The disconnect between biology and philosophers

I sometimes worry about the lack of attention philosophers pay to actual biology, settling instead for purely verbal arguments. I am travelling right now so I don't have time to carefully critique Jerry Fodor's latest attack on "Darwinism", but it seems that he is actually making the argument that natural selection is not selection because there's no agent doing selection, and it's just a metaphor and hence bad science. At the same time as this debate is occurring over on PhilPapers, this piece comes out showing that selection on trout size by bear predation is directional and hasn't yet settled on an equilibrium. Does Fodor really think that because he can't make an account by which intentionality of "selection for" can occur without agency that somehow selection is a myth? That's an excruciatingly bad argument, worthy of the worst of David Stove.

If I get time when I get back, I may do a more detailed account, but they just called my flight...

More like this

I looked at the article too - Fodor's argument is "not even wrong" (thanks Wolfgang Pauli!). Gosh, it's a word soup with sprinklings of drivel!

Your analysis of what he says about selection can't be selection because there is no mind seems to convey exactly his meaning. And you wonder how someone who has worked in the field of linguistics could make such a huge error! We call it natural selection to distinguish it from artificial selection - artificial means (pretty much) made by a person. So the term natural explicitly indicates that no mind is involved.

Perhaps he needs a semiotician to smack him around the head and explain the difference between a sign and what it signifies.

Fodor thinks that because drift can occur, all natural selection is nonsense. That's as daft as me saying because Fodor writes rubbish here, everything he writes is rubbish.

He burbles on about "Darwinism" - a sure sign of ignorance as we know that scientists who study living forms within an evolutionary paradigm are called simply "biologists". Much of the discussion at PhilPapers seems to make the classic philosophers' mistake of assuming that biologists are followers of Darwin in the same way as Marxist philosophers are followers (or worshippers?!) of Marx. As scientists, we respect Darwin's work and are in awe of his achievements, but in no way is his work regarded as the last word on the subject. We don't ask "hmm... what does Darwin say about this?" or "what would Darwin do?" when we consider a problem.

Fodor is clearly a mouth in search of a brain.

I'm not convinced that Fodor's burbling blahing inanity is worthy of your time in writing a more detailed account. As I said, it's not even wrong.

Its interesting that Fodor's argument would have different weight (or even simply not exist) if this entire discussion (including Darwin's Origin and most of the relevant literature) was being carried out in, say, Hopi.

"Does Fodor really think that because he can't make an account by which intentionality of "selection for" can occur without agency that somehow selection is a myth?"

I gather the answer is emphatically "no". Selection occurs on Fodor's account as I understand it, but it is explained by historical particularities plus basic physical laws and not by macrobiological covering laws. Fodor is arguing for reductionism not creationism. I take his point to be that genuine macrobiological explanations will be particular and would not ultimately cite non-basic physical laws. Wilson's point, in the philpapers link, that Fodor is strawmanning biologists might be a good one. I don't know enough biology to adjudicate. However, imputing to Fodor the view that selection does not occur seems equally a strawman.

Also (turning to Sam C), I thought that Fodor was reasonably clear. Maybe wrong, but pretty clear. I don't quite get the "word soup" objection. The distinctions between intentional (with a "t"), intensional (with an "s"), and extensional contexts are all perfectly familiar, at least to philosophers. Also, I don't doubt for a second that Fodor knows the difference between a sign and what it signifies. That difference is, in fact, crucial to distinguishing intensional from extensional contexts. Extensional contexts, after all, are simply those in which the substitution of co-referring terms (i.e., signs that same-signify) preserves truth value. Arrogant and dismissive bluster is a poor substitute for actually responding to an argument.

This may be helpful. A classic illustration of the difference between extensional and intensional contexts:

Valid argument:

Twain = Clemens.
Twain wrote Huck Finn.
________________________
Clemens wrote Huck Finn.

Invalid argument:

Twain = Clemens.
Mary believes that Twain wrote Huck Finn.
________________________
Mary believes that Clemens wrote Huck Finn.

