Rediscovering Intelligent Design

I’ve referred this to PZ Myers, at Pharyngula.

Even if the anomalies are easily identifiable by contemporary biologists, I’d say they’re still likely to remain entirely obscure to the average “functional-equivalent of the Discovery Institute” — for the obvious reasons.

Without any more than an high school education in biology, my guess is going to be “no.” Because in the time it would take for this future species to evolve, all these genetically engineered species are going to evolve, cross-pollinate, etc, etc and it won’t be as simple as locating one species of really awesome corn or some such. Any trace of genetic interference by some intelligent designer will likely be erased over the millions of years it would take for the next human-ish species to come around.

Really depends on just how “exotic” they actually are. There’s a huge amount of horizontal gene transfer that goes on naturally (especially in the prokaryotic world where it has been going on at such a clip since the origin of life that molecular taxonomists have pretty well given up on being able to draw a “tree of life” with a single root; it’s more like a very complicated mesh.) Thus the kinds of examples you give of organisms with a single “exotic” transgene might not raise the future intelligent being’s equivalent of eyebrows that much.

There’s actually something more interesting embedded in the question, and that is the idea that “intelligent” beings would once again evolve after our demise. Being the vain species that we are we like to kid ourselves that the appearance of the trait we most value is somehow inevitable. But it’s quite possible that our big and intermittently functional brains are a one-off evolutionary accident (and perhaps dead end.)

lb,
That sort of thing happens all the time in evolution. Horizontal gene flow is not uncommon.

On the other hand, if vanity genomes were to become all the rage, design might be inferred from sequence data alone. I.e. sequences that encode “Healy was here” or other non-biological information. Of course, there would have to be a lot of it cropping up in genomes in order for it to be recognized that such sequences carried non-biological information in the absence of external knowledge of the encoding language. Some genomes might be recognized as carrying a unique class of information. However, without a biological function and selection the information wouldn’t stick around very long. In which case the tires would probably still be around as well.

Well, for one thing, there wouldn’t be much evidence of the tinkering. Most of the species that have been subject to transgenic work are either lab-bred (and doomed if they were to escape) or agricultural (requiring careful tending). It’s going to be nothing but an occasional weird exception that won’t obscure the nested hierarchy of relationships at all.

But for another, nature does this stuff all the time! We’ve got viral inserts in our genome, for instance. Here’s an example of a paper describing horizontal transfer of a transposable element from the squamates (lizards and snakes) to the ruminants (cows and that sort of thing) that occured 40 or 50 million years ago.

http://www.pnas.org/cgi/content/full/95/18/10704

We really aren’t surprised to see that sort of scrambling of DNA on a small scale, and we don’t see that and assume aliens visited earth and were trying to generate snake-cow hybrids*. Future intelligences who found the fragmentary evidence of our dabbling in genomes would certainly assume similar natural causes.

*Although, if they did, this would explain where snow comes from. Or cake.

9: Yeah, I was thinking of a gene complex showing up only in a set of unrelated plants whose common factor was not that they were related but that they were useful. Bt in corn by itself wouldn’t tell you much, but Bt in corn and cotton and strawberries, but not in close relatives of corn and cotton and strawberries seems like it would be a tipoff.

But this relies on ‘corn and cotton and strawberries’ or whatever, still being identifiable, and the Bt gene not having moved too far from the plants we engineered it into.

PZ Myers is right that most artificially produced GM varieties would probably die out or lose their unusual genes to natural selection (lactating silk proteins is not going to help transgenic goats survive after the Fall of Man; crop plants would be supplanted by the natural climax community flora in the area, and so on) or just random genetic drift.

However, I suspect that, if some GM varieties did survive, future molecular biologists would be tipped off by irregularities in the gene sequence around the insertion site. I’m not up on the latest methods, but many certainly used to use a “marker” gene or gene sequence that is transferred along with the gene you are actually interested in inserting.

For example, you would put an antibiotic resistance gene next to your insulin gene when inserting it into a bacterium, because that provides you with an easy way to select for modified colonies. In mice I believe they use a sequence with an unusually high frequency of recombination to help insert genes at a particular site.

