Wednesday, June 21, 2017

Two things to read

Here are a couple of readables that have entertained me recently.

The first is a NYT report (here) on what is taken to be an iconoclastic view of the role of animal aesthetics in evolution. According to the article, a female’s aesthetic preferences can drive evolutionary change. This, apparently, was Darwin’s view but it appears to be largely out of favor today. More utilitarian/mundane conceptions are favored. Here’s the mainstream view as per the NYT:

All biologists recognize that birds choose mates, but the mainstream view now is that the mate chosen is the fittest in terms of health and good genes. Any ornaments or patterns simply reflect signs of fitness.

The old/new view wants to allow for forces based on fluffier considerations:

The idea is that when they are choosing mates — and in birds it’s mostly the females who choose — animals make choices that can only be called aesthetic. They perceive a kind of beauty. Dr. Prum defines it as “co-evolved attraction.” They desire that beauty, often in the form of fancy feathers, and their desires change the course of evolution.

The bio world contrasts these two approaches, favoring the more “objective” utility-based one over the more “subjective” aesthetic one.  Why? I suspect because the former seems so much more hard-headed and, thus, “scientific.” After all, why would any animal prefer something on aesthetic grounds! If there is no cash value to be had, clearly there is not value to be had at all! (Though this reminds one of the saying about knowing the price of everything and the value of nothing).

An aside: I suspect that this preference hangs on importing the common sense understanding of ‘fitness’ into the technical term. The technical term dubs fit any animal that sends more of its genes into the next generation whatever the reason for this. So, if being a weak effete pretty boy allows greater reproductive success than being a tough successful but ugly looking tough guy than pretty boyhood is fitter than ugly tough guy even if the latter appears to eat more, control more territory and fight harder. Pretty boys may be less fit on the colloquial sense, but they are not less fit technically if they can get more of their genes into the next generation. So, strictly speaking appealing to a female’s aesthetics (if there is such a thing) in such a way as to make you more alluring to her and making it more likely that your genes will mix with hers makes you more fit even if you are slower, weaker and more pusillanimous (i.e. less fit in common parlance).  

Putting the aside aside, focusing on the less fluffy virtues may seem compelling when it comes to animals, though even here the story gets a bit involved and slightly incredulous.  So for example here’s one story: peahens prefer peacocks with big tails because if a peacock can make it in peacock world despite schlepping around a whopping big tail that makes doing anything at all a real (ahem) challenge, then that peacock must be really really really fit (i.e. stronger, tougher, etc.) and so any rational peahen would want its genes for its own offspring. The evaluation is purely utilitarian and the preference for the badly engineered results (big clumsly tail) are actually the hidden manifestations of a truer utilitarian calculus (really really fit because even with handicap it succeeds).

And what would the alternative be? Well, here’s a simple possibility: peahens find big tails hot and are attracted to showy males because they prefer hotties with big tails. There is nothing underneath the aesthetic judgment. It is not beautiful because of some implied utility. It’s simple lust for beauty driving the train. Beauty despite the engineering-wise grotesque baggage. Of course, believing that there is something like beauty that is not reducible to having a certain (biological) price is a belief that can land you in the poorly paid Arts faculty and exiled from the hard headed Science side of campus. Thus, they are unlikely to be happily entertained. However, it is worth noting how little there often is behind the hard headed view beside the (supposed) “self evident” fact that it is hard headed.  Nonetheless, that appears to be the debate being played out in the bio world as reported by the NYT, and, as regards animals, maybe ascribing aesthetics to them is indulgent anthropomorphism.

Why do I mention this? Because what is going on here is similar to what goes on in evo accounts concerning humans as well. The point is discussed in a terrific Jerry Fodor review of Pinker’s How the Mind Works in the LRB about 20 years ago (here). If you’ve never read it, go there now and delight yourself. It is Fodor at his acerbic (and analytical) best.

At any rate, he makes the following point in discussing Pinker’s attempt to “explain” human preferences for fiction, friends, games, etc. in more prudential (adaptationist/ selectionist) terms.

