7b. Cauchoix, M., & Chaine, A. S. (2016). How can we study the evolution of animal minds? Frontiers in Psychology, 7, 358.
During the last 50 years, comparative cognition and neurosciences have improved our understanding of animal minds while evolutionary ecology has revealed how selection acts on traits through evolutionary time. We describe how cognition can be subject to natural selection like any other biological trait and how this evolutionary approach can be used to understand the evolution of animal cognition. We recount how comparative and fitness methods have been used to understand the evolution of cognition and outline how these approaches could extend our understanding of cognition. The fitness approach, in particular, offers unprecedented opportunities to study the evolutionary mechanisms responsible for variation in cognition within species and could allow us to investigate both proximate (i.e., neural and developmental) and ultimate (i.e., ecological and evolutionary) underpinnings of animal cognition together. We highlight recent studies that have successfully shown that cognitive traits can be under selection, in particular by linking individual variation in cognition to fitness. To bridge the gap between cognitive variation and fitness consequences and to better understand why and how selection can occur on cognition, we end this review by proposing a more integrative approach to study contemporary selection on cognitive traits combining socio-ecological data, minimally invasive neuroscience methods and measurement of ecologically relevant behaviors linked to fitness. Our overall goal in this review is to build a bridge between cognitive neuroscientists and evolutionary biologists, illustrate how their research could be complementary, and encourage evolutionary ecologists to include explicit attention to cognitive processes in their studies of behavior.
During the last 50 years, comparative cognition and neurosciences have improved our understanding of animal minds while evolutionary ecology has revealed how selection acts on traits through evolutionary time. We describe how cognition can be subject to natural selection like any other biological trait and how this evolutionary approach can be used to understand the evolution of animal cognition. We recount how comparative and fitness methods have been used to understand the evolution of cognition and outline how these approaches could extend our understanding of cognition. The fitness approach, in particular, offers unprecedented opportunities to study the evolutionary mechanisms responsible for variation in cognition within species and could allow us to investigate both proximate (i.e., neural and developmental) and ultimate (i.e., ecological and evolutionary) underpinnings of animal cognition together. We highlight recent studies that have successfully shown that cognitive traits can be under selection, in particular by linking individual variation in cognition to fitness. To bridge the gap between cognitive variation and fitness consequences and to better understand why and how selection can occur on cognition, we end this review by proposing a more integrative approach to study contemporary selection on cognitive traits combining socio-ecological data, minimally invasive neuroscience methods and measurement of ecologically relevant behaviors linked to fitness. Our overall goal in this review is to build a bridge between cognitive neuroscientists and evolutionary biologists, illustrate how their research could be complementary, and encourage evolutionary ecologists to include explicit attention to cognitive processes in their studies of behavior.
“Still, the ability to disentangle effects that contribute to the evolution of cognition (Figure 3) and experimentally test hypotheses to determine causative effects make the fitness approach an under-utilized method that would nicely complement comparative studies.” (From the section “Cognitive Mechanisms of Host–Parasite Arm Races in Brood Parasites”)
ReplyDeleteI was initially unsure why we had been assigned a reading about evolution because I didn’t understand how it was related to anything we had been learning like T3 robots. This quote in particular gave me an idea of why we are going to be discussing evolution. Mostly we have been discussing the ‘how’ aspect of the easy problem, and my impression is that computation and categorization are at least partial answers to this. Importantly, the easy problem also asks us ‘why’ we do what we can do, and from reading this paper it seems that evolution may provide some answers. As the quote discusses, studies can be done to isolate the causal effects of fitness on cognition. One example the paper uses could be something like food storage requiring a good memory, so organisms with a good memory will have a higher fitness and be able to pass those genes on to the next generation.
One thing that the YouTube video for this week really helped me understand though, is that my specific cognitive capacities may not have had specific selection pressures. By this I mean there seems to be a great deal of environmental influence as well. I doubt that there was a specific and very strong selection pressure to be able to categorize mushrooms, because not all environments and circumstances demand this skill. Maybe there was a selective advantage for individuals with the general learning abilities for categorization, and this has led to different people having different cognitive capacities (e.g. be an expert on categorizing the periodic table, or categorizing mushrooms, or categorizing wines etc.).
Hi Stephanie,
DeleteIt does seem at first glance like we are getting closer to the "why" for what we're doing. Evolutionary psychology, in sum, says we do what we do to ensure the survival of our genes. While there are lots of ways to do that, be it by not eating kin, laying our eggs in others' nests, or categorizing mushrooms, that seems to be the end goal.
However, even that doesn't seem to do the trick, because we can go on to the question of how - how is the want to propagate my genes written into my DNA? As far as I know, no sequence of amino acids has spelled out the mission statement of the genome. Nor do I hold much hopes of finding one.
