Some more thoughts about cultural relativity and perception.
Click here to download pdf file
Over the last couple of years—and in more concentrated fashion in recent months—I have been giving attention to the serious1 psychological literature on cultural relativity/relativism (CR), and especially to reports of significant differences between Western Caucasians and East Asians, with respect to visual perception, discrimination and categorization, and to moral and aesthetic judgments. The familial relevance of such issues should be obvious: if cultural relativism is more than skin deep, then it is important for me to have a greater appreciation of how Japanese people see the world and organize experience if I want to understand those close to me, and help my bicultural children reconcile what are allegedly quite distinct world-views (not that this appears to be a great stretch for them).
What I’d like to believe about CR is that it is a largely cosmetic phenomenon—deeper than make-up or fashion perhaps, or even than politeness conventions or signals of aggression, but superficial compared to the “stuff that counts”: that beneath the variegated surface lies a universal perceptual system and universal higher cognition (including a common Language of Thought and moral psychology). I want to believe that although we may be strongly influenced by our language and culture to see the world in particular ways, to form culture- or language-specific conceptual categories and to organize experience in terms of varying analytic and moral judgments in unreflective situations, nevertheless—fundamentally, essentially—we are all equally capable of the same types of perceptual, moral, and aesthetic discrimination, and bound by the same genetically-constrained limits on personality and cognition: there may be moral or aesthetic parameters of variation (Dwyer 1999), but there are no societies or groups of individuals who lack essentially the same moral or aesthetic codes that inform my own perception and behaviour. No-one, barring individual pathology, is “beyond The (moral) Pale,” nor is anyone perceptually or cognitively blessed, in virtue of their cultural inheritance.
Of course, this is not to deny the vast range of diverse cultural values and behaviours, from culturally-sanctioned homicide, honour killings, geronticide, cannibalism and state-sponsored executions, through physical mutilation of various kinds—male and female circumcision, neck or lip stretching, piercing, tattooing—to different dietary regimes (eating people, eating monkeys, eating whales, eating pigs and cows) and forms of dress, and ways of dealing with litter. A basic course in anthropology—or a couple of evenings watching National Geographic—will tell us all about these bizarre practices (bizarre to us, of course).2
But this brief list simply illustrates a range of behaviours and displays of which we are all capable, and which we might sanction given a particular set of social conditions. Each society may have its own set of preference rules, ‘rankings’ of moral and social imperatives—loyalty to one’s own above or below compassion for others, for example—Yet, though the rankings may be different, in an Optimality theoretic way perhaps, we can still think that the principles are the same. The Universalist assumes, for instance, that no cultural group shows complete equinamity in the face of any of the last six of the Ten Commandments: killing people, stealing stuff, committing adultery, bearing false witness, and coveting various kinds of chattels, are all basically wrong for everyone, even if we can imagine mitigating circumstances. Eating people, too (as Malcolm Bradbury pointed out in 1959)3…
In addition to my essentialism, as a card-carrying Fodorean modularist, I also want to believe that low-level attentional and perceptual processes—such as visual discrimination and visual memory—are immune to the ‘top-down’—influences of higher-level cognition, including moral and aesthetic beliefs and desires: as suggested by visual illusions (Müller-Lyer, Neckar Cube, etc), I’d like to think we are obliged to see what we see—what our visual system constructs for us—even when we know that it is not real. As Fodor points out in the Modularity of Mind, there would be good evolutionary reasons for such cognitive impenetrability, if it were true: individuals whose perception is informed by hopes and expectations rather than on bottom-up sense-data tend to be eaten first (If you see a panther, don’t anther!).