I suppose it could turn out that one’s interest in having friends, or in reading fictions, or in Wagner’s operas, is really at heart prudential. But the claim affronts a robust, and I should think salubrious, intuition that there are lots and lots of things that we care about simply for themselves. Reductionism about this plurality of goals, when not Philistine or cheaply cynical, often sounds simply funny. Thus the joke about the lawyer who is offered sex by a beautiful girl. ‘Well, I guess so,’ he replies, ‘but what’s in it for me?’ Does wanting to have a beautiful woman – or, for that matter, a good read – really require a further motive to explain it? Pinker duly supplies the explanation that you wouldn’t have thought that you needed. ‘Both sexes want a spouse who has developed normally and is free of infection … We haven’t evolved stethoscopes or tongue-depressors, but an eye for beauty does some of the same things … Luxuriant hair is always pleasing, possibly because … long hair implies a long history of good health.’

Read the piece: the discussion of why we love literature and want friends is quite funny. But the serious point is that aside from being delightfully obtuse, the more hard headed “Darwinian” account ends up sounding unbelievably silly. Just so stories indeed! But that’s what you get when in the end you demand that all values reduce to their cash equivalent.

So, the debate rages.

The second piece is on birdsongs in species that don’t bring up their own kids (here). Cowbirds are brood parasites. They are also songbirds. And they are songbirds that learn the cowbird song and not that of their “adoptive” hosts. The question is how do they manage to learn their own song and not that of their hosts (i.e. ignore the song of their hosts and zero in on that of their conspecifics? The answer seems to be the following:

…a young parasite recognizes conspecifics when it encounters a particular species-specific signal or "password" -- a vocalization, behavior, or some other characteristic -- that triggers detailed learning of the password-giver’s unique traits.

So, there is a certain vocal signal (a “password” (PW)) that the young cowbird waits for and this allows it to identify its conspecific and this triggers the song learning that allows the non cowbird raised bird to learn the cowbird song. In other words, it looks a very specific call (the “chatter call”) triggers the song learning part of the brain when it is heard. As the article puts it:

Professor Hauber's "password hypothesis" proposes that young brood parasites first recognize a particular signal, which acts as a password that identifies conspecifics, and the parasites learn other species­-specific characters only after encountering that password. One of the important features of the password hypothesis is that the password must be innate and familiar to the animal from a very early age. This suggests that encountering the password triggers specific neural responses early in development -- neural responses can actually be seen and measured.

It seems that some of the biochemistry behind this PW triggering process has been identified.

…cowbirds' brains change after the birds hear the chatter call by rapidly increasing production of a protein known as ZENK. This protein is ephemeral; it is produced in neurons after exposure to a new stimuli, but disappears only a few hours later, and it is not produced again if the same stimuli is encountered. The production of ZENK occurs in the neurons in the auditory forebrain, which are regions in the songbird brain that respond to learned vocalizations, such as songs, and also to specific unlearned calls.

So, hear PW, get ZENKed, get song. It gets you to think: why don’t humans have the same thing wrt language? Why aren’t there PWs for English, Chinese etc? Or more exactly, why isn’t it the case that humans come biologically differentiated so that they are triggered to learn different languages? Or, why is it that any child can acquire any language in the same way as any other child?  You might think that if language evolved with a P&P architecture and that different Gs were simply different settings of the same parameters with different values (i.e. each G was a different vector of such values) that evolution might have found it useful to give those most likely to grow up speaking Piraha or Hungarian a leg up by prepopulating their parameter space with Piraha or Hungarian values. Or at least endowing these offspring with PWs that when encountered triggered the relevant G values. Why don’t we see this?

Here’s one non-starter of an answer: there’s not enough time for this to have happened. Wrong! If we can have gone from lactose intolerant to lactose tolerant in 5,000 years then why couldn’t evo give some kids PWs in that time? Maybe too much intermixing of populations? But we know that there have been long stretches of time during which populations were quite isolated (right?). So this could have happened, and indeed it did with cowbirds. So why not with us? [1]

At any rate, it is not hard to imagine what the cowbird linguistic equivalent would be. Hear a sentence like “Mr Phelps, should you agree to take this mission then, as you know, should you or any of your team be captured, the government will disavow any knowledge of your activities” and poof, out pops English. Just think of how much easier second language acquisition would be. Just a matter of finding the right PWs.  But this is not, it seems, how it works with us. We are not cowbirds. Why not?

