More Options

Very Like a… Machine?

Robert Camp

November 15, 2005

“One thing that literature would be greatly the better for
Would be a more restricted employment by the authors of simile and metaphor.”

These lines from Ogden Nash’s clever poem Very Like a Whale begin an irreverent piece about the overly florid tone of some poetry. Rejecting excessive use of such comparative devices as “simile and metaphor,” Nash neatly skewers ostentation and delights the reader with clever rhymes and silly verse as he calls for the clarity of simple observation,

“Authors of all races, be they Greeks, Romans, Teutons or Celts,
Can't seem just to say that anything is the thing it is but have to go out of their way to say that it is like something else.”

It’s likely that Nash, having been born a hundred years after William Paley died, and a hundred years before “intelligent design” (ID) became a household phrase, intended his remonstrations strictly for the poetic arts, but his points are well taken as they apply to certain aspects of the current debate over biological evolution. After all, both sides do employ comparative devices and analogies (e.g. biologists often refer to cellular systems as machines). Of course when a cell biologist says a particular protein structure is a little machine he is hoping to elucidate its structure or function. However, when ID proponents, from Paley to present, use the argument, they are hoping to imply much more. They mean to illuminate a structure’s purposive creation.

William Paley was, of course, the nineteenth century philosopher who made famous the “watch on the heath” argument. This was his contention that after discovering a pocket-watch lost in the verge, and considering its origins (presumably while on the way to the pawn shop), one would not be inclined to conclude that the watch was a part of the natural world. One would recognize, as a result of it’s complicated and obviously designed configuration, that the watch was the product of an intelligent agency. Paley was trying to logically support the corollary that man, owing to his own complex and interdependent innards, is also a product of design, albeit of the transcendental sort.

So it happens that Paley has, if not overused, then certainly overstretched a simile, in this case - men are like watches, therefore therein, ipso facto, and so on and so on. The modern “intelligent design” movement broadens this teleological analogy to include “order” and “information,” and de-emphasizes its application to gross morphology. In essence, proponents of ID have tried to update the man/watch metaphor in the hope it will be more scientifically, and legally, palatable. Thus, as regards the contemporary controversy, the important question is not whether someone like Paley can be forgiven this device on poetic grounds but whether the “intelligent design” argument can be so relieved on the merits of logic. In short, does the analogy hold up?

On this the relevant scientific evidence is clear, there are no data derived from investigation of large or small-scale evolutionary processes that require an inference to heretofore-unknown intelligent causal agency. However, “intelligent design” advocates, not satisfied with the provisional nature of scientific inquiry, believe that their “theory” can, and should be extended beyond the data to include some (particularly weak) inductive arguments. And in doing so, they end up commingling incredulity and intuition with a smattering of ambiguous but technical sounding jargon, producing a methodology that can sometimes, at a hasty, sideways glance, resemble science. But due to the presumed conclusions they are ideologically bound to observe, their “theory” never meets the standards of legitimate natural inquiry.

“Obvious,” or oblivious?

One of the most egregious examples of weak ID inductions masquerading as reason is the argument from analogy used in an attempt to establish the validity of a “design” inference. Biological systems and structures are routinely compared to casually recognized artifacts from the broad scope of human activity. The sculpted faces of Mt. Rushmore are an oft-employed example used to argue that inference of design is obvious and uncomplicated. An ID proponent’s hope is to persuade that there is no need to look more deeply into the comparison, that recognizing “design” on a biomolecular level is “very like” the immediate, instinctive identification of Abe Lincoln and his humongous pals as a product of intelligent engineering. In this vein Michael Behe once opined that,

“Design should not be overlooked simply because it’s so obvious.” — Michael Behe, NY Times, 2005

But, of course, this argument is quite flawed. Just a brief evaluation of such a suggestion — that transcendent (non-human) design can be inferred with the same ease with which we reach that conclusion regarding human design - reveals that it is little more than a request for exemption from those rules of investigation that would establish the validity of the proposition itself. It presumes exactly that which requires evidence.

In fact, why worry about evidence at all when there are so many convenient gaps in our knowledge into which a “design” explanation neatly fits? According to Behe,

“ the absence of any convincing non-design explanation, we are justified in thinking that real intelligent design was involved in life.” - Behe, NY Times, 2005

In other words, all of this is tantamount to saying “look, we know design when we see it, and you have no specific scientific explanation anyway, so we need not fret over producing evidence for “intelligent design, we’ll just assume it.”

But the deficiencies of the design analogy go deeper than mere misapplication. Consider, for a moment, the difference between making a carefully limited statement about a particular shared characteristic, and specifying a shared essential nature. An example will help make the point. In the process of proposing an analogy between my dog and my car, I might observe that both appear to be aligned on a head to tail axis with a dorsal and ventral (top and bottom) orientation. Likewise I might note that both have side-to-side symmetry. These observations are collectively referred to, in biology, as bilateral symmetry. It is a character used to classify organisms at very basal levels.

