orgtheory.net

universal reality (almost): the case of categories and colors

I’m interested in the nature of reality and particularly the boundaries and scope of the social construction of reality.  I think social construction clearly plays an important role, but the question is, how “strong” is that role?   For example, I think the performativity argument (and associated “strong programme”) pushes the social construction argument way too far.

But let’s get more specific: what role do categories, language and naming play in the construction of reality?

One empirical setting for actually studying this question is the case of color categories and color naming, an active area of research in linguistics, computer science and psychology.  Scholars in this space have looked at whether the extant categories and names of colors of particular languages impact what individuals actually see and remember.  The famous Sapir-Whorf thesis of course argued, broadly, that language, categories and culture strongly determine perception and reality.  But, the color research shows otherwise.  Languages with highly fine-grained distinctions for individual colors, as well as languages with relatively few (or even no!) distinctions and names for color, lead to the same perceptions and experiences of color.  (Check out the citations below to see the clever way in which this is empirically tested.)

Well, almost. Recent work is making some important qualifications to the argument (articulating a middle ground, of sorts, between universality and strong construction), and there clearly is a very active debate in this space.

Here are some links to this literature:

Now, I don’t, by any means, think that the color research necessarily is a knock-down argument against social construction.  But I do think this research definitely questions the “strong” form of construction — I have opportunistically cited and referred to these and other findings to make that point.  And another, perhaps unfair, knockdown argument is that no matter what linguistic categories a color-blind person has, it simply won’t matter in the perception of color.

There is of course much debate in the color literature as well and some of the work points toward a particular, softer form of construction.   And, the color research of course is just one setting, and the findings may not generalize to other settings.  But I do like the fact that the color research actually allows us to more rigorously say some things — with the usual qualifications and questions — about the specific role that language (as well as categories, culture etc) plays in the way we perceive the world.

Advertisements

Written by teppo

January 11, 2011 at 7:50 pm

30 Responses

Subscribe to comments with RSS.

  1. Lera Boroditsky at Stanford has a ton of work in this domain worth checking out. One of her recent studies that got popular attention showed that a person’s native color categorization affected aspect of visual perception (I think it was something like the more “natural” categories for a color you had – many different words for types of blue, rather than just “blue” “light blue” etc. – the better you were at correctly identifying lines drawn in that color of ambiguous length.)

    http://www-psych.stanford.edu/~lera/papers/

    Like

    MDH

    January 11, 2011 at 9:08 pm

  2. MDH: Thanks for the link. There’s also (and Boroditsky also seems to have work in this space) on other types of linguistic categories (number systems, spatial representation, conceptions of time etc) and associated impact (or not) on perceptions of reality.

    Like

    teppo

    January 12, 2011 at 1:12 am

  3. I agree that the color research undermines the strongest version of linguistic relativity. However, it’s quite a leap from this to the idea that, say “fair” and “unfair” or even “beautiful” and “ugly” are similarly universal.

    Like

    andrewperrin

    January 12, 2011 at 4:00 am

  4. Sure, sure. The beauty, aesthetics stuff is a whole other, (very) heated debate (there you’ve also got universals type arguments, e.g., Hyman’s Objective Eye relates, or Peter Kivy etc).

    Like

    teppo

    January 12, 2011 at 4:32 am

  5. I actually had lunch with Terry Regier yesterday and asked him about a lot of these issues. I opened with “isn’t this Whorfian stuff way overblown?” But he responded with a pretty nuanced account of the whole thing. Turns out that, like most things, the answer is pretty complicated. Some things seem robust to linguistic influence and other things don’t. But he did point out that cross-cultural variability only makes sense against the backdrop of some constants and thus we could completely agree that “strong constructionism” — that is, a fundamentalist antiessentialism — is indeed a dead end.

    I came away thinking that it all depends on your audience. For soc undergrads who think that they see the world just exactly as it is, a strong constructionist antidote may make sense just to attempt to bring balance. On the other hand, to a crowd of sociologists, a strong dose of realism is probably a needed corrective to the social-construction-of-everything trump card.

    Like

    Steve Vaisey

    January 12, 2011 at 7:00 am

  6. The “strong constructivist” camp is often used as a phantom villain. I keep asking who are these people anyway?
    I would not get overexcited about the research being able to show that ‘language matters’. Clearly, language is a social practice that evolves over time as people find vocabularies and meanings useful for them. Use of langauge also requires plenty of non-verbal skills — e.g. distinguishing different words for colors requires perceptiveness. A community which has a lot of words for colors probably incentivizes people to become more perceptive. Just as having more categories for social movements requires scholars to be more perceptive about the organizations they observe.
    None of this in any way hurts social constructivist arguments. Clearly nobody has a big power stake in color vocabulary whereas the vocabulary of e.g. ethnicities is subject to use of power.
    There is a somewhat related paper on color in ASR 2010 october issue, “Stratification by Skin Color in Contemporary Mexico” by Andrés Villarreal. The point is that without specific _racial_ categories, people distinguish degrees of skin color. I didn’t find this surprising at all since obviously people are not imbecile and always use categories (light/dark) to order their perceptions of the social reality. The interesting (speculative) point is that since the use of certain racial categories is illegitimate in Mexico it is very difficult to mobilize a social movement to fight for greater equality. Reminds me of France’s official decision not to collecting ethnical information making it impossible to produce statistics.