I think a smart molecular bio-archaeologist might be able to detect the results of genetic engineering. As you point out, I think the actual old-fashioned archaeologists might find the layer rich in plastic to be a bit of a clue; not to mention all the stuff we’ve left on the Moon.

So, I will posit an experiment that I have considered for a long time, that is somewhat related. If anybody ever steals this and does it, please let me know, and at least acknowledge me (better yet, if somebody wants to collaborate…):

In brief: Take an FPGA. Come up with a problem for it to solve. Evolve a solution to the problem using a genetic programming technique. Also, design a solution to the problem. Repeat a lot. Now, use network measurement metrics like network radius, connectivity, etc, to find simple metrics that can be used to distinguish designed from evolved objects. Now apply similar metrics to biological objects. Voila! We can now determine whether biological objects are designed or evolved (assuming that such metrics exist, and we look for them in the right ways). To zeroth order, this is a test of intelligent design. But, more usefully, it is a way of determining what sorts of properties of evolved objects make them different from designed objects, and perhaps give them some of the neat-o properties that we like in evolved objects (like robustness.)

Why consider genetic engineering? Wouldn’t it be obvious to future biologists that cows, chickens, corn, cotton, marijuana, artichokes, dogs, and all sorts of other crazy improbable plants and animals were in close co-evolution with some other species? Just as we’ve been able to infer the existence of North American cheetahs in the recent past from the speed with which pronghorns run, these creatures would be evidence of our existence. Whether they’d be clearly distinguishable from things like rats, raccoons, and pigeons (not to mention cane toads and the like) in terms of having been at least somewhat designed for their characteristics, is less clear.

I would think that transgenic things, if they survived, would probably be even slightly more obvious, once you look at the actual gene sequence, but I don’t know about that.

What if they analyze some cabbage gene and find a video of Hitler opening the 1936 Olympics encrypted in it? And they have no knowledge of the previous civilization. Now what?

abb1,
That would be a big gene. The problem for the perpetrators of that sort of cosmic joke would be making the genetic encoding sufficiently stable to last a while. They could build in some sort of error-correction mechanism and couple it redundantly to some essential biological process.

Wasn’t it O.E. Wilson who thought that genes are the dominant life force on Earth and that species are evolved to meet the changing conditions of our environment? The genes will carry on in whatever living container that can survive a constantly changing (over hundreds or millions of years) set of ecological parameters. It wouldn’t be survival of species that matters, it would be survival of the genes.

Just another way of looking at the problem.

I have no reason to doubt that evolution on this planet has been a purely natural process, at least until the advent of genetic engineering. However, I wonder what evidence most Darwinists would accept as falsifying the thesis that there has been no intelligent design in the past — apart from the by-now tiresome fossil hominid in the Precambrian. From Popper’s standpoint, the falsifiability of a theory doesn’t depend simply on its making just any old testable predictions (including predictions of what won’t be found); these have to be risky predictions, and modern creationists of the ID variety, like Michael Behe, no more expect to find fossil hominids in the Precambrian than Richard Dawkins does. The temptation will always be to rule out ID on principle and simply say, “Well, this structure must have resulted from natural selection somehow or other.” I think most Darwinists would, in the end, much rather opt for space aliens or inhabitants of Atlantis than some god. And they might be right to do so. I wouldn’t rule out the possibility that this particular universe is a virtual reality game created by human beings (à la eXistenZ) — but how would we know? At the end of Carl Sagan’s Contact, irrefutable evidence of intelligent design is found buried in the number pi. Would it take something like that to convince everyone?

Also not a biologist, but in re horizontal gene transfer muddying up the picture:

Surely future intelligent species, if as intelligent as we are, would be able to come up with models of how fast these things happen, and project how far in the past the common ancestor of plants (or whatever) containing our inserted gene was (while, possibly, being puzzled that these plants did not otherwise seem to share a common ancestor). Surely some future-biologist, knowing this date, would run across a future-archaeologist who said, gosh, that date is the same as with all these rubber piles we have.

Biology wouldn’t be the only evidence of our existence, and the dating “coincidences” would eventually get obvious.