What is important here is the nature of the conclusions I am entitled to draw from these observations. I could say, for example, that this analogy allows me to infer that for both the dog and the car there may be a preferred direction (i.e. forward), that different processes take place at the opposing ends, and that both have a geotactic (up or down) obligation. These would be reasonable inferences based upon the number of characters taken into account. But I would hardly be justified in extrapolating from this that the dog and car are both animals. To make the argument that the car is an “animal” I need to provide far more than speculation about a couple of shared similarities with animals. I need to compare many, many characters across multiple examples. I need to marshal vast amounts of evidence (including replicated observations) that convincingly make the case. Otherwise I have stretched the usage of “animal” such that it is meaningless for purposes of comparison with “non-animals.” If a car can be an animal, then so can nearly anything else, leaving “animal” as a category without any real distinction, and leaving my analogical conclusion very weakly supported.

Of course if I compared two of the family pets, the dog and a turtle, I would note once again the bilateral symmetry. But in addition to this I could observe that both of their bodies appear to be based on a similar architecture, that of a head, a torso with four limbs, and a tail. Even more, both have integument-enclosed internal systems. There are many additional qualities shared here, and consequently my analogy can be used to infer greater relatedness. They’re both biological organisms, they’re vertebrates; they share spatial orientation characteristics etc., etc. But although the dog and turtle can be considered animals, I cannot draw from this analogy either that they are both turtles, or both dogs. The number of shared similarities doesn’t justify going that far.

The point is that the strength of any analogy depends, in significant measure, upon the nature and number of shared qualities between the phenomena involved. The more similarities between the object of interest and the object to which it is compared the more powerful the analogy, and the more persuasive are the arguments based upon it.

Yet there is an even more glaring flaw in the “machine” analogy. It suffers from a veritable dearth of relevant examples. Just as an increased number of compared similarities improves an analogy so does an increased number of instances of comparison.

Returning to the example of the dog and the car, if my analogy is used to establish that the dog is a human-engineered machine then a comparison with human engineered machines will serve my purpose. But if I’m trying to establish that the dog is “designed” by an other-than-human intelligent agency then I must compare it not only to a car but also to a broader set of intelligently designed artifacts, a set that includes examples of non-human design. In this way, those characteristics diagnostic of “intelligent design,” not simply human design, can be understood and compared. But for obvious reasons ID proponents rely exclusively on the many examples of human design, including mountainous heads and other HEDs (human-engineered devices) of bewildering variety, even though these can hardly be used in the service of an analogy meant to evidence a kind of “design” the existence of which is yet to be demonstrated.

Thus, it doesn’t matter how many metaphors for “machines” we glean from biomolecular structures. The only machines we know (barring some few possibilities from the natural world which do not serve the “design” argument anyway) are human designed. To draw from the machine analogy an inference of “intelligent design,” rather than “human intelligent design,” one needs to provide other examples that sustain this comparison. A selection of extraterrestrial machines would be a good start. But for the best analogical dollar-value, an empirically verified supernaturally designed artifact would really pop the cork. This collection of exemplars, all sharing established “intelligent design” characteristics, would give the ID machine analogy some substance.

What is an “obvious” inductive argument to Behe and other design “theorists” is, to most of the rest of us, an overarching inference that challenges the way we approach natural investigation. Sweeping comparative inductions demand a sweeping set of comparisons. Absent that level of support, the design analogy is a specious appeal to incredulity.

Why settle for half a loaf?

So, do ID proponents accept the limits of the analogy, and merely suggest that some biological structures share some similarities with machines? Heck, no. Take, for example, this from Michael Behe,

“I mean, literally, there are real machines inside everybody’s cells and this is what they are called by all biologists who work in the field, molecular machines. They’re little trucks and busses that run around the cell that takes supplies from one end of the cell to the other. They’re little traffic signals to regulate the flow. They’re signposts to tell them when they get to the right destination. They’re little outboard motors that allow some cells to swim. If you look at the parts of these, they’re remarkably like the machineries that we use in our everyday world.” — Behe, Understanding Creation, Evolution and Intelligent Design, 2005

Or this from Jonathan Wells,

“In the electron microscope, centrioles look like tiny turbines. Using TOPS as my guide, I concluded that if centrioles look like turbines they might actually be turbines.” — Jonathan Wells, Using Intelligent Design Theory to Guide Scientific Research, 2005

Or this from William Dembski,

“There are structures in the cell that don’t just resemble humanly built machines-they actually are machines in every sense of the word.” — William Dembski, Five Questions a Darwinist Would Rather Dodge, 2004

“Every sense of the word?” We don’t see a lot of equivocation or restraint in any of the above statements. But does Dembski really believe that certain cellular structures are machines in “every sense of the word?” Of course not. He wants his reader to take from this comment the sense that for him is most important, that these things had to have been transcendentally designed. Does Wells believe that calling a centriole a turbine (rather than simply noting that the centriole acts like a turbine) is an act of rigorous scientific observation? How could he? A “turbine” is a specific man-made device. Regardless of the degree to which centrioles act like turbines they cannot be said to “actually be turbines” unless one is proposing either that centrioles are designed and built by humans, or that the word “turbine” has no specific meaning. It is semantic sophistry. It is partisan rhetoric, not science. Wells clearly hopes that this suggestive terminology will communicate the need for his “designer.”

