7a. Lewis, D. M., Al-Shawaf, L., Conroy-Beam, D., Asao, K., & Buss, D. M. (2017). Evolutionary psychology: A how-to guide. American Psychologist, 72(4), 353-373.
Researchers in the social and behavioral sciences are increasingly using evolutionary insights to test novel hypotheses about human psychology. Because evolutionary perspectives are relatively new to psychology and most researchers do not receive formal training in this endeavor, there remains ambiguity about “best practices” for implementing evolutionary principles. This article provides researchers with a practical guide for using evolutionary perspectives in their research programs and for avoiding common pitfalls in doing so. We outline essential elements of an evolutionarily informed research program at 3 central phases: (a) generating testable hypotheses, (b) testing empirical predictions, and (c) interpreting results. We elaborate key conceptual tools, including task analysis, psychological mecha- nisms, design features, universality, and cost-benefit analysis. Researchers can use these tools to generate hypotheses about universal psychological mechanisms, social and cultural inputs that amplify or attenuate the activation of these mechanisms, and cross-culturally variable behavior that these mechanisms can produce. We hope that this guide inspires theoretically and methodologically rigorous research that more cogently integrates knowledge from the psychological and life sciences.
Researchers in the social and behavioral sciences are increasingly using evolutionary insights to test novel hypotheses about human psychology. Because evolutionary perspectives are relatively new to psychology and most researchers do not receive formal training in this endeavor, there remains ambiguity about “best practices” for implementing evolutionary principles. This article provides researchers with a practical guide for using evolutionary perspectives in their research programs and for avoiding common pitfalls in doing so. We outline essential elements of an evolutionarily informed research program at 3 central phases: (a) generating testable hypotheses, (b) testing empirical predictions, and (c) interpreting results. We elaborate key conceptual tools, including task analysis, psychological mecha- nisms, design features, universality, and cost-benefit analysis. Researchers can use these tools to generate hypotheses about universal psychological mechanisms, social and cultural inputs that amplify or attenuate the activation of these mechanisms, and cross-culturally variable behavior that these mechanisms can produce. We hope that this guide inspires theoretically and methodologically rigorous research that more cogently integrates knowledge from the psychological and life sciences.
While evolutionary psychology might be great at pinpointing which pressures lead. to which adaptations, I would ask the same question that Fodor asked neuroscientists: what are you hoping to achieve? The same way that neuroimaging took snapshots of the brain, it seems like evolutionary psychology is trying to put together a picture, except this time it's more like putting together the pieces of a puzzle.
ReplyDeleteAdmittedly, evolutionary psychology has one thing neuroscience didn't as much: a mechanism, namely those of genes. However, genes are as useful as action potentials, since we don't quite know what about pressures drives changes in those, nor how genes result in how we cognize - except the fact that it helps form our neurons, hormones, and all other structures. When it comes to evolutionary psychology, we are still far from solving the hard problem, and I'm not convinced we've come much closer to solving the easy one.
Neuroscience studies genes quite frequently and they can tell us a lot in terms of the easy problem. An example of this would be when neuroscientists experiment with Gene deletion in mice/rats. Deleting one gene would seem rather inconsequential given the thousands of other genes that make up their genome, but this is not the case.
DeleteOne gene deletion experiment that is notable is when they knocked-out (deleted) the gene that encodes the enzyme that makes dopamine (Tyrosine Hydroxylase -- without this enzyme, neurons cannot synthesize dopamine, and the mouse has no dopamine in its body). The result of this was that the mice could not move, eat or drink water and would starve to death because of their inability to initiate any form of purposeful movement (I.e. walking to a bowl of food and eating it).
This all happened because one gene was deleted. If the animal-data could be translated to human-data and there is not much discrepancy between the two, then this would be telling a large part of the story in terms of the easy problem (in fact, we can study low levels of dopamine in the human population with Parkinson's patients and yield some similar results to the animal-data). The data here is also causal and not correlational, therefore the results paint a clear picture as to how this gene plays a role in behaviour.
So, I think it is safe to say that studying genes can help us understand many aspects of the easy problem, the complication arises when we have to explain how the genes function and interact with each other as a whole. Studying the combinatorial function of 20,000 genes is no easy task and the data that comes out of it is very complicated and very unclear at times.
Despite this, I don't think we can say that it cannot help us understand the easy problem, because genes make up who we are as a species -- thus solving the easy problem involves understanding the underlying role of genetics in influencing our behaviour.
Here is the study I referred to: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3932907/
Deletecoming at neuroscience, @julian fight me
Delete(also, I don't believe anything we study in this class will be looking to solve the hard problem. Everything is going to be talking about the easy problem because nobody has any idea how to even begin to solve the hard problem)
Julian, I’m afraid I’m going to have to disagree you here. I have to side more with the opinions of Claire and Ishika, and others – that evolutionary psychology does in some capacity provides us some of the tools to solve the easy problem (although the hard problem is off the table). Recall that the easy problem of consciousness is both *how* and *why* we do the way we do. If we take a look at what Fodor was saying, he claims that localisation of certain cognitive functions in the brain does not provide us a causal explanation for the soft problem, that it does not tell us how or why we think the way we do; rather, we are only provided with correlations. In contrast, evolutionary psychology attempts to get at the origin of psychological behaviours, i.e. to describe *why* evolution can scientifically (causally) explain the existence of cognitive processes. (Other statements such as Robert’s and Ishika’s point out to how it can answer the how question, too). While I admit that the article doesn’t provide us with much evidence of how cognition can broadly stem from evolutionary pressures, I think that the study of evolutionary psychology can actually provide us with some pretty broad theories. For example, the frugivory vision theory suggests that colour vision stems from the increased fitness from being able to distinguish fruit, tinged more with red and more nutritious, from their leaves, more often green.
DeleteJuilan, in other areas of biology, genetic explanations are certainly explanatory, but in evolutionary psychology there is little genetic explanation, and even less reverse-engineering. It is mostly about fear of spiders and about sex. These are important matters, but they certainly don’t generalize to most of human cognition. The most challenging cogsci question in evopsych is the origin of language — but evopsych only offers spider/sex-level hypotheses.
DeleteRobert, do you think knocking out dopamine may have some clinical relevance but does it contribute to reverse-engineering cognition? (Evidence about what is learnable vs. what is inborn is certainly relevant.)
William, yes, the “why” question in the easy problem is about the adaptive advantages and the evolutionary origins of cognitive capacities, but it is not clear whether spider/sex evopsych provides answers. (It certainly says nothing about the hard problem, Lyla.)
@Prof I think the contribution that knocking out Dopamine has towards reverse-engineering cognition is pretty low on the grand scheme of things. However, I do think that dopamine in particular does raise some interesting concerns when looking at how humans interpret rewards and how do we predict things based off of dopaminergic signaling (Dopamine prediction error).
