Taylor on cognition, teleology, and God

In his book Metaphysics , a classic brief and lucid introduction to the subject, Richard Taylor devotes a chapter to the topic of God.  Most of the attention philosophers have paid to it seems to focus on his version of the cosmological argument, which is indeed a fine brief exposition and defense.  But Taylor also presents a second argument, of a broadly teleological sort.  It is decidedly not a variation on Paley’s design argument (of which, as longtime readers know, I am not a fan).  It is much more interesting and metaphysically deep than that, and at least in a general way closer to the spirit of Aquinas’s Fifth Way(even if it is not quite the same as Aquinas’s argument either).

Stones and semantics

Taylor begins by asking us to consider a couple of scenarios.  Suppose you are traveling by train through the UK and, peering out the window, you see an arrangement of stones in a pattern that looks like this: THE BRITISH RAILWAYS WELCOMES YOU TO WALES.  You would naturally assume that the stones had been deliberately arranged that way by someone, in order to convey the message that you are entering Wales.  Now, it is possible in theory that the stones got into that arrangement in a very different way, through the operation of impersonal and purposeless natural causes.  Perhaps, over the course of centuries, the stones gradually tumbled down a nearby hill, and each one stopped in a way that generated just that pattern.  This is, of course, extremely improbable, but that is irrelevant to Taylor’s point and he allows for the sake of argument that it could happen.

Taylor’s point is rather this.  Even if you could reasonably entertain the latter possibility, what you could not reasonably do is both accept it as the correct explanation of the arrangement of stones and at the same time continue to regard that arrangement as conveying the message that you are entering Wales.  The arrangement could intelligibly be conveying that message only if there is some intelligencebehind its origin, which brought it about for the purpose of conveying the message.  If, instead, the arrangement came about through unintelligent and purposeless causes, then it cannot intelligibly be said to convey that message, because it could not in that case intelligibly be conveying any message at all.

Or, to take Taylor’s second example, suppose a rock were dug up from the ground and found to be covered with an interesting set of marks, of roughly the same size and arrangement that the letters and sentences of a book might exhibit.  One explanation of the marks might be that they had been formed by some impersonal and purposeless natural process, such as glaciation or volcanic activity, which simply happened by chance to result in a pattern that looked like writing.  Whether or not this is likely is, again, not to the point, and Taylor allows for the sake of argument that it might be a perfectly reasonable explanation. 

Another possible explanation, of course, is that it really is writing.  Suppose some scholar studied the stone on this assumption, and proposed that the correct translation of the marks would be: HERE KIMON FELL LEADING A BAND OF ATHENIANS AGAINST THE FORCES OF XERXES.  Now, Taylor allows for the sake of argument that you could opt either for the first explanation or the second.  But what you cannot reasonably do is suppose both that the marks arose through an entirely impersonal and purposeless natural process and at the same time that they really do convey the message represented by the proposed translation.  For if they arose through an impersonal and purposeless process, they cannot convey any message at all.

I hasten to emphasize again that Taylor’s point has nothing whatsoever to do with probabilities, and in particular nothing to do with how likely or unlikely it is for arrangements of the kind in question to form via natural processes.  He allows, for the purposes of the argument, that that could happen.  His point is rather that, no matter how complex and orderly are the arrangements of physical components that might be generated by purely impersonal and purposeless natural processes, they could never by themselves generate something with intentional or semantic content.  (This way of putting things is mine rather than Taylor’s.)  This is not a point about probabilities, but rather a conceptual and metaphysical truth.  Neither stones nor marks on a rock have any inherent connection with any semantic content we might decide to convey through them.  The content they might have must derive from a mind which uses them for the purpose of conveying such content.  Delete such a mind from an explanation of the arrangements of stones or marks, and you delete the semantic content along with it.

Minds and meaning

Taylor then asks us to consider our perceptual and cognitive faculties.  These too we take to have intentional or semantic content.  We have visual experiences such as the perception that there is a cat on the mat, auditory experiences such as the perception that someone has just rung the doorbell, and so on.  We have the thought that there will be rain tomorrow, the thought that two and two make four, and so on.  We take it that a visual experience like the one in question is not merely the presentation to the mind of an array of colors and shapes, and that the auditory experience in question is not merely a sequence of sounds, but that the experiences convey the messages that the cat is on the mat and that someone is at the door.  Of course, we might be misperceiving things, but that is not to the point.  The point is that the experiences do convey those messages, whether or not the messages are accurate.  Similarly, we take it that when we “see” or “hear” a sentence like “There will be rain tomorrow” as it passes through our imaginations, this is not a mere string of internally apprehended sounds or shapes, but conveys the meaning that there will be rain tomorrow.