The temptation will always be to rule out ID on principle and simply say, “Well, this structure must have resulted from natural selection somehow or other.”

Ah, no, that’s not what biologists do AT ALL. They seek to find, and to support from observations and experiments, actual pathways by which “this structure” could have evolved. You’re confusing them with IDiots, who DO in fact throw up their hands and claim things were “somehow” designed by a who-knows-what that we’re not even supposed to ask questions about.

@matthew gordon:
Go for it!

I personally regard the FPGA evolutin stuff as pretty compelling (indirect!) evidence for evolution as the main force behind living organisms’ present form.

A friend-of-a-friend (no really!) did some stuff on evolving FPGAs to do relatively simple tasks, and it was a nightmare – you had to keep cycling the designs from one FPGA to another, or it’d evolve a design that critically depended on array-specific properties and so would only work on the board it’d evolved on.

When he tried to unpick how the evolved designs were working, it was a total mess of half-useful, half-bonkers kludges all on top of each other that somehow magically combine to produce a system of breathtaking complexity that does what it needs to.

At the time I was a neophyte biochemist getting my head round metabolic pathways, and the similarities in design style were painfully striking. No intelligence would, or could, design something like that.

(Except by getting evolution to do it for them, of course … which is the omphalos argument.)

At the time I was a neophyte biochemist getting my head round metabolic pathways, and the similarities in design style were painfully striking. No intelligence would, or could, design something like that.

Larry Moran recently discussed a nice example on his blog.

Without going into “Will the GM organisms survive in the wild”—suppose a packet of seeds ends up frozen in a glacier or something. The markers that many of y’all are talking about (evidence of actual gene-splicing trickery) have nothing in common with the “markers” that Behe et. al. claim to see.

IDers claim to see protein structures that simply look funny. They look at the flagellum and say, “whoa, there are, like, twenty pieces”; they look at the centrosome and say, “whoa, it looks sort of like a jet engine”. Humans may be designing our own organisms, but we are absolutely not designing custom proteins with irreducible mechanical arrangements. GMO proteins look and behave just like natural proteins … they’re the same blobby enzyme-y things with pokey active sites and inscrutable folds. For the most part, they’re developed, not by a Designer sitting around fitting parts together, but by mutation and selection—artificial, forced, or targeted mutation (sometimes!) and selection in a petri dish (usually!). I’m willing to bet that the “Round-Up Ready” gene produces some surface protein whose structure and modus operandi were not known until after it was found to cause Round-Up resistance. A future scientist looking at the structure of this protein would conclude, at best, “This gene evolved in a species which was exposed to thus-and-such a toxin; it evolved very rapidly, so must have been under intense selection.” He will not be able to say, “This thing has a totally different design principle than any other protein.”

The only artificial, engineering-like step is where the scientists decide to move the genes from one organism to another. If that’s detectable or not, I dunno—“this enzyme seems to be a trivial modification of an Icelandic hot-spring bacterium’s gene, including all of the translation-invariant choices. What the hell is it doing in wild North American cabbage?”. If there is any evidence of Behe seeing, or even looking for, such a thing, I’d love to hear about it.

Horizontal gene transfer might be invoked to explain one or two such oddities. If I were a a future crustacean version of Michael Behe, I might notice a systematic pattern of: lots of “apparent” horizontal-transfer events; all apparently under rapid selection, with sources (apparently) randomly scattered over the world but with targets concentrated in plains and pasture species; all occurring at about the same time (“Just before the late Holocene mass extinction”) as the peak of an ill-understood material culture. Weird.

38. Evolved systems and networks tend to have “scale free” like statistical properties that might distinguish them from most designed systems.
Limbs are arguably more versatile than wheels. Rolling behavior has probably evolved many times.

See Michael Mouse @ #33. Evolution tinkers, it doesn’t “design” from scratch. Wheels could be ever so efficient, but if you can’t get there from here using the genetic toolkit present in ancestral tetrapods, they still wouldn’t happen.

Dan Simon,
I’m not so sure. We are part of biological evolution to be sure, but that doesn’t mean our inventions can not have features or obey principles that distinguish them from other types of biological artifact. Statistical analysis has been used to distinguish forged from natural data for example. That doesn’t mean our artifacts are special, just constrained in ways that set them apart.