DeleteReverse-engineering cognition is centered around figuring out how we do what we do, therefore we must be able to understand how rewards come into play for our behaviour and how it affects our behaviour. If a T3 robot cannot interpret and react to reward in the same manner that we do, then we can't say that we have successfully reverse-engineered cognition, as if a T3 robot has improper reward functioning, then its behaviour as a whole will not be representative.
I am not saying that we would need to have a T3 or T4 robot to have dopaminergic functioning in order to be able to do what we do, but that it would have to be able to replicate what dopamine does in order to satisfy this behavioural component that we innately carry.
I took a course with one of the authors of the assigned paper, Daniel Conroy-Beam, while on exchange. The course was about relationships, so there was indeed talk of sex. I took a course on conspiracy theories at the same time, which talked a lot about the evolution of pattern detection, so that covered the spiders.
DeleteHowever, I still feel the need to push back a bit on your characterization of evolutionary psych. Professor Conroy-Beam’s course began with sex, but it gradually discussed more and more complex human interaction, such as friendship. Obviously, the mechanism in question is always natural selection, so reproduction (sex) and survival (spiders) will be prominent, but I think that evolutionary psychology can go beyond this. However, it is likely true that there is a limit to the behaviour that evopsych will be able to explain, given that so much of what we do as humans seems only very distantly related to our reproductive success.
Despite this, I can’t help but bring up a question that often came up in response to the Fodor reading, which is: where else should we look? As a defense of neuroscience, which is trying to answer the “how” question, it is possible to counter that we can also look to computer science, which can replicate elements of cognition. However, as you said above, the “why” question is specifically about evolutionary origins. So from what I can tell, there is no other option but to turn to evolutionary psychology, however unsatisfying the answers provided may be.
I understand that the central criticism here is that evolutionary psychology is not particularly useful in terms of reverse-engineering. However, I must say that I have this nagging frustration with the fact that findings that directly address our question, “how” and “why”, seem to be dismissed on the basis that they do not use/support this particular methodology. It seems to me that any answers to cognitive science’s question are valuable in and of themselves.
The easy problem is how and why can we do what we can do. I’m going to try to tie this reading to that. Evolutionary psychology argues that traits are favorable when they make the trait haver more likely to survive and reproduce and then pass the favorable trait on genetically. The paper argues that traits evolve to solve an adaptive problem. Applying this logic to cognition, evolutionary psychologists would say our ability to cognize is solving an adaptive problem. Sticking with the evolutionary perspective, I’d say each of our cognitive functions is a solution to their own adaptive problem. For example, cavegirl Lyla kept forgetting where her cave was and ended up dying by getting eaten by a lion. The people who didn’t get eaten were the ones who managed to remember where their cave is and they got to pass on this trait. The authors would call this bottom up approach, I observed the phenomenon we have now (in this case memory) and tried to come up with an adaptive problem it would have solved. Of course, I’m sure an evolutionary psychologist would do a much better job than I did, but they’d probably explain how each cognitive function we currently have is solving an adaptive problem. Now does this solve the easy problem? I personally don’t think so (although I’ll be looking for skywritings that give me a reason to think otherwise). Telling us we evolved to think makes sense; we know that our brains are more complex, we developed frontal lobes that are heavily folded to fit all the complexities. But evolution as an explanation of how we can think tells us just as much as the shape of our weirdly folded brains does: not much.
ReplyDeleteIt is not evopsych that teaches us that adaptive traits are the ones that get passed on to the next generation, it’s Darwin!
DeleteEvolutionary explanations (“why”) are more difficult to discover and test than mechanistic ones (“how”), and certainly our cognitive capacities must have had selective advantages. Evolutionary hypotheses risk turning out to be just “Just So” stories (Smith 2016), ad-hoc speculations. But let’s say there are ways to surmount that (as in evolutionary computational modelling).
The question to ask yourself is whether the paper by Buss’s group has provided a substantive hypothesis for the origins and adaptive advantage of cognitive capacities at all (other than spider phobias and sexual preferences and practices).
Hint: Evolution is “lazy”. It is more reliable, economical and flexible to offload a learnable behavioral pattern onto a (genetic) capacity for learning it, rather than precoding it genetically. This is the real challenge of evo(cog)psych.
Smith, R. J. (2016). Explanations for adaptations, just‐so stories, and limitations on evidence in evolutionary biology. Evolutionary Anthropology: Issues, News, and Reviews, 25(6), 276-287.
Abstract Explanations of the historical origin of specific individual traits are a key part of the research program in paleontology and evolutionary biology. Why did bipedalism evolve in the human lineage? Why did some dinosaurs and related species have head crests? Why did viviparity evolve in some reptiles? Why did the common ancestor of primates evolve stereoscopic vision, grasping hands and feet, nails instead of claws, and large brains? These are difficult questions. To varying degrees, an explanation must grapple with (1) judgments about changes in fitness that might follow from a change in morphology – without actually observing behavior or measuring reproductive success, (2) the relationship between genes and traits, (3) limitations on doing relevant experiments, (4) the interpretation of causes that are almost certainly contingent, multifactorial, interactive, hierarchical, nonlinear, emergent, and probabilistic rather than deterministic, (5) limited information about variation and ontogeny, (6) a dataset based on the random fortunes of the historical record, including only partial hard‐tissue morphology and no soft‐tissue morphology, (7) an equally partial and problematic (for example, time‐averaged) record of the environment, (8) the compression of all data into a geological time scale that is likely to miss biologically important events or fluctuations, (9) dependence on a process that can only be inferred (“form and even behavior may leave fossil traces, but forces like natural selection do not”, 1:130) and finally, (10) the assumption of the “adaptationist programme”2 that the trait in question is in fact an adaptation rather than a consequence of genetic drift, correlated evolution, pleiotropy, exaptation, or other mechanisms.
I had the thought that it seems as if categorization may be at the root of many of the examples evolutionary psychology provided throughout the paper. When discussing how researchers must articulate the logical equipment that could solve a given adaptive problem, the authors write:
ReplyDelete“This includes sensory, perceptual and physiological systems that detect cues to the problem (inputs); computational machinery that processes these cues (algorithms); and behaviors, emotions, and cognitions mobilized by these computations in orders to solve the relevant problems (outputs)” (355)
In the paper, the inputs are sometimes features of the environment or people, as seen in the example about kin recognition and aggression recognition. Interestingly, the authors do not discuss categorization at all. In my mind, it appears that categorization could be a grand overarching mechanism that is responsible for potentially all evolutionary psychology.
There are clear relations between this paper and paper from last week. Namely the notion of Evolved Category Perception, which is the idea that sensory category detectors are ‘biased’ by evolution for things like color and speech sounds. The spectrum of perception is then ‘warped’ through within-category compression and between-category separations.
