« The Concept of Design | Main | Asics »

Monday, May 25, 2009

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Bill,

I have a quibble with point 2. How do we know there is 'no such agent involved in natural selection'? I understand that the theory itself may make no reference to such an agent, but I can give quite a lot of explanation of how a computer works without needing to make any reference to an agent either.

It seems to me that natural selection is not teleological, but it's also not "not teleological" - at least as far as science goes. If there exists an agent who is using evolution as a means to an end (and a model example is given with 'artificial selection'), then natural selection is teleological. Or, perhaps, "natural selection" understood the way Mayr sees it, does not exist - all that exists is artificial selection on various levels.

I don't deny that someone can approach the question of evolution with an atheistic view and imagine natural selection as non-teleological (though 'teleological but the designing agent is Mother Nature' seems to be one option sometimes chosen), but at that point we're out of Ernst Mayr's field (science) and into another (philosophy/theology), even if there's some overlap.

I don't think this is just a quibble about words, but I also think there might be good reason to "loosen up" our use of a variety of terms that are associated with telelogy.

As a first step on a long and winding road, we could start with the notion of a natural function, as distinct from artificial functions. Artificial functions are the product of conscious design directed toward the realization of goals represented in counsciousness. Because it is directed toward the realization of goals, the process of design is clearly teleological.

The process of natural selection, in contrast, is not goal directed -- it's not teleological. But we have good reasons to think that a process of natural selection can, in principle, result in the genesis of things to which we attribute natural functions, things like wings, hearts, eyes, and so forth. If we can abide natural functions, it would not seem a great stretch to talk about the natural purposes of things that possess the former, the realization of their functional capacities being the "what it is for" of such things. Now some would say that "what it is for" is itself a teleological locution. If so, I'd be inclined to say that what we've uncovered is a reason to speak of natural teleology as applied to the products of a non-teleological process.

This is a fascinating post. From my readings of more recent (meaning, still living biologists) Mayr is no longer taken as seriously as he was. Larry Moran, for example, thinks natural selection is overated as the 'more important' mechanism of evolution (and John Wilkins has had a lot to say about Mayr's views as well).

None of which is to take issue with your main point, as Dawkins pretty much came of age, if I'm correct, whem Mayr's views were more dominant.

When we humans design, we go through a process that has a lot in common with evolution.

First, we imagine the desired goal. Then, we brainstorm. We mix up ideas (as if at random), giving some priority to ideas that have worked in the past. Then we simulate the function of each combination of ideas, weeding out the ideas that plainly can't meet the goal. Finally, we implement and test the ideas we think have a good chance of working to see if they meet our original goal.

Evolution is very similar to this process except that there's no simulation, and no imagination of the final product. Evolution has to implement every combination, even the ones that won't work, because it cannot tell ahead of time what is going to work. Furthermore, because there's no "goal comparator" to see if the final product meets the goal, there's only one goal that's compatible with mindless evolution: survival.

I agree that evolution is not teleological, and does not "design". For a process to be labeled "design", the process must have some ability to simulate for fitness relative to a goal. Evolution doesn't have that.

You say... "The products of evolutionary processes, therefore, exhibit no design."

That's an interesting phrase. IMO, no product "exhibits design" per se. Rather, a product may exhibit utility relative to a goal. (For example, fruit bats exhibit superb utility relative to their survival niche. A pocket watch exhibits great utility for accurately measuring time and displaying the time visually.)

However, idiomatically, we often refer to "products-with-utility" as "designs". I think Dawkins and others do this with evolved products because evolution is a mechanism that creates products with survival utility. When scientists refer to evolved products as designs, I read it as an idiom. When scientists say that rattlesnakes are a good design, they mean that rattlesnakes are products with good survival utility.

Hi Bill,

I'd like to draw attention to a point that is frequently overlooked (though it may not be overlooked by you). The point is this: that a process is not teleological does not justify us in claiming that the products of that process are not teleological. For instance, the evolution of language may not be teleological, but even if that is so, that does not justify us in claiming that language resulting from this process is without purpose or end, or that specific sentences within given languages cannot be evaluated based on their approximation to certain ends.

*Isn't* this just a terminological quibble? Evolutionists use the word "design" because there is simply no better word in English for what natural selection does. The products of natural selection typically exhibit "engineering" of mind-boggling ingenuity and refinement. Bob has it right, I think: non-teleological processes can still lead to things that are "for" something (a point I have been making a nuisance of myself about, in these pages, for years). The key idea is that all living things can be seen to have "interests" (even if they don't "know" it themselves).

As far as I am concerned, the word "design" captures this perfectly. If we knew that a *human* engineer had drafted the plan for a eukaryotic cell, or a mammalian kidney, we would be agog at the brilliance of his "design". We refrain from applying the term, I think, entirely because of our anthropocentric insistence that only purposeful agents can produce "design" (an atavistic relic of a pre-Darwinian mindset, it seems to me, but I certainly don't wish to offend anyone). I would simply expand the meaning of "design" to include the evolutionary "design" of living things; indeed I think most evolutionary scientists already have.

But if we really cannot get ourselves to accept the word "design" for the products of natural selection, then all that means is that we need a new word: one that signifies "'design' that is not of teleological origin".

I'm happy to give up the word "design" if it is to be irredeemably freighted with so much baggage, but certainly a process that can give rise to such "engineering" marvels as the albatross's wing and the human nervous system deserves a little credit, and a suitably descriptive word.

Bill,

I posted the previous comment before seeing your earlier post on the same topic.

I quite agree that it is important to be clear about teleology, and you correctly remind us that evolution by natural selection is non-teleological. As Court Merrigan pointed out in his comment to your previous post, the problem, really, is that it is hard to speak, in a language whose very fabric "requires us to speak in terms of intention, purpose, agency, etc.", of a non-teleological process of "design".

So let's, then, have a new word for what evolution does, so we can get out of this ditch we're in.

It's a quibble which uncovers a vastly larger issue - namely, that "these processes accomplish all these things without any foresight or intention" is a statement of assumption, of axiom, rather than a result of examining the evidence or scientific investigation. And that means all the difference in the world when discussing teleology, especially when the name of the game is relaying the discoveries of science. If science doesn't discover a lack of teleology, but rather some adherents presuppose it, it changes the discussion dramatically.