Increasingly, however, these dual beliefs—in essentialism and cognitive modularity—are being challenged by scientific reports that claim to show deep-seated (low-level) effects of cultural diversity: articles that document differences between East Asians and Westerners in highly implicit perceptual tasks, where the relevant dependant measures are entirely unconscious responses such as eye-movements or involuntary physical gestures. And such results are very unsettling, a real cause for concern, since they imply that there may in fact be no necessarily common ground in how we perceive, internally represent and thus explain reality (either to ourselves or to others). To adapt Sapir’s famous comments regarding Linguistic Relativism:
In front of me is the most recent article on this topic by Dr. Rachel Jack and her colleagues of the University of Glasgow, published in the latest issue of Journal of Experimental Psychology. Prosaically if accurately entitled Internal Representations Reveal Cultural Diversity in Expectations of Facial Expressions of Emotion, the article documents a set of experiments investigating differences between East Asians (EA) and Western Caucasians (unselfconsciously abbreviated to WCs) with respect to which facial features are judged most relevant for attributing different emotional states; in other words, what features do we focus on to tell if someone is happy or sad, fearful or angry, disgusted or surprised? The authors constructed a very nice experiment, in which they presented observers with the same race-, gender- and emotion-neutral face overlaid by a randomly-ordered series of “white noise masks”, in which white pixels added to particular regions changed the apparent shape of the mouth, eye-brows, the shape of eyes and/or the gaze direction. Incredibly, each observer was presented with 12,000 trials; for each trial they were asked to categorize the facial expressions in terms of six emotional categories (significantly, there was also a can’t tell category).
On the face of it (!), Jack et al’s findings are impressive—which is presumably why the paper got published. The results show a clear divergence with respect to which specific facial features were deemed most important—by implication, most revealing to the observer—in attributing emotional states: whereas WCs were more sensitive than EAs to manipulated changes in the eye-brows and mouth regions, EAs paid more attention to the eyes themselves—in the case of surprise and anger, a lot more attention than did WCs. In addition, manipulated change of gaze direction was highly correlated with non-neutral emotions for East Asians, but not for WCs.
The statistical results and conclusions are limited by the experimental design—the chi-squared results (p <.05 for location, p <.0001, for change of gaze) show only that WCs and EAs patterns of categorization are different across the whole range of locations, and not, for example, which features are most important for everyone, which emotions are more or less distinguished by the two groups, or what kinds of within-group variation are significant. Nevevertheless, the research points up an important subconscious difference between the two groups: East Asians and Westerners ‘read emotions’ differently.
Yet, as surprising this finding might be, the big question is not whether but why: why do we find this difference, and has it anything to do with culture?
In the commendable, if dull, tradition of mainstream British psychology (and SOCO investigations), Jack et al do not really commit themselves to any particular interpretation of the observed differences, cultural or otherwise: as they might say: ‘here are the facts, you deal with them.’ However, at the outset to the paper, they allude to work by (smongst others) Richard Nisbett, who—as readers of this blog will know—has a particular view of the ways in which ‘Asians think differently from Westerners, and why’ that I have taken issue with in the past, and continue to resist. Specifically, Jack et al contextualize their paper by means of the following paragraphs:
The implication of the first cited paragraph is that that there is not simply a correlation, but a causal relationship, between higher-order cultural values and visual perception: the correlations between historical, philosophical and social tendencies and areal culture—Western society being more individualistic and analytic, Asian society more collectivist and holistic, at least traditionally—have given rise to distinct cognitive-processing strategies. Asians and Westerners literally “see the world differently” because of their cultural traditions.
Once again—as is the case in all of Nisbett’s papers too, as well as I can judge—the case for a causal relationship rests entirely on circumstantial evidence. In the absence of any specification of cognitive mechanisms for tranducing higher-order information about cultural values into low-level sensory-motor instructions, this is simply handwaving. And it doesn’t matter how often or deftly the hand is waved at culture, or how many experiments show that Asians perceive things differently from Westerners: without a proposed mechanism, it is an empty and—to me, for one—an irritating gesture.
Ironically (since most social psychologists are offended by nativism) the only way I can see the CR story to work is if it’s genetic: Asians are genetically predisposed toward a different kind of sensory perception, acquired through millenia of selection for this cognitive trait. It is the biological fate of Westerners to attend to focal objects, to ignore the context, to explain the world in terms of essential properties not contingencies of the situation, just as it is the fate of Asians not to see the wood for the trees, to attend to the context, etc. Indeed, it seems a good deal more plausible to think of differences in perception as driving higher-order beliefs—such as individualism vs. collectivism—rather than the other way around.
I’m not going to develop this idea, though, not only because—as someone with mixed-race children—the implications are just too disturbing, but because I think it’s nonsense: the arrow of causation flies no more cleanly in this direction than the other.