Dan, there’s a difference between saying the brain is a product of a Darwinian process and saying that what the brain does is a Darwinian process. The computer I’m using right now is the product of an industrial process. The activity of the computer in acting as a medium for my thoughts is not an industrial process, but a cognitive one and a technological one on its (or cognitive, if you like to think of computers that way). Said cognitive process relies on biological material that given by a Darwinian process and skills (English, typing, etc.) given by a cultural one.

If cognition is a Darwinian process, this means the brain generates ideas purely at random, has some means of determining fitness, and the ideas that “survive” are the ones you have. There may be a bit of that going on. But Darwinism claims comprehensiveness – even Behe concedes that natural selection exists – so for it to be a Darwinian process, what I just gave has to be a complete description, albeit at a low level of detail. Cognition appears to have much more to do with the emergent properties of structure, than with competition among atomic units, such as species’ or genes. There is competition in the sense that, over time, neurons not much used get fewer connections that those greatly used, but this seems to be how habits of cognition are formed, not how specific instances emerge.

As for spiderwebs, whale songs, etc. I have no trouble regarding these as creative acts in the same sense as human creative acts – they are products of internal cognitive structures of those animals. Whoever this “we” is that holds human experience so special, I am not in it.

drm,

Do you think there are problems that minds can solve that Darwinian processes cannot, for example, because there is no incremental path to them? Certainly, evolved and written software seem quite different, and the latter so far better for most purposes. I don’t know if the fact that less complex species than humans can use wheels necessarily solves the problem; the dung beetle wheel is still the product of a cognitive structure, it’s just a less difficult problem than our “discovery of the wheel” trumpeting would suggest. Evolution never solved the basic problem of getting energy from the sun as efficiently as photovoltaics do, so it does seems that a cognitive apparatus can get you places random variation with selection cannot.

Incidentally, I understand that some adaptationists do argue that religion (and therefore the concept of intelligent design) is a product of a Darwinian process.

“Do you think there are problems that minds can solve that Darwinian processes cannot, for example, because there is no incremental path to them?”

In my opinion, the “incremental path” argument against evolved wheels is flawed and inadvertently provides ammunition to ID proponents. Basically, I think the flaw is in thinking that we can reliably tell when something could not come about via an incremental path. After all, that error is essentially the entire opposition to evolution.

I’ve always thought that the real problem of an evolved wheel is that it’s such a weird structure, from a biological viewpoint. The wheel has to turn independently of its axis, which would be very complicated to achieve for living tissue. Alternatively, a wheel could be formed and then the living tissue could retreat to the axis. But then you’d have to ask exactly what kind of selective pressure could exist that would prefer this structure over others?

It’s not so much that there’s no incremental process that could allow the evolution of the wheel, it’s that it’s just so unlikely relative to others that would achieve roughly the same result. We humans are confounded by this example because we see the simplicity of the wheel and erroneously think that it is a simple solution to a simple problem in evolutionary terms. But it’s probably not.

And thus here I find I disagree with those who are saying that there’s really no way to detect the difference between designed and evolved structures. I agree that there’s probably not any strong qualitative distinction between the two (in terms of the structures, of course). But there’s a whole bunch of quantitative differences you can easily see. Designed structures have little general redundancy and they are usually “brittle”. That is, they easily fail completely outside of their intended environment. The key in that last sentence is to realize just how narrow “intended environment” is for designed objects as compared to biological structures. It may seem that an ecological niche is “narrow”, but not in the same sense as the “niche” in which an airplane operates, for example.

Arguably, however, these distinctions rely upon how an intelligent creature prefers to design! One can imagine a much more biologically-oriented worldview that both prefers explicit designs that mimic evolved biological structures, and/or prefers to use evolutionary processes for design purposes. In those cases, it would obviously be much harder to differentiate design from intelligence-independent evolution.

All of which argues that this notion of “design” versus “evolved” may be a bit of a red herring, distracting us from the qualitative distinctions that are more descriptive, as science should be. Creationism and ID really are regressive science because they explicitly embrace teleology. It’s not so much that teleology is essentially unscientific, as it is the case that teleology is so often unhelpful or misleading for scientific purposes. Just so with creationism and ID.