Joseph,

"If science doesn't discover a lack of teleology, but rather some adherents presuppose it, it changes the discussion dramatically."

There is no teleology, and we can be virtually certain of it. Here's why.

Suppose I have a deck of cards, and I tell you that the deck is either shuffled or sorted by suit and rank. I start dealing cards off the top of the deck. 2, 3 ,4, 5, 6 of clubs. Is the deck sorted or shuffled?

It would be irrational to conclude that the deck is shuffled. It is possible (one part in 310 million, I think), but incredibly unlikely.

More generally, when we're considering two possible scenarios, and there are many outcomes consistent with one scenario, and only the observed outcome consistent with the other, we're rationally going to side with the latter. And there are many (trillions at least) ways of designing life instead of evolving it. First of all, a designer doesn't need descent, let alone common descent. We don't breed cars, we manufacture them. Second, a designer can use completely different architectures and raw materials from one design to another. Rabbits could be plastic, and deer could be nuclear powered, but they're not. They're blood relatives of ours, sharing most of our DNA. Finally, the utility of life is, as far as we can tell, survival, which is the one utility consistent with evolution.

This means that Darwin's theory of evolution for all practical purposes rules out teleology. And it's not just a wishy-washy philosophical argument. It's an argument that has an extremely high probability of being true. Theistic evolution is implausible.

doc logic - I think you overstate your case, at least a bit. The theory of evolution by natural selection does not "rule out" teleology. Instead, it renders it superfluous within the domain addressed by the theory of natural selection. In other words, we have no need for the hypothesis of a designer to explain the origin of adaptations that could, in principle, have arisen by a process of natural selection.

Bob,

I disagree. Although my argument would not apply to say, the ultimate laws of physics, it does apply to anything and everything that leads to us specifically. In other words, we are accidents.

And I think the argument goes way beyond saying that design is superfluous. That's like saying that, in the card case, the lucky shuffler is superfluous. We would not say that the Sorter theory merely makes the Shuffler theory superfluous. We would say that the Shuffler theory is all but ruled out.

doc logic - I think that if we were to pursue our disagreement to the bitter end, it would come down to a question about when the absence of evidence is evidence of absence. That's a thorny issue itself.

Doc Logic,

I'm sorry, but it absolutely is a wishy-washy argument.

You say there are "trillions of ways" of designing life instead of evolving it. My response is that there are "trillions of ways" to design life by using evolution. You give examples where we don't 'evolve' cars - even if I were to grant that, there are plenty of examples where evolutionary processes are put to use by humans by design. Evolutionary algorithms in computer programming, procedural design concepts, artificial selection, etc.

Really, all you've told me here is that biological common descent is true (I agree - so what?), and that any purposes of evolution are difficult to discern (I agree - science is helpless in this regard, and is incapable of finding design or its lack on this level). Sadly, that's not at all persuasive. The only way evolution is consistent with a non-teleological worldview is by sheer force of imagination, rhetoric, and metaphysical pre-commitment - which has a place in philosophy and theology, but it's absolutely not science. What's all the worse for the non-teleological position is that, as we've seen, we know that minds are capable of accomplishing quite a lot with evolutionary processes. We have no such evidence that utterly mindless processes even exist, much less what they're capable of - those are assumptions. In science, it's a more pragmatic one, but an assumption nonetheless.

Bobkoepp,

The problem I have with your view is that even it goes too far. Both a designer as well as the lack of a designer is superfluous to a scientific understanding of natural selection. Yes, we can imagine (or say we imagine) these things 'just happening' - but how useful is that? I can imagine a full-blown YEC scenario 'just happening' as well (Indeed, once we start getting into many-worlds scenarios, it becomes even easier to imagine that.)

Keep in mind here, my objection here was in the scope of science itself - when Mayr starts talking about the presence or lack of teleology, he's stepped outside the bounds of science. There's simply no way to test for what he's supposing; the only way there is to assume it. And I'm not denying him the right to assume such things or have that worldview. We just should not pretend it's a finding of science.

Joseph - You say, "Both a designer as well as the lack of a designer is superflouous to a scientific understanding of natural selection." I'm not sure how this relates to my earlier remarks, so I don't understand where or how I go "too far." Can you clarify?

Bobkoepp,

Perhaps I misspoke with saying you go 'too far'. Rather, I'd just point out that both a designer and a lack of a designer is superfluous to the theory, insofar as the theory remains scientific. Along the lines of how I can take an artifact that is certainly designed (say, an artificial heart) and get very, very far in describing it without needing to make reference to a designer. Assume the thing popped into the world uncaused and ex nihilo - you're still going to be able to say quite a lot about it.

Joseph,

"My response is that there are "trillions of ways" to design life by using evolution."

There are trillions of ways that evolution could proceed, but for every one of them, there are far more design alternatives.

Consider the step from amphibians to reptiles. A designer does not need to use common descent, and so the designer does not need to incrementally modify the construction and DNA of amphibians. For every survival solution found by evolution, there are a huge number of design alternatives - different materials, different utilities, different architecture, manufacturing, etc. For example, there's no need for rabbits to be related to us. There's no need for horses even to be related to early proto-horses. So, even accounting for the large number of ways evolution can proceed, the number of ways of designing life is still many orders of magnitude greater.

The challenge to you is to show, for each evolutionary history, there are as few or fewer design pathways. Given that evolution imposes extreme constraints on pathways, I don't see how you can accomplish that. The number of design pathways simply dwarfs the number of evolutionary ones.

"there are plenty of examples where evolutionary processes are put to use by humans by design. Evolutionary algorithms in computer programming, procedural design concepts, artificial selection, etc."

Genetic algorithms are used in an extremely small number of applications, and they're typically used when we humans don't have an alternative design methodology, i.e., when we're not smart enough to do the alternative. Artificial selection isn't evolution, even if it operates on evolutionary products. And, even if we bend the definition of evolution to include artificial selection, we only do artificial selection because we are incapable of designing (or constructing) the ideal domesticated animal from scratch. Finally, we're interested in evolving something to meet our own needs, e.g., for food, clothing, radio efficiency, etc. Generic survival is not something we deliberately evolve for. We don't breed dogs or cattle for survival.

