Tuesday, September 06, 2005

CogBlog - Tomasello: Chapter 2

I have commented on Tomasello’s Chapter 1 before. Second chapter is much longer and somewhat disjointed, but I would like to write some of my own first impressions now (also long and disjointed), before I read what other members of the reading group have written. As usual, I will make the post contrarian and critical, in the good tradition of blog-writing, but that does not mean I dismiss Tomasello’s hypothesis altogether or do not look forward to reading the rest of the book. Read reviews by other group members for other angles and perspectives. Finally, let me point out that this post is aimed mainly at the first half of Chapter 2 (roughly up to p.36), as I feel my expertise is lacking to evaluate properly the second part (from p.37 on).


Possibly the most influential and cited paper in the history of behavioral biology is the 1963 paper by Niko Tinbergen (Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift fur Tierpsychologie, 20: 410-433. Reprinted in L. D. Houck and L. C. Drickamer (Eds.) Foundations of Animal Behavior: Classic papers with commentaries (pp. 114-137). Chicago, IL: The University of Chicago Press.), which, surprisingly, is not available online, though the Journal is a relatively small and obscure one and the paper is pretty ancient.

The main gist of the paper (it is a long paper with lots of “gists”) is that a proper understanding of behavior requires answering four big questions: mechanism, ontogeny, history and function. The four angles of inquiry inform each other, inspire each other, and place checks on each other. Thus, answers to all four are needed before a problem can be deemed “solved”.

By mechanism, Tinbergen means the underlying physiology – how the sensory structures, brain, hormones and muscles work together to produce behavior. Today, one would definitely include molecular and cellular aspects of mechanism.

Ontogeny includes both prenatal and postnatal development of the behavior and the development of its underlying anatomy and physiology. Most importantly, it also focuses on the ways environmental influences affect the development of the behavior. This is probably the line of questioning most readily associated with ‘classical ethology’ as much work was done in this vein before modern techniques allowed one to study details of molecular and neuroendocrine mechanisms underlying behavior.

History refers to the study of phylogeny of the behavior. What is usually meant by this is correlating variations in behavior between related species to their ecological demands. This line of study provides the answer to the question what aspects of behavior are present due to evolutionary history (ancestors had it, too) and which aspects are due to selective pressures of the immediate environment on a particular species (novel traits). This methodology is often called Comparative Biology.

Function is adaptive function. Insights into the adaptiveness of a behavior can provide clues into details of evolutionary mechanisms that produced the behavior, i.e., was it straightforward natural selection, or sexual selection, niche-construction, perhaps selection on multiple levels, etc.

The first two questions are also called Proximate Causes and the latter two Ultimate Causes of behavior. Combining all four approaches in one’s research is called Integrative Biology and has spread from ethology to many other areas of biology, e.g, physiology, anatomy, embryology, etc.

How does Tomasello’s account so far stack up against the four big questions?

There is, so far, not a word about mechanism, i.e., the brain, and there is no hint that there will be more in the latter part of the book. Ontogeny is taken very narrowly – only the post-natal cognitive capabilities and environmental influences on them, and there is nothing about the pre-natal or post-natal development of the underlying neural architecture. Thus, no Proximate Causes seem likely to be discussed in the book. Without the study of proximate causes, hypotheses about ultimate causes are speculative – Just So Stories.

On the front of History, Tomasello in this chapter looks at differences in cognitive capabilities of humans in comparison to non-human primates, as well as comparing non-human primates to non-primate mammals. A for Function, the hypothesis is that Theory Of Mind is what separates humans from non-humans and that that the rest of the differences we see today are a result of cultural evolution.


Reading Chapter 2, especially the first few pages of it, reminded me that I have read something by the author before: a chapter in a book called Evolution of Cognition. What I remember from it is how narrowly anthropocentric his chapter appeared right next to chapters by Big-Picture people like Sara Shettleworth, Nikki Clayton, David Sloan Wilson or Dan McShea.

Although I have read a little bit about primate cognition, e.g., Wild Minds by Marc Hauser (and a wonderfully astute critique: If A Lion Could Talk by Stephen Budiansky), this is not my area of expertise and I am not sure how much to believe Tomasello’s assertions.

However, I am inclined to distrust him after I have read his first couple of “lists” of differences between mammals and other animals, and between primates and other mammals. Every cognitive capability on the first list, assumed by Tomasello to be a privilege of only mammals, has been found and studied in a number of other animals, including non-mammalian vertebrates and, in some cases, even in invertebrates. Studies of animal cognition in a variety of species of birds, e.g., Clark’s Nutcrackers, Mexican Jays, Pinyon Jays, Scrub Jays, pigeons, quail, chickens, hummingbirds, parrots, crows and ravens have described examples of either fully-developed or incipient instances of all the items on the list. Some of the “mammal-only” (according to Tomasello) capabilities can be found even in honeybees and jumping spiders. I wish students of primate/human cognition would familiarize themselves with the latest research in animal cognition (read, for instance Cognition, Evolution, and Behavior by Sara Shettleworth, Animal Cognition in Nature, Animal Minds by Donald Griffin and Mind of the Raven by Bernd Heinrich). Just look at what Alex is capable of.