Moreover, there’s a much simpler, local explanation (as anyone who knows East Asians could point out), and my graduate students—Japanese themselves—did. It is this: compared to typical Westerners, typical Asians are considered and consider themselves rather deadpan, they don’t betray their emotions as obviously as less continent Westerners, who delight in more elaborate facial gestures. Doesn’t everyone know this?! (see footnote 1).
Now, supposing this simple idea to be true (i.e., experimentally validated), it should be clear why Westerners look to the eye-brows and mouth, and East Asians focus on the eyes. To use an apposite metaphor (cf. Masuda & Nisbett 2001) if you want to catch a fish, you don’t stare at a dead pond: you cast your line—in this case, your eyes—to where you’ve seen fish jumping in the past—in htis case, their eyes: experience has told Asian adults a long time ago in their childhood developments that mouths are not a likely source of information about emotions, but the eyes don’t lie…
Of course, this explanation is cultural in its own way, but that way is not philosophically disturbing to my essentialism:4 If it is the right way of thinking about things, then with respect to these phenomena CR is not skin-deep..but nor is it fatal. And I can sleep at night.
Notes
1. By ‘serious’ is meant peer-reviewed articles appearing in reputable scientific journals: I ignore—at any rate, try to ignore—the copious anecdotal first-hand reports of such differences, which may be suggestive, but are for the most part unreliable and unvalidated. That includes my own observations. Some radical Cultural Relativists would no doubt see this distinction—at least, the preference for scientific data over naïve, first-hand reporting—as itself culturally determined: see the Catch 22 article (forthcoming). Whether or not this is a valid objection, I’ll continue to discount anything that has not passed peer review…but see below.
2. And, for a different perspective, showing nicely how narrow of our view of normal is—especially in American Psychology—see Heinrich, Heine & Norenzayan (and commentaries). (What a happy collocation of authors’ names!)
3. This is a book that I haven’t yet read, for my sins. The following relevant quotes suggest that I should get on with it:
Click here to download pdf file
Over the last couple of years—and in more concentrated fashion in recent months—I have been giving attention to the serious
What I’d like to believe about CR is that it is a largely cosmetic phenomenon—deeper than make-up or fashion perhaps, or even than politeness conventions or signals of aggression, but superficial compared to the “stuff that counts”: that beneath the variegated surface lies a universal perceptual system and universal higher cognition (including a common Language of Thought and moral psychology). I want to believe that although we may be strongly influenced by our language and culture to see the world in particular ways, to form culture- or language-specific conceptual categories and to organize experience in terms of varying analytic and moral judgments in unreflective situations, nevertheless—fundamentally, essentially—we are all equally capable of the same types of perceptual, moral, and aesthetic discrimination, and bound by the same genetically-constrained limits on personality and cognition: there may be moral or aesthetic parameters of variation (Dwyer 1999), but there are no societies or groups of individuals who lack essentially the same moral or aesthetic codes that inform my own perception and behaviour. No-one, barring individual pathology, is “beyond The (moral) Pale,” nor is anyone perceptually or cognitively blessed, in virtue of their cultural inheritance.
