WRT: Writer Response Theory

Recently I talked about the idea of biowriting and the interesting category of DNA bioart. Here is one example of digital DNA biowriting in practice.

One of the central issues of representation in DNA biowriting is transcoding the alphabet into DNA. In digital DNA biowriting, this means transcoding ASCII or Unicode into the nucleotides adenine, cytosine, guanine, and thymine - going for example from 8-bit ASCII, via binary, to base-4 DNA.

Eduardo Kac’s “Move 36” is a bioart and video installation piece composed of a chessboard, a genetically modified plant, and a projection. The plant grows from an earth-and-sand chessboard, planted on the square where chess computer Deep Blue made a crucial move against chess world champion Gary Kasparov in 1997. Its leaves curl due to an introduced genetic modification, coupled with a DNA string that spells out Descarte’s “cogito ergo sum” (I think therefor I am).

The “Cartesian gene” was produced according to a new code I created especially for the work. In 8-bit ASCII, the letter C, for example, is: 01000011. Thus, the gene is created by the following associations between genetic bases and binary digits:

A = 00
C = 01
G = 10
T = 11

The result is the following gene with fifty-two bases:
CAATCATTCACTCAGCCCCACATTCACCCCAGCACTCATTCCATCCCCCATC

Tags: bioart, biowriting, video

Beginning from the example, we can see that “c” = 01000011 = CAAT… which is the first string in the base sequence. With 4 bases per ASCII letter, 52 bases can encode 13 letters - the number of letters in “c o g i t o e r g o s u m” and coincidentally the number of letters in “c a r t e s i a n g e n e”.

Each letter of our message (“c”) is written in 8 bits of ASCII (“01000011”) and then spelled in 4 letters of base-4 DNA (“01-00-00-11 = C-A-A-T”). The final sequence of 52 base nucleotides encodes 13 ASCII letters. We could work this out by hand using an ASCII / binary lookup table or take a shortcut with machine translation such as the ASCII encoder / decoder. The result is Decarte’s quote:

CAAT = 01000011 = c
CATT = 01001111 = o
CACT = 01000111 = g
CAGC = 01001001 = i
CCCA = 01010100 = t
CATT = 01001111 = o
CACC = 01000101 = e
CCAG = 01010010 = r
CACT = 01000111 = g
CATT = 01001111 = o
CCAT = 01010011 = s
CCCC = 01010101 = u
CATC = 01001101 = m

That Kac’s project encodes 8-bit ASCII makes sense considering the way his piece reflects on machine intelligence. It is important to note that, were it not for the context of binary computing, this is by no mean an intuitive way to map an alphabet (base-26 in English - base-23 in Latin) against DNA (base-4), which would only require groups of 3 nucleotides to provide 64 possible values - hardly Unicode, but more than sufficient for a message format that doesn’t encode punctuation or even spaces between words.

AAA = a
AAC = b
AAG = c
AAT = d
ACA = e
ACC = f
ACG = g
…

Thus reimagined, Kac’s “cogito, ergo sum” would read:

AAG = c
ATG = o
ACG = g
AGA = i
CAT = t
ATG = o
ACA = e
CAC = r
ACG = g
ATG = o
CAG = s
CCA = u
ATA = m

although we could imagine cheating and declare a substitution cipher for the purposes of this message alone:

AA = c
AC = o
AG = g
AT = i
CA = t
AC = o
CC = e
CG = r
AG = g
AC = o
CT = s
GA = u
GC = m

So what do we gain by examining the code of Kac’s project and and imagining it encoded differently? Are these strings really that different, especially given that they are totally invisible to the human eye, and only related to the ruffling leaf by proximity?

CAAT-CATT-CACT-CAGC-CCCA-CATT-CACC-CCAG-CACT-CATT-CCAT-CCCC-CATC

AAG-ATG-ACG-AGA-CAT-ATG-ACA-CAC-ACG-ATG-CAG-CCA-ATA

AA-AC-AG-AT-CA-AC-CC-CG-AG-AC-CT-GA-GC

While the message loses nothing, the context of the message loses much. Kac’s original message has been encoded in a way that is both general purpose (although it will never contain the many other characters in the ASCII character-space) and optimized for processing (although it will never be processed).

The difference has everything to do with the world that “Move 36” inhabits - a world where biowriting is not chemical cryptography, but a process enabled by computation - and the message is the voice of a machine, whose binary language is the lingua franca that stands between Decarte’s Latin and the plant DNA, transmuting one into the other.

[Move 36 via rhizome.org]