I must have been taking a nap a couple of years ago. I just found this interesting discussion of EP by a psychologist, and I agree very much with it.
Evolutionary psychologists believe that the human mind works much like the body… that it is an information-processing system, with pre-specified psychological programs (or environmentally-triggered ones), adapted much like the rest of the body, to meet specific problems in our evolutionary past. Others, including myself, disagree with this definition of the human mind. While I would certainly agree that evolution had a profound role in shaping lower-level modular systems, including autonomic nervous system responses, reflex arcs, immune systems, complex motor control, systems related to sexual arousal, and so on, it does not make sense for us to assume that our more complex social behaviors were shaped in the same way, or that they would even depend on lower-level domain-specific systems that evolutionary psychologists so frequently assume to be the ‘ultimate’ causes of behavior. Neurobiologists Panksepp and Panksepp point out that while evolutionary psychologists may interpret psychological data in a way to fit with their preferred theory, the philosophical assumptions that are the foundation of that theory are not at all consistent with what we know about human neurophysiology.
There’s also a longer paper associated with the argument. I liked this bit:
Evolutionary psychologists appear to be living in the Land of Oz –implicitly suggesting that when our genetic sciences mature, we will someday look behind the Wizard’s curtain to find DNA proof supporting their modular hypotheses. However, there is reason to suspect that we will uncover what we should have always guessed—and this is where the computer analogy does ring true—it was not nature that selected these modules, but humans who put them there, crafting stories that were so good, they would even fool themselves of the truth.
I still shake my head in disbelief every time I run across an evolutionary psychology paper. Fortunately, they don’t get published much in the kind of journals I read, so I don’t encounter them often.
” it does not make sense for us to assume that our more complex social behaviors were shaped in the same way, or that they would even depend on lower-level domain-specific systems ”
If he is implying that neocortical functions are divorced from lower level subcortical/brainstem functions, then this is a poorly reasoned critique of EP. Even Panksepp, Damasio, Kandel, Edelman…….cognitive neuroscientists et al. would absolutely disagree.
And, like with a computer – you can a) change the OS entirely, to some extent, or b) replace the bad, buggy, code with better. The one thing that sticks in my craw with EP is that, if true, it would be like a scientific version of predestination religion – you can’t change any of the screwed up, stupid, things we do, as a culture, because its all “hardwired” in. Its also why its wrong, we can find cases that don’t fit their silly, “this is fixed and unchangeable”, nonsense, and they lack any sound way to explain, even as “variations”, never mind what they should be, which is “impossible” if many/most of their claims where true.
Hmmm. Why not, may I ask?
Is there no heritable variation in the populations for social traits?
Is there no complex social behavior in “low-level” organisms which lack our glorious mind and cortex? (say, birds)
While I can’t speak for the authors, I didn’t get that sense at all. For two reasons, here is the sentence: it does not make sense for us to assume that our more complex social behaviors were shaped in the same way, or that they would even depend on lower-level domain-specific systems.
First, he isn’t saying that the lower-level domain specific systems have no affect. He is saying that the more complex social behaviors don’t depend on the lower-level domain specific systems. And he is basing this judgement on his knowledge of human psychology, which I believe is further supported in another section of the quoted material;
Which suggests to me that he is saying that even if the modules are there, they are created by culture and can at times supersede the directives from lower-level domain specific systems.
The second reason is, in my opinion, the more damning critique of EP,
In which he says that the underlying assumption of EP is that our complex social behaviors are based on lower-level domain specific systems without having any evidence to show that assumption might be true.
Strange. It dropped one of my blockquotes. Maybe I formatted it incorrectly.
the human mind works much like the body
Doesn’t that presuppose that “mind” and “body” are separable? I’d have thought you wouldn’t agree with that, PZ.
Irene @ 6:
I didn’t read it that way, it’s pointing out similarity in function. Mind doesn’t equal Brain.
As someone who disagrees with the tenets of EP (ie. domain specific genetically determined modules aka Swiss Army knife phrenologic cognitive architecture), there is substantial evidence that higher level cognitive functions do indeed depend on lower level systems. The reward networks in the brainstem have a direct impact on the social decision making involved in the orbitofrontal prefrontal neocortices. The pathways were nicely detailed for me during my Neurology residency at Iowa City. The plastic neocortex is very much dependent on these lower domain specific regions.