Of course, this is not to deny the vast range of diverse cultural values and behaviours, from culturally-sanctioned homicide, honour killings, geronticide, cannibalism and state-sponsored executions, through physical mutilation of various kinds—male and female circumcision, neck or lip stretching, piercing, tattooing—to different dietary regimes (eating people, eating monkeys, eating whales, eating pigs and cows) and forms of dress, and ways of dealing with litter. A basic course in anthropology—or a couple of evenings watching National Geographic—will tell us all about these bizarre practices (bizarre to us, of course).2
But this brief list simply illustrates a range of behaviours and displays of which we are all capable, and which we might sanction given a particular set of social conditions. Each society may have its own set of preference rules, ‘rankings’ of moral and social imperatives—loyalty to one’s own above or below compassion for others, for example—Yet, though the rankings may be different, in an Optimality theoretic way perhaps, we can still think that the principles are the same. The Universalist assumes, for instance, that no cultural group shows complete equinamity in the face of any of the last six of the Ten Commandments: killing people, stealing stuff, committing adultery, bearing false witness, and coveting various kinds of chattels, are all basically wrong for everyone, even if we can imagine mitigating circumstances. Eating people, too (as Malcolm Bradbury pointed out in 1959)3…
Increasingly, however, these dual beliefs—in essentialism and cognitive modularity—are being challenged by scientific reports that claim to show deep-seated (low-level) effects of cultural diversity: articles that document differences between East Asians and Westerners in highly implicit perceptual tasks, where the relevant dependant measures are entirely unconscious responses such as eye-movements or involuntary physical gestures. And such results are very unsettling, a real cause for concern, since they imply that there may in fact be no necessarily common ground in how we perceive, internally represent and thus explain reality (either to ourselves or to others). To adapt Sapir’s famous comments regarding Linguistic Relativism:
The fact of the matter is that the ‘real world’ is to a large extent unconsciously built upon the [cultural] habits of the group…No two [cultures] are ever sufficiently similar to be considered as representing the same social reality. The worlds in which different societies live are distinct worlds, not merely the same world with different [cultural] labels attached (modified from Sapir 1929)
On the face of it (!), Jack et al’s findings are impressive—which is presumably why the paper got published. The results show a clear divergence with respect to which specific facial features were deemed most important—by implication, most revealing to the observer—in attributing emotional states: whereas WCs were more sensitive than EAs to manipulated changes in the eye-brows and mouth regions, EAs paid more attention to the eyes themselves—in the case of surprise and anger, a lot more attention than did WCs. In addition, manipulated change of gaze direction was highly correlated with non-neutral emotions for East Asians, but not for WCs.
The statistical results and conclusions are limited by the experimental design—the chi-squared results (p <.05 for location, p <.0001, for change of gaze) show only that WCs and EAs patterns of categorization are different across the whole range of locations, and not, for example, which features are most important for everyone, which emotions are more or less distinguished by the two groups, or what kinds of within-group variation are significant. Nevevertheless, the research points up an important subconscious difference between the two groups: East Asians and Westerners ‘read emotions’ differently.
Yet, as surprising this finding might be, the big question is not whether but why: why do we find this difference, and has it anything to do with culture?
In the commendable, if dull, tradition of mainstream British psychology (and SOCO investigations), Jack et al do not really commit themselves to any particular interpretation of the observed differences, cultural or otherwise: as they might say: ‘here are the facts, you deal with them.’ However, at the outset to the paper, they allude to work by (smongst others) Richard Nisbett, who—as readers of this blog will know—has a particular view of the ways in which ‘Asians think differently from Westerners, and why’ that I have taken issue with in the past, and continue to resist. Specifically, Jack et al contextualize their paper by means of the following paragraphs:
How could culture exert such an influence on the production and perception of basic emotion signals? Each culture embraces a specific conceptual framework of beliefs, values, and knowledge, which shapes thought and action. Culture-specific ideologies could exert powerful top-down influences on the perception of the visual environment by imposing particular cognitive styles. For example, individualistic (e.g., Western) cultures could generate tendencies to adopt local feature-processing strategies, whereas collectivist (e.g., East Asian) cultures may promote the use of global processing strategies, as suggested by relative size judgments (Davidoff, Fonteneau, & Goldstein, 2008), categorical reasoning styles (Norenzayan, Smith, Kim, Nisbett, 2002), change blindness sensitivities (Masuda & Nisbett, 2006), and eye movements (Blais, Jack, Scheepers, Fiset, & Caldara, 2008; Caldara, Zhou, & Miellet, 2010; Kelly, Miellet, & Caldara, 2010). By using distinct cognitive processing strategies, observers likely acquire culture-specific perceptual experiences of the visual environment, including facial expression signals.
Similarly, ideological concepts underlying societal functioning (e.g., Triandis, 1989) highlight important cultural differences, which likely influence the production of facial expressions. For example, individualistic versus collectivist cultures may adopt different display rules that govern when, how, and to whom emotions are expressed (e.g., Matsumoto, Seung Hee, & Fontaine, 2008), thus diminishing, enhancing, or altering facial expression signals. As a result, cultural differences in the expectations of expressive signals could give rise to the reported cultural confusions (e.g., Biehl et al., 1997; Ekman et al., 1987; Ekman et al., 1969; Jack et al., 2009; Matsumoto, 1992; Matsumoto & Ekman, 1989; Moriguchi et al., 2005) when expectations are not met.