To see that the latter argument is not valid, let us suppose that Mary was a fan of Twain's work. Suppose Mary encounters a nice gentleman at a bar, who introduces himself as Samuel Clemens. It is clearly possible that she not believe that the man she has just met wrote the books that she loves.

It may well be the case that some philosophers don't pay close enough attention to biology. However, it should be noted that the recent trend in philosophy of science has been toward greater specialization and expertise in the particular sciences. In any case, it might equally be true that some scientists are sometimes too quick to judge philosophical distinctions as invidious, unclear, or "merely verbal".

Fodor doesn't happen to call himself "Backspace" while online, does he?

I quite literally laughed out loud.

Thanks for the morning pick-me-up!

jrshipley:

Arrogant and dismissive bluster is a poor substitute for actually responding to an argument.

OK, my response is dismissive, and I admit to little interest in in/exten[st]ionality issues. That sort of philosophy seems to be navel-gazing arguments about the meanings of words rather than about the inter-play of interesting ideas and their implications for real-world actions. I'm from a scientific background, methodological naturalism is my bent ("show me"!).

But Fodor's paper is built on a foundation of either accidental incomprehension or deliberate misrepresentation about what he calls "Darwinism". Either invalidates the paper. And certainly reduces the need for a fully reasoned response: if Fodor is arguing A therefore B therefore C, when A is wrong, there is no need to address B or C. His is an "if pigs can fly, this means that..." argument.

I'm inclined to think that Fodor is the arrogant one here: I am a nobody commenting to a blog post and my words will be rightly lost in the compost heap of the blogosphere with no lasting impact or value, Fodor is a senior academic writing a more lasting and authorative paper in which he implies that biologists don't understand the core principles of their own subject, but that he does.

Fodor should stick to subjects that he understands where I'm sure his skills would be appreciated. But he's dissing my boys, and I don't like it.

I heard Fodor present an early version of this paper a couple years ago, and in the ensuing discussion it seemed that his problems with selectionism (if you can't abide 'darwinism') stem from a particular view about the nature of "scientific laws" and their role in "scientific explantions". He thinks there can't be "laws of selection" that could "explain" evolutionary change of the sort manifested in organic lineages. So I think that of the comments appearing here, those of jrshipley are probably closest to the mark. The rest seems to evidence a bit of jerkiness about the knees.

By bob koepp (not verified) on 20 Apr 2009 #permalink

John, like you I'm puzzled by the lack of attention to biology. Is there a reason?

I think there is. Biology is a lot more messy and unpredictable than, say, physics. Philosophers don't like that very much because it's difficult to construct grand theories of epistemology that apply to biology.

Some philosophers don't appreciate the difficulty of biology so they wade right in as though their understanding of the subject was sufficient. You and I have seen that many times.

With respect to evolution, I'm reminded of something that Jacques Monod said over thirty years ago ...

Another curious aspect of the theory of evolution is that everybody thinks he understands it. I mean philosophers, social scientists, and so on. While in fact very few people understand it, actually as it stands, even as it stood when Darwin expressed it, and even less as we now may be able to understand it in biology.

The debate on PhilPapers suggests that nothing has changed since then. That's why we have philosophers like Michael Ruse and Daniel Dennett whose concept of a "dangerous idea" is still rooted in a Victorian mindset.

The work of Griffiths and Stotz on defining a simple term like "gene" should have alerted all philosophers to the messiness, and difficulty, of biology. Your own work on a "simple" concept like "species" should also serve a valuable lesson for those philosophers who think they understand biology.

I guess there's a silver lining here for theologians who are used to being the ones raked over the coals for not understanding evolution.

Now they've got some company.

;)

The opacity of reference thing seems to me rather irrelevant, but it does look like I have misread Fodor. At least, the first three pages. I'm in Lisbon right now after 36 hours of travel and not sleeping, but it's being introduced as an argument against there being intentionality of psychological content (the "aboutness" of, say, a frog's response to a fly moving across its field of vision, an old standard example in this debate). He seems to be saying that selection doesn't give that aboutness. Which I think is just fine. The Millikan line that selection is what gives intentionality of functions is, I think, a case of ascribing to the world that which we hold as semantic convention.