Also, there is a degree of domain specificity in the cerebral cortex. Face recognition is a classic example. Those fusiform gyri are online at birth as you recognize mom and dad. What is plastic is that these regions are active when say a surgeon recognizes specific tools in the OR. Domain specific regions for face recognition are co-opted for recognizing different objects.
I fail to see how falsely severing complex behaviors from lower domain specific functions (eg. brainstem reward systems), disproves EP. It only adds an additional layer of pseudoscience that no honest cognitive neuroscientist would ever accept.
I get the sense that some view higher brain function as being independent of body function. Recall Robin Williams’ (RIP) character of the Moon King in The Adventures of Baron Von Munchausen. His character’s head which is dedicated to art, music, science, etc.. often detaches from his body which seeks physical gratification, farting, eating, etc… This is a false view of the mind. Body functions involved with homeostasis are mapped in the brain (brainstem and hypothalamus to be specific) and is called the proto-self. The moment to moment interaction of the protoself with the environment is mapped in the cerebral cortex to form what is called the core self. With the development of the neocortex via evolution for language, memory and decision making, this culminates in what is called the extended self. This has all been proposed via lesion analysis among patients who have suffered neurological illness. Now there is growing evidence that this embodiment of mind is not merely a proposal but an established canon in cognitive neuroscience. You are not just the neocortex. Indeed, you are the manifestation of an interaction between a brain that is dependent on the body. Am I preaching to the choir here?
The “mind” is just shorthand for what the brain does, and the brain is a body part. (Well, actually, that’s not quite true either, given that emotional responses and so forth occur throughout the body, so “mind” can’t be limited to just things happening inside the skull.) There is no separate ghost in the machine. I see people (including many otherwise respectable scientists) trying to use “mind” or “brain” the same way people used to use “soul” all the time. It’s partly repeating habitual phrases, and partly that we still see things in terms of a separable “self” who is not the body.
Al Dente says
I’m possibly talking out of my ass but I do see a few indications that human behavior is affected by something like EP. We’re social animals. Our basic social behavior and ethics are things like: Don’t kill other members of the family/tribe/other social group unnecessarily; share food and other possessions; don’t take things belonging to others you don’t have a claim to; protect younger generations; etc. When people do violate these mores, the violation gets noticed and often punished, either formally or informally. I think that evolution does have some impact on humanity’s success as social beings.
It’s a pretty good critique, but I have a few issues with it.
I’ve read the Panksepp & Panksepp paper that Brad Peters references. It’s a really good paper, and one of the best criticisms of evolutionary psychology I’ve ever read. But, it was published in 2000, and we’ve learned a few things about the brain in the last 15 years that we didn’t know then. There’s considerably more evidence for neuroanatomical modularity of certain cognitive functions than there was then.
In addition to the highly specialized areas for processing faces that ragdish mentioned, human neuroimaging has identified specific parts of the anterior cingulate cortex, orbitofrontal cortex, and dorsolateral prefrontal cortex which seem to participate solving particular social problems. Homologous areas performing similar functions have been identified in monkeys, suggesting that this neuroanatomical organization of function has been conserved across primate evolution.
Also, while I have great respect for Jaak Panksepp as a scientist, I don’t think his understanding of brain evolution is very sophisticated or up-to-date, and I wouldn’t necessarily rely on him as a reference on that front. I attended a talk he gave a few years ago in which he endorsed the “Triune Brain” hypothesis (according to which there is a “reptilian brain” which evolved first, a “paleomammalian brain” which evolved later, and a “neomammalian brain” which evolved last, each supporting more complex functions), which was originally proposed based on no data and which comparative neuroanatomists have known to be completely incompatible with the evidence for decades, although it is unfortunately still taught in many introductory psychology classes. Panksepp’s criticisms of evolutionary psychology with respect to its internal consistency and misuse of evolutionary theory generally are spot-on, but his criticisms based on brain evolution are in my opinion a little off-the-mark.
So basically, I think Peters overstates the case that there’s no or weak evidence for modularity in neural or cognitive function, and that complex social (or other cognitive) behaviors can’t be built from and depend on “lower-level ‘domain-specific’ systems.” To the contrary, I think there’s actually considerable evidence coming out from neuroscience research that this is exactly the case. However, the manner in which this seems to be true still doesn’t help the evolutionary psychology approach.