Joseph - Since evolutionary theory is explicitly about the origins of various organic forms, we can't just assume a studied ignorance about such matters. The theory of natural selection affirms the "facts" of trait heritabilty, variability in heritable traits among members of a population, and the causal relevance of trait variability to differential survival/reproduction. It doesn't affirm or deny anything beyond what can be inferred from those basic premises. It isn't even part of the theory proper to claim that there are no other factors that are relevant to the history of organic evolution (i.e., there is no "closure clause"). If that's what you mean when you say, "... both a designer and a lack of a designer is superfluous to the theory," then, OK.

Doctor Logic,

"There are trillions of ways that evolution could proceed, but for every one of them, there are far more design alternatives."

And there isn't a single way evolution could proceed that could not itself be the desired outcome of an agent. There is nothing - honestly, truly nothing - about evolution that demonstrates the lack of agency or guidance while still being a purely scientific examination. The claim of 'no teleology' or 'no agency' is an unverifiable assumption that adds nothing to the science. It can be discarded.

Your mistake here is that you, frankly, are assuming what you have to prove. You say that 'that are trillions of ways that evolution can proceed, but for every one of them there are far more design alternatives'. First, that presupposes the intentions of the designer - if the process is part of what the designer wants to implement (I don't just want a dog, I want a dog that comes about by these methods), that's that. Second, the whole argument here is that 'evolution' IS a 'design alternative'. Trying to make evolution compete with design when the claim is that evolution itself is a design process gets nowhere.

"The challenge to you is to show, for each evolutionary history, there are as few or fewer design pathways."

No - the challenge to you is to show how to determine that evolutionary events in history are not the result of agency, teleology, or intention. And what's more, to do it while staying within the scientific realm. I'm not arguing that science proves design - I'm pointing out that science is incapable of settling the question in either direction, *even when evolution and common descent is taken as a given*.

"Genetic algorithms are used in an extremely small number of applications, and they're typically used when..."

Sorry, but questioning when we humans choose to use such methods gains you nothing. We can choose to use certain methods for the sake of art rather than efficiency - we can and do make choices that are more involved and for which more efficient means to a particular end are available. It all comes down to what we intend, how we intend it, what resources are available, etc. Once more - this is an area science crumbles at. If you want to insist, say, that design is not possible because there's no way any God would choose to create a dung beetle, that's your prerogative. If you want to say that the proof of dung beetles is scientific evidence that God is not involved in nature, the abundant problems with that claim are easily exposed.

Bobkoepp,

I think we're on the same page here. The problem is that, clearly, this is not how the theory is popularly taken.

I think there might be an overlooked point in this discussion. While evolution fails to proceed by any overt, overarching teleological "forward pull" ala-Hegel, I would argue that it makes sense only on a teleological picture. We take organism A, which is in its native environment. There is a whole population of organism A in this environment, doing quite well for themselves. All of a sudden, a gigantic meteor falls from the sky. As a result, much of the native habitat is transformed into an alien space for organism A's population. Let's offer that, given natural selection, organism A has some random mutation "f1" which allows it to survive in the modified habitat. Similarly, many other organisms have this "f1." But what allows organism A to survive? Is it truly "chance?" Of course not; it is "f1," or more precisely, A's possession of "f1" that allows A to survive his potential meteoric demise. But this strikes me as plainly teleological, explaining the survival of organism A in light of a salient feature or function.

This does not bring in a designer question, but we can easily use this notion of its teleology to justify use of phrases like "the marvelous workings and design of organism A" to explain its particular formal character. It has all sorts of functions which operate, truly, "in harmony." This strikes me as a non-controversial "naturalistic" teleology of evolution.

Joseph,

I don't think I'm making the assumption you're accusing me of.

You're saying that it's not impossible for a designer to use evolution to accomplish a design goal. I totally agree.

Dealing 5 cards in order off the top of the deck doesn't prove with absolute certainty the deck wasn't shuffled. Maybe the preparer did indeed shuffled the cards. Science cannot tell the difference between a shuffled deck and a sorted one when the card sequence is identical. However, science can say whether or not it is *likely* to have been shuffled.

Likewise, we can come up with a scenario in which a designer uses evolution as the mechanism of design. However, we have to create a fine-tuned model to do this.

Here's what I mean by fine-tuning. By analogy, when the shuffling proponent says that the deck was shuffled, the proponent is arbitrarily fine-tuning his shuffling model to refer only to those cases of his shuffling model in which the first 5 cards on the shuffled deck turn out to be 2, 3, 4, 5, 6 of clubs. The proponent is right that the deck is possibly shuffled. However, this fine-tuning of the solution space doesn't escape the probability argument at all. Just because the deck is possibly shuffled, and we can't tell a shuffled deck from a non-shuffled one (when the card orders are identical), doesn't mean we don't know (with high probability) the deck wasn't shuffled. We do know the deck is very probably sorted.

Fine-tuning can beat the odds when the fine-tuned model successfully makes predictions that are unlikely to be true a priori. Evolution has done that because descent (and even more so, common descent) is not likely to be true a priori. In contrast, design advocates are happy to fine-tune their model of a designer, but they can't make any predictions to justify the fine-tuning.

Doctor Logic,

"Science cannot tell the difference between a shuffled deck and a sorted one when the card sequence is identical. However, science can say whether or not it is *likely* to have been shuffled."

No, DL - it cannot. Not when science is in no position to know what the intentions of the designer are, what degree/power of intervention/foresight/etc said designer may have had available and how much was exercised. A 2 3 4 5 6 of clubs isn't the only order a designer may specify - they can specify and arrange any order they please. Your shuffling example doesn't even begin to do the job you need it to.