Furthermore, it would seem that the observation-driven, bottom-up approach is what we have used so far when we discuss category perception, especially when discussing color perception. We observe distinct colors in a rainbow and then postulate why is it that we see color this way. As the authors put it, we ask what psychological adaptation may be responsible for producing that phenomenon and what adaptive problem that mechanism may be designed to solve it (362). We reverse-engineer. Now we need to turn around apply the top-down approach to make testable predictions. Would this be what we are doing with neural nets?
"A key first step in solving this adaptive problem is to identify potential aggressors. To do this, the organism must detect cues to aggression. Based on this line of reasoning alone, the researcher already has generated the hypothesis that humans may have evolved psychological mechanisms designed to detect probabilistic cues to aggression."
DeleteIn this extract, the author attributes the evolved psychological mechanisms to a type of categorization. In the previous paper we read (To Cognize is to Categorize: Cognition is Categorization), the idea of innate categories is mostly refuted. Evolutionary psychology as explained in this paper supports the idea that we have (innate) evolved psychological mechanisms dedicated to this task, rather than innate categories. So categorization is presumably a context dependent feature (what we categoriza depends on what we are exposed to in our environment) that we developed at some point during evolution.
This paper also supports the idea of the necessity of a sensorimotor system in order for a machine to pass the Turing Test: the ability to "detect" these cues requires the ability to interact with the world.
Matt, “categorizing” (“doing the right thing with the right kind of thing”) covers a huge behavioral territory — but to say it is so is not yet to reverse-engineer it.
DeleteEvoPsych assumes (as we all do) that the things organisms can do from birth without having to learn what and how, are inborn, hence genetically encoded, and usually as a result of an evolutionary advantage they conferred in ancestors. To say this is just say you are a Darwinian (just as saying the brain causes the capacity to do and feel is just to say that you are a biologist).
The same goes for organisms’ (many) capacities to learn to “do the right thing with the right kind of thing” — but that’s where it gets more challenging and cognitive: how do these learning capacities work, causally, and why do organisms have the ones they do? To answer this, it is not informative to reply “because they conferred an evolutionary advantage to their ancestors”!
What was the evolutionary advantage, and how did they evolve?
Easy to answer for things like spider-phobia and sex-preferences, which are closer to the survival/reproduction selective factors in evolution; it becomes much more demanding for learning, and especially language.
Yes, inborn categories have inborn feature-detectors: how did they evolve? And, much harder, how did our capacity to learn to detect the features that distinguish members from nonmembers of categories evolve — and become, in the human species, the capacity to learn those features verbally? Spiders and sex don’t help.
Ishika, “innate categories” means innate (hence evolved) feature-detectors. Learned categories means learning to detect features (hence an evolved capacity to learn). I know how neural nets might do something like that. But I have no idea what “psychological mechanisms" would be. Did the paper propose one?
I really liked the comprehensive proposed methodology of the authors in their “how-to guide” because I thought it was a good attempt at both answering how and why we do what we do. I will just take the example of the kin selection middle-level theory (pp. 354-355).
ReplyDeleteIn this case, the cognitive capacity that they are trying to explain is “how do we decide whether or not to help another individual?”. They propose an expression rB>C as an answer which basically is a weighting of the degree of relatedness and the benefits against the costs. I thought of this a bit like an algorithm. In this case it is a computation that takes many different input cues from linguistic to environmental to social, and then can calculate some measure of relatedness with the other, can calculate the cost, and can calculate the benefits, and in this way answers the question. As we have discussed the entire “how” can’t be computation, but this seems to me a possible partial explanation. There are also some details missing as far as exactly how these input cues are filtered (I agree with @Matt that perhaps categorization could help explain this!) As far as the “why do we decide whether or not to help another individual” their explanation is that our kin share the same genes as us and this is why we are also are interested in their wellbeing.
The main problem I have with this explanation is it is extremely specific. We have so many cognitive abilities that I am skeptical evolutionary psychology could propose such an elegant explanation for each and every one of them. On page 362 they discuss “byproduct explanations” which I understand to an extent but I find this very unsatisfying because the “why” is just…”random noise”? I do not think this is informative for reverse-engineering the brain.
Degree of relatedness (kin selection, inclusive fitness) was the answer by the great theoretical biologist, WD Hamilton, to the question of why an organism would ever waste the energy or resources to help any other organism than itself, since "fitness" is a trait of an individual and its own genes. All else is competition.
DeleteThe obvious answer was that each organism's offspring are the "carriers" of an organism’s genes, and the genes are the determiners of their evolutionary fitness. So don't eat your own children.
But once that has been noticed, it's also follows that other relatives carry many of those same genes too. And it's the genes that survive across generations, not the carrier. So the notion of a “selective advantage” is about the genes' success in making it into the next generation, not the carrier's (who just has to survive long enough to reproduce).
So the simplest answer to the question of why an organism would ever do anything for anyone but itself is “inclusive fitness”: the more genes you share, the more you (or, rather, your genes) should be motivated to help (your own genes, in) other organisms.
Trouble is, we don’t have good direct indicators of genetic relatedness to others. We have to go by appearances (similarity) and circumstances (correlations) (rather like in the TT and mind-reading). But that’s ok, because it’s predictable (in mammals) that the one to whom you give birth is your own offspring and that the ones you grow up with are your close kin. Evolution, being lazy, is happy to go by those reasonably reliable and predictive circumstantial cues to give you only the capacity to learn from them who are your kin, and hence whom you should feed rather than eat.
The kinship “equation” in this week’s paper by Lewis et al. (all co-workers of David Buss) is about an innate detector of these “lazy” cues of degree of relatedness.
Ask me in class why Richard Dawkins (not quite the stature of Hamilton, but probably not pygmy stature like Buss et al!) thinks we cannot have more exact detectors of relatedness than the lazy, circumstantial ones.
And ask me also why our mammalian inclusive-fitness legacy might be the only source of hope for other species that will ever show mercy to them…
I thought Julian's comment about this week's reading resonates with me the most. The more I read (and re-read) this article, the more... angry I got? The never left so many "so what?" questions in my head. These are some of the issues I took:
ReplyDelete1. The article presents the top-down approach first, thereby stating its importance. However, the authors then admit that it's almost impossible to do that, because we are now "too cognitively advanced" to visualize and understand what survival/reproduction problems our ancestors faced. So, we have to resort to a bottom-up approach. What if the theory that we came up with from observations is not correct? Also, why does it matter that we have to know what our ancestors did that is no longer applicable to us?