    Like

    Henri

    January 12, 2011 at 7:13 am

  7. Steve: Right — it depends is frequently the answer to the extremes. Sapir-Whorf-like extremes, though, appear to be alive and well (in a variety of forms and re-packagings) in the social sciences and that is why I think a reality check (aided by findings from nearby disciplines — though different context: colors) can be helpful in helping us think through things.

    Like

    teppo

    January 12, 2011 at 7:20 am

  8. Henri: Strong constructivism a phantom villain? Again, I think the performativity argument (and associated theories) is precisely in the ‘strong’ domain, and direct derivations show up in other literatures as well —- best example (this piece gets lots of citations/love and also received the best paper award — we’ve discussed this one before, when Fabrizio guest blogged):

    Ferraro, Pfeffer, & Sutton, 2005. Economics language and assumptions: How theories can become self-fulfilling. Academy of Management Review, 30: 8-24.

    Like

    teppo

    January 12, 2011 at 7:48 am

  9. Teppo, I fail to understand your argument, you need to elaborate a bit. How is the self-fulfilling theory/performativity argument invalidated by assuming some shared external reality? I do not see this. In fact, the argument Ferraro et al. put forward assumes something called “empirical validity” both exists and is perceivable ex post (see abstract). The perspective would not be acceptable to the phantom villain constructivist as it assumes a common objective standpoint: it claims that initial acceptance of theories has further non-discursive outcomes.

    It is hardly a strong constructivist view that complex bodies of knowledge are subject to social processes and the acceptance of claims are influenced by rhetoric, culture, etc. It does not follow that since individuals in every society obtain the same kind of color vocabulary, the individuals in every society would also obtain the same kind of knowledge of e.g. particle physics.

    With regards to the second part of the paper, I am not certain how assuming (a) individuals behavior to be influenced by their beliefs and (b) behavior of individuals in the aggregate to influence social reality are somehow alien to the existence of some shared reality?

    On the performativity front, would claiming that Marx’s views of historical materiality created the communist revolution be a strong social constructivist claim?

    In short, I think performativity is a process theory that succeeds or fails based on empirical justification, irrespective of one’s realist, weak/strong constructivist, or pragmatist viewpoint.

    Like

    Henri

    January 12, 2011 at 8:11 am

  10. Henri: OK, I can’t resist (and I guess I shamefully teed this up) — but Nicolai Foss and I answer your exact questions (about empirical validity, ex ante and post reality, etc) in these two papers (Ferraro, Pfeffer and Sutton’s response sandwiched in between):

    Felin, T. & Foss, N.J. (2009). Social reality, the boundaries of self-fulfilling prophecy, and economics. Organization Science.

    Now, note FPS’s response (in my mind, they refute their own argument by further anchoring on performativity and like theories): Ferraro, F., Pfeffer, J., & Sutton, R. (2009). How and why theories matter: A comment on Felin and Foss. Organization Science.

    Felin, T. & Foss, N.J. (2009). Performativity of theory, arbitrary conventions, and possible worlds. Organization Science.

    A more careful and direct response to your points later (offline now for sleep and teaching).

    Like

    teppo

    January 12, 2011 at 9:03 am

  11. […] that universal reality (almost): the case of categories and colors […]

    Like

  12. Teppo, with all due respect, I think your papers make exactly the same mistake you’ve been making here. I don’t see the performativity trend as claiming anything even approaching strong constructionism. The “Social Reality” article attacks a straw man in which performativity claims that any theory can create any reality, then invokes an assumed, undemonstrated concrete social reality as the scope condition for constructionism.

    I agree with Steve that a kind of “nuclear option” strong constructionism is great for undergrads but a silly position to try to defend among social scientists. In contrast, performativity contains a set of claims about technical causality, social/technical arrangements (“agencement“), and the mechanisms by which convincingness and establishment of theory produce its truth. These are empirical/genealogical claims, subject to evidential examination; they do not constitute an arbitrary relationship to “reality”, as your arguments imply.

    Like

    andrewperrin

    January 12, 2011 at 1:28 pm

  13. The recent double issue of Interdisciplinary Science Reviews on G.E.R. Lloyd’s book *Cognitive Variations* may be of some use in moving beyond various straw men: http://www.maney.co.uk/index.php/books/history_and_human_nature/.

    Like

    Brookes

    January 12, 2011 at 2:43 pm

  14. Andrew: I share Teppo’s reading of the performativity “trend.” Sure, there is often lip service paid to boundaries on social construction. But it is lip service, and ad hoc. There is no serious attempt to understand objective bounds that constrain construction, and no serious attempt to try to subject claims of performativity to an empirical test (see pp. 3-4 of this paper– http://web.mit.edu/ewzucker/www/handbook_chapter_on_market_efficiency_draft%202%20Dec_22_2010.pdf for a proposed criterion).