I will revisit this later, when things settle.

That's why we have philosophers like Michael Ruse and Daniel Dennett whose concept of a "dangerous idea" is still rooted in a Victorian mindset.

Larry, John, do you think Dennett's book is worth reading? Reason I ask is, it's still staring at me from my shelf of "haven't gotten to yets" ...

The selective agent is the Second Law of Thermodynamics: âNatural selection for least actionâ, Kaila and Annila (doi:10.1098/rspa.2008.0178).

He's partly right that the theory is "frightfully complex"... but that's only in the details of how the principles lead to particular results, not the basic principles themselves.

He's also right that selection is a bit of an oversimplification, but I really don't think he wants to learn the mathematics of non-uniform random walks in n-dimensional spaces.

Read Dennett's book. You'll either be convinced, or not, and if not, you'll have to do some work to figure out why he's wrong. Either way it's worthwhile. I do not agree with Dennett on a host of things, but I have had to do some serious work to figure out why.

Abb3w, I totally agree. In my view it's all about the use of exergy. Selection is a sorting process driven by that or the lack of it. I think that selection is merely one dynamic of a complex of dynamics, and that the wider model (as given by, say, Gavrilets) is the real theory; that includes drift and non-selective biases. But I also think that any real population is maintained at a very high level of fitness, and drift occurs only on a few alleles anyway. Larry can correct me.

But I also think that any real population is maintained at a very high level of fitness, and drift occurs only on a few alleles anyway. Larry can correct me.

Here's how I would put it.

For any given population more than 99% of all segregating alleles are neutral or nearly neutral and their eventual fixation or loss will be due to random genetic drift.

Is that what you meant to say? :-)

Also, I not sure what you mean by a "very high level of fitness." Do you think that humans fall into this category?

If you think about adaptive peaks, where the peak is perfect adaptation to the current environment, then where would you place most species? I think most of them are well below the tree line.

How well are you adapting to Lisbon? :-)

Are you flying back through Toronto? Why not?

So many questions so little functioning forebrain...

How many genes are allelic? As a fraction of the total genome, I mean...

Adaptive peaks are a fiction. Adaptive corridors or ridges, on the other hand. And they are not optimal, but optimal enough for government work.

Lisbon is fine, but they screwed up the time zone. It should be nine hours later. And the trip is being paid by mine hosts, and they couldn't make the return trip open. Later this year...

John,

That paper has been floating around the tubes for a while, though I guess there isn't much else to respond to since the publication date of Fodor's book seems to keep getting pushed back. When his LRB article came out, I see to recall it was slated to hit the following Spring, which was a year ago.

Fodor seems to be angling for the long view here. Going after Saint Darwin is hardly the path of least resistance, especially this year, especially when the creationists are still throwing their back into their own attempts at a take down. He could easily make the same critique of Natural Selection by praising Darwin's diligence and creative intuition. Unfortunately he picked the worst possible way to get his point across by picking at this particular scab. Even the less reactionary biologists, like Coyne and Rose, have their hackles up over this, which is too bad, because some version of Fodor's critique will ultimately prevail. I prefer the approach of John Dupre or Evelyn Fox Keller, each of which are more nuanced and biologically informed.

But Fodor's observation is a fairly modest one: "Selection" doesn't have anywhere near the predictive or explanatory power it's advertised to have (notwithstanding Dennett's attempts to prop it up as a universal deus ex machina, er, acid.) We know too much, today, about how organisms and environments interact to be able to support the passive and reactive "selection" metaphor for much longer. Natural selection had a huge impact on our understanding of biology, and nobody can take that away from Charles Darwin, but we're not doing his memory any favors by clinging to it long past its utility for scientific understanding.

But Fodor's observation is a fairly modest one: "Selection" doesn't have anywhere near the predictive or explanatory power it's advertised to have (notwithstanding Dennett's attempts to prop it up as a universal deus ex machina, er, acid.)