2. "Environmental evolutionary adaptedness" is importantly defined as not a particular epoch or area, but a set of selective pressures responsible for shaping a certain adaptation. Doesn't this seem vague? We will never know exactly what shaped a certain adaptation because there are way too many different forces at hand. Answering one hypothesis is just a drop in the ocean to what could have actually happened.
3. I think other comments have also touched on the fact that evolutionary psychology is just an attempt at explaining how we cognize, but it does not answer any substantial questions. So... what now? How many more easy problems can we solve until we are exhausted of funding and energy, until we throw in the towel and admit that the hard problem cannot be solved?
Good points, but it's not about the hard vs easy problem. The question is whether evopsych can help with the "why" part of the easy problem: "How and why can organisms do what they can do?" (I think it can, but not through spider/sex thinking!)
DeleteThe notion of "Top-down" is pretty meaningless in this context (and most others). The right way to put it is that the kind of survival/reproduction reasoning that works for spiders and sex does not work on higher cognitive skills, starting with learning, and all the way to language. And that's mainly because the capacity for learning and language, once involved, take over many of the functions that selection by success in survival/reproduction originally performed. With learning and language we've more or less left the EEA for just about everything except spiders and sex.
But that's all still about the easy problem (of explaining doing-capacity) and not the hard problem (of explaining feeling-capacity).
I think this article proposed a potential answer to the easy problem of cognitive science.
ReplyDeleteWhy do we do what we do (cognize)? A potential answer that may be proposed by evolutionary psychology : cognition increases our chances of survival. Human language capacities come to mind, as we’ve been talking about it in class. I think that language would have increased the efficiency with which humans were able to address the adaptive problems mentioned in the article. Finding nutritious food for example would be made easier for humans with the help of language. They would be able to communicate with each other about food location as well as verbally barter and trade in order to obtain it.
Evolutionary psychology also provides a potential answer for how we developed this capacity : genes that allowed humans to communicate more effectively would have been selected for and become increasingly prevalent in the population.
I don’t see how evolutionary psychology can attempt to answer the question of “how we do what we do?”, as has already been mentioned in the above skywritings. I thought Julian’s link to Fodor was insightful, although I also agree with Robert that genes make up who we are as a species which makes them worthy of study.
@Wendy I had a similar “so what” response to many of the points in this article. It seems intuitive to me that cognitive capacities would have evolved to improve the survival of our species. What does this add to our discussion of cognitive science?
In response to your point about evolutionary psychology not being able to answer "how we do what we do?":
DeleteCognition is categorization. Evolutionary psychology seems to be attributing a lot of our behavior to categorization. In this paper, the author mentions the "distal level of analysis, which addresses the distant historical origins of how and why the psychological structures responsible for producing the behavior or mental process evolved in the first place".
Although evolutionary psychology may not answer the easy problem directly, its attempt to explain how and why the structures that seem to be responsible for categorization evolved may be the key to learning how to reverse-engineer the brain.
That makes sense — thanks for explaining!
DeleteClaire, yes, evopsych is meant to help answer the "why" of the easy problem, but does it? Don't the answers sound like empty generic textbook generalities about biological evolution rather than anything specific to cognitive capacities?
DeleteIshika, the answer to evopsych's own question of how (genetic) traits evolved would be part of cogsci's question of why we have them. But although plausible for spider/sex-related traits, do they tell us anything specific (yet) about categorization capacities except that "they must have evolved for something"?
The question of whether this paper provides an answer to the easy problem is overwhelming ‘no, it does not’. I agree with Prof. Harnad when he asks if evopsych simply gives “empty generic textbook generalities… rather than anything specific to cognitive science”. This paper, despite trying to add specificity to the subject, is still too vague. All the answers to evolutionary questions such as why there are kinship benefits and mating preferences are nowhere near the level of detail needed to reverse engineer anything in my opinion. How can we take any of this paper and apply it to a T3 robot? Perhaps I am expecting too much from the paper. In general, I applaud the effort to incorporate proper evopsych into the sciences.
DeleteIn this article by Lewis et al., we learned about evolutionary psychology and how researchers generate their hypotheses, test their predictions, and interpret their study results.
ReplyDeleteThere are two ways evolutionary psychologists formulate their hypotheses.
1. Top-down approach: a researcher recognizes a problem and suggests an adaptation that could have solved that problem.
2. Bottom-up approach: a researcher perceives a behavioural phenomenon and suggests a mechanism responsible for that phenomenon from a reverse-task analysis.
Two weaknesses of evolutionary psychology are:
- Underdetermination for evolutionary hypotheses and explanations of human behaviour. There may be more than one variable or explanation that influences a mechanism’s activation.
- If the goal of natural selection is to optimize human adaptations, why has it not eliminated genes for common, heritable mental disorders, such as BD, schizophrenia and autism?
I agree with many of the statements above that the goal of evolutionary psychology is not to solve the hard problem (how and why we feel what we feel). Rather, it can give us some insight into the easy problem (how and why we do what we do) by answering how and why we have certain cognitive adaptations (such as language, learning, etc.).
To be eliminated by natural selection, traits would have to be bad enough to kill us or make us unable to reproduce. (And biomedical science is partly compensating for that too.)
DeleteAll empirical hyotheses are underdetermined, but historical hypotheses are especially handicapped by the fact that they cannot be tested experimentally ("history cannot be repeated"), although computer modelling can get around that to a certain extent. Also history sometimes leaves evidence (and not just fossils!).
Evopsych certainly can't solve the hard problem any more than cogsci can. But it is part of the homework of both..
(1) " In 57% of runs, the frequency of the murder-immunity gene rapidly rose from 1% to over 99% in the population. That the murder-immunity trait evolved to be species-typical— present in virtually every member of the population—is particularly striking because those individuals who were susceptible to murder faced less than a 1% chance of being murdered over the course of their entire lifetime."
ReplyDeleteI found this idea to be similar to that of Chomsky's UG in the sense that there was poverty of stimulus. In the same vein as the simulated population in the example above appeared to evolve a trait because although it was low frequency, it had a high magnitude, perhaps we learn languages for the same reason? Knowing a language and being able to communicate effectively with fellow members of our community are essential to our survival (specially in childhood), and so of a high magnitude. Thus, even in the absence of stimulus, we have evolved mechanisms that are able to replicate the grammar used by those around us.
(2) I was wondering whether (and if so, how) the idea that "psychological adaptations may produce functionless outputs as incidental byproducts of their evolved information-processing design" can be related to the explanation proposed for mirror neurons in a previous reading? Mirror neurons (according to our reading) did not have a specific evolutionary purpose. They are forged by associative learning processes. However, in contrast to the extract above, mirror neurons appear to have psychological functions.
(3) Do we use the bottom-up approach to gather sufficient information to reverse-engineer the brain, and then use the top-down approach to formulate testable predictions in order to pass the Turing Test? Are those testable predictions themselves the Turing Test?