Again you make the artificial contrast of 'evolution' versus 'design advocates'. But no 'design advocate' needs to reject evolution - and when restricted to actual science (rather than metaphysical presuppositions and speculation that simply reference scientific findings), the theory of evolution offers no evidence or argument against design whatsoever. Let me repeat: The truth of common descent and the truth of evolutionary mechanisms (even while noting that the power of 'natural selection' is under dispute even by orthodox evolutionary biologists - I'm willing to let that slide for the purposes of the discussion) does nada, nothing, zero, zip to demonstrate a lack of teleology, intention, mind, or design at work in natural history. It can neither affirm nor deny these things and still remain science.

And, one more time: I don't think ID is scientific. But the dark and dirty secret is that ID's opposite, no-ID, is not scientific either. Both make claims that go beyond the bounds of science, are extraneous, and thus are removable when discussing the science and research purely.

Malcolm writes,

>>*Isn't* this just a terminological quibble? Evolutionists use the word "design" because there is simply no better word in English for what natural selection does. The products of natural selection typically exhibit "engineering" of mind-boggling ingenuity and refinement.<<

Malcolm, you are just not getting it, and this is why I see no point is discussing this or related issues with you. This is not about words but about the concept of design. I just quoted an expert in the field who makes it clear that natural selection is not teleological. He refutes your view, so you just ignored him. I also made it clear in an adjacent post that it is nonsense to speak of a design without a designer, as nonsensical as speaking of a divorced person who was not previously married. But you don't get it. Nor do you respond to what I say. You simply repeat yourself, opposing what I say without supplying any reason to think that what I say is incorrect.

Can you understand that simply opposing what someone says does no good?

Now I know you are not stupid, but you are coming across like a sophist. You speak of "engineering." But you enclose the word in sneer quotes which signal that you don't mean it literally. So are you joking? Playing some linguistic game? There is no engineering without an engineer (or group of enginners). If you don't agree to that then there is no point in discussing these matters with you. A fruitful discussion cannot be had with someone who denies the self-evident.

Is there "engineering" without an engineer? Yes, in the following sense: something can appear AS IF it had been engineered even without an engineer. Similarly, something could appear AS IF designed even with no designer. But that is not the issue. The issue is whether natural things have been designed.

The very fact that you cannot make your points without using sneer or scare quotes shows what is wrong. You want to play fast and loose with words like 'design' and I'm not going to let you get away with it. That's the issue in a nutshell. You are playing the same miserable game that Dawkins plays ans I exposed in my post above. You want to inflate the word 'design.' But inflated terminology like inflated currency becomes worthless. Would you inflate the word 'divorced' to cover both those who have been previously married and those who have not? But why not? It might have the good effect of increasing the membership in divorced persons clubs.

Take a fruit fly. Observe it carefully. One might be moved to exclaim, "What a wonderfully complex piece of engineering! So complex and yet so tiny!" But, for a naturalist like you, that cannot mean that there are one or more intelligent beings who engineered fruit flies for some purpose. And if you say that Mother Nature did the job, or natural selection, or evolution, then you are committing the fallacy of hypostatization. There is no agent called 'Natural Selection' that designs things like fruit flies.

What you can't see is that you talking bullshit without even knowing it. And you don't seem to grasp that the whole thrust of evolutionary theory is to expell final causes from nature. Final causes come into the picture only with purposive beings like us. We have purposes and we can then look at birds' wings and such and think of them in teleological terms: the wings are for flying; the nests are for sheltering, etc. Teleological talk can be provisionally fruitful, but in the end the naturalist must be able to reduce it to non-teleological talk.

Diversity in species is a good, and evolution in unintelligible except as a process that gives rise to a diversity in species.

see here:

http://newadvent.org/summa/1047.htm#article1

The reason why we can't see the obvious end of evolution is because we assume the erroneous metaphysical position that a multiplicity of things is only accidental, and not a real perfection.

Doctor logic,

The reliability of science to judge whether a deck of cards has been shuffled or sorted presupposes the intentions of an agent. Without previous knowledge of how decks of cards are typically ordered by their creators, we would be in absolutely no position to discern whether or not any particular deck showed signs of intentionality. All your example proves, it seems to me, is the completely trivial truth that we can reliably recognise certain examples of teleology when we know what to expect.

The real question is not if science is competent to decide whether any particular teleological model is likely to be true. In certain circumstances, it can adjudicate with some reliability, such as in the case of the deck of cards, whereas in others it can't, such as in circumstances where the agent's intentions are unpredictable or inscrutable.

What's at issue here is whether science is in principle able to decide, once and for all, whether there is any teleology in nature. I simply cannot see how this is possible. The only proper basis for the wholesale banishment of teleology from nature is through the adoption of a metaphysical system that is incompatible with it, the most obvious candidate being some form of materialism. But that leads us outside of science and into the realm of philosophy.

***

On a naturalist worldview that denies teleology in all its forms, it isn't enough to simply say that natural selection is non-intentional, and go on talking about organisms as if they engaged in genuine intentional behaviour. The bird doesn't build the nest for it's chick, because nothing can be said to behave FOR anything; the lion didn't fail in its intention to tackle the gazelle, because the lion didn't INTEND anything; Napoleon didn't plan to invade Russia, because human beings are non-intentional, material objects, and cannot be said to PLAN anything, to act FOR some outcome.

As David Stove said (Bill quoted him elsewhere), for the sake of honesty and clear thinking, those who aver a non-teleological interpretation of evolution owe us a vocabulary that is consistent enough that it avoids presuming intentionality in its descriptions and explanations, and that doesn't end up obfuscating the strange, deeply unintuitive consequences that attend such an uncompromising denial of teleology.

I agree with Bill's puzzlement over Dawkins' statement. Under the assumption that nature is comprehensively non-teleological, as Dawkins' believes, by what standard can something be said to take on the 'appearance' or 'efficiency' of design? How can intentional design have a special, recognisable form of 'efficiency' if it doesn't exist, if it is identical to the thing with which it is being contrasted? (By the way, I'm well aware that 'intentional design' is a tautology!)

Hi Bill - This comment is really an aside about an error in Mayr's thinking about teleology and selection -- it doesn't affect the main point that natural selection is not goal directed.

Under point 4. Natural selection is not teleological, we read, "Indeed, how could an elimination process be teleological?" and your own aside, "[It can't be, my man!]". But of course, a process of elimination can be teleological, even if the process of natural selection is not. Examples are easy to conjure.