The implication of the first cited paragraph is that that there is not simply a correlation, but a causal relationship, between higher-order cultural values and visual perception: the correlations between historical, philosophical and social tendencies and areal culture—Western society being more individualistic and analytic, Asian society more collectivist and holistic, at least traditionally—have given rise to distinct cognitive-processing strategies. Asians and Westerners literally “see the world differently” because of their cultural traditions.
Once again—as is the case in all of Nisbett’s papers too, as well as I can judge—the case for a causal relationship rests entirely on circumstantial evidence. In the absence of any specification of cognitive mechanisms for tranducing higher-order information about cultural values into low-level sensory-motor instructions, this is simply handwaving. And it doesn’t matter how often or deftly the hand is waved at culture, or how many experiments show that Asians perceive things differently from Westerners: without a proposed mechanism, it is an empty and—to me, for one—an irritating gesture.
Ironically (since most social psychologists are offended by nativism) the only way I can see the CR story to work is if it’s genetic: Asians are genetically predisposed toward a different kind of sensory perception, acquired through millenia of selection for this cognitive trait. It is the biological fate of Westerners to attend to focal objects, to ignore the context, to explain the world in terms of essential properties not contingencies of the situation, just as it is the fate of Asians not to see the wood for the trees, to attend to the context, etc. Indeed, it seems a good deal more plausible to think of differences in perception as driving higher-order beliefs—such as individualism vs. collectivism—rather than the other way around.
I’m not going to develop this idea, though, not only because—as someone with mixed-race children—the implications are just too disturbing, but because I think it’s nonsense: the arrow of causation flies no more cleanly in this direction than the other.
Moreover, there’s a much simpler, local explanation (as anyone who knows East Asians could point out), and my graduate students—Japanese themselves—did. It is this: compared to typical Westerners, typical Asians are considered and consider themselves rather deadpan, they don’t betray their emotions as obviously as less continent Westerners, who delight in more elaborate facial gestures. Doesn’t everyone know this?! (see footnote 1).
Now, supposing this simple idea to be true (i.e., experimentally validated), it should be clear why Westerners look to the eye-brows and mouth, and East Asians focus on the eyes. To use an apposite metaphor (cf. Masuda & Nisbett 2001) if you want to catch a fish, you don’t stare at a dead pond: you cast your line—in this case, your eyes—to where you’ve seen fish jumping in the past—in htis case, their eyes: experience has told Asian adults a long time ago in their childhood developments that mouths are not a likely source of information about emotions, but the eyes don’t lie…
Of course, this explanation is cultural in its own way, but that way is not philosophically disturbing to my essentialism:4 If it is the right way of thinking about things, then with respect to these phenomena CR is not skin-deep..but nor is it fatal. And I can sleep at night.
Notes
1. By ‘serious’ is meant peer-reviewed articles appearing in reputable scientific journals: I ignore—at any rate, try to ignore—the copious anecdotal first-hand reports of such differences, which may be suggestive, but are for the most part unreliable and unvalidated. That includes my own observations. Some radical Cultural Relativists would no doubt see this distinction—at least, the preference for scientific data over naïve, first-hand reporting—as itself culturally determined: see the Catch 22 article (forthcoming). Whether or not this is a valid objection, I’ll continue to discount anything that has not passed peer review…but see below.
2. And, for a different perspective, showing nicely how narrow of our view of normal is—especially in American Psychology—see Heinrich, Heine & Norenzayan (and commentaries). (What a happy collocation of authors’ names!)
3. This is a book that I haven’t yet read, for my sins. The following relevant quotes suggest that I should get on with it:
I like the English. They have the most rigid code of immorality in the world. Ch. 5
With sociology one can do anything and call it work. Ch. 7
I've often thought that my scruples about stealing books were the only thing that stood in the way of my being a really great scholar. Chap. 8.
More can be found here.4. And it’s relatively testable, given a good experiment with infants. That’s the thing to do next…
No comments:
Post a Comment