(1) Interesting analogy, but little-effects/big-consequences is a learnable contingency, whereas the "poverty of the stimulus" for Chomsky's UG is like the example I gave of "laylek": it's not "few negative example" but no negative examples. (And UG is just about grammar rather than language as a whole. If, on the other hand, it could be demonstrated that there could not be a language without UG, then that would explain why UG has to be innate if there is to be language. But that has not been demonstrated.)
Delete(2) The capacity to imitate certainly has adaptive value! But whether the capacity is wholly innate or (more likely) partly learned depends only on how lazy evolution is in this case. (The lazier it can be, the better: Why?)
It remains true that cogsci has not reverse-engineered imitation capacity (MNs) nor has evopsych given us a clue of a clue of what a "psychological function" is, or how and what it explains.
(3) What is the "bottom-up" and "top-down" approach?
The article mentions a distinction between the "top-down" and "bottom-up" approaches to evolutionary psychology. The top-down approach focuses on a problem that earlier humans would have needed to overcome for survival, and proposes an adaptation that would have evolved to address this problem. On the other hand, the bottom-up approach observes a behavioural or "cognitive" adaptation and seeks to identify the adaptive problem that contributed to its evolution. Importantly, the bottom-up approach necessitates reverse engineering the adaptive mechanism in order to generate "novel testable predictions based on the hypothesized psychological mechanism."
ReplyDeleteWhile I don't think evolutionary psychology provides us with actual answers to the hard problem or the easy problem, I do think that it is attempting to address both with the top-down and bottom-up approaches. The hard problem has to do with "why". Why do we have feelings? Why do we have the capacity to understand and to feel? Based on the reading it seems that the top-down approach is mostly concerned with the question of 'why', and in the context of the hard problem the top-down approach would answer as follows:
"Why do we have feelings?"
"Because there is a specific adaptive problem that necessitated the development of feelings and consciousness in order to ensure human survival."
Perhaps a better question, and one that the top-down approach cannot so easily answer, is: what specific adaptive problem actually caused us to evolve feelings and consciousness?
Evopsych's "top-down" is just to speculate about the ancestral environment (EEA) as in the sweet-tooth example (sugar rare, predators many, energy for escape crucial) and what would have been needed to solve it (children who have more appetite for sugar survive more).
Delete"Bottom-up" is to ask why children today over-eat sugar, even though it is abundant, they don't need it, and it is bad for health.
The two are obviously opposite faces of the same coin. Neither "approach" reverse-engineers the mechanism dictated by the gene (though in the case of the sweet-tooth it's easier than in the case of imitation capacity -- let alone learning, or language).
Neither approach -- nor evopsych itself, nor cogsci -- has anything to say about the hard problem. (Evopsych would give us the empty response "if sugar didn't taste sweet, why would we eat it." Can you see why that misses the hard problem completely?)
The article mentions on several occasion the goal of finding causal mechanisms which explain how adaptation to certain problems would explain why we think the way we do. But I have a hard time believing that these evolutionary explanations could ever be more than just correlational and hypothetical. In the same way that thinking about a teapot and imaging the brain doesn’t show the causal existence of a teapot area, showing that men’s testosterone is lower when they become fathers does not show causality. I’m not sure it could ever be proven beyond a doubt that these adaptations are causal. In which case, I would agree with @Julian that this is no closer to solving the easy problem than neuroscience is. We can postulate why our behaviours as humans would have changed over time due to selection pressures, but suggesting that answering these questions would be the same as answering why we think and are able to act on these thoughts is a reach.
ReplyDeleteI think you are underestimating the causal link between testosterone and aggression. That's well within the scope of evopsych's spider/sex mechanisms (although evopsych never actually describes the causal mechanism).
Delete"Proven beyond a doubt" (i.e., proving something to be necessarily true) is only possible in maths. Science (and ordinary practical explanation) settles for high probability.
Evopsych does a reasonable job on why we have the spider/sex behaviors (and thoughts!) we do, but of course not how, because evopsych does not do reverse-engineering: cogsci does.
And evopsych does no job at all on most of our most of our cognitive capacities, and most of what we think and believe (and do).
And that's without even mentioning the hard problem -- which weighs as heavily on evopsych as it does on cogsci.
This article piqued my curiosity and so did the comments above. I cannot say I totally agree with the general conscensus that evolutionary psychology, as described in the article, is not a useful endeavour, for several reasons. The easy problem is about reverse-engineering. So is the bottom-up/top-down combination approach in some sense. In this approach, observation of cognitive phenomena is the first step. What Lewis et al. propose it that following such observations, reverse task analysis through the top-down approach will allow psychologists to explain the mechanism of the cognitive phenomenon. The approach in this article is an attempt to generate a framework for evolutionary psychology that fulfills the requirements of empirical research. Namely, formulating hypotheses on psychological mechanisms based on central tenets of evolutionary theory can allow for testability, falsification, plausibility confirmation and crucially alternative hypothesis consideration when necessary.
ReplyDeleteDuring the last lecture, we talked about a certain species (bird?) that was shown to have categorical speech perception although clearly it did not evolve for speech. These kinds of observations allowed us to determine that categorical perception is not special to speech (or humans) and might be a more fundamental mechanism shared by many species. Evolutionary psychology methods as described in this article may allow for the formulation of testable hypotheses that could answer questions related to the how and why of cognitive phenomena. Namely, if we can determine what kinds of problems this capacity evolved to solve, we can then propose mechanistic explanations of how the adaptive mechanism allowed organisms to solve the problems posed by the adaptation EEA. A crucial component of this process is the testing part: this is where we can determine if the proposed mechanism is plausible. Once that is determined, this mechanistic explanation can be added to the armementarium of cognitive science (i.e. step by step reverse-engineering). Cognitive science may as a result become one small step closer to fully reverse-engineering cognitive capacity.
Furthermore, I think that the methods proposed by Lewis et al. may be useful to determine what kinds of phenomena are merely incidental by-products. This seems crucial to me. For instance, mathematical ability seems to some extent incidental. Why? Well because there is no clear reason to believe that hunter-gatherers needed this kind of ability to survive. So the real question becomes: what cognitive mechanisms were developed as a result of evolutionary pressures (and what evolutionary pressures led to such developments) that may have, as a by-product, given us generic mathematical ability. If such mechanisms can be identified, then cognitive science may then be able to reverse-engineer them.