I'd also note that a process of elimination can often be viewed equally as a process of preservation. This was recognized by Darwin himself in the full title of his major work, i.e., "On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life." I mention this because of the tired objection that natural selection is a "purely negative" process that can't "produce" anything. It can.

Dr Vallicella,

Mayr writes that selection is not teleological because it is "really a process of elimination". But it is a process of elimination based on fitness ("the possession of characteristics that made them particularly well adapted for the prevailing environmental conditions"). It seems to me that fitness cannot be defined without teleological language, but I'd be interested to hear what you think. Can fitness be defined non-teleologically?

Bill,

What I am doing is not "bullshitting" at all, and it saddens and surprises me that you suggest this is what I am doing.

What view of mine do you suppose Ernst Mayr refutes? Nowhere have I asserted that evolution is teleological; I would never say such a thing. What I *am* maintaining is that the language we are forced to use is inadequate for an accurate description of the phenomena. A bird's body is an exquisite flying machine: how did it arise? Its form exemplifies aeronautical optimizations that would be the envy of a human designer, yet we cannot use the word "design" without bringing in a lot of extra, and entirely unwanted, baggage about teleology. That is why I used the quotes around such freighted words as "design" and "engineering" in my comment - it was simply an effort to indicate that I realized the words, as commonly understood, are not ideally matched to the concept that needs to be expressed. It had nothing whatsoever to do with sneering, or bullshitting, but was rather just an acknowledgment of the inadequacy of the few terms we have available to refer to such things. I thought that if I did NOT put them in quotes you would object once again that design implies agency, so I used the quotes only to indicate that I knew the word was not quite the right one to convey my meaning. I can't win, it seems.

All of the words we have available to us to describe the evolution of the bodies and minds of living things: "design", "engineering", "for, "about", et cetera -- are charged with these notions of conscious agency. It is easy enough to see why this would be so, but they are currently the only words we have, and all I am trying to do in these comments is to point out that we lack suitable terminology, and are thus forced by the inadequacy of the language into constant (and infinitely tedious and entirely unlooked-for) arguments about agency. Certainly it is plain enough that a bird's wing is "for" flying (otherwise, why do they have wings at all? Why wings, and not paddles?), but we need, and conspicuously lack, a word that expresses that idea of "forness" - that expresses the fact that a bird has wings because they enable it to fly, and if they didn't, it wouldn't have them - without dragging in the extra notion that some conscious agent *intended* it to be that way. Because we lack such a word, however, we are stuck with "for"; it's the best we can do. You make it very clear that this is a problem, and noone is about to disagree. But it would be far more productive to be able to move forward with some agreement about what IS an appropriate word for this critically important concept, instead of just having a big row about "for" every time it comes up.

I wholeheartedly agree that there is no *agent* called Natural Selection that designs fruit flies, and I have never suggested that there is. What I have said is that there is a natural *process* called Natural Selection that creates fruit flies, and that for lack of a better word for the marvelous construction of living things we have to use the word "design". ("Construction" in that previous sentence is not good enough, either; it lacks the sense of the clever solution to a problem that "design", or "engineering", connotes.)

Let me be very clear: I am not playing linguistic games, nor am I trying to "play fast and loose" with language (indeed, that is exactly what I am trying NOT to do). I have never considered evolution to be in any way a teleological process, I have never said anything to suggest that I do, I certainly need no persuading that there are no final causes in Nature, and I assure you that if there were a decent English word for the important naturalistic concept of "bottom-up design" or "design-without-agency", I, and every other Darwinist in the English-speaking world, would be using it, and gratefully. But there isn't, and so we are forced to say "design", only to be pounced on again and again by those who have a chip on their shoulders about naturalism. Frankly, I think it's beneath you to do so, Bill. I know that we have our axiomatic differences (though I will say we also do agree about many things), but this is not one of them, and I entered this discussion in a respectful, serious and amicable spirit of earnest inquiry. This latest jab from you just seems inexplicably crabby and vindictive, and, what is worse, completely misses the point.

Joseph,

I find it a very queer notion that science cannot be applied to agents.

A person's character consists of regularity about intentions, powers, foresight, etc. We don't normally apply science to personal character, but it is applicable because we can construct statistical models of an individual. That's why we can say that a person is acting out of character, or counter to past goals. That's why stories make sense when the hero predicts the next move of the villain based on the villain's goals and personality. Personality is predictive.

"Not when science is in no position to know what the intentions of the designer are, what degree/power of intervention/foresight/etc said designer may have had available and how much was exercised."

In the case where there are multiple possibilities, we are rational to assign those possibilities equal weight. That means considering all possible intentions, powers, foresights, etc. to have equal weight.

For some designers, it is likely that they would evolve life on Earth, but this misses the point. I shouldn't assume the designer is someone who would do this, when there are so many other options that are, a priori, equally likely. Why shouldn't I equally well assume that sophisticated designers would never ever evolve stuff? I have no reason to prefer one to the other, so I weight them equally in my consideration.

If we had a successful, predictive theory that explains why we ought to give more weight to a designer whose designs are indistinguishable from unguided evolution, then we could say that finding something that looks unguided is not decisive. Of course, we lack even a theoretical model like that, let alone a predictive or successfully predictive model.

"The truth of common descent and the truth of evolutionary mechanisms... does nada, nothing, zero, zip to demonstrate a lack of teleology, intention, mind, or design at work in natural history."

You're just stating your conclusion. And my argument explains why. If teleology were involved, as with the cards, it is, a priori, far more likely the world would look different. Do you have a counterargument or not?

I don't need to make this problem artificially sterile to reach my conclusion. I just have to treat like as like. There's no rational reason to prefer each evolver god to the relative googol of manufacturer gods and hybrid manufacturer-evolver gods. So, I am rationally justified in concluding that life was not designed. This is argument is subject to revision, should reason for a different statistical weighting come along. However, rationality does not wait for different statistical weights. Rationality must proceed based on the simplest weighting consistent with the data.

We don't assume it is equally likely that the Sun will rise as not rise tomorrow. Why? How do we know the distribution isn't flat until next Thursday, where the distribution becomes zero and the Sun doesn't rise? It could be a peculiar distribution. However, it is not rational to assume a peculiar distribution when a non-peculiar one will do.