The second list, itemizing “primate-only” cognitive capabilities also does not make me happy. Decades of research on dogs, whales and dolphins has uncovered either fully-developed or incipient instances of all of the items on the list (Check these for instance: Domestication of Social Cognition in Dogs(PDF) and Did domestication make dogs smarter?).

So, when we get to the list of “human-only” cognitive abilities, I mistrust Tomasello. He describes studies in which primates exhibited certain human-only capabilities, but dismisses them due to the need for thousands of repetitions before the task is learned. But, nobody is expecting a chimp to perform like Einstein. They are chimps, after all. But, they CAN DO IT, which means that they possess the capability even if it is not as well developed as in humans. It is therefore not honest to place such capabilities on a “human-only” list. The anthropocentrism is the one of the most important mistakes of Evolutionary Psychology (capitalized), so why does Tomasello persist in this error?

Imitation vs. Emulation

Tomasello also makes big deal about the difference between imitation and emulation. The former incorporates the understanding of the intentions of the individual you are learning from, and the latter only contains the obvious utility of the result of the action for the learner who is observing. He argues that only humans imitate and all the others only emulate. Is that true? Even if it is, is it important?

Have you ever heard of ‘cribbing’? This is something horses do and it is deemed a vice, as it diminishes the appetite and thus prevents building of muscle mass needed for optimal performance (in sports, for instance). A horse places his teeth on a fence or gate, leans on it somewhat with half-open mouth and, with a burping sound, swallows air.

Now, horses do all sorts of idiosyncratic things, from head-waving and tossing, to assuming funny poses or making strange sounds. Yet, those remain idiosyncratic – no other horse picks up on the habit.

Occasionally a horse becomes a master of opening complicated latches on doors and gates, yet 60 other horses watching this happen several times a day, every day, for years, NEVER learn by imitation how to undo the locks.

But, if one horse discovers cribbing, within a couple of weeks the WHOLE barn will be cribbing. The owners fight it really hard. You put tight leather strap around the throat to make swallowing air difficult, you cover the tops of gates and fences with metal which does not feel as comfortable on teeth as wood does, you apply a thick coat of used motor oil (or jalapeno sauce) on top of all edges and surfaces that are conducive to cribbing in order to make them unpalatable, you exercise the hell out of the horse every day, you let him stay outside in the field as long as possible, but most importantly, you isolate the horse from other horses, both visually and auditory (blaring music has been heard in some stables) because cribbing spreads through barns like wildfire.

Now, imagine that you are a horse (Horse A) who has never cribbed and never seen or heard cribbing before. One day, a new horse moves into a nearby stall (Horse B) and soon starts cribbing. You see him and hear him. Why on Earth would you try to do it yourself? There is no obvious utility to it (unlike unlocking latches which lets you free to run outside or get into the food bin). Well, cribbing is apparently pleasurable to horses (endorphins are released in the brain, producing a ‘high’). But how do you know that cribbing is pleasurable? A bunch of your stable-buddies do all sorts of crazy things and you never even bother trying to imitate them, yet you immediately start cribbing yourself. How do you know that cribbing is pleasure? Do horses have a Theory of Mind? They certainly appear to be able to discern the emotional state of other horses in connection to their actions and then try to imitate those actions IN ORDER to achieve the same emotional state. Is there any other explanation?

Hey, even insects can imitate (or is it emulate?).

But, is recognition of intentionality really that important? I used an ‘elephant-with-an-axe’ example last time. Why is recognition of intentionality necessary for cumulative culture? Primatologists have observed chimp cultures for only a couple of generations. The civilization we see around us is the result of at leas 100,000 human generations, and most of the recognizible culture developed only within the last 10,000 generations. The modern civilization developed during the past 100 generations and the Enlightement (what most of us think of when we think of human civilization) is the product of the past 10 generations or so. Perhaps chimp culture would develop in a cumulative fashion in the future, in a million years or so. Perhaps not due to minor differences in cognition. I don't think Tomasello has a case for a big cognitive difference between the two species.

Well, what is it?

Chapter One and the second half of Chapter Two push for a single small difference in "biology" and the rest being dependent on cultural evolution. But the first half of Chapter Two appears to argue for several large differences between humans and the closests relatives to humans. Why the discrepancy? Where does the turth lie? Why is Tomasello undermining his own hypothesis?

With all this Katrina blogging plus Dissertation writing, I never made it to Chapter 3. I am looking forward to it.

See more on
CogBlogGroup Technorati tag, CogBlogGroup Del.icio.us. tag and Culturaloriginsofhumancognition Del.icio.us. tag

Technorati tag:

posted by Bora Zivkovic @ 11:01 PM | permalink | (0 comments) | Post a Comment | permalink