I could only brainstorm this answer - that when molecules got together, they did so randomly a few hundred million years ago, some of them survived, only because the host didn't die, and so they got passed on. It's not so much that the genome is there to let the genome survive, it's there *because* the genome survived. Survival isn't the product of purpose (why), but the product of physics (what).
Or, if that's all there it is, then maybe we found our answer: when we ask "why" our genes did what they did, our only answer can be "what" our genes did what they did. But that gets us nowhere really. So, the easy problem persists.
@Stephanie, I enjoyed reading your last point about the environmental influences on specific cognitive capacities. I wonder if this could be applied to the development of specific talents. Some people are artistically inclined — perhaps their ancestors had increased visual capacities because they were in charge of foraging? Others have impressive hand-eye coordination — perhaps their ancestors were accomplished hunters?
Delete"Multiple selection pressures on the same cognitive trait would drastically change their evolutionary dynamics (eg. speed of evolution) and trajectories (sequence of trait values on the path to an optimum" (7). This quote brought language to mind. Why is it that only humans have evolved to communicate verbally, with language? What selection pressures were different for us than all other beings on the planet? Animals have their own unique ways of communicating, but I wonder why ours evolved the way it did.
@Claire I think that there is definitely some sense of balance between genetic predispositions and the actual phenotype observed. Even certain health conditions that I would associate with a strong genetic basis can be greatly mediated by the environment of that individual. So I wonder if in this case it is more that our genetic code has afforded us certain sensorimotor abilities, but then through our lived experiences (and a combination of unsupervised, supervised, and instructional learning of categories) we can develop those talents. This is why an infant could learn any human language at birth. I am sure we will discuss language more in the coming weeks, but there seems to be some innate mechanism that is very accommodating of various inputs it could be given. I think that a great point you bring up is why we do what we do seems to be context dependent. If these evolutionary and selective pressures are always changing with the environment, is it possible that there is no static answer to the why we do what we do question and that is what makes it difficult to answer?
DeleteStephanie, yes, evolutionary hunches and evidence address the “why” question (for both the easy and the hard problem).
DeleteThe “laziness” of evolution (which really just means that it would take longer and use more resources, and be much more rigid, to code in the genes things that can be learned from experience. Better to code only the capacity to learn them. And not only is genetic variation important in evolution (to give survival and reproductive success something to “select” from), it’s important during an organism’s lifetime to have lots of different potential talents that might turn out to be useful to develop in different circumstances and environments.
Julian, there is a difference between “distal” & “proximal” causes in evolution.
Ask me in class about the standard “sweet tooth” example usually used (now probably wrong and just a “just-so” story, but the same point could be made with many other traits):
Why do children have a “sweet tooth”? i.e., why do they like sugar so much? Because unlike today (when, because of their sweet tooth, there is a candy machine on every corner), in the human “ancestral” environment, sugar was rare, and predators were many, and fast. So because sugar gives you energy, the children who liked it and ate as much as they of it could when it was available escaped predators better than children who were indifferent to sugar.
Today, when there is sugar everywhere (and few predators), and we know it causes tooth decay and obesity, it’s not beneficial any more (but not harmful enough to make the sweet tooth die out).
But the important thing to note in all this is that children do not eat sugar because they want to raise their blood glucose levels so as to escape predators. They eat sugar because they love the taste; the disposition is genetic. Ancestors who did not have the genetic disposition got eaten by predators, so they did not pass on their genetic indifference to sugar.
The (unconscious, unfelt) “distal” cause of the human sweet tooth is that it gives you energy to escape predators; the (felt, conscious) “proximal” cause is that it tastes good to us. In the ancestral environment, that was an adaptive advantage; in the current environment, not so much.
Claire, not all genetic traits are adaptive: why are some people tall and some short? Height is (mostly) a genetic trait. Lazy evolution is again better for adapting to variable circumstances. Height variance (in a population) is more adaptive, on average, than one-size-fits all. Same is true for talents: we don’t learn to be tall, but we can learn to develop latent talents if we have them and if the environment makes it adaptive, such as running speed (to escape fast predators, or to get into the Olympics).
But these are mostly just cartoon evopsych stories for cartoon ancestral environment scenarios. The big evopsych question is how and why language evolved.
I wanted to highlight some common misconceptions about evolution and natural selection that I noticed in both Julian and Claire’s posts. First, although it gets resolved at the end of the post, the beginning of Julian’s post shows the tendency that we often have to think of natural selection as “goal-oriented”. Natural selection simply refers to the fact that genes have different degrees of reproductive success, leading to changes in the frequencies of genes over time. There is no “mission statement” or “want” encoded in your genome – it is a completely passive process. There can never be the question “why did our genes do what they do”. Genes do not “do” things - organisms do things, based on a combination of genetic and environmental factors. Asking “why” in the field of evolutionary psychology amounts to asking whether an observed behaviour can be explained by a genetic adaptation (or to what extent it can be), and if so, what may have been the selection pressure that produced that adaptation.