"even while noting that the power of 'natural selection' is under dispute even by orthodox evolutionary biologists "

No, it isn't. Not in any relevant way. But, hey, if you are making a case that evolutionary biologists think the power of NS is a threat to the theory of evolution, please feel free to include references.

Doctor Logic,

"I find it a very queer notion that science cannot be applied to agents."

It absolutely can - when you have the information available that I stipulated. A Designer on the level we're talking about? It's simply out of the question. You may as well try to develop a scientific theory to test the truth of solipsism. Or even panpsychism for that matter. Hell, something as basic as the problem of other minds comes across as a question philosophy can gain traction on, but science has to run on assumption.

Science has limits. Sad, but true.

"You're just stating your conclusion. And my argument explains why. If teleology were involved, as with the cards, it is, a priori, far more likely the world would look different. Do you have a counterargument or not?"

You have no argument, DL. You have a very contrived example involving cards and human agents - otherwise your statements here boil down to "I can imagine a Designer using a googleplex number of alternate methods to create things, and there's a priori reason God would choose evolution, ergo there was no design". I may as well say 'I can imagine a googleplex number of alternate ways an undesigned world would look, and there's no a priori reason it should look like this, ergo the world is designed.' There's simply not much meat to what you're saying.

What we do know is that agents can and do use evolutionary processes to achieve ends. If we 'mere humans' can do it (and with increasing power, no less), then suddenly we have a good reason to view our own evolutionary history as the subject of design - it's a rational conclusion, and better supported than the 'poof' alternative.

"No, it isn't. Not in any relevant way."

Spare me. I mentioned that there were challenges to NS by orthodox evolutionary biologists, and put it aside. Go argue with Margulis and others if that upsets you - I in no way relied on the falsity or lesser utility of NS to make my point. Quite the opposite.

Eric Westman asks, "Can fitness be defined non-teleologically?" 'Fitness' as used in natural selectionist explanations is difficult to define precisely -- there's a huge literature on this topic -- but so far as I know, none of the difficulties involve a teleological element in the definition. In a rough and ready way, the fitness of a trait relative to an environmnet is a measure of the causal contribution it makes to the survival and reproduction of organisms which possess the trait in question in the relevant environment.

Joseph,

"I may as well say 'I can imagine a googleplex number of alternate ways an undesigned world would look, and there's no a priori reason it should look like this, ergo the world is designed.' "

The fact that you are saying this means you have not understood any aspect of my argument.

There are at least two important points you've missed.

1) How many natural alternatives are there versus designed alternatives? As I explained, for every unguided evolutionary pathway, there are many many more design alternatives. The evolutionary possibilities are a tiny (*really* tiny, 1:N) subset of the possibilities open to a designer. That means, weighting members of each class equally, unguided evolution is far more probable once we know that evolution occurred. This means we have some 1 in N likelihood that the world was designed.

2) Yes, there are countless ways that unguided evolution might have unfolded. Why believe any particular one? Wouldn't that be fine-tuning? Yes. But, as I explained, fine-tuning is okay as long as you get a payoff for the fine-tuning in the form of predictions. Evolutionary biology does just that.

3) In principle, a design theory could overcome this hurdle by making a prediction that has M:1 leverage, M>N. As we know, "design theorists" don't want to talk about the designer, so they have no hope of ever making any predictions.

"A Designer on the level we're talking about? It's simply out of the question."

Argumentum ad simply out of the question? :)

Brodie,

"The reliability of science to judge whether a deck of cards has been shuffled or sorted presupposes the intentions of an agent. Without previous knowledge of how decks of cards are typically ordered by their creators, we would be in absolutely no position to discern whether or not any particular deck showed signs of intentionality."

I think you're missing my analogy. Intentionality isn't part of the card deck analogy. The card deck simply illustrates reasoning with probability. The way I am applying it is as follows. In model 1, there is an agent with some unspecified intent. In model 2, there is no agent , but there is the prediction of common descent, etc. Clearly, the space of a zillion different intents in model 1 includes results consistent with model 2. However, there are a zillion different intents in model 1, and we have no reason a priori to suppose that "unspecified intent" should be weighted towards producing common descent versus the other zillion possibilities.

Again, there's no way to tell the difference between unguided evolution and an agent who intended to create an apparently unguided evolutionary pathway. But this is irrelevant because it has not been established that we should fine tune our intents to creating an apparently unguided evolutionary history. This is precisely analogous to the shuffling case. Yes, shuffling could produce an ordered deck, and we would be none the wiser. However, it is incredibly unlikely this would happen unless we have reason to say why shuffling (in this particular case) would lead to a sorted deck.

Everyone here, whether they are for teleology or against it, assumes that the telos of evolution, if it existed, must be the good of some particular species. This confuses universal and particular causes. The telos of evolution (if it exists) is not for the perfection of some particular species as particular; just as the telos of rainfall (if it exists) is not for this particular plant, as particular. Aristotle made this same point when arguing for teleology in all natural processes, even though he was well aware that rainfall sometimes made wheat grow, and other times made it rot (like when collected in barns). Rainfall is a part of a larger hydrological cycle, and any good that might be discovered in this cycle as a cycle is not going to be found by subordinating the whole process to one dandelion that might happen to get rained on.

Everyone agrees that whatever cause or constellation of causes accounts for evolution operates on a more global level than causing this species or that one; even though this species or that one is an inevitable result. The telos of evolution cannot therefore be sought at the level of particular species, and therefore to deny any goal of the process at this level does not decide whether evolution is a teleological process. I will concede that bird wings did not evolve for the sake of flying, but this observation is not decisive in determining whether evolution is teleological or not.

Doctor Logic,

"The fact that you are saying this means you have not understood any aspect of my argument."

You're making no argument, and you repeatedly work unwarranted assumptions into the conversation. For example:

"How many natural alternatives are there versus designed alternatives?"