Evolutionary psychology may be a useful endeavour, but I agree with a lot of the comments above in that it doesn’t provide a satisfying answer to the easy problem. Evolution is all about finding an advantage to survive and pass on genes, so how does cognition fit into this? After all, if we were all zombies without consciousness we could theoretically exhibit the same behaviours like fear of spiders, or disgust towards rotten foods - we wouldn’t necessarily need to be conscious of them for these behaviours to evolve. It may be possible for the “by-products” you talk about to be explained mechanistically by evolutionary psychology, but you would not get an explanation for the cognitive phenomena. There is no clear reason why hunter-gatherers would need mathematical ability to survive, but I also think that there is no clear reason why hunter-gatherers would need conscious decision-making to survive. It makes sense that we have evolved to fear spiders but a zombie could also exhibit a fear of spiders and gain an evolutionary advantage that would still allow the zombie pass on its genes. Of course evolutionary psychology is a useful tool to explain human behaviour, but I don’t think it will provide a useful answer to the how and why of cognitive phenomena.
Delete@Solim I am intrigued by your point about some of our abilities being incidental by-products. I am not sure I see mathematical ability as just incidental though, I think I would classify it more as indirect. I would agree that mathematical abilities may not help one for example find berries or determine whether or not those berries are poisonous, but I wouldn't exclude the possibility that mathematical abilities could be selectively advantageous. Maybe in a group situation where food and resources must be shared, having mathematical abilities to divide items equally could be helpful. That being said, what I have just proposed is a nice idea but it is really difficult to test my hypothesis because it is an indirect effect and my proposal is fairly vague. Furthermore, because mathematical ability is such a broad cognitive capacity, it is really difficult to quantify its selective advantage given I think it could be useful in also building simple structures or shelters, some forms of timekeeping maybe for growing seasons etc., all of which could make it more likely an individual is healthy. It is very unclear exactly how much each of these improve fitness though, and an additional complexity is how much mathematical ability is involved in each of these practices. I think it is difficult to fit such a complex cognitive capacity as mathematics into the framework these authors propose because there are so many factors with undefined values and weightings at play.
DeleteIn this way I see evolutionary psychology struggling to fully explain these indirect or incidental cognitive capacities. As you mentioned also in your post this framework that these authors propose I believe will generate empirical and convincing evidence of "why", but only if there is a pretty clear direct link between the cognitive ability and fitness.
@ Alessia I think you are conflating the easy problem with the hard problem. Your zombie example seems to me to be an other minds problem. We can't foreseeably solve that one. Mechanistic explanations of cognitive phenomena is all there is to reverse-engineering cognition. Explaining "conscious decision-making" is the hard problem. So if we agree that evolutionary psychology may be useful for that, then we also agree that it may provide useful answers to the how and why of cognitive phenomena.
Delete@Stephanie The incidental by-products idea is mentioned in the article as an alternative explanation that may present itself when testing a hypothesis fails. I totally agree with you that the abilities you mentioned have adaptive advantages. I don't think any of those require the ability to do calculus though (although the average person can understand it with enough work). If you took a hunter-gatherer child from before the agricultural revolution and brought him to today, he would theoretically be able to learn all the things we learn, such as manipulating functions and playing around with them in the Cartesian plane. I think that this ability is incidental. We didn't specifically evolve to solve those problems. Yet we have the capacity to. I took a random example and it might not be the most relevant one for reverse-engineering cognition but there may other capacities that are incidental by-products that may be absolutely necessary to reverse-engineer cognition. If that is the case, then we may need evolutionary psychology to determine what the underlying capacity is and what problem it evolved to solve.
Delete@Alessia The *that* that I was referring to in that last sentence is reverse-engineering, not explaining consciousness... Sorry about the misunderstanding.
DeleteSolim, "what cognitive mechanisms were developed as a result of evolutionary pressures (and what evolutionary pressures led to such developments) that may have, as a by-product, given us generic mathematical ability. If such mechanisms can be identified, then cognitive science may then be able to reverse-engineer them."
DeleteYou've been a little taken in by evopsych's weasel-word "mechanisms." They are not talking about mechanisms at all! "Mechanisms" means "causal mechanisms." That's what cogsci is trying to discover by reverse-engineering, simulation, and T-testing. But what evopsych calls "mechanisms" is capacities and behaviors with no causal mechanism proposed. In other words, just the data that the causal mechanism is then supposed to explain.
You are right that reflecting (and preferably gathering new evidence) on what we can do, and what it might have been selected for in the ancestral environment might possibly give some clues as to how the brain (or anything) does it. But it it more likely that it will only give a clue as to why.
And "byproducts" is the key to the fuzzy border between evopsych thinking and cogsci: Whatever was the environmental pressure that led to the evolution of language, not only mathematics and science but everything we can say and communicate to others came with the territory, as a byproduct. ("Stevan says" nothing like evopsych's standard spider/sex "adaptive" explanation will have any relevance to this.)
Alessia "if we were all zombies without consciousness we could theoretically exhibit the same behaviours like fear of spiders, or disgust towards rotten foods - we wouldn’t necessarily need to be conscious of them for these behaviours to evolve."
Yep. Which just shows again that neither evopsych nor cogsci has a leg up on the hard problem. But the question here is about what contribution evopsych makes to the easy problem of reverse engineering the causal mechanisms underlying our doing-capacities,
Stephanie, there are many components to "mathematical ability," including "numerosity perception" (being able to say at a glance whether there is one thing, two, three [maybe 4-7 too, but that depends on subgrouping patterns], also counting (which is different, and can be done visually in space and auditorily in time). Many other species can do it up to low numbers. We may be the only ones who can understand recursion (that you can just keep adding more and more); and a lot of maths is also based on logical ability, which may depend on the capacity to distinguish negative and positive feedback, which also goes way back in evolution. Some of it, though, is clearly peculiar to language, and understanding propositions with truth-value true or false -- and the explicit formal reasoning (including computation) that all that makes possible.
None of this can be explained with anything faintly resembling spider/sex adaptationism.
(cont'd) *Cognitive science may then be able to reverse-engineer* incidental by-products such as generic mathematical ability, say, only if we understand the more fundamental processes that evolved to solve specific adaptation problems (not mathematical ones). Evolutionary psychology may be able to provide an answer to that question. We may not really be interested in the fundamental process in and of itself, but we still need it to generate mathematical ability or other incidental by-products. I suspect that at least some of what we consider cognition may be incidental by-products, which really is just to say that we are not describing the phenomena of interest at the right explanatory level. Evolutionary psychology methods proposed by Lewis et al. might allow us to do that.
ReplyDeleteOne interesting "byproduct" question that many (starting with Stephen Jay Gould) raised is "spandrels." Please ask me about it in class. It has most often been proposed in connection with Chomsky's UG, which we will be discussing in the next two weeks. Here's a bit of skywriting about it from a quarter of a century ago! (My long-ago student at Southampton, Judy Chatwin, has since then no doubt become a mother, and eventually a grandmother, so her grandchildren will be seeing her discussion of spandrels...)
DeleteAn anatomical example of a spandrel: the pseudo-penis of the female spotted hyena, a byproduct of hormonal adaptations.