#49-
“I’ve always thought that the real problem of an evolved wheel is that it’s such a weird structure, from a biological viewpoint. The wheel has to turn independently of its axis, which would be very complicated to achieve for living tissue.”
Spend a lot of time in the woods and you’ll eventually see a tree that’s rotted from the interior out, and kept enough of its outer layers to be coherent, round and rollable. At least for a little while and distance.
All you have to surmise is finding something like that, rolling it around, not usefully as conveyance, but as a means of moving the wood itself, then eventually somebody jamming a large enough stick up the middle of something it and yoking the ends. The cylinder turns, the yoked axle doesn’t. Still not useful, really, but there’s your concept. We’re talking about people who spent all their time in the woods, generation after generation.
It seems pretty obvious the first wheels wouldn’t have been as two-dimensional as now, but cylindrical more. An axle wouldn’t be so bizarre in that context, especially after enough time and usage had occurred.

A significant portion of GMO crops contain a gene called “the terminator” and as a result the seeds are infertile (which helps maintain profit margins for the seed companies). I suspect that most of the plant transgenics will have disappeared by the time of your hypothetic new insect overlords.

Bakho points out that corn (maize for our foreign guests) is incapable of reproducing on its own. Several other crops are in similar situations. The varieties of bananas that we eat need human cultivation to spread (they lack seeds to begin with). I don’t expect any evidence of human tinkering with genetics to survive the first century after humans go poof, let alone survive to irritate future intelligent species.

I’m pretty sure if you engineered an organism so that one of its internal organs had pigmented regions such they read “hey, nerdlinger! suck my balls!” then the scientists of the future would be able to figure it out

How wide spread will engineered genetics be in 10, 20, 40, 200 years? If it grows at the rates the electronics industry has grown I suspect the foot print on the ecology will be much larger than the occasional natural horizontal gene transfer. Meanwhile, slightly off topic, I found the concerns raised regarding genetic engineering in the book Normal Accidents sobering.

Can’t a ball and socket achieve this?

Yes, in theory, but the kicker is that you’d still need some way of carrying fluid streams (with nutrients, waste and operational gubbins) across the joint. Otherwise the structure at the far end will die and rot in short order. A few hose-like structures would do it at a bare minimum. (Think blood vessels or xylem/phloem.) But those’ll get irretrievably tangled as the wheel rotates, so it’s not an option.

I reckon you could probably manage it with dead wheels, particularly if you could find some way of replacing them as they rotted. There’s a quadruped species in a Phillip Pullman novel (mulefa) that do something like this with a symbiotic plant, which is far-fetched but not, to my mind, biologically implausible in the way that living wheels *seem* to be. I stress ‘seem’ – I’m very wary of saying things are implausible – still less impossible – biologically.

This is a key point to bear in mind, whether you’re talking evolution or elsewhere. “I’ve never heard a good model of how it could possibly work” is very far from rock-solid proof that it it can’t, even with an added, “despite having spent happy hours with biochemists and biologists musing about how you could do it.” But it is jolly unlikely.

[Of course the obvious option – which you do find all over the place – is to turn the entire organism in to a single wheel. There are plenty of small, round organisms, and many larger organisms can form themselves in to a roundish shape and roll if the need arises.]

Is it unheard-of for living organisms to produce structures that are “dead” enough not to need nutrient streams? What about shells? What would be the fundamental obstacle to producing a hard inert substance, keeping in mind that it doesn’t have to last forever, just for the lifetime of the organism, so “rotting” is not necessarily a concern in the relevant timeframe. Calcium structures can do this, no? I realize that bones are living, but even when dead, they seem to persist an awful long time. I suppose this could be brittleness – if your wheel is dead, it will be harder to repair if it breaks. I also wonder why nutrients and waste could not pass between two lubricated mucous membranes for that matter, although to look at the question that specifically, we have to stipulate more specifically what we are talking about, I think, e.g. a large or a small organism, etc.

Agreed about never regarding ignorance, even expert ignorance, of how to do it as proof that it is impossible.