You simply do not get it: Calling evolution the "natural alternative" will not fly when the point is being made that the "natural alternative" *is* (and for a Designer, 'could be') a "designed alternative". Further, any specific chosen pathway is itself a "tiny subset of the possibilities open to a designer". You say evolution is unguided, but there's your metaphysics again: Science cannot determine whether or not an agent above or behind nature intervened in advance (front-loaded) or on the spot (intervention) so that evolutionary event X took place. The fact that you may clench your fists and say "oh yes it can" with all your might won't change that. But by all means, do try to formulate a scientific research program whereby you attempt to investigate the capabilities of a designer above or behind nature. Maybe you can cite Frank Tipler and Nick Bostrom. Kurzweil too. ;)

2) Yes, evolutionary biology - in a very broad sense - does that. But you're trying to equate evolutionary biology itself with the 'no design' position. Sorry, but no. That's the mistake Dawkins makes, and the mistake Mayr similarly makes. Speculating on whether or not evolutionary history was itself the result of guidance, non-guidance, or even a mix is beyond science. We can use science to reasonably investigate and theorize about certain physical (even admitting that what comprises the "physical" remains a shaky question) aspects/models of what may or may not have taken place in natural history. That's wonderful, but it doesn't begin to touch on questions of guidance or agency behind/above nature. Believing otherwise is a modern myth.

3) Who is talking about design theory? I'm talking about the distinction between science and philosophy - and knowing the difference between what scientific conclusions we can draw from observation and research, and at what point we're entering the realm of philosophy. I've pointed out that Mayr's claim of "no teleology" leaves behind science every bit as much as, say, Denton's front-loaded evolutionary idea does. ID theorists wouldn't like what I'm pointing out here, because I'm making the same criticism everyone from Eugenie Scott to Ken Miller has made - their claims are philosophy, not science. The difference is I'm pointing out that the claims of ID opponents fall in exactly the same category.

Sorry, but science has limits. Why not respect them, instead of bastardizing science to try and make it do what it clearly cannot?

James,

Speaking for myself, I'm not making that assumption at all - and I think I agree with what you're saying in large part. I'm simply pointing out the practical limitations of proper science in this discussion - and that philosophical speculation (rather than scientific conclusion) enters the game far earlier than many people seem to realize, when it comes to natural history and evolution.

I would however agree that even the scope of 'teleology', and how it relates to evolution, is deeper than is being discussed here. "Evolution of limbs took place so that I personally have hands" is an oversimplification of the teleological to say the least.

Joseph,

You say
"Calling evolution the "natural alternative" will not fly when the point is being made that the "natural alternative" *is* (and for a Designer, 'could be') a "designed alternative"."

This is *precisely* analogous to the sorted versus shuffled analogy. We cannot tell whether a deck was sorted into order or shuffled and happened to fall into order. Right? I mean, all you know is that the deck before you is sorted. Does that mean that we can't be confident that it was sorted versus shuffled? No. We can be almost (but not quite) certain that the deck was sorted.

Why? Because "apparently-sorted" accounts for a teeny tiny number of outcomes of shuffling, whereas "apparently-sorted" accounts for 100% of the outcomes from sorting.

Likewise, when I look at the species before me, I cannot distinguish the case (i) where the life was evolved unguided and (ii) where an agent designed us using evolution. However, "apparently-unguided evolution" is only one teeny tiny subset of "designed". And, "apparently-unguided evolution" accounts for roughly 100% of unguided evolution.

Maybe you're getting hung up on the term science. You'll note that I don't use the term in the argument above. My argument relies on rationality, i.e., deductive and inductive inference, theorizing and probability. Call it philosophy, if you like. It won't change the argument at all.

James Chastek - Sorry to disagree (well, not really...), but it simply is not true that I assume "the telos of evolution, if it existed, must be the good of some particular species." In fact, I'm curious about how you think you can pin such an assumption on _anybody_ who has contributed to this discussion based on what they've said here. Can you clarify?

The telos of evolution cannot therefore be sought at the level of particular species, and therefore to deny any goal of the process at this level does not decide whether evolution is a teleological process.

Exactly.

Hmm. James quote that I posted above was supposed to be in italics, but obviously the html didn't take.

Doctor Logic,

"Likewise, when I look at the species before me, I cannot distinguish the case (i) where the life was evolved unguided and (ii) where an agent designed us using evolution. However, "apparently-unguided evolution" is only one teeny tiny subset of "designed". And, "apparently-unguided evolution" accounts for roughly 100% of unguided evolution."

This is so arbitrary a distinction that I'm surprised you're basing your argument on it. You're making what is at absolute best a subjective judgment call ("Evolution sure looks unguided!" - Sorry, but to me it looks guided.), making reference to a 'googleplex' (oh boy, hyperbole) of alternative design options that would subjectively 'look guided' to you, ignoring the fact that there are also a 'googleplex' number of design processes (evolutionary, hybrid, and otherwise) that could/would also 'look unguided' or even look inexplicable to you, and then pointing out that an unguided process will also look unguided (But does it? Denton doesn't think so. Nor does Behe. Nor do many others of varying views.)

Your argument is weak to say the least. First, because it relies on assumptions and tremendous subjectivity - your judgment call of what "looks unguided" means little to me, just as my judgment call of "looks guided" likely means little to you. Second, your off the cuff (and that's putting it nicely) estimations of the space of possibilities for creative methods that 'look designed' and 'look undesigned' are broken - go beyond googleplexes and say that there's an infinite number of possible design choices available that (again, subjectively) 'look designed' if you like. There's also an infinite number that 'look undesigned' or 'look like a mix' - so not only is the call subjective (subjective while based on tremendously incomplete information, mind you) but even your estimations are invalid. Third, we have definite and clear examples of (limited, human) agents using evolution and evolutionary design principles towards intentional ends. Not only does that put stress on your subjective judgment call (at the very least highlights that very subjectivity, showing how things that 'look undesigned' can be designed, either in part or in whole), but it highlights one important point: Design is always an option that can be explored and tested. Undesigned must always be assumed, because to demonstrate the power of an 'undesigned process' is to demonstrate the power of design. Which is why science is limited to demonstrating the possibilities of processes, period - and 'designed' or 'undesigned', 'teleological' or 'non-teleological' falls to philosophy.