Probably a peculiar (and wrong) idiosyncrasy of mine, but I've often had an instinctive tendency to view selection negatively rather than positively. In other words, individuals are "selected" for termination, instead of survival and replication. Viewed in this context, selection establishes some minimal baseline requirements, and anything goes after that. When those requirements are satisfied, adaptive fitness may become less significant and compete with other factors.

jeff - Whether or not viewed "negatively" (and mean that in a positive sense;-), what you describe as "minimal baseline requirements," beyond which "anything goes" is what Bruce Wallace has long touted as 'soft selection.' Where we encounter soft selection regimes, employing a "satisficing" rather than a "maximizing" version of selection theory would be in order. The math gets trickier and trickier...

By bob koepp (not verified) on 21 Apr 2009 #permalink

Jeff,

You're getting close to my own thinking. "Elimination" may be a better word than "selection". Artificial breeders often accomplish their goals by culling (eliminating) all the individuals that deviate from the desired phenotype. In nature, there are far more ways to fail than to succeed, and the vast majority of propagules leave no descendants (in the Philpapers discussion, Wilson refers to the 19th-century recognition of "waste"). It's the ones left standing that create the next generation.

Some individuals succeed because of heritable traits, others are just lucky. Sometimes it's easy to say which is which, more often not. Some heritable traits create their own luck: 65 mya, it probably helped to be a member of a fossorial species when the asteroid struck. In most cases though, all we know is that survivors survive.

By Mal Adapted (not verified) on 21 Apr 2009 #permalink

Yeah, maximizing fitness doesn't seem realistic. It's probably quite subjective, but I suspect if one really tried, one could find far more non-adaptive than adaptive properties in almost any species.

Survival is the default. Existence is given for free. Individuals are not non-existent patterns striving for existence (selection) - they are existing patterns striving to avoid non-existence (elimination). To me, that's a crucial philosophical difference.

Jeff and Mal: you are making the mistake that Wallace made. Selection is only tangentially about survival. It's rather more about doing well in the longer term with what Fisher called "reproductive investment". The makeup of a population changes as varieties do better or worse comparatively. Existence is not given for free - it costs a lot in resources to invest in progeny to maturation.

"He's also right that selection is a bit of an oversimplification, but I really don't think he wants to learn the mathematics of non-uniform random walks in n-dimensional spaces"

The point is not what he is willing to learn, but rather the terminology the discourse ought to be carried out in. His point is that the selection metaphor carries with it a lot of serious misapprehensions that affect not only less-knowing observers, but the technical discourse itself.

I'll have to give this a good long re-read because there's an awful lot by way of philosophic trappings in here which I didn't get first time, but I do think this represents a real challenge to the kind & degree of explanatory power many selectionists are claiming at the moment.

The main failing of the metaphor, by my reading, is that "darwinian" explanations tend to assume a kind of rhetorical authority they don't really have. Which, I think, has been a perfectly legitimate and damning complaint about EP for quite a while now: that it is a lot more like those fields that it claims to replace--psychology, sociology, history--than it cares to admit and that the "covering theories" really serve to cover up a lot of argumentative sins.

I sometimes worry about the lack of attention philosophers pay to actual biology, settling instead for purely verbal arguments.

"The difference between a philosopher and a physicist is that the philosopher hasn't got a lab."

Selection occurs on Fodor's account as I understand it, but it is explained by historical particularities plus basic physical laws and not by macrobiological covering laws. Fodor is arguing for reductionism not creationism. I take his point to be that genuine macrobiological explanations will be particular and would not ultimately cite non-basic physical laws.

Did anyone ever argue there are separate laws of physics in natural selection? I hope not⦠Is that what you mean?

selectionism (if you can't abide 'darwinism')

Can't abide that either: it still sounds like an ideology, or at least like a philosophy.

Probably a peculiar (and wrong) idiosyncrasy of mine, but I've often had an instinctive tendency to view selection negatively rather than positively. In other words, individuals are "selected" for termination, instead of survival and replication. Viewed in this context, selection establishes some minimal baseline requirements, and anything goes after that. When those requirements are satisfied, adaptive fitness may become less significant and compete with other factors.