DeleteConley, A., Place, N. J., Legacki, E. L., Hammond, G. L., Cunha, G. R., Drea, C. M., ... & Glickman, S. E. (2020). Spotted hyaenas and the sexual spectrum: reproductive endocrinology and development. Journal of Endocrinology, 247(1), R27-R44.
In reading this paper, I was really interested in the small section that was devoted to cultural differences (pg. 366). The authors are writing in defense of evolutionary psychology, which is quite a bit different from where we are with trying to reverse-engineer cognitive science, but I think that there are still interesting parallels that can be drawn.
ReplyDeleteOkay, so the authors want to argue that cultural differences do not represent a roadblock to understanding universal psychological mechanisms. In fact, what I understood from this passage is that comparing culturally different inputs and outputs can actually help lead the researcher to uncover the actual universal psychological mechanism that is responsible for these differences.
This reminds me of our much earlier discussions in how to consider individual differences when looking at cognition. What if, instead of focusing on these individual differences and drawing different mechanisms for each ones, researchers instead conducted many more comparative analyses?
Cognition is categorization. And categorization is doing the right thing with the right kind of thing. If we think that reverse-engineering cognition means reverse-engineering the underlying universal mechanisms behind categorization, then maybe lots can be achieved by comparing multiple cultural perspectives right from the beginning can help highlight what kind of similarities there might be. So, for example, if we're doing a study on how categorization occurs for visual textures, I wonder if there might be different (or similar!) data if we were to compare that study being done to North American (WEIRD) university students vs. rural communities in Mongolia (I don't know, just trying to think of a polar opposite culture).
Individual differences and cultural differences are of course interesting and important (especially for identifying cognitive universals, such as Universal Grammar (UG)). If those universals help in reverse engineering cognitive capacity and passing TT -- or even if the individual and cultural differences do -- then they are a useful asset. But remember that we are trying to reverse-engineer generic cognitive capacity, not that of a particular individual or members of a particular culture.
Delete“evolutionary hypotheses make claims about both the distal origins of psychological mechanisms as well as the proximate causes of the phenomenon.” So, from my understanding evolutionary psychology is trying to determine the distal causes for a psychological phenomenon (why a mechanism evolved), as well as the proximate causes (how the mechanism operates). This why/how business sounds a lot like the easy problem of cognitive science (why and how we do what we do). The question is whether we only need to turn to evolutionary psychology to solve the easy problem. How much does this field help? Is it like neuroimaging which promised so much but which has so far let us down?
ReplyDeleteEvolutionary psychology won’t answer the easy problem, at least not entirely. Obviously, unless you don’t believe in Darwinism, it is clear that evolution shaped us into who we are today. Evolutionary psychology focuses on how/why mental traits and behaviours evolved. But we are complex beings, and evolution is lazy. As long as someone’s genes are encoding their behaviour such that the individual doesn’t die, humanity and evolution will continue. Genes don’t generally encode more complex (yet still helpful) behaviours, these are usually learnt. My parents taught me to be polite. This is a helpful behaviour of mine which wasn’t evolved. Being rude won’t kill someone, so this behaviour hasn’t been encoded in my genes. In other words, it wasn't evolved. However, politeness is still something I do, so cognitive scientists want it explained. Evolutionary psychology cannot explain behaviours like this.
Some of the concepts behind evolutionary psychology seem very reasonable, and other applications still give me great pause - either because they seem far fetched or because I see no way to separate what element is the cause and what is the effect. For example the paper discusses women's increased selectivity in the realm of sexual partners. However It does not mention the societal environment we live in and have lived in surrounding this. Not only is there the reproductive cost, but just considering the sexual relationship women tend to be far more more punished by the societies and groups they live in for having non-monogamous sexual relationships. In 1700s England for example, many wealthy married men had official mistresses, and yet married women could be ruined completely if an extramarital affair they had partaken in was revealed. And this double standard still exists in a less structured format. But how can we say with surety that our societal standards were caused by this evolutionary trend and not that our societal structure shaped how women choose sexual partners? I supposed this comes back to the classic nature v nurture debate.
ReplyDelete“Because genetic relatives carry copies of one’s genes, the preconditions necessary for the evolution of altruism can be met if this altruism is directed toward one’s genetic relatives”
ReplyDeleteI found a really interesting video which explains kin selection and the selfish gene very clearly with computer simulations. A purely altruistic individual who helps everyone regardless of whether or not they are related doesn’t help their own genes replicate. In fact, from the simulations in the video, we can see that altruistic individuals like this are selected against. “The problem with the gene for purely unconditional altruism is that it helps copies of competing genes as much as it helps copies of itself.” So, in order to be successful, the gene needs to find some way of helping its own copies more than other copies. How does it do this? Unless the gene that encodes altruism also encodes some sort of marker (like the red dot on the forehead), there is no way for an individual to be sure that someone else shares their genes. In nature, animals try help the animals that are related to them (although sometimes they are incorrect in guessing who their relatives are). Sterile worker bees help their queen because she carries the same genes they do, humans help their family members, etc. Everything we do is done to increase the chance of our genes’ survival. There is still a problem, however. Unless the gene that encodes altruism also encodes some sort of marker (like the red dot on the forehead), there is no way for an individual to be sure that someone else shares their genes.
Here’s the video: https://www.youtube.com/watch?v=lFEgohhfxOA&ab_channel=Primer
On page 356, the authors elaborate on the process of identifying "adaptive problems" to investigate and research evolutionary psychology. They write, "All adaptive problems share the following key feature: The genes of the individuals who solved that problem had greater replicative success than the genes of those who did not solve that problem."
ReplyDeleteThis type of question, of evolutionary adaptability, seems almost impossible to answer. We can identify possible issues that have to be overcome where maybe some people were genetically better equipped to do because of their psychological propensity for xyz traits and have thus survived better and reproduced more plentifully, but we don’t exactly have access to the psychological traits of those who did not survive; we cannot observe or analyze the traits we want to say were not naturally selected. How can Evolutionary psychology generate actual, testable hypotheses, if we can only identify the categories of beneficial and nonbeneficial traits based on what is beneficial and not on empirical evidence of what is not? We don’t have access to any negative examples of the category “cognitive survival traits” to really define it.
I also just absolutely must pick the bone with the example given on page 363, for how we generate testable hypotheses regarding what traits were/are more conducive to survival and reproduction, because it is based on entirely cis-hetero-patriarchal assumptions. One of the purported findings, “men express a greater desire for a variety of sex partners” used as evidence for, well, anything, is problematic because of (a) its underpinning gender essentialism (denying the existence of trans and nonbinary identities) and (b) ignorance of nuanced effects producing such evidence - like social pressures discourage women from expressing sexuality and desire as much or in the same ways as men are encouraged to do. That doesn’t tell me anything about how these differences may have developed and transformed through human evolution leading us to carry certain traits today – also because sexuality and gender are not innate categories that hold consistent meaning across and within cultures or time periods.