DeleteI felt there was a bit of unresolved confusion in your post as well, Claire. I think you may have misunderstood what Stephanie meant by “environmental influences” on cognitive capacities. This doesn’t mean that highly specific abilities are encoded into our DNA, it’s actually sort of the other way around. The environment selects for broad capabilities, such as categorization, which form the basis for more specific cognitive abilities based on the needs of the environment, such as the ability to categorize mushrooms.
I also wanted to point out that you do not develop genetic adaptations by doing tasks, such as hunting or foraging. Environmental pressures weed out certain genes and favour others, meaning that genes that facilitated hunting and foraging “survived”, and indeed influence our present-day abilities. However, most of these adaptations are shared by all of us.
Evolutionary psychology manages to correlate variations of different brain regions to different tasks, but they aren’t explaining how this can be done, they’re simply suggesting X species does it better because x part of the brain is bigger. They explain this by relying on selection of behaviors acting on genes to modulate cognitive capacity. But the genetic code could be thought of as a computer, nothing but squiggles and squoggles used to form the body and the brain according to specific rules encoded in it, not understanding by our bodies involved. Also, if we consider cognition as categorization (doing the right thing with the right kind of thing), then most if not all animals cognize. All of these different species have different behaviours and are under different selection pressures, both ecological and social, that likely change aspects of their cognition over time. And yet they all still cognize. We would not have gotten closer to the how nor the why by studying how cognition changes. I think that evolutionary psychologists’ attempt to link behavior, fitness, and cognitive ability would require so many assumptions and pure correlational interpretations that this is probably a dead-end field of research as far as the easy problem is concerned.
ReplyDeleteYou're right that evopsych does not explain causal mechanisms; just point out that they probably have an adaptive advantage. (That's why they're often called "just-so stories." But they're not all just-so stories. For morphology (being able to fly, or to swim) evolutionary biology (evopsych) has some causal scenarios that are probably right -- and with drosophila (fruit-flies) they can sometimes identify the genes. The causal scenarios for fear of spiders and sex preferences are probably sometimes right too. But when it comes to language, not only is it much harder to explain its evolution but language also seems to change the rules of the game, with (say) internet memes taking the place of genes. Evopsych is not able to explain memes (as in "mirror neurons"); cogsci will be able to explain our capacity to create and use memes, but not predict what we decide to do with that capacity.
DeleteInterestingly, this article did not convince me as much that the methods it outlines are useful to reverse-engineer a T3 passing robot. As opposed to the previous article, the focus was more on cognitive ecology, merely correlating neuroscientific data and gene expression with evolved cognitive abilities. For example, a bigger brain might or might not suggest certain abilities. The comparative approach is admittedly useless for cognitive science since its main drawback is that causality cannot be determined which we desperately need for any kind of reverse-engineering... On the other hand, the fitness approach may have some usefulness in identifying the ultimate causes of cognitive evolution. As I suggested in my previous post, hypothesizing about such causes might be useful for empirical testability of mechanistic hypotheses informed by these ultimate causes. It could also potentially shift the focus from explanatory levels that are not useful for reverse-engineering or just wrong to more fundamental processes that when explained will allow for reverse-engineering of relevant capacities.
ReplyDeleteI wholeheartedly agree with you that this article (and in my opinion, the last one also) is not useful instructions on how to reverse-engineer a T3 passing robot. Also, the authors did admit that while the comparative approach only takes us so far, the fitness approach also fruitlessly assumes that what we can contemporarily observe is representative of certain pressures that select for certain traits.
DeleteI definitely like your suggestion of shifting our explanatory levels. Maybe certain studies (cognitive ecology, for example), is not meant to be used as a tool for reverse engineering? It's like saying knowing the history of humanity will allow us to accurately predict the future...
A little tangent: I also like Claire's comment above because it proves that genes and traits are not everything that determines our behaviours. It is also our immediate environments that trigger and shape how we behave, given the genes that we possess. So maybe knowing how we evolved is just trivial, since our current behaviours are subject to change depending on our environment?
@Solim, I certainly feel the same way as you and @Aylish when it comes to the correlative nature of evolutionary psychology. Like you say, without causation we cannot reverse-engineer. I think like a few people who have already commented, I am still grappling with how we relate this to our previous topics, considering these facts. Even if we identify “the ultimate causes of cognitive evolution” and create empirical tests of mechanistic hypotheses, how can this be more useful? Is it possible for you to Kid-sib your idea?
DeleteDoes the fitness approach not also have the drawback of being correlative in nature?