That's not wrong at all.

It should just be mentioned that how minimal the "minimal basement requirements" are depends on the environment and can thus fluctuate pretty wildly.

(Oh, and don't emphasize survival over replication; survival is pretty uninteresting except where it helps replication.)

65 mya, it probably helped to be a member of a fossorial species when the asteroid struck.

Apparently not, according to a talk given last year at the Society of Vertebrate Paleontology meeting.

By David MarjanoviÄ (not verified) on 21 Apr 2009 #permalink

The main failing of the metaphor, by my reading, is that "darwinian" explanations tend to assume a kind of rhetorical authority they don't really have. Which, I think, has been a perfectly legitimate and damning complaint about EP for quite a while now:

Do you mean evolutionary psychology?

If so, you've changed the topic. While it could be done as a science, it isn't done that way. Instead it consists mostly of arguments from ignorance (like laughable assumptions that just about any behavior is utterly totally unique to humans â I've seen an "explanation" for why humans yawn; never mind that, at the very least, all jawed vertebrates yawn!) and arguments from personal incredulity (amounting to "I haven't thought of this possibility, so it doesn't exist").

By David MarjanoviÄ (not verified) on 21 Apr 2009 #permalink

Do you mean evolutionary psychology?

If so, you've changed the topic.

The first sentence of Fodor's paper states that it began as a critique of EP. Though it has grown into something broader than that, EP is still a more than appropriate example of the tendencies he's complaining about.

Thanks, btw, for drawing my attention to all of this. It's fascinating stuff.

That philosophy PhD message board discussion is VERY interesting. I'm glad they let the public see and not contribute.

Seems to me Fodor may not have given up on the tautological argument, because the direction he's forcing Harnad to run now seems to make them susceptible to such an argument.

Existence is not given for free - it costs a lot in resources to invest in progeny to maturation.

I meant from the perspective of the individual - you didn't pay anything to come into existence. Survival is the default state.

Selection is only tangentially about survival. It's rather more about doing well in the longer term with what Fisher called "reproductive investment". The makeup of a population changes as varieties do better or worse comparatively.

I think I understand, but the problem with this view is that almost every abstract concept favoring the replication of specific properties can then called "selection", and evolution becomes tautological. Even drift and population dynamics could be rationalized in hindsight as a form of positive selection. If you restrict the meaning of selection to the negative, you don't have this problem. And even concepts like sexual selection can be viewed in the negative (i.e. mates not chosen are "selected" for termination).

Having skimmed the paper, Fodor seems to be threshing about with teleology and intention as the motor of natural selection, possibly parodying the overuse of the analogy and other analogies such as the blind watchmaker. It's unconvincing as he doesn't seem to want to appreciate that natural selection is an effect of the existence of life, not a purpose driving organisms. The sole purpose is successful replication, and that need be no more than something that happens when materials get together in a certain way. Replication is something that living organisms do because that's what makes them living organisms, and if they're unsuccessful their descendants don't live.

So, all this gibberish about a frog having intention or purpose in eating flies is either missing the point or parodying missing the point. Who can tell?

By dave souza (not verified) on 22 Apr 2009 #permalink

John,

I'm not a philosopher -- hey, I don't even play one on TV 8^) -- but I did spend two years in the Ph.D. program in Ecology and Evolution at Stony Brook University in the early 1980s. This was not long after Gould and Lewontin published "The Spandrels of San Marco and the Panglossian Paradigm", with which I'm sure we're all familiar. There was much interest in my department at that time in making adaptationist explanations more testable. That was also shortly after the impact hypothesis for the end-Cretaceous extinctions was proposed, emphasizing the role of luck in long-term evolutionary trends. I didn't complete the program, but I did come away thinking that natural selection is much better at explaining the apparent adaptation we observe, than it is at predicting future winners of the "struggle for existence." After following developments in E & E through the ensuing decades, I still think so.

