Culture Shapes How We Look at Faces

Constructivism. Determinism. It is all a bunch of hooey.

ResearchBlogging.orgA recent paper published by PLoS (Culture Shapes How We Look at Faces) throws a sopping wet blanket on widely held deterministic models of human behavior. In addition, the work underscores the sometimes spooky cultural differences that can emerge in how people see things, even how people think.


The following is from a PLoS press release:

Because face recognition is effortlessly achieved by people from all different cultures it was considered to be a basic mechanism universal among humans. However, by using analyses inspired by novel brain imaging technology, researchers at the University of Glasgow have discovered that cultural differences cause us to look at faces differently.

Lead researcher Dr Roberto Caldara said: “In a series of eye-movement studies, we showed that social experience has an impact on how people look at faces. Specifically we noticed a striking difference in eye movements in Westerners and East Asian observers. We found that Westerners tend to look at specific features on an individual’s face such as the eyes and mouth whereas East Asian observers tend to focus on the nose or the centre of the face which allows a more general view of all the features. One possible cause of this could be that direct or excessive eye contact may be considered rude in East Asian cultures.”

The results of the study, funded by the Economic and Social Research Council and the Medical Research Council, provide novel insights into why non verbal communication between people from different cultures is sometimes problematic, in an age where globalisation has dramatically increased interdependence, integration and interaction among people and corporations from all over the world. Western societies are generally more individualistic, whereas East Asian societies are collectivistic; Westerners appear to think and perceive focally and Easterners globally.

Dr Caldara continued: “By disproving the long-held assumption that face processing is universally achieved we have highlighted that the external environment, including the society in which we develop, is very influential in basic human mechanisms and caution should be taken when generalising findings to the entire human population.”

There is a general perception that behavioral variation emerges from a genetic substrate in a kind of inverted pyramid … at the genetic level there is not much variation, but at the “surface” (or beyond) there is potentially quite a bit of variation. It is more or less true that everyone believes this, and that people argue over the amount of variation that is determined way down at the genetic level vs. constructed at the surface. Even the most extreme construtivists will acknowledge that genes determine the basic neural circuits that give us, say, the ability to speak, even if the same constructivists will then argue that genes have nothing to do with what we say or how we make or interpret meaning.

So you have genes, then you have basic primitive neural mechanisms and neural-hormonal circuits, and these building blocks are put together to make more complex behaviors which are increasingly shaped by context (culture, etc.) at increasingly derived levels.

But this is a naive and inaccurate view. To the extent that genetic information is involved in human behavior, there is no strong evidence that genetic determination is manifest mainly in one vs. another developmental stage, behavioral substrate, or context. Yes, yes, you will see positivist statements asserting a link between developmental depth and genetic determination, but these are almost never based on conclusions from evidence, but rather, these statements are almost always assumptions used to interpret observations. Indeed, in the paper at hand, we see this:

Face processing,… is thought to be invariant across all humans. …Here we monitored the eye movements of Western Caucasian and East Asian observers … Contrary to intuition, East Asian observers focused more on the central region of the face.

I’ve added emphasis to the phrase “Contrary to intuition.” The “intuition” here is ‘what everybody assumed to be true.’

The constructivist I mention above who admits that the basic neural circuitry is genetic but that the nature of our thoughts is cultural has it backwards and upside down! The neural circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The preservation of this circuitry in humans occurs because we grow up as babies in a linguistic environment. Culture determines that we have the ability to speak (with word, sign language, whatever). On the other hand, the nature of our thoughts may in some cases be determined by hormone levels. If you don’t think that is true than you hav never met, or been, a teenager.

Constructivism. Determinism. It is all a bunch of hooey.

Caroline Blais, Rachael E. Jack, Christoph Scheepers, Daniel Fiset, Roberto Caldara, Alex O. Holcombe (2008). Culture Shapes How We Look at Faces PLoS ONE, 3 (8) DOI: 10.1371/journal.pone.0003022


Get the paper HERE.

Comments

  1. #1 Tony Jeremiah
    October 4, 2008

    I’ve added emphasis to the phrase “Contrary to intuition.” The “intuition” here is ‘what everybody assumed to be true.’