@Wendy @Matt Human history doesn't allow us to predict the future that's for sure. Nonetheless, I doubt that we can assert that "knowing how we evolved is trivial". Our current behaviours are subject to our environment but the possible set of responses to our environment is constrained by past evolutionary pressures that selected for the cognitive capacities we currently have. Ultimate causes as described in the articles are the selective pressures that gave rise to cognitive adaptations. I think these might be important because knowing what the ultimate causes are may allow cognitive scientists to propose hypotheses about the cognitive mechanisms that allowed us to solve the adaptation problems. Cognitive science seems to be riddled with vague terminology like memory, perception, etc. While identifying these general cognitive phenomena is relevant, reverse-engineering necessitates more precise definitions and explanations.
DeleteTake for instance categorical perception: it was thought specific to speech but is now known to be present in birds (who clearly did not evolve for speech) when they hear speech sounds (as we saw last lecture). Birds and humans clearly don't solve the same problems in their respective environments. So why do we both have categorical perception which seems to be important for language? What problem does it solve for them? Did this capacity evolve at some common point in our phylogenetic trees (and how far down that tree)? If so, can we hypothesize what the selective pressures were at that timepoint for it to evolve? If not, why is it common to two radically different species? I think that by looking at answers to these questions, using evolutionary psychology, we may come closer to finding answers to what problems the evolved mechanisms for categorical perception solved. The idea is that knowing what certain cognitive capacities aim to solve can give a clearer idea of how to test for potential mechanisms that solve these problems. From what I understand the fitness approach aims to uncover evolutionary change and speciation. The human species developed a set of abilities to solve problems posed by the EEA. Identifying the EEAs may help determine the critical cognitive abilities we have (not the by-products - although I think we may be interested in some of those). Explaining those abilities is easier if you can hypoothesize and test them which is what the evolutionary psychology framework allows. In my opinion, fhe first paper outlines this process more clearly (and its usefulness) and the second one shows us what kind of evolutionary science might not be so useful for reverse-engineering cognition.
Solim, evolutionary explanations are useful in cogsci if they explain why organisms have the cognitive capacities they have — but it requires something much more specific than that they (somehow) aided (ancestral or current) survival and reproduction. Your synthesis at the end (comment 2) is mostly right but the first paper is simplistic in trying to explain higher cognitive capacities in much the same way as spider/sex capacities. There is some sort of fuzzy boundary past which that sort of reckoning becomes absurd.
DeleteWendy, yes, evolutionary theory and evidence do not provide the causal mechanisms, just hypotheses about why they evolved. And of course using our (evolved) learning capacity and experience is a huge part of what we are and become.
Matt, rarely do evolutionary hypotheses help explain how the brain (or the genes) causally generate those capacities (i.e., they rarely help reverse-engineer them). The challenge is to give a valid evolutionary explanation of how and why the capacities evolved (and not just a “just-so” story); and that in turn becomes the “why” part of the solution to the easy problem.
T-testing tests the “how” in the explanation of our cognitive capacity, turning correlations into causal explanation. For testing evolutionary hypotheses there is also evolutionary modelling in virtual life simulations (Strong C/T Thesis), because history does not repeat itself, so is not open to direct testing.
In this article by Cauchoix and Chaine, we learned about evolutionary ecology and the comparative and fitness methods used to study the evolution of animal minds.
ReplyDeleteThe comparative approach: a researcher answers questions about the evolution of adaptive cognitive traits through comparison of existent species and fossils (evolutionary history). For example, researchers determined that sensorimotor intelligence and tool use skills in cerbus monkeys are adaptations for feeding.
Disadvantages:
- As mentioned in Aylish’s post, evolutionary ecology is the study of genetic and phylogenic correlations. A correlation does not mean causation.
- Researchers have to make many assumptions because they cannot be sure of past environmental pressures.
Fitness approach: a researcher answers questions about the evolution of adaptive cognitive traits through the measurement of contemporary selection. For example, researchers determined that problem-solving performance in great tits is an adaptation for multiple life-history traits determining reproduction.
Advantage (and disadvantage):
- Contemporary selection considers many factors (proximate and ultimate causes) to explain the evolution of adaptive cognitive traits. (Underdetermination: There is more than one variable or explanation.)
This week’s readings got me thinking: Just like humans, a T3 robot might need to be able to evolve too!
P.S. Did you know there’s a subfield of robotics called evolutionary robotics? You can read more about this here: https://www.frontiersin.org/articles/10.3389/frobt.2015.00004/full (I found section 6.1.2 especially interesting.)
The evolutionary robotics paper is a good supplement. (It works both ways, as in rocket science: you learn to build robots from computer simulations of robots and their environments; and from real robots and their interactions you learn more about what properties need to be simulated in the virtual robotics.) Unlike learning, though, I don't think evolving is part of T3. (But some learning algorithms are similar to evolutionary algorithms.)
Delete“First and foremost, analyses should begin to compare specific regions of the brain, neural structure, or brain function.”