Now, Taylor is happy to allow for the sake of argument that, as with the arrangement of stones you see out the train window and as with the series of marks on the rock that has been dug up, our sensory organs and neural structures may have arisen through entirely impersonal and purposeless natural processes, such as evolution by natural selection.  He is not interested in challenging the probability of such explanations.  He writes:

The mere complexity, refinement, and seemingly purposeful arrangement of our sense organs do not, accordingly, constitute any conclusive reason for supposing that they are the outcome of any purposeful activity.  A natural, nonpurposeful explanation of them is possible, and has been attempted – successfully, in the opinion of many.  (p. 117 of the second edition)

Notice that he includes the “seemingly purposeful arrangement” of our sense organs as among the considerations that do not suffice to show real purpose.  The arrangement of stones you see out the train window and the marks you see on the rock seem purposeful, but Taylor allows that that may be an illusion.  Similarly, he allows that our sense organs could seem to have a purposeful arrangement and yet be purposeless for all that.  His argument has nothing at all to do with how likely or unlikely it is that the appearance of purpose could arise from purposeless impersonal process.

What he is concerned about instead is the case where we suppose our sense organs and neural processes to have genuinepurpose, and in particular where we suppose our perceptual experiences and thoughts to have genuine intentional or semantic content.  And he wants to make a point that parallels the point he made about the arrangement of stones and the marks on the rock.  We could take the deliverances of our sense organs and neural states to have genuine intentional or semantic content.  Or we could take those organs and states to have arisen through entirely impersonal and purposeless natural processes.  What we cannot reasonably do is both of these things at once.  In particular, we cannot intelligibly both take these cognitive faculties to have arisen through entirely impersonal and purposeless processes and at the same time regard them as having genuine intentional or semantic content – as conveying any message about cats on mats, the ringing of doorbells, rain tomorrow, or anything else. 

The cases, Taylor argues, are in all relevant respects parallel.  Delete mind and purpose from your account of the origin of the arrangement of stones or the marks on the rock, and you delete any semantic content along with them.  Similarly, if you delete mind and purpose from your account of the origin of our cognitive faculties, then you delete any intentional or semantic content along with them.  You can have one or the other account, but not both. 

Now, our cognitive faculties do in fact have intentional or semantic content.  We really do have perceptual experiences with the content that the cat is on the mat, thoughts with the content that it will rain tomorrow, and so on.  Since this is intelligible only on the supposition that our cognitive faculties originated via some mind and its purposes, there must be some intelligent being that brought us about with the aim of having our cognitive faculties convey to us information about the world around us. 

Taylor does not elaborate further.  Presumably he would identify this mind with the necessary being whose existence he argues for earlier in the chapter, by way of a version of the cosmological argument.  But if so, he does not explain how.  Indeed, he makes only modest claims for his arguments, and leaves it an open question what relevance they might have for religion.  If the chapter can be said to defend theism, it is a purely philosophical theism that does not entail (though it also does not rule out) a specifically Jewish, Christian, or Muslim conception of God and his relationship to the world.

Some bad objections

Taylor does address several objections that he suggests some readers might take to be obvious, but which in fact simply miss the point.  For example, some may point out that our cognitive faculties are not always reliable.  As Taylor says, this is irrelevant.  The point isn’t that our cognitive faculties convey accurate messages, but rather that they convey any message at all.  Consider once again the case of the arrangement of stones that you see out the train window.  If you suppose that it arose through purely impersonal and purposeless processes, then it isn’t just that you can’t regard it as accuratelytelling you where you are.  The point is rather that you can’t intelligibly regard it as telling you anything at all.  By the same token, Taylor argues, if you suppose that your cognitive faculties arose through purely impersonal and purposeless processes, then what would follow is not merely that they don’t accurately represent the world but rather that they don’t represent anything at all, whether accurately or inaccurately.

A second bad objection would be to suggest that Taylor is presenting an inductive argument from analogy, which is no stronger in this case than when Paley presents such an argument.  As Taylor emphasizes, he is not, in the relevant sense, presenting an argument from analogy.  In particular, he is not making a point about how improbable it is that a certain complex natural structure arose apart from intelligent design, given how similar it is to human artifacts.  On the contrary, he explicitly concedes that the complexity, refinement, and appearance of purpose that our sensory and cognitive faculties exhibit could have arisen through unintelligent processes, just as the arrangement of stones could have.