    Developmental psychologists have shown that when exploring human faces, 1-month-old children focus on the edges of the face (e.g., hairline, chin, ears). Around 2-3 months, when their contrast sensitivity improves, they begin to explore internal facial features (eyes, nose, mouth). I skimmed the cited study, and it looks like the study participants were all adults. To be certain that there isn’t a genetically programmed eye movement pattern that is eventually overriden by cultural factors, the study would need to be repeated with infants from Western and Eastern cultures who are 1-3 months of age.

  2. #2 Greg Laden
    October 4, 2008

    Tony: Your statement could be interpreted as a perfect example of what I’m talking about. You are telling me that you KNOW that there are genetically programmed eye movement patterns in infants, yet you cannot produce information on what genes are involved. You assume they are there because you have a model of early = genetic (early in lifespan, early in developmental sequence, or basal to a more complex behavior as in the present study). If you assume that this is true and never test the assumption, and use any circumstantial evidence that comes along to support it, then the very simplistic inverted pyramid model will become very strongly ‘supported’ without ever being tested.

  3. #3 gudden
    October 4, 2008

    The neural circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The preservation of this circuitry in humans occurs because we grow up as babies in a linguistic environment.

    This is a really surprising assertion to me, a non-primate neuroscientist. So surprising that one of the following must be true: 1) you’re quite mistaken, or 2) you know something really interesting that I don’t know. If 2 is the case, I’d be really grateful (by “really,” I mean “I’m not being snotty here”) for a reference or something to lead me toward the really interesting thing that I don’t know.

  4. #4 Greg Laden
    October 4, 2008

    Gudden: Good question! I assure you that this is a number 2 situation. And you already know this if you are a neruoscientist, unless you don’t work with organisms with a cortex or are very very specialized and skipped your intro classes or something. Most likely, though, you are just not thinking of it this way. I’ll be happy to supply some references (see below).

    Cortical connections (cell number, axon growth and connections) in mammals are significantly overproduced, and later culled … not expanded … culled to produce functioning major circuits. This culling results from use patterns. So a reasonably isolated function that is not used in a given individual may in fact not develop at all.

    An adult human has a number of critically important tracts enervating the diaphragm, larynx, pharynx, tongue, and some other head muscles that are unique among adult primates and that are critical for speech. These are different when compared to the same circuitry in non-linguistic primates (which also enervate these areas, of course).

    Newborn macaques also have these circuits, but they are culled out to form the standard non-speech primate pattern. This is how circuitry is generally formed in mammal cortex, and beyond (beyond cortex, beyond mammals … but I don’t know a lot about non-mammals so I don’t want to go into that.)

    I’m reaching a bit when I say all primates. Really, I should say that this is likely true in a very wide range of primates.

    This is following what is generally known about these things, but the specific work done on primates that I’m referring to was done by Terrence Deacon. He’s written a number of papers and at least one book that lays out this research.

    Deacon, T 2000. Evolutionary perspectives on language and brain plasticity. J. of Communication Disorders 33:4 (273-291).

    See also C. Holden’s review in Science “The Origin of Speech” (Science 303:5662(1316-1319).

    Deacon’s book is: The Symbolic Species: The Co-Evolution of Language and the Brain

  5. #5 Tony Jeremiah
    October 4, 2008

    You are telling me that you KNOW that there are genetically programmed eye movement patterns in infants

    That would probably be the interpretation from the following:

    “Developmental psychologists have shown that when exploring human faces, 1-month-old children focus on the edges of the face (e.g., hairline, chin, ears). Around 2-3 months, when their contrast sensitivity improves, they begin to explore internal facial features (eyes, nose, mouth).”

    But with this:

    “To be certain that there isn’t a genetically programmed eye movement pattern that is eventually overriden by cultural factors, the study would need to be repeated with infants from Western and Eastern cultures who are 1-3 months of age.”