ReplyDeleteJerry Fodor disagrees. He thought that studying the brain with our current rudimentary imaging techniques wouldn’t reveal anything we don’t already know. When trying to solve the easy problem, neuroimaging won’t help. I wonder what his response to this paper would be. Cauchoix, M., & Chaine, A. S. think that this technique could show us how the mind evolved and help us understand cognition. I wonder if knowing whether “specific cognitive abilities usually coevolve with specific neural structures” would actually help us answer the easy problem. It certainty doesn’t explain consciousness. While I do think evolution can help us study cognition (such as looking at causal effects of fitness on cognition), I think the authors were wrong to emphasize neuroimaging, after reading Fodors paper.
At least human noninvasive neural imaging is harmless; a lot of comparative neurobiology across species is not. But you're right that that where/when data, whether within individuals or between them, though it might be useful clinically, is unlikely to help reverse-engineer the underlying causal mechanisms.
DeleteHi Katherine! I absolutely agree with you point and wanted to expand upon it. As noted by Professor Harnad and many other students in this skywriting and the previous one, evolutionary psychology is a discipline that is best suited to answer the why part of the easy problem of cognitive science. The authors of the first paper for this week, in my opinion, fully embraced this role and only really described the main ways in which a researcher can create and test hypotheses related to discovering the functional purpose of cognition. (As noted in class, however, this is a very broad field of study, and it’s questionable whether it can beyond ‘sex and spiders.’) This reading tried to extend this theory to the how aspect of the easy problem of cognitive science, to discover a mechanism. Some of the ways that they do this as by taking very basic measures of the brain, e.g. size and volume, and comparing it with behaviour. As Solim mentioned, this is a correlation, not a causation, which the how part of the easy question is asking for. Though they do acknowledge this methodology is limited, their only solution is to present more narrow correlations instead such as neuroimaging, as you mentioned, or correlating genetic variance with behaviour. In essence, this is what Fodor was protesting against, but I’d argue even less precise than in human trials.
DeleteIt's true that evolutionary psychology has an interesting take on the "why" components of the easy and hard problems however, as others have stated, I'm not sure if it has actually gotten us anywhere in terms of specifying a mechanism or a ‘how’ or if it's mainly intriguing on the basis of its potential. On the fitness approach, the article states:
ReplyDelete“the fitness approach provides opportunities to integrate our proximate understanding of cognition with new findings on the ultimate causes of cognitive evolution”
This is an interesting approach since it examines the fitness of a trait in a contemporary context. For example, the article talks about brood parasites and how the hosts are under pressure to identify the parasites while avoiding errors whereas the parasites are under pressure to integrate and be mistaken for host eggs. The article goes on to say that this parasite-host relationship is a driving force behind the “recognition” systems of the host. This case study seems to satisfy the ‘why’ of the easy problem but doesn’t explain how it happens or what the ‘mechanism’ of recognition (I do have to say that this article uses quite a few ‘loaded’ or ‘weasel’ words like categorization, perception, recognition without really explaining. If this course has emphasized anything to me so far, it’s that you have to break a concept down to its component parts or you run the risk of simply deferring your explanatory debt) . I feel no more capable of building a South American coot with intact recognition abilities than I did before I read the article.
It is possible, as someone said above, that having a solid answer to the question of ‘why’ could be a good starting point for getting at the how but it remains to be seen if evolutionary psychology can make those connections.
Maybe I should give evolutionary psychology a little more credit since I understand that these types of approaches aren't necessarily for figuring out the question of how we do what we do (reverse engineering). Perhaps I've have paired the 'how' and 'why' of the easy problem two closely. A methodology that only attempts to answer one of them can still be very valuable.
DeleteIf we think back to the course so far, no other approach we've seen has given us a concrete answer to the question of 'why can we do what we do' except evolutionary psychology. Turing gave us a way to know if we've sufficiently answered the 'how' question (Turing Test), Fodor showed us that cognition cannot be all computation because it cannot lead to understanding and we've seen that categorization is a huge part of our general cognitive capacities, but no one thus far has really looked at the question of why except for evolutionary psychology, which says that we have certain cognitive capacities in order to address specific evolutionary problems.
Good job: I can't add anything!
DeleteMy immediate instinct, like many others I think, is to criticize evolutionary psychology for not addressing the “how we do what we do” part of the easy problem although why is still a part of it. In this course, we’re specifically looking to reverse engineer cognition, but that doesn’t mean that every other field exploring cognition is looking to do the same. Evolutionary psychology is looking for what adaptive pressures gave rise to the adaptations we have, in this case, cognition. So, in that sense it does contribute to the easy problem. In my previous skywriting, I said that knowing we can cognize because we evolved to do so does not get us any closer to reverse engineering cognition. I still stand by this, but there are other aspects of understanding cognition that don’t involve reverse engineering, that’s just our (and Turing)’s goal. I will say that I find it unsettling to think that at some point, pre-adaptive pressure, humans were cognition-less. For some reason, I even find it hard to even imagine a human that does not cognize. In true Halloween fashion I imagined zombie cavemen, but I don’t feel like that’s right. Anyway, I digress. Evolutionary psychology’s contributions aren’t useless in solving the easy problem, they’re just not particularly useful in reverse engineering cognition (solving easy problem a la Turing).