Adaptationist arguments can be rigorous. I remember a colloquium in which George C. Williams, who was on the Stony Brook faculty, talked about ways that a Just So Story like "How the Elephant's Child Got Its Trunk" could be tested. For example, you could look for close relatives of the elephant that had noses only as big as a boot; you could examine the musculature of the elephant's trunk to see if it appeared to have been stretched; you could check for crocodile tooth marks in the trunk's tip; and so forth. Of course, by the time he'd got that far we were all ROTFLOAO! His point was made, however. Some adaptationist arguments have met the testability criterion, and it seems likely that more will over time.

Nevertheless, no one can argue that chance and contingency haven't had a huge influence as well, even in the long term. That's why predictions about the future course of evolution can only be speculative, albeit sometimes inspired (Dougal Dixon is great at it). We can never really know which genotypes are "fittest" until we count their descendants.

Does that make the whole idea of adaptation by natural selection tautological? If I understood Dennett correctly, "Darwin's Dangerous Idea" was that, given heritable variation among a surplus of propagules, natural selection must happen, and that, acting over generations, it can explain why the organisms we see are apparently adapted. Particular adapationist arguments should be rigorously tested, but the underlying theory itself is ineluctable, tautology or no.

By Mal Adapted (not verified) on 22 Apr 2009 #permalink

Does that make the whole idea of adaptation by natural selection tautological? If I understood Dennett correctly, "Darwin's Dangerous Idea" was that, given heritable variation among a surplus of propagules, natural selection must happen, and that, acting over generations, it can explain why the organisms we see are apparently adapted. Particular adapationist arguments should be rigorously tested, but the underlying theory itself is ineluctable, tautology or no.

Mal,

Isn't this a little bit like saying that if we release an object from a certain height, it will either fall, or rise, or move to the side, or hang motionless in space, or fall for a while, then reverse course and rise, or some other variant of these possibilities, depending on "chance and contingency"? The laws of physics don't explain why the objects on my desk are there, rather than some other place, although the full historical explanation of the their arriving there in their present array would include the laws of physics, and would find no need to call upon, say, divine intervention to fill in the gaps.

"Adaptation" is in the eye of the beholder. The earliest organisms, or very their close relatives, still exist; in fact they rule the earth. For every giraffe there is a gazelle, for every zebra, a horse, for every cryptically-patterned butterfly, a cabbage moth (even in the famous and "testable" case of the peppered moth, the unpigmented variant did not die off, but still thrives.) As Bertalannfy once said, if differential reproduction were the only issue, than why didn't evolution end with the rabbit and the herring? He also said (and this is more to the point):

If selection is taken as an axiomatic and a priori principle, it is always possible to imagine auxiliary hypotheses ... to make it work in any case. But this procedure corresponds exactly to that of epicycles in the Ptolemaic system: if planetary motion is a priori cyclic, then any orbit, however seemingly irregular, can be explained by introducing more epicycles. Some adaptive value can always be construed or imagined. But [this] proves neither the Aristotlean-Ptolemeic nor neo-Darwinian system.

In the real world, organisms and environments are interconnected with such intricacy that our claims to understand their causal relationships with confidence are almost comical. The only effective "test" would be to run the whole enterprise over again, each time isolating a new variable. ("This time, let's give the lichen a Pantone color of 358C, and with the headwinds from the West this time, OK, Action!")

Most importantly, the math doesn't work. Genes (to the extent they can be defined at all) don't function in the one-to-one relationship with phenotypes that the adaptationist model requires. Epigenetics matters. Gene expression matters. The locust and grasshopper are two "phases" of the same species. Which phase is manifest has everything to do with the actual experience of individual Acrididae (specifically, how much rubbing they experience upon their hind legs, a sign of overcrowding).

"...Fodor doesn't happen to call himself "Backspace" while online, does he?..." Cory Albrecht | April 20, 2009 11:55 PM

Nope, I am Stephanus Rensburg(backspace) but it is though as if Fodor has been reading my posts. All I would add to his article is that Darwin never said NS is an effect, he used it as a cause. Fodor assumed Darwin used NS as an effect.

By Stephanus Rensburg (not verified) on 30 Apr 2009 #permalink