    I’m suggesting that a further experiment needs to be conducted to rule out the possibility that there is a genetically-based (by that I mean environmentally independent and not literally gene-based) phenomenon responsible for the (intuitive) eye movements. Here, intuitive eye movements means the frequently observed triangular sequence of facial recognition involving the shifting of attention from one eye, to the mouth, and then to the other eye, rather than the non-intuitive eye movement that involves (primarily) fixating attention around the nose.

    There are two main results I think that are possible by repeating the study with 1-3 month olds, with one suggesting an environmentally-independent eye movement pattern. If the results show that 1-3 month olds from both Western and Eastern cultures show the triangular movement pattern, but then changes to the pattern observed in the study cited, that would suggest a biologically-based phenomenon overriden by cultural factors. OTOH, if the results show that 1-3 month olds from each culture show the pattern observed in adults from each culture, that would suggest a very early cultural/environmental influence on eye movement behavior that is consistent with experience-dependent neural development.

    …yet you cannot produce information on what genes are involved.

    I’m not quite buying this argument. It seems analogous to saying that if someone hears a particular song on the radio that they recognize, they are lying when they give the name of the song because they are unfamiliar with the specific sequence of notes that make up the song.

  6. #6 llewelly
    October 4, 2008

    gudden:

    This is a really surprising assertion to me, a non-primate neuroscientist.

    You’re a non-primate? Really? I guess it’s true what they say. On the internet, no-one knows you’re a dog.

  7. #7 Greg Laden
    October 4, 2008

    Tony? Again, to which genes are you referring? What is the protein that the gene codes for?

    Also, you are leaving something out in your analysis. The face.

  8. #8 Tony Jeremiah
    October 4, 2008

    Tony? Again, to which genes are you referring? What is the protein that the gene codes for? Also, you are leaving something out in your analysis. The face.

    You’re going to have to seek an expert for specifics. What you are asking me is well out of my expertise area, and probably just generally modern genetics’ ability to answer at the present time. As I understand it, we’ve only just finished mapping the genome, and are now only starting to understand epigenetic phenomena (which I suspect is probably the explanation for very early environmentally-independent patterns–so perhaps not even alleles are involved).

    The best speculation I can provide, is based on a neuropsychological framework from a former professor of mine called the levels of analysis framework. The framework was actually proposed as an organizational tool that would help conglomerate several research domains which may study particular phenomena at macroscopic and microscopic levels. Proceeding from macroscopic to microscopic, these levels of analysis include: Performance Domain, Cognitive Domain, Mental Operations, Neural Systems, and Cellular Domain. Perhaps allelic domain could be added.

    In this instance, the phenomena in the performance domain would be the particular eye movement behaviors observed (i.e., the triangular scanning pattern vs. the nose fixation pattern). I don’t really have a clue as to what the cognitive domain and mental operations would entail–that would involve making some assumptions about the thoughts accompanying why persons would have preference for staring at eyes rather than noses. At the neural systems level, we would most likely be talking about areas associated with the visual attention network such as the fusiform face area, lateral geniculate nucleus, and dorsolateral prefrontal cortex, and more obvious structures such as the eyes and the muscles attached to them. When I start to think about how the characteristics of the aforementioned structures are constructured at the cellular and genetic levels, that’s where things get out of my expertise area. But my guess is, is that particular eye movement patterns associated with scanning of faces, are an emergent property of cellular and genetic processes.

    You might have better luck getting your questions answered with experts who study phenomena at the cellular and genetic levels.

  9. #9 The Nerd
    October 7, 2008

    Thank you! It’s refreshing to see someone out there who doesn’t believe in predestination determinism.

  10. #10 gudden
    October 10, 2008

    You’re a non-primate? Really? I guess it’s true what they say. On the internet, no-one knows you’re a dog.

    Tail wagging is a constant hazard around the lab.

    Now that the conversation has long since moved on (I had a trying week with respect to interets access, sigh), I took a look at one of the Deacon references (Deacon, 2000) you listed as well as a couple others that came up in the PubMed search, and while I admit to skimming through them, I don’t think they support the statement that I questioned originally.

    That statement, again:

    The neural circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The preservation of this circuitry in humans occurs because we grow up as babies in a linguistic environment.

