Books

Short works

Books : reviews

Stefan Helmreich.
Silicon Second Nature.
U California Press. 1998

rating : 4.5 : passes the time
review : 29 September 2001

In the early 1990s, Helmreich conducted an "anthropological study" of the Artificial Life researchers at the Santa Fe institute, and chronicles his results herein.

I had a lot of problems with this book. So many so that I started also carrying a pad of paper with me on the train (where I was reading it during a daily commute) so that I could scribble down all the points I wanted to argue with.

As I understand it, Helmreich contends that everything the ALife researchers do is polluted by their own implicit racial, cultural and gender biases, because they are all white Western middle-class heterosexual males (except for the few who aren't). Well, yes, to some degree, certainly. We are all creatures of our environment and upbringing. And this is not exactly a criticism that hasn't been levelled at scientists, and science, before. Yet, as a scientist myself (albeit not a male one), I simply don't recognise the people Helmreich is describing. The inferred mind-set seems very off-kilter, as if viewed through a set of distorting mirrors that filters out all the scientific curiosity and questioning, and grotesquely magnifies the biases. (But then, I would say that, wouldn't I?)

And how much should we believe these are real biases, anyway, when they are contradictory? For example, consider the critique of ALife models of reproduction. First Helmreich describes the "heterosexual" bias of genetic algorithms thus:

p146. Humans are a familiar example of such animals, and for this reason, culturally built notions of gender, sexuality, and family frequently wind their way into artificial worlds. Affairs might be different if asexual plants were the model of choice or if polyploid sexual organisms like maize were the preferred casts for life-as-it-could-be.
...
There are a number of ways we might understand the exchange of bits between strings, but the metaphor of productive heterosex is gleefully emphasized by most authors.

But later he talks of Holland's Echo system:

p168. The Echo agent is cut after a masculine individual that masquerades as a universal organism. Keller writes of the biological individual copied in Echo, "Much as the atomic individual in political and economic discourse is simultaneously divested of sex and invested with the attributes of the 'universal man' (as if equality can prevail only in the absence of sexual differentiation), so too, the biological individual is undifferentiated, anonymous, and autonomous-assumed even to be capable ... of reproducing itself" .... Echo, like the mathematical ecology Keller is writing about, treats all organisms in a species as interchangeable.

So, if a model of reproduction is the culturally "familiar" one, its author is criticised for letting their heterosexual biases blind them to asexual alternatives. And if a model is "divested of sex", its author is criticised for letting their masculine biases blind them to heterosexual alternatives! Damned if you do, damned if you don't.

Another example that struck me as off-key was Helmreich's discussion of the use of the word "seed", in computational contexts such as random number seed and seeding an initial population. He claims this is deeply masculine imagery, because "seed" = "sperm", and all our imagery of "planting active seeds" in "passive, nurturing soil" is masculine Judeo-Christian imagery. Well, I'm sorry, but the imagery that raises for me is so old-fashioned as to border on the risible. (The bit about "Onan spilling his seed on the ground" has always given me a mental image of a sower dropping a bag of grain more than anything else.) Tom Ray picks up on this point, too, in his summary of Helmreich's position (based on reading an earlier draft), which expresses well much of what I'm trying to say:

p239. I find Helmreich's work to be a poor example of anthropology, because rather than trying to understand his subjects, he forces them into the template of anti-heroes in his own politically motivated world view. We could be replaced by anyone and his thesis would be essentially the same. Heimreich advances his own political agenda, the character of which is revealed by the literature that he copiously cites. His analysis has the feeling of a societal equivalent of a Freudian analysis. Rather than our behaviors being determined by hidden events in our infancy, he asserts that they are determined primarily by obscure events in the history of our culture, long before we were born. He devotes several pages to what he sees as the politically incorrect meanings and uses of the word "seed," meanings which he claims are inextricably and historically tied to the work and workers of Artificial Life. But Stefan never learned the meaning of "seed" which is in my mind. For his purposes he didn't need to know. Rather, he introduces a medieval image of rotten semen incubating in horse dung as though it somehow reveals something about his subjects rather than himself. I am tempted to conclude that Helmreich has not learned from his predecessor, Gusterson, who concluded that how researchers use language is "ultimately unconvincing as a key to their inner psychology" (to quote Helmreich's own summary of this position). As a result, Helmreich's book is more about the world according to Stefan, his own mind and world view and the politically incorrect demons that haunt him, than it is about his putative anthropological subjects. It is sad that anthropology has come to this.

In the very passage where Helmreich is discussing the imagery of the term "seed", he says

p115. The choice of the word seed--and of the word ancestor--is important, for it plants the idea that from the word tierra we are witnessing nothing less than life-forms in realized potential.

I draw attention to the use of the word "plants" in this extract. Is Helmreich guilty of using masculine Judeo-Christian imagery to thrust his potent ideas into our receptive open minds? Or is he merely using a common vegetative metaphor deeply rooted in our organic language?

Or maybe he is merely punning? I think not, for there is very little that is punning or playful about the style of the rest of the book. (Not necessarily a criticism, merely an observation.) But could that explain why he doesn't seem to "get" scientists' language? Scientists are always using terminology and metaphors from everyday use, usually playfully (strange, coloured quarks, anyone?). Earlier, there was the extract "the metaphor ... is gleefully emphasized", with a negative connotation. I would say rather "playfully emphasised". Just because scientists use familiar terminology in new situations, even deliberately metaphorically, doesn't mean they use these terms literally in their original senses. I admit that the choice of metaphor is influenced by culture (it could hardly be otherwise), and even admit that the original sense colours the eventual meaning, but not to the extent that Helmreich seems to be suggesting.

As another example of what looks like a misunderstanding of the very playfulness of scientists' terminology, on p186 we are shown, amidst a diversion into Westernised Zen and other religions, "evidence" of the fact that science fills a religion-shaped hole in the soul: Chris Langton is routinely referred to as the ALife "guru". Well, I'm sorry again, but guru is a common term in computing circles -- if it ever had any religious overtones, nowadays those have been lost. The word means, to most of us who use it, an expert, especially one who helps others, rather than keeping their arcane knowledge to themselves. Indeed:

guru n. 1. [UNIX] An expert. Implies not only wizard skill but also a history of being a knowledge resource for others. Less often, used (with a qualifier) for other experts on other systems, as in VMS guru. See source of all good bits. 2. Amiga equivalent of 'panic' in UNIX. When the system crashes, a cryptic message "GURU MEDITATION #XXXXXXXX.YYYYYYYY" appears, indicating what the problem was. An Amiga guru can figure things out from the numbers. Generally a guru event must be followed by a Vulcan nerve pinch.
The New Hacker's Dictionary, 1991]

One is rather left wondering why the use of "guru" is picked up on as indicating religious connections, but the equally common use of "wizard" is ignored? Why is there no suggestion that science stands in for our need for magic? Is it because that would just be silly?

I'm not hostile to all the ideas in here. But this kind of picking through the facts to support a position seems to me like reaching. I would instead have preferred a book that analysed how the researchers' racial, cultural and gender biases might be reflected in their work, how it affects what they study, the modelling assumptions they make, and their interpretation of the results. I would instead have liked to see how ALife research might be broadened, enriched and generalised by the inclusion of different biases. But then I would, wouldn't I?