For example, the “visual word form area” (VWFA) is a part of human cortex which is involved in processing written language. It’s been found to have a very consistent location regardless of the individual, language, or writing system. Yet despite being a readily identifiable neuroanatomical “module,” this area obviously cannot have evolved specifically to support the function of reading, since written language was only developed 6000 years ago, and for many human populations was introduced only very recently. (E.g., the Cherokee writing system was invented in 1821 by one individual, Sequoyah, whose only exposure to writing was a bible which he himself could not read, and the Cherokee people subsequently achieved near-universal literacy within a single generation.) So the VWFA provides evidence that preadaptation/exaption is an important evolutionary mechanism for explaining human cognition, rather than the typical evolutionary psychology position which presumes adaptation is universally the best explanation.
And from another angle, in 2011 a research group at Oxford published a paper in which they showed that when monkeys were housed in social groups of various sizes, those housed in larger groups showed increases in cortical thickness in very specific prefrontal cortical subregions, which were already thought to play a role in social cognition in humans and other primates. This highlights that while there is indeed some functional modularity in the primate (including human) brain, even for complex social behavior, those functional modules are highly sensitive to individual experience, making it clear that a simple-minded notion of “cognitive modules” which are inflexibly genetically programmed is inconsistent with the highly plastic nature of the brain, even if there are in fact neural modules showing some correspondence to cognitive ones.
I would like to quote one section where I think Peters absolutely nails it, though, in distinguishing two separate ideas that defenders of evolutionary psychology rely upon:
You took the words right out of my mouth. I think the next key step is to define what a domain specific module actually is in terms of biological neural networks. Indeed, there is overwhelming evidence that they are not Jerry Fodor’s information encapsulated genetically defined cognitive programs. But, connectionist theories have largely been abandoned. I’m more sympathetic to Damasio’s convergence zones theory which posits that “modules” are the manifestation of binding regions in association cortices that link primary motor and sensory regions. Broca and Wernicke’s areas are good examples of convergence zones.
“Those fusiform gyri are online at birth as you recognize mom and dad. What is plastic is that these regions are active when say a surgeon recognizes specific tools in the OR. Domain specific regions for face recognition are co-opted for recognizing different objects.”
You introduced this by saying this was for face recognition. I am curious as to how neuroscientists know that these parts of the brain are not instead processing information about moveable parts of the surroundings. In infancy, given the limited mobility and ability to move objects themselves, infants would consistently see parents moving about more than anything else. And the faces would be prominent, distinctive and undisguised by clothing. Knowing the fusiform gyri are “for” facial recognition rather than processing a particular kind of information seems to be kind of tricky? (Alternatively, the fusiform gyri might be considered as performing one key algorithm/computational procedure that happens to be essential to “facial recognition” in infancy? A kind of processing modularity rather than goal modularity?)
biogeo@12 Perhaps I’m wrong, but I thought the quick and dirty takeaway from your examples, for lay people at least, should be that “learning rewires the brain.” Or to put it more polysylllabically, the presumption should be that differences in brain function are due to development (including learning,) but interpreting brain modularity as genetically pre-programmed must be demonstrated by more than a fMRI version of “post hoc, ergo propter hoc.”
Or is this too influenced by computational models of mind? I must say I’ve been thinking the difference between digital and analog computation is more drastic than the usual impression conveyed, but I’ve thought that on some level, the computational model does highlight the fact that in some sense there are computations of a kind that must be solved. In some sense, a brain does “solve” equations for the trajectory of projectiles. And thus we have a handle on the amount of work/information processing that the brain must accomplish in its own way, however different from the way a person does consciously.