    As I read it, a literal interpretation of this would suggest that vocal primates raised among verbal humans might acquire language. But this experiment has already been done, at least with gorillas and chimpanzees, and the nonhuman primates did not acquire spoken language. Furthermore, while some primates can learn a limited sort of sign language when taught by human instructors, two or more deaf-mute children raised together will spontaneosly develop their own sign language if left to their own devices without instruction.

    I’m not sure that you meant it quite so literally though. The closest version I could get on board with as a hypothesis worth testing would be something like, “The basal cortical circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The specific organization of that circuitry such that a capacity for language results, occurs when the unique human morphological adaptations that support language production develop in the context of a linguistic environment.”

    The plastic qualities of the cerebral cortex that you describe are certainly true, and don’t necessarily even require the developmental context – cortical areas formerly associated with amputated limbs, for instance, will become “colonized” by representations of other, nearby parts of the body. But, first of all, the subcortical neural circuitry between language-related cortex and language-producing effectors isn’t just support staff. The developmental plan that makes a human larynx is part of the developmental plan that makes human language, and topographical representations of the effectors and the sensory epithelia involved in language production are iterated many times between the periphery and the cortex, and without them, the cortical representations are impoverished at best. Second, even at the level of the cortex, as far as I can tell from my quickie read of Deacon and what little I already knew about this subject, we don’t actually know to what degree specific cortical adaptations involved in language production are determined by a developmental plan independent of activity. It’s not just a matter of which cortical neurons are hooked up to what, but also why distinct cytoarchitectures are associated with cortical language areas. Would a human raised without language have the same distinct cortical asymmetries? I have no idea, and I don’t think anyone else does either.

    So, unless I’ve failed to understand you, I’m still leaning toward the “quite mistaken” option on this particular point, but more broadly, I do mostly agree with the argument I think you were trying to forward with this post. The degree to which behaviors and their neural correlates are NOT determined by developmental plan can be astounding. My favorite story of this sort argues both in favor of your position, and for its limitations: the blind cavefish A. mexicanus is eyeless as an adult and visual areas of its brain are reduced relative to a closely related, surface-dwelling form. If you take an embryo of the eyeless variety, and transplant into it one of the eyes (or just the lens) of the surface-dwelling variety, the transplanted eye will “take” in the cavefish, and the visual brain on the contralateral side will develop like that of the sighted fish (retinal projections are completely crossed in fish). So, on the one hand, an animal that has evolved not to grow eyes or a visual nervous system will develop both with complete functionality if you put somebody else’s eye in there. On the other hand, this species had to have had a developmental plan for growing eyes and a visual brain in its phylogenetic history in order for this to be true.

    The lesson of this, and of the study in your post, I think, is that a belief in genetic determinism of behavior, particularly in its more strongly-stated form, is dumb. A presumption of determinism in general, on the other hand, is almost a prerequisite for doing behavioral science – any behavior is necessarily some combination of genetic heritage and environmental variables, and if you like, some value for a stochastic influence. The problem is always one of discovering which environmental variables are determinant, and what the actual constraints of the genetic heritage are. The dumb kind of genetic determinism, I think, amounts to a failure to appreciate that we usually know less about the latter than we think we do, which allows us to know less about the former than we’d like to.

  11. #11 gudden
    October 10, 2008

    You’re a non-primate? Really? I guess it’s true what they say. On the internet, no-one knows you’re a dog.

    Tail wagging is a constant hazard around the lab.

    Now that the conversation has long since moved on (I had a trying week with respect to interets access, sigh), I took a look at one of the Deacon references (Deacon, 2000) you listed as well as a couple others that came up in the PubMed search, and while I admit to skimming through them, I don’t think they support the statement that I questioned originally.

    That statement, again:

    The neural circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The preservation of this circuitry in humans occurs because we grow up as babies in a linguistic environment.

    As I read it, a literal interpretation of this would suggest that vocal primates raised among verbal humans might acquire language. But this experiment has already been done, at least with gorillas and chimpanzees, and the nonhuman primates did not acquire spoken language. Furthermore, while some primates can learn a limited sort of sign language when taught by human instructors, two or more deaf-mute children raised together will spontaneosly develop their own sign language if left to their own devices without instruction.

