their first time hearing the two radically differing pieces:

“… The experience of listening to these parallel compositions, each developed independently in two different locations (Clark in Texas and Dunn in Michigan), but with the same protein data… gave more insight into the astounding depth of structure Nature has built into Her art.” 3 (Dunn and Clark)

In this case, Clark and Dunn give reason for the two Lysozyme C pieces

being so different to the openness of the DNA sequences themselves.

Therefore, practical aesthetic choices become the only limitation on such an

open structure which, of course, is the case for all types of data sonification. In

this mindset, if a DNA musical score is without any type of compositional

control, the large amounts of data that can be collected from the multitude of

available DNA and protein sequences appears too versatile to be truthfully

represented. Most composers who practice DNA sonification share this view

and sometimes use it to justify radical aesthetic decisions. Three researchers, X.

J. Shi, Y. Y. Cai and C. W. Chan looked closely at the representative issues of

DNA sonification to its original data sets and have come to some rather steep

conclusions. In discussing the transposition of DNA data into the acoustic realm,

they write:

“This is like translating English into Chinese. In cases where direct translation cannot express the original message clearly, the translator needs to look at the context and decide on a similar Chinese expression. This flexibility is equally necessary in translation between proteins and music. …Instead of sonifying the amino acid according to their solubility, we sonified according to aesthetic appeal…” 5 (Shi, Cai and Chan)

When this method of sonification is applied toward DNA sequence data,

then there exists no connection between the composer’s results and what the

source information actually represents. If this connection is forgotten then there

is no reason to claim that results of this type are representations of the materials

supplied to us by nature.

Since the information that we can collect from DNA and protein

sequences are so open ended in terms of musical interpretation, it is difficult to

imagine a truly successful realistic representation. If a composer’s intention is to

represent the source of data from which a musical piece is sonified, then the

creator must consider a realistic (or even scientific in the case of DNA)

connection between the two. Artists such as Susan Alexjander and Peter Gena

are some of the few composers who have focused on this approach of a more

truthful representation by utilizing science and math.

Susan Alexjander utilizes a unique set of data found in DNA sequences

that help to create a more physical attachment to the music that the information

produces. By calculating the light absorption frequencies of bases, Alexjander is

able to connect the data to sound through their shared measurement; cycles per

second. Through the Law of the Octave, these extremely high light frequencies

are repeatedly divided in half until their frequency resides within the range of

human hearing.7 Dividing these frequencies does not result in the “comfortable”

tonal scale that we have been conditioned to recognize, but instead creates

microtonal pitches. These microtonal pitches are not rounded in order to match

any particular scale, in turn resulting in a more truthful representation of the

original measurements.

Along with Susan Alexjander, Peter Gena’s DNA sonification work

depends significantly on the idea of truthful representation by utilizing complex

algorithms. Gena, with help of geneticist Charles Strom, has created computer

software he calls The DNA Mixer, which can translate the information of DNA

sequences into sound in real-time. The composer himself best describes his

intentions:

“…I believe that the musical reading of DNA ought to be rendered literally. As the sequences represent life of many sorts, I am reluctant to tamper with the ‘score’. The DNA mixer can realize sequences as digital sound and/or prints them out in musical notation. …The Physio-musical conversation of DNA sequences takes place via a series of formulae that were worked out in a manner based on physical properties of DNA and musical parameters.” 8 (Gena and Strom)

Gena’s method goes beyond just representing different bases as different

frequencies; he uses myriad physical characteristics of amino acids in order to

recreate not only pitch, but also timber, amplitude and duration. His algorithms

for generating the musical information involve data obtained from sequences not

commonly being utilized by other composers. He takes into account Pk(a)

acidity measurements, molecular weight and chemical properties of each amino

acid represented. Just as in John Cage’s sonification of star charts, Gena

composes for more than just one musical dimension. By taking timbre and

dynamics into consideration, he has discovered a way to more truthfully

represent the data supplied by DNA and maintain aesthetically interesting

results.

The composers and scientists presented here only represent a small

portion of the total who are exploring DNA sonification, but they demonstrate

the vastly differing approaches and results that can come about through the

process. Though there always exists a role the artist must play in the sonification

of such an extensive data structure as DNA, there also exists the possibility of

creating strong physical and conceptual attachments to the original source

material. Artists such as Susan Alexjander and Peter Gena show in their work

that this connection is not only appropriate in representing something of such

scientific importance as DNA, but also have the ability to generate aesthetically

interesting results. What is most interesting in the sonification process, most

particularly that of DNA data, is the hybridization of artist and scientist. These

roles have become blurred as our access to scientific data becomes

increasingly available. As scientific information and understanding develops, so

will the approaches and techniques that further push the connections between

science and the methods in which artists choose to interpret its information.

These approaches, designed by artists and scientists alike, are transforming the

conceptual connections an artist can explore between the two practices. DNA

sonification only represents one methodology for artists to utilize within the

endless possibilities for the artistic interpretation of science.

Works Cited

1. International Community for Auditory Display. Sonification Report:Status

of the Field and Research Agenda. (Prepared for The National Science

Foundation, 1997)

2. Andi Schoon and Florian Dombois. Sonification in Music. Proceedings of

the 15th International Conference on Auditory Display. Copenhagen 2009.

3. M. A. Clark and John Dunn. “Life Music: The Sonification of Proteins”.

Leonardo, Vol. 32, No. 1. (1999).

4. Kenshi Hayashi and Nobuo Munakata. “Basically Musical”. Nature. 310

(1984).

5. X. J. Shi, Y. Y. Cai and C. W. Chan. “Electronic Music for Bio-Molecules

Using Short Musical Phrases”. Leonardo, Vol. 40, No. 2 (2007)

6. Nobuo Munakata. Gene Music and Sangen Studio.

http://www.toshima.ne.jp/~edogiku/index.html#WhatIsGM

7. Susan Alexjander. “Microcosmic Music – A New Level of Intensity”.

Online Article. Our Sound Universe.

http://www.oursounduniverse.com/articles/microcosmic.html

8. Peter Gena and Charles Strom. “A Physiological Approach to DNA

Music.” Online Article. http://www.petergena.com/docs/gena-strom-

DNA.pdf (2001)