ReplyDeleteThis article, much like the other for this week, agitates my frustration at the division we have created in the academic/scientific realms. I understand the desire to categorize is strong and is a major component to our understanding of the world (to harken back to last week) but when done poorly or to too great an extent can be harmful. This can even be seen in the discord between different fields using different measurements or terms for the same exact thing. This lack of cooperation is just immensely frustrating and I would hazard a guess that it has likely prevented further progress in these realms of study that have such obvious common ground - and even more so in fields where there may be much less obvious connections.
ReplyDeleteIs the difference between cognitive ecology and evolutionary psychology in the previous reading that cognitive ecology seeks to explain both the easy problem (how) and origins of cognition (why), while evopsych focuses on the question of why we cognize the way we do?
ReplyDeleteAs others have mentioned before me, I am taken by the approach that evolutionary psychology has because it is the first time in the course that we seem to begin answering the easy problem in terms of "why". Answering the question of why we cognize may be a part of answering how we cognize. It is true that the studies mentioned in this paper have recourse to correlation in a similar (albeit more unethical and cruel) way as neuroimaging in humans does - and we have seen from our Fodor reading why that is useless in our attempts to explain cognition.
Cognitive ecology as explained in this paper seems to be asking the correct questions for reverse-engineering, but the methodology seems to still be focusing on correlation as opposed to causation. That being said, I do feel like research funds would be better spent on answering the "why" of cognition through evopsych research rather than neuroimaging in its present state that solely looks at correlation without explaining why behavior occurs/how it came to be. As Cauchoix and Chaine mentioned, "the fitness approach provides opportunities to integrate our proximate understanding of cognition with new findings on the ultimate causes of cognitive evolution." Maybe the fitness approach is then where we should be directing our efforts, at least until we find a way to investigate the "how" of cognition.
- As many of the comments before me have mentioned, evolutionary psychology might be one (if not currently the only) viable method to understand the “why” of the easy problem. But, although evopsyc might be really helpful for understanding many different cognitive functions, I wonder if it can be just as helpful when understanding the “why” for language? Of the two methods described, we can’t use the phylogenetic comparative methods, since there are no other extant species that have developed language yet. That leaves us then with the fitness approach.
ReplyDeleteThe authors write that one of the advantages to the fitness approach is that it “ideally includes identification of the agent of selection or the specific social or ecological challenges that favour” the trait (i.e. language). But in what ways is the fitness approach different from a “just-so” explanation?
I think we have concluded that although evolutionary psychology has not provided any answer to the how of cognition, it has attempted to provide an answer to the why. This is not necessarily a failing of evolutionary psych, it just means that this method approaches the mystery of cognition from a different angle. However, I think this reading still leaves something left to be desired when it comes to the “why” question of language evolution. If we are to draw a link between animal cognition and human cognition, then why haven’t animals also developed a predicate based language? And are we so sure that they haven’t? Bees are able to communicate the location of pollen sources with intricate dances - is this not analogous to a sentence with a truth value? Or vervet monkeys who have separate alarm calls for “leopard” and “eagle” - can we not consider this at least a primitive form of a language that has predicates with truth values? I am still a bit confused as to why the human language is considered to be so far removed from animal communication.
ReplyDeleteI think the main difference between human language and animal language is that we use propositions with truth values whereas animal communication doesn't have truth valuess. Take the examples you gave. What the bee is communicating (according to itself) is a place (kinda like saying "the store") which is neither true nor false. The same applies for the monkey alarms I think. I see wheree you're coming from in that I also phrase it as a sentence when thinking about it but the animals don't (and can't learn to, I think) which is why human language is considered so different. What this makes me wonder is to what extent the evolutionary mechanisms of our cognitive processes is different from those of other animals, since we have this nuclear superpower of language and I feel like it must have impacted everything that evolved after it at least.
DeleteCauchoix, M., & Chaine, A. S. (2016). How can we study the evolution of animal minds?