    I’m not sure that you meant it quite so literally though. The closest version I could get on board with as a hypothesis worth testing would be something like, “The basal cortical circuitry that determines our ability to speak is present in all primates, even those that don’t speak. The specific organization of that circuitry such that a capacity for language results, occurs when the unique human morphological adaptations that support language production develop in the context of a linguistic environment.”

    The plastic qualities of the cerebral cortex that you describe are certainly true, and don’t necessarily even require the developmental context – cortical areas formerly associated with amputated limbs, for instance, will become “colonized” by representations of other, nearby parts of the body. But, first of all, the subcortical neural circuitry between language-related cortex and language-producing effectors isn’t just support staff. The developmental plan that makes a human larynx is part of the developmental plan that makes human language, and topographical representations of the effectors and the sensory epithelia involved in language production are iterated many times between the periphery and the cortex, and without them, the cortical representations are impoverished at best. Second, even at the level of the cortex, as far as I can tell from my quickie read of Deacon and what little I already knew about this subject, we don’t actually know to what degree specific cortical adaptations involved in language production are determined by a developmental plan independent of activity. It’s not just a matter of which cortical neurons are hooked up to what, but also why distinct cytoarchitectures are associated with cortical language areas. Would a human raised without language have the same distinct cortical asymmetries? I have no idea, and I don’t think anyone else does either.

    So, unless I’ve failed to understand you, I’m still leaning toward the “quite mistaken” option on this particular point, but more broadly, I do mostly agree with the argument I think you were trying to forward with this post. The degree to which behaviors and their neural correlates are NOT determined by developmental plan can be astounding. My favorite story of this sort argues both in favor of your position, and for its limitations: the blind cavefish A. mexicanus is eyeless as an adult and visual areas of its brain are reduced relative to a closely related, surface-dwelling form. If you take an embryo of the eyeless variety, and transplant into it one of the eyes (or just the lens) of the surface-dwelling variety, the transplanted eye will “take” in the cavefish, and the visual brain on the contralateral side will develop like that of the sighted fish (retinal projections are completely crossed in fish). So, on the one hand, an animal that has evolved not to grow eyes or a visual nervous system will develop both with complete functionality if you put somebody else’s eye in there. On the other hand, this species had to have had a developmental plan for growing eyes and a visual brain in its phylogenetic history in order for this to be true.

    The lesson of this, and of the study in your post, I think, is that a belief in genetic determinism of behavior, particularly in its more strongly-stated form, is dumb. A presumption of determinism in general, on the other hand, is almost a prerequisite for doing behavioral science – any behavior is necessarily some combination of genetic heritage and environmental variables, and if you like, some value for a stochastic influence. The problem is always one of discovering which environmental variables are determinant, and what the actual constraints of the genetic heritage are. The dumb kind of genetic determinism, I think, amounts to a failure to appreciate that we usually know less about the latter than we think we do, which allows us to know less about the former than we’d like to.

  12. #12 gudden
    October 10, 2008

    Oh, lord. Sorry about that ridiculously long double-post.

  13. #13 Greg Laden
    October 11, 2008

    As I read it, a literal interpretation of this would suggest that vocal primates raised among verbal humans might acquire language

    The fact is that these neural connections are there. That does not, however, imply what you are saying here. Being raised among verbal humans is absolutely essential to get normal verbal speech in humans, but clearly is not sufficient for primates more broadly. They can’t “catch” language or language.

    For one thing, speech and language are not the same thing. We are speaking hear about speech (the neural circuitry for speech) but language would be a necessary pre-requisite for speech in an individual.

    You can’t read Deacon quickly and get it right. I worked with him for years and I assure you this is true! Read it again much more carefully, separate out language and speech, and you’ll have it.

    Look at his book … there is also a paper I can’t put my hands on right now but that I’ve used a zillion times in teaching this material that summarizes it very well, actually a chapter in a collection … that covers the cave fish, as well as moles and other examples which, as you point out, are very instructive.

Current ye@r *