And does Michael Behe really think that there are little vehicles with wheels and expletive-shouting drivers tooling down the cellular interstate? I sure hope not. But he apparently feels this analogy is inductively justified, and because of his precommitment to a belief that is scientifically irrelevant he is willing to stake his reputation on such foolish argumentation. The logical grounds for such an inference are woefully inadequate to his purposes. We cannot say that both a dog and a car share an essential nature called “animal” from which we draw inferences as to their ontogeny (what, the car developed from an embryo?). And neither can we infer that both are “machines,” birthed at the behest of a “designing” creator. The same proscription applies to any similar comparison using biomolecular systems.

So what’s the big deal?

There is a vast reservoir of data and documentation pertaining to the biological disciplines. There are many, many journals, textbooks, scholarly papers and databases full of observations, methods and conclusions drawn from the experiences of biologists. Even a cursory examination of this work demonstrates that similes, metaphors, casual analogies and teleological references are ubiquitous in the literature.

Professors often refer to evolutionary processes using terms such as “try” and “goal” (e.g. “the organism’s goal is to pass on its genes to the next generation) as if biology acts in accord with a plan. And scientists frequently use similes to characterize the functions of particular structures,

“Why do we call the large protein assemblies that underlie cell function protein machines? Precisely because, like machines invented by humans, these protein assemblies contain highly coordinated living parts.” — Bruce Alberts, Cell, The Cell as a Collection of Protein Machines, 1998

Though there may well be a few biologists who use this terminology because they feel it reflects higher order purpose, for the vast majority of scientists it is simply a device of convenience, a way of avoiding the awkward phraseology that would come from having to launder their language of all shorthand teleological references.

And this clearly leaves the door wide open to nearly limitless opportunities for quote-mining and the dissemination of deliberate misinterpretation. These efforts certainly do nothing to undergird the empirical framework of ID “theory,” nor are they really meant to. Comments such as those from Behe et al above do not reflect thoughtful scientific inquiry; they reveal the bias of an agenda. They are slogans intended to bolster the confidence of the faithful and nudge the less educated fence-sitters.

Some biologists find the situation intolerable. In a more serious echo of Ogden Nash’s sentiments, Indiana University professor Rudy Raff wrote recently,

“...let us not play into the hands of ID propagandists. For instance, be careful about using teleological words to describe biological entities in our teaching and writing. Calling cells “machines that do X” or describing biological structures as “well designed to do Y” will be duly cited in ID propaganda as one more biologist-supporting design.” — Raff, Stand up for evolution, Evolution and Development, 2005

Although I am sympathetic with Raff’s goals here (and it certainly would not hurt for most people writing and thinking about these issues to order and communicate their thoughts more appropriately) it’s hard to imagine his admonition ever really taking hold. These types of references are simply too convenient, and natural for humans who view the world through pattern seeking eyes, to be easily weeded out of the language or the literature.

The fault lies with those who would twist these innocuous semantic devices to their own ends. It happens with enlightening infrequency that someone misapprehends the contemplation of a physicist who opines that the universe “needs” dark matter, or the chemist who describes negative ions as “wanting” to bond with positive ions. That this convenient phraseology is so often abused by those who wish to attack evolution says something about the motives of the individuals involved, not about the nature of the science.

It is not the intent of “intelligent design” proponents who push the machine metaphor (and the selective quotation of scientists who use it) merely to highlight particular similarities or even to dramatize biological characters for the purpose of clarity. Their goal is to appeal to intuition by way of provocative imagery, and so persuade without presentation of evidence.

Thus it is clear that the use of the “machine” analogy is simply another ID marketing strategy. As with other for-public-consumption arguments such as “teach the controversy” and “academic freedom” and “ID is science just like SETI,” their goal is the cynical manipulation of popular opinion.

Biological systems may indeed be considered similar to machines in some broad senses, but they are clearly not machines in the way design proponents (who often display computer crafted machine-like graphics of structures such as the flagellum) wish to imply.

“If you actually look at these “machines", they don’t resemble anything at all from our macroscopic world. Well, except maybe these: [Picture of pop-bead toys] Remember those? Maybe you played with them as a kid, or have kids who have tinkered with them. They're just little beads of various shapes and colors with a knob on one side and a socket on the other, and you can string them together by popping the knob on one into the socket on another. Look closely at those things Behe is calling “trucks” and “busses” and “traffic signals", and what you find are pop bead necklaces, or proteins.” — Paul Myers, Genes, machines, mutations

The analogy with machines is meant to lead the credulous past those pesky empirical details Myers describes to a position where their natural inclination to see patterns and purposes in their existence kicks in. It is typical of how ID argumentation appeals to emotion and vanity.

And it is illustrative of how the ID movement virtually ignores research in favor of pursuit of public sentiment, demonstrating further that “intelligent design” theory is not “very like” science at all.

Robert Camp

Robert Camp is a freelance writer living in San Juan Capistrano, California. A selection of his work can be found at the Nightlight Blog.