ReplyDeleteIn the introduction, the authors described two different complimentary approaches to the study of animal behavior: the proximate and ultimate. The former is the cognitive science approach which
studies behaviors as specific motor responses to cognitive tests. The latter is the evolutionary ecology approach which studies “behavior” as complex responses to social and ecological situations. Behavior in this latter sense is said to be an amalgamation of many cognitive behaviors. It is interesting that animal behavior in cognitive science is defined as specific motor response because it highlights one of the objects of study in cognitive science: the robotic capacities (T3) or what we can do and more specifically, sensory motor responses in the case of animals. The paper highlights the importance of studying evolutionary ecology using the proximate approach of cognitive testing to reveal a better understanding of how the mind evolves as very little is known about the ecological pressures that shape it. The paper emphasises the importance of studying our learning cognitives capacities. The fact that evolution is “lazy” makes it plausible that learning cognitive processes would have evolved instead of full blown complex ready to go innate cognitive skills.
Relating the previous lecture and your response about distal and proximal causes on the blog:
Can we say that distal causes in evolution are obtained from the top-down approach?
And that proximal causes are based on the bottom-up approach?
This article was really interesting, covering so many pros and cons of the comparative vs fitness approaches to studying evolutionary psychology, in the end recommending their combined complementary functions. I am confused by “contemporary selection.” It seems to refer to emergent patterns identified in “short term selection” which is, according to Darwinism, “the primary cause of evolutionary change and speciation” (Cauchoix and Chaine 7). I am struggling to see the use in such analyses for reverse engineering or further understanding human or animal cognition. Are we looking at evolutionary perspectives because the selection process has theoretically led to the dominance of certain cognitive traits over others? I’m finding it really hard to believe that we could through any experimental paradigm actually isolate and “identify agent(s) of selection” (13) because it seems like there are too many variables that are inextricable. Cauchoix and Chaine acknowledge the multiplicity of possible paths for selection to have taken, and of traits, proposing that “cognitive ecology” could be a promising field for linking studies of both social- and natural- selection, but they don’t suggest any improvement to the collective assumption in evolutionary psychology that “contemporary selection…is representative of selection pressures that have led to current expression of cognition in a species” (7).
ReplyDeleteIn lecture, we discussed the difference between proximate and distal causes of behaviour. Distal causes are defined as the O.G. (original) causes of a behaviour or trait (e.g. kids liking sugar because it provides lots of energy and was scarce), while proximal causes are 'why' a behaviour happens on the level of experience/biological mechanism (e.g. sugar tastes good, sex feels good etc.). Evolutionary pressures can also be divided into those that existed in our original environment (the grasslands somewhere, presumably) and those that exist in current environments (urban centers, college campuses, the COVID 19 pandemic).
ReplyDeleteAs I posited in lecture, I think for contemporary (and perhaps all) human behaviour to be captured by evolutionary models, we need to consider a fifth sort/source of evolutionary pressure/fitness: the level of culture. Cultural pressures and norms surely shape how successful individuals are at reproducing - even if we conclude that the sexual/romantic preference for older men and younger women in heterosexual couplings is 'universal', having been replicated across many cultures, it is not only the factor of age that confers who gets partnered with and who doesn't. Many other factors go into what makes someone a suitable mate, and many of these factors (or features, if we take a categorizing view) depend on cultural context. Consider beauty standards, star signs, eye colour, and fashion choices.
Each of these categories are filled with arbitrary symbols: their arbitrary-ness can be shown by their extreme variation; what makes you cool in one cultural context or moment makes you strange, lame, dangerous, or simply dull in another. Some of these categories may be ground-able in distal/ultimate causes: women tend to want men who can demonstrate themselves to be successful/loyal because their investment in reproduction is intrinsically higher; yet what symbolizes 'success' looks radically different in different cultures; its symbols are formal (their shape can be anything so long as it attaches to the right meaning), not innate.
The question then becomes, how do these formal/arbitrary shapes get imbued with meaning; how do piercings become hot, or popped collars 'douchey'? Comparing this process to other forms of meaning-making we explored in Blondin Masse et al's paper on language (the mushroom sim), two forms of transmission come to mind: 1) observing members of one's group doing the right thing with the right thing (ignoring people with x trait, befriending or sleeping with others with y traits) or 2) receiving these categorizations through language. These processes can explain not only positive selections (making friends and finding partners who share our style, habits, interest, etc.), it also explains the categorizations underlying our prejudices, xenophobia, and racism. Categorization is thus constitutive of culture and context-specific cognitions; not just the 'universal' elements and processes humans are capable of.
I originally thought this paper was off to a good start, explaining various waves of darwinism and layering the ground for threefold path to tracing the evolution of cognition and thus hopefully of categorisation. Early on it explains that brain and cognitive functions are subject to natural selection if and only if (A) there is variability in cognition between individuals (B) that this variability is inheritable (C) that the variation in question is related to variance in fitness under environmental conditions. We will see that although attempting to answer these three conditions could provide us with insight in solving the easy problem, the way it is tackled in this paper is too limited to hint at any clues to reverse engineer cognition.