So, enjoy the pieces, they amused me. Hopefully they will amuse you too.

[1] This would particularly apposite given Mark Baker’s speculations (here; p. 23): could be that linguistic diversity has the desirable function of making it hard for a greedy or dangerous outsider to join your group and get access to your resources and skills. You are less vulnerable to manipulation or deception by a would-be exploiter who cannot communicate with you easily.

In this context, a PW for offspring might be just what Dr Darwing might have ordered. But it appears not to exist.


  1. With respect to the password idea, I thought you were going to go a different direction. Namely, why do humans acquire language, and not birdsong, or dog barking, or whatever? Because there is some property of human language that triggers acquisition of it as opposed to other kinds of vocal or gestural systems.

    1. I wasn't thinking of that, but nice point. What I was curious about is why specific languages are not pass-worded. Say there are parameters and say that one cares about one's group (the point Baker mooted), then one can now see how this might be coded genetically. However, as far as we know this did not happen. So why not? Why isn't there a genetic preference for SPECIFIC languages? I have no idea.

      Note that this its related to another question: why are there parameters to begin with? Why don't we just all speak one language? One syntax, one phonology, one morphology etc. The only "answer" I have been able to come up with is that coding for the entirety of G facilities is too expensive. Putting info in the genes is costly and so you put in as little as possible to allow the whole thing to get off the ground. If PLD sufficient to acquiring a G is reliably available and we have some kind of general learning mechanism that can use this info to fill in the blanks in the genetic program then the blanks will remain. I have no idea if this is right, but I can understand it. But, then given that there are blanks, why didn't different sub-communities of humans fill some of them in over time to give learners of that G a leg up? Dunno.

    2. Well the point Norbert raises here (if I understand it at all) is exactly Dan Everett's argument against there being a lot of genetically encoded UG. But it seems to me that the most potentially convincing experiments to show that there is zero genetic adaptation to prevailing grammatical structures never could have been performed in a sound way, even before there was any such thing as research ethics. Such as, grab a bunch of very young kids from villages in mainland SE Asia whose ancestors have been speaking 'isolating' languages for quite some time, and shift some them to pre-contact Navajo country to see how they go with Navajo verb inflection, to pre-contact Mornington Island to check them out on Kayardild case-stacking, etc. etc. But the confounding factors of different appearance and unusual provenence of the children would make such experiments unethical.

      Given that humans clearly adapt to climate and parasites in particular areas, the apparent lack of adaption to apparently stable and widespread grammatical properties is a bit puzzling.

    3. Yes, but the natural experiments of adoption seem to be convincing, even if the original languages may not be so carefully selected.

      One answer is the Chater, Christiansen one : namely that the time scale of language change is so much shorter than that of evolution that there isn't any coupling.

    4. But have the kids been tested enough for absolutely full proficiency in piling the case-markers, etc? For diseases, some 'racial' groups are more susceptible to certain diseases than others, but none are truly immune, afaik.

    5. Also, as I failed to include above, to really demonstrate the irrelevance of heredity to specific features of a supposed UG, you'ld need to test reasonably large groups of transplanted kids, similarly to what is done for epidemiological research, and there's clearly no way to do it, neither ethically nor practically.

    6. Agreed the definitive experiments will be hard to do. But at this time there is no evidence suggesting that the idealization (viz. that any kid can learn any language in the same way) is incorrect. Form what we can tell, it is roughly correct. So, the puzzle is prima facie reasonable. Given that we don't have just one G (there are differences) why aren't we differentially primed biologically to learn some more easily than others.

      Re Alex's point re Chater et al: not sure that ALL properties of a G change quickly. So though Gs may change over time some properties may be quite long lived. If so the problem reappears. Prima facie, the more parameters that are fixed biologically the fewer that need to be acquired and this should make the process easier. If so...

    7. But it is worth remembering that there is a lot of variability in the course and rate of child linguistic development and indeed in the level of adult performance. So there is some space for such effects to be hidden from us.