True, he does draw an analogy between the arrangement of stones on the one hand and our cognitive faculties on the other.  But the point of the analogy is simply to illustrate the general principle that it is metaphysically impossible for something to have actual intentional or semantic content (as opposed to the mere appearance of such) if it arose entirely from impersonal and purposeless processes.  He is not giving an inductive “argument from analogy” of the form: “A is like B, so the cause of A is probably like the cause of B.”  Again, probabilities are not what is in question.

A third bad reply, Taylor says, would be to suggest that, even if our faculties arose through entirely impersonal and purposeless processes, we have good inductive grounds for taking them to be reliable.  One problem with this, as he points out, is that such an argument would be circular.  For in order to get such an inductive argument off the ground, you’d have to rely on your cognitive faculties, and whether they are reliable in that case is precisely what is in question.

But the problem is deeper than that (and I think that in his response to this particular objection Taylor could have made the point clearer).  For as I have said, the point is not merely that our cognitive faculties would not reliablyconvey messages if they arose via purely impersonal and purposeless processes.  The point is that they would not convey any messages at all, that they would be as utterly devoid of intentional or semantic content as an arrangement of stones that formed via impersonal and purposeless processes.  And they would first have to have such content for us to be able to get any inductive argument, or any argument at all, off the ground.  Hence this third objection to Taylor simply misses the point.

A fourth bad objection addressed by Taylor would be to appeal to survival value as a reason to think that our cognitive faculties are reliable.  Taylor notes in response that the deliverances of our sensory and cognitive faculties far outstrip anything that could plausibly be said to have survival value.  But here too I think he could and should have made a stronger point, which is that the appeal to survival value also misses the point.

Suppose that the power to become immaterial at will would afford my descendants tremendous survival value.  (I’ve seen this example used by someone else, by the way, but I do not recall where.)  Does it follow that, through mutation, natural selection, and the like, such a power might arise in the generations that follow me?  No, because there is no plausible mechanism by which that power, specifically, might arise through such processes.  Or suppose that an organism would gain tremendous survival value from being a round square.  Does it follow that round squares might evolve via mutation, natural selection, and the like?  Of course not, because round squares are logically impossible.  Appeal to “survival value” is mere hand-waving without some plausible process by which a property or power could actually arise.

Now, the deep point behind Taylor’s argument is that you simply aren’t going to get intentional or semantic content from entirely impersonal and purposeless processes any more than you are going to get immateriality or round squares out of them, so that appeal to the survival value of having such content is a red herring. 

Evolution and cognition

Notice that, though Taylor is not explicit about it, this is compatible with an evolutionary story about the origin of our cognitive faculties.  It just isn’t compatible with a materialistic-and-mechanisticevolutionary story about the origin of our cognitive faculties – one that entirely excludes mind and purpose from the story. 

Daniel Dennett characterizes Darwinian evolutionary theory as a style of explanation that excludes any “mind first” account of the world, viz. an account that takes mind to be a fundamental feature of reality, rather than one that derives from non-mental phenomena.  But while this is true of Dennett’s preferred style of evolutionary explanation, it is not true of evolutionary explanations as such, not even Darwinian ones.  (See chapter 6 of Aristotle’s Revenge for discussion of this issue.)  And Taylor’s point is that whether or not evolutionary processes are part of the story of the origin of our cognitive faculties, we must in fact affirm a “mind first” account of the world.  There would simply be no minds like ours at all if there were not a more fundamental kind of mind to bring them into being.

Some readers will have noted that Taylor’s argument is reminiscent of the “Evolutionary Argument Against Naturalism” that Alvin Plantinga would develop years later (and which I have discussed in other places, such as this one).  But there are some important differences, and differences that, in my view, make Taylor’s argument the metaphysically deeper and more interesting one.  In Plantinga’s argument, there is a lot of heavy-going about probabilities.  But as I’ve been emphasizing, the point really has nothing essentially to do with probabilities at all, and in my opinion Plantinga’s emphasis on them just muddies the waters.  Second, Plantinga focuses on the question of whether our cognitive faculties would be reliable if they arose through entirely impersonal and purposeless natural processes.  But the deeper question is whether they would have any intentional or semantic content at all (whether reliable or not) if they arose in that manner.