Yes and no. Certainly that is part of it, but should be paired with the idea that the “rewiring” happens in predictable ways, which may be partially explainable in terms of evolutionary mechanisms including adaptation, exaption (co-opting a mechanism evolved for one function to a new function), and phylogenetic constraints (once a lineage has evolved a particular mechanism it may be hard to lose it). I think it’s appropriate to think of neural plasticity as extended development, which means various concepts from evo-devo apply. It is the whole developmental program, not just the adult form, which is under selection. So the effect of learning on the brain is not just arbitrary rewiring, but relatively specific changes depending on the kind of learning that’s happening. This is why the VWFA exists; presumably something about the kinds of functions it performs predisposes it to be capable of taking on the function of processing written language, in ways that other parts of the brain could not do, or could do so less effectively. The flip side is that while it is true that we see this kind of specific modularity in response to learning, neural plasticity is a sufficiently powerful process to overcome this natural modularity when necessary. For example, in blind individuals, much of the “visual” cortex can be remapped to process information from other senses, while still performing similar functions on that information. So in my opinion, the takeaway should be, “The brain is really complicated, and both specificity and flexibility are important principles in understanding its organization.”
Well, to be fully up-front about my biases, I’m a pretty strong computationalist with regards to the brain and the mind. I think you’ve absolutely hit the nail on the head: digital and analog computation are very different from each other in important ways, and the brain is fundamentally an analog computational system. (Though this is a point of contention; some computational neuroscientists think that the brain really is a digital computational system, because the main form of communication in the brain, action potentials, are all-or-nothing events, like binary 1’s and 0’s. I think this is wrong because there are a variety of communicative and computational principles at play in addition to action potentials, and even for action potentials it’s unclear to what extent individual events are meaningful rather than the analog rate at which they occur. But this is an open topic in the field.) I think that considering the brain, and subsystems of the brain, as an analog computational system is absolutely critical to understanding its function. Furthermore, in many cases it may be the computations which are performed rather than the cognitive or behavioral functions supported by those computations which are highly conserved in evolution, and so I think the computationalist approach has the potential to give us insight into the evolutionary history of the brain, and vice versa.
Also, I fear in my comments I’m at risk of propagating a common misconception about the brain, that cortex = cognition. I’ve talked about a number of cortical brain areas, but many subcortical structures also play an important role in cognition. In fact, some neuroscientists (including myself to a certain degree) think that the function of the cortex cannot be dissociated from its strong interconnections with subcortical structures. For example, one of the fundamental motifs of the organization of the brain is the “cortico- basal ganglia loop,” which is a circuit running from the cortex, through subcortical structures called the striatum and pallidum, to a very important structure called the thalamus, and back to the cortex; it may be that this entire loop is a fundamental computational unit in the brain for supporting cognition, and the cortex is only one node in that unit.
IIRC reading doesn’t really map onto face recognition. People with local brain damage who lose face recognition, don’t lose reading abilities. Face recognition seems to be a very specific brain module. On the other hand, in the sequel to “The gods must be crazy” (a movie and sequel about the reaction of Kalahari Bushmen to traces of “civilization”, first a coke bottle dropped from a plane and then a couple of clueless lost smugglers IIRC. Sympathetic and funny) In the second movie the Bushmen protagonists were shocked to reluctantly understand that the lost smugglers couldn’t read tracks. They referred to them as(translated) “illiterate”. I think that the translation was very apt. First for the emotional reaction as we would have, a combination of pity and disgust. And that it seems to be IMHO that reading might be based on the same module or ability that developed for tracking.
I asked a similar question to my supervisor during residency. If you were born in an environment wherein there are no faces (ie. humans, animals, etc), what will those regions in the brain that serve face recognition be doing? Do they shut off? On the contrary, those neural regions are good examples of exaptation. They will serve another visual function. What if a person is blind? Then those neural regions will serve another sensory modality for recognition. The key is that those regions involved in face recognition are involved with the task of recognition. It just so happens that from birth until death we are bombarded with faces and thus, the fusiform gyri function towards that end. It is speculation as to why those neural regions evolved in the first place. Evolutionary psychologists dogmatically claim that those areas evolved via natural selection to enable us to recognize predator from prey and therefore increased our chances of survival. That may be true or it may be a “just so story”. But I will concede that face recognition is a “module” (not in the Fodorian sense) that is partly hardwired via genes but can be rewired for different tasks (eg. surgeon recognizing specific surgical instruments).
ck, the Irate Lump says
Trying to explain the mind entirely through evolutionary biology is like trying to explain art entirely through chemistry. There is plenty that the latter can tell you about the former, but there will always be limits. When so much of evo psych can be overturned by simply researching history or other cultures, you know the discipline is attempting to prove far too much.
Interesting new findings about the brain that might be relevant to the debate:
ragdish@18 Thanks for the information.