ReplyDeleteUnfortunately, cognitive variability in cognition is never quite properly defined and would more adequately be renamed cognitive capacity for the sake of this paper. Cognitive variability is here often limited to performance tests such as IQ tests. Human intelligence tests do not explain the capacities being tested, they only measure the capacity to cognize. Furthermore, when not limited by tests, cognitive variability is often merely assimilated to brain size, but explaining an increase in brain volume does not account for what mechanisms of the brain are involved or even selected for cognition. And if even if more specific areas of the brain were mentioned (which occurs occasionally), we recall Fodor who reminds us that where something occurs in the brain barely get us any closer to how it occurs in the brain…
I’m being very cynical with my above criticism, but the take-home for me is that this paper was on the right track in laying out the questions that evolutionary psych had to answer in order to play its part in reverse engineering cognition, that is:
1. Prove varibaility between individual
2. Prove inheribility of this variability
3. Prove that this variance is related to the variance in fitness
But it also missed its chance in begging the question for each of those problems as related to how we’ve defined cognition.
However, if were we to apply this heuristic beyond reproduction and survival, say with categorisation and language, then evolutionary psychology may give us some important hints on how to reverse-engineer cognition.
In the paper, "How Can We Study the Evolution of Animal Minds?" the authors distinguish between the "proximate" and the "ultimate" approaches to studying animal behaviour: the proximate approach is, as Maximilien points out in this thread, the cognitive science-based approach, where we try to understand the underlying causal mechanisms which allow animals to do what they do; the ultimate approach, on the other hand, focusses on the evolutionary history of traits and how traits that were selected for could drive behaviour.
ReplyDeleteIn my last response (in 7a), I argued that evolutionary psychology helps us understand the "why" of the easy problem but not the "how." Reading this paper, however, made me think that maybe taking an "ultimate" (evolutionary psychology-based) approach in conjunction with a "proximate" (cognitive science-based) approach could help shed light on the "how" as well. In other words, perhaps trying to gain insight into the distal cause of a behaviour could help us understand the mechanisms driving the proximal cause of the behaviour. For example, if an ultimate approach tells us that a given trait is adaptive, then it might be possible to identify the gene associated with that trait. If we can start to pair genes with traits, we might be able to build a partial causal explanation of the behaviour.
One thing that Prof Harnad has pointed out is how difficult it is for evolutionary psychology to answer the how/why questions of the easy problem when it comes to language. Could this be because even though language is clearly helpful to us in so many ways, it is so broad, complex, and has so many different functions that it would be hard to nail down a single environment where language specifically was adaptive? Wouldn't language be adaptive in all environments?
I really enjoyed reading this paper, and found it to be rather informative, though very limiting at the same time.
ReplyDeleteTo give credit, where credit is due, I think that Evo-Psych should be recognized for giving a strong foundation for how to interpret why things are they way they are (e.g. different hippocampal structures resulting in different memory capacities). This is extremely important, as it helps narrow down our focus as to what is the function of these regions, and we can then ponder upon the questions as to how this function could have been selected in its environment through selection pressures. It also helps that we can see these variations of structure and function manifest themselves through different forms of behaviour (elicited by the animal) in its respective environment. This serves as form of confirmation as to why these structures and functions have been selected by nature and have allowed said animal to propagate its genes.
On the contrary, Evo-Psych's Achilles heel (at least for cog. sci purposes) is that it begs the question as to how did it (natural selection) engender these functions or traits? This limitation is one that places a significant detour in our quest to reverse-engineer cognition. If my understanding is correct, it seems that Evo-Psych's problem is very similar to the UG problem; where we don't know how UG really functions, or where it comes from, all that we know is that it is generally right and works as an explanation to linguistic issues (i.e. the complexities around syntax). In Evo-Psych we have a good explanation as to why things may have evolved and they can provide us a clear picture as to why certain species have different structures in their hippocampus, yet it still lacks depth, specificity and like UG, begs the question as to how things are they way they are. There is no clear causal explanation as to how a certain environmental factor (or factors) could force an animal to undergo specific morphological changes that help it survive in its environment better than it could without these specific changes, and then pass on its genes.
As a result, Evo-Psych's explanation of "why" functions may give us intimations to "how" functions, but exclusively studying Evo-Psych will not yield satisfactory results in that department. Its lack of specificity hinders our ability to really hone in how selection pressures resulted in different hippocampal functions in animals, but also doesn't even come close as to answering the more difficult questions in humans concerning: how did language evolve? Where Does UG come from? It can offer very basic explanations such as "black-capped chickadees (Poecile atricapilla) living in harsh winter climates (i.e., Alaska) cache more food, have higher spatial memory capabilities, and have a larger hippocampus [...] than individuals of the same species in populations from milder climates (i.e., Colorado)." But this is only a surface level explanation, and does not provide us with any specific mechanisms that explain how black-capped chickadees have a bigger hippocampus -- and unfourtunately -- these specifics are necessary to help us understand how to reverse-engineer cognition.