Am I saying you're irrational for not believing in design? Honestly, I don't care to speculate. I'm just showing why evolution is no barrier to belief in design/teleology, why in fact it could bolster the case for such a belief, and the rationality of believing in design. And most of all, one more time: Showing that 'science' turns into 'philosophy' quicker than Mayr and others would care to admit.

I didn't see Bill as for teleology or against it, but just for consistency. I don't quite see how Malcolm is answering this with appeals to the inadequacy of language. Everyone agrees that language is not always a good guide to reality (not all bluefish are blue). Bill's claim is more radical, it seems to me: the denial of any kind of teleology commits one to saying that eyes aren't really for seeing; ears aren't really for hearing.

So what say you, Malcolm? Are eyes for seeing? Or is this simply a trick of language?

Full disclosure: I think eyes are for the good of seeing and evolution is for the good of the diversity and multiplicity of species. I agree that evolution does not explain the teleology of the eyes or of any member of an evoloved species, but this does not make it non-teleological, only joined to the wrong end.

Bob,

I'll stick to your claim. Your original claim, as I read it, is that one does not need to invoke a designer to explain species trait X, and therefore evolution is not teleological. I respond that the telos of evolution (if it exists) is not proportionate the arising of traits, any more than the telos of rainfall (if there is such) is not proportionate to watering uninhabited forests, even though such forests would not be watered if not for the rainfall.

Bob,

oops "is proportionate"

Joseph,

I think we're making progress. You dispute my estimation of the numbers (e.g., you think my estimates or classifications are subjective), but you seem to agree with the basic structure of the probability argument.

"You're making what is at absolute best a subjective judgment call ("Evolution sure looks unguided!" - Sorry, but to me it looks guided.),"

No, it's not subjective. When we consider an unguided process of natural selection combined with random genetic modification, we get common descent, common architecture and common construction. Evolution doesn't subjectively look like this model, it *objectively* looks like this model.

"Third, we have definite and clear examples of (limited, human) agents using evolution and evolutionary design principles towards intentional ends."

Sure. Suppose I want to design a better radio receiver using a genetic algorithm. Well, for starters, I'll create an environment with a fitness function that favors receivers with higher S/N ratio, sensitivity, etc. Then I'll run my evolution sim and get a design for my receiver. Then, I'll take that design and pair it with a power supply that I designed using traditional methods. Also, I won't populate my sim with independently evolving predators that eat my receivers.

As for the relative size of the spaces, consider this.

At every point, I have options. I can switch from an evolutionary algorithm to a traditional design algorithm. I can re-implement an evolved algorithm in a different substrate. I can also add manufacturing facilities to my world. I can stock the world with completely new, traditionally designed entities. I can eliminate species at will, and render others invulnerable.

That's a lot of different options available to a designer that are not available to unguided evolution at EVERY step. At every step I can switch from evolving to manufacturing, switch from evolving one kind of species to another, and switch from one substrate or technology to another, etc. I can also choose to change my goals from time to time.

What does that mean in biological history? The designer could decide that the platypus will be implemented in silicon nanotech instead of organic molecules using non-evolved components. Or the designer might decide that hedgehogs will not breed but be manufactured at thousands of miniature factories across the globe. There is no evidence for such design choices at all.

In other words, look at the issue from an atomic point of view. At each step in a naturalistic evolutionary process, only certain things are allowed. With design, virtually anything is allowed *at each step*. This isn't subjective, it's objective.

The number of possible worlds consistent with design becomes vastly bigger at every moment in time relative to the number of possible worlds consistent with evolution.

In theistic language, a god can intervene (or not) at every instant in time, and veer away from what appears to be unguided evolution. At every instant in time, the number of possible theistic worlds is vastly greater than the number of possible naturalistic worlds. Therefore, the total number of worlds consistent with theism is vastly greater than the number of worlds consistent with naturalism. And the observed world is one of the worlds consistent with naturalism.

James,

The point I was making, or was trying to make until getting chewed out by Bill, is that we need to be clear about words like "for", and if we can't agree on what they mean, we need a new vocabulary.

Wings are obviously, in an important sense, "for" flying. If birds couldn't fly with their wings, there wouldn't *be* any wings. The only reason wings exist at all is that they conferred a selective advantage on their owners by enabling them to fly. And over time wings got to be more and more superbly conformed (I am desperately fishing around for a non-teleological word to use here) for flying, until by now they are astonishingly well-perfected for the task, far better in many ways than the designs we humans could have come up with. But there is *no telology* here; it all works from the bottom up.

Now the word I want to use for all of this, rather than "conformed", is "designed". But as we have seen, that is proving to be horrendously troublesome, because in its customary usage it implies an intentional designer and a final cause, neither of which is evident in the natural evolution of living things.

What has happened here is that evolutionary biologists have co-opted words like "for", "designed" and "engineered" as technical jargon. When Richard Dawkins talks about "design", he certainly isn't trying to pull a fast one, and the furthest thing from his mind is sneaking in some teleology. Evolutionary biologists simply are not thinking of final causes when using the word "design" to refer to what evolution does, and they understand one another perfectly. Evolution by natural selection is, with very few exceptions, universally assumed by evolutionary theorists not to be teleological in any way at all.

This is no different from the many ordinary words that have become technical terms in (to take a handy example) philosophy; when philosophers of mind talk of "zombies", for example, the cultural anthropologists don't get apoplectic about the word "zombies" referring only to re-animated dead people. Usually, when it is explained what the word means to the technical specialists, then everyone understands, and can move on. But that doesn't seem to be working out so well in this case.

So it all depends on what you mean by "for". When someone says "the denial of any kind of teleology commits one to saying that eyes aren't really for seeing", he is using "for" in the ordinary, intentional, pre-Darwinian, non-technical sense. If that's all it can ever mean, then we should come up with some new language for discussing evolution, that's all. But it's hard to coin new words, so old ones tend to get co-opted to apply to new concepts. Non-teleological "design" is a concept that is only 150 years old, and the language simply hasn't had time to catch up yet.

The comments to this entry are closed.

My Photo
Blog powered by Typepad
Member since 10/2008

Categories

Categories

October 2020

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31
Blog powered by Typepad