From what I understand from the paper, you're supposed to test the hypothesis by developing a model assuming your hypothesis is true and showing that the conditions required for your model to start have been satisfied sometime in history (like if you're assuming a drought 100 thousand years ago, showing that there actually was one then) and showing that your model actually leads to traits that we currently possess.
DeleteHowever, I agree with you that I'm not entirely satisfied by the method of developing/testing hypotheses. I feel like the method to test the hypothesis is semi-open-ended (I know models are powerful tools but I keep considering the examples in class (i.e. incest is evolutionary preferable if your family is the richest family in the community and ...) and thinking that someone proposed these, in what I'm assuming to be actual peer editted journals so they must have been able to "prove" these using the above method, and so the model/resulting traits must be too open-ended to be definitive criteria) and so considering the complexity of the problem we're trying to solve, wouldn't be any more convincing of a proof then saying look here's my reasoning and here's why it makes sense.
Lewis et al (2017) Evolutionary Psychology
ReplyDeleteEvolutionary psychology allocates little place to genetic explanation and reverse-engineering; it offers spider/sex-level hypotheses. Because evolutionary hypothesises are harder to test and discover than mechanistic explanations, it will turn out that often, evolutionary psychology is a lot of speculations and it will often offer “Just So” stories. Evolutionary psychology offers explanations of the type: this cognitive capacity of learning x and y exist because it has conferred an evolutionary advantage (which is about the gene) to ancestors developing this capacity. This may be valuable, but it does not give a mechanism of “how” the cognitive capacity is generated or of “how” we learn x and y. EvoPsych is thus not helpful to understand the “how” of the easy problem. Nevertheless, it might help for the “why” part of the easy problem. It might help if we stay away from the same kinds of survival/reproduction explanations (that work for spiders and sex) when explaining higher cognitive functions. I have also learned in the reading that the “top-down” approach is speculating about conditions and adaptivity in our ancestors’ environment (hopefully with evidence) and that the “Bottom-up” approach is to ask why we still do certain things today (eating sugar example).
I believe that using the top-down approach may possibly give us some clues as to how the brain works. For example, the order in which certain cognitive capacities (in language development for example) have evolved may help to guide our thinking because it would offer restrictions for possible mechanisms that would explain the cognitive capacities (that we could do this before we were able to do that might mean that the latter depends on the former). Another problem still remains; it is hard to differentiate between “intended products” of evolution and “by-products” of evolution and some of our most complex cognitive capacities might just be “byproducts”.
While I agree that the top down approach is useful - tracing causality almost always is - I also got frustrated reading this paper. In many ways, it was convincing - it flowed, the categories which the authors created made sense - yet in other ways, perhaps as a result of this class, I have become suspicion of smoothness. I think the problem with evolutionary psychology - which does not make it useless, merely incomplete - is that it offers one story or explanation for how things happened. Each of these theories - as I believe the authors may even acknowledge at some point - is based on correlation, and temporal relationships: these things happened at the same time and then this happened, therefore this depends on these things. The leap to causation is sizable, and it is dangerous (dramatic, I know).
DeleteWhile these theories may scaffold some possibilities for reverse engineering, and propose places to start, i.e. with a machine that can learn, and collaborates with other machines if you are trying to produce a babbling machine, if followed to strictly they also foreclose possibilities. This, I think, is their danger to cognitive science: they advocate for evolutionary hard equivalence - of environments, pressures, etc. This demand for strong equivalence is limiting along similar lines as a demand for T4 equivalence in the Turing test: it pares down the paths open to scientists trying to create and replicated thinking organisms.
I really appreciated the intention of this paper - presenting good practices and efficient heuristics in order to avoid asserting the ad-hoc conjectures that have given evo-psych such a bad rep. However, as I try to make sense of this paper among other readings, I realise, as pointed out in numerous above skywritings, that there is a dissonance between assessing the why and the how of categorisation in order to reverse engineer cognition. One ask why categorisation was an adaptive trait while missing the how of the mechanisms that allow categorization and communication.
ReplyDeleteI nevertheless wish to give evolutionary psych and this paper a little bit of credit, first in that reverse-engineering cognition is not necessarily the end)goal of evolutionary psychology nor of this paper. Second in that it does bring into light and support (without giving aa complete answer) how some traits of categorisation have come about and why they are likely to function in one way (the fittest way) rather than another.
Lets recall our debate about innate versus learned categorisation which lead to a consensus that both were complementary and necessary for creating our T3 passing robot. Evolutionary psych gives us the tools to ask whether “innate” categorisation is inborn in that it’s a universal law/trait (just as light going 300,000km/s is a universal law), or whether innate categorisation refers more to “evolved” trait. We discussed as an example properties of the visual system such as color which are not learned but inborn and which are intimately related to evolution. It is pretty trivial to say that the eye and it functioning is an evolved component serving our categorical perception and not an innate one that suddenly evolved along amoebas. We perceive between 400-700nm because it was more advantageous for us to distinguish between yellow and red berries rather than perceive in ultraviolet or possess echolocation. So I would definitely argue that evolutionary psych may play a role in clarifying some innate elements of categorization, even though most might only ad hoc conjectures. However, whether or not all innate categroziation is evolved, that I do not know.
The paper "Evolutionary Psychology: A How-To Guide," presents some basic methodology of evolutionary psychology. The authors suggest that an effective way to do evolutionary psychology is first to identify an adaptive problem––a problem that would have made it more difficult for our ancestors to survive and reproduce––and second to propose a psychology-based adaptation––a psychological trait that would have helped our ancestors solve the adaptive problem (thereby helping explain why we now have this psychological trait).
ReplyDeleteReflecting on this paper, I remembered that cognitive science is generally interested in solving the easy problem––that is, how and why we can do what we do. In this class, I would argue that we have been more interested in the "how" part of the easy problem, as we have learned that the number one goal of cognitive science is to reverse engineer a T3 robot which could provide a potential causal explanation as to how we do what we do. Evolutionary psychology, it seems to me, can give us insight into the "why" part of the easy problem but not the "how." If we propose an adaptive psychological trait––like being afraid of spiders––and say that this fear kept our ancestors alive, then we have helped explain why we have this fear. However, we have not explained how this fear works––what the underlying causal mechanisms are that give rise to this fear. It does not seem to me that knowing, for example, that enjoying sex is adaptive would help us at all to reverse engineer a T3 robot with our sexual preferences or capacities. Hence, evolutionary psychology can only give insight into one part of the easy problem (the "why"), and unfortunately that is the part we are less interested in.