    Teppo: Thanks for the cites to this work on color. Very interesting. (And thanks for remindning me of your work with Foss on these issues in org theory.)

    P.S. I’ll read any further comments on this thread with interest, but won’t be commenting further.

    Like

    ezrazuckerman

    January 12, 2011 at 6:55 pm

  15. Ezra: “I share Teppo’s reading of the performativity ‘trend.’ Sure, there is often lip service paid to boundaries on social construction. But it is lip service, and ad hoc. There is no serious attempt to understand objective bounds that constrain construction . . .”

    Would you put your criticism of performativity in the context of ‘money’ or ‘the dollar’? In light of how the dollar has changed since 1971 (and before), I wonder if it is not a strong empirical case for the performativity perspective (although not the without ‘bounds’ side of things).

    Like

    Austen

    January 12, 2011 at 7:18 pm

  16. “In short, I think performativity is a process theory that succeeds or fails based on empirical justification, irrespective of one’s realist, weak/strong constructivist, or pragmatist viewpoint.”

    Henri: It’s impossible to assess whether “performativity” succeeds or fails. That is the problem. Specifically, performativity shows how something takes effect (say, Black-Scholes model), fine, but that’s just a story, not a theory. In other words, the theory does not offer up the set of possible, ex ante alternatives (theories, language, models, whatever) that might have been considered and taken effect, or not (so, what else could fulfill itself, but doesn’t – and, how does “correction” occur), but opportunistically points out how a given reality instantiates itself (the one that happened to work). Importantly, actors do not blindly follow extant scripts (a point that Ferraro et al make), but instead (under uncertainty) deliberate, construct beliefs and make choices based on assessments about the nature of reality (social psych is wrestling with the same problem, specifically, see the references to Jussim, Kenny, and Funder in our paper). Are there mistakes? Sure — but actors dynamically correct their assessments of reality based on feedback, and of course influence it. So of course there is some measure of construction, no question, I just don’t buy the strong construction implied by performativity, at all.

    “I don’t see the performativity trend as claiming anything even approaching strong constructionism.”

    Andrew: Well, the whole performativity effort has its direct roots in the strong programme (Barnes, Bloor, etc), essentially importing science studies intuition (e.g., Latour) into the domain of markets, management and economics. The STS angle can be interesting (Laboratory Life stuff etc), but seems all-too heroic, strong construction at the expense of any reality (I’ll spare the quotations on this). The problem with strong construction is that we, I guess, move to an advocacy-based model of science (as there is no meaningfully reality to start with — ex ante reality) rather than science as prediction, or attempts to understand the world, etc. This has even been made more explicit, Callon argues that we “no longer have to choose between interpreting the world and transforming it” (Callon 1998, p.352). Fine. At the extreme, we might as well construct anything — but there seem to be some basic, stable realities (e.g., human nature). OK, perhaps that’s an extreme (and sure, we all hope that science impacts reality), but that extreme also shows in the focus (tracing a given performance, ex post) and assumptions (ex ante reality, human nature) of the work in this vein. Furthermore, it feels all-too convenient to have something perform, and then if it doesn’t, it counter-performs — story-telling. The accounts are of course interesting (the McKenzie book is fascinating), but the theory is, IMHO, wrong.

    Beyond the above, it’s simply inconsistent and self-refuting to, on the one hand, say that some theories (models) are “false” (Ferraro et al focus on economics), but then, on the other hand, utilize a theory that does not, at all, allow for meaningful judgments about whether something is true or false (only ex post, if the model is realized). Furthermore, why are only some aspects of models realized (say, self-interest) but not others (say, omniscience)? I thought the authors would somehow catch the self-refuting problem in their response, but they instead anchored even further on the performativity argument.

    Overall, I don’t see it as radical or extreme, at all, to postulate that there are some important boundaries to the social construction process — Nicolai and I specifically discuss 1) objective reality and 2) human nature. I think the mechanisms and underlying processes of the social construction process definitely deserve much additional work vis-à-vis markets (which is why I am fascinated with the performativity of markets idea). I think the social aspects of markets remain woefully understudied.

    Like

    teppo

    January 12, 2011 at 11:48 pm

  17. [Fixed the link to Ezra’s piece above.]

    Like

    teppo

    January 13, 2011 at 1:36 am

  18. Teppo, rather than further hijack this thread I’m going to post some thoughts on your performativity article over at scatterplot. Watch for them soon. Baseline: of course it’s not radical or extreme to “postulate that there are some important boundaries to the social construction process.” But what those boundaries are and how they work remain thorny, unsolved problems. More on scatterplot, hopefully today if the schedule allows.

    Like

    andrewperrin

    January 14, 2011 at 1:26 pm

  19. Andrew: Sounds good. (Don’t worry about hijacking thread – good fun; witness Fabio’s endless pomo thread.)

    “But what those boundaries are and how they work remain thorny, unsolved problems.”

    No question, I think there is much yet to be written, studied in this space — I think a fascinating area.

    Like

    teppo

    January 14, 2011 at 3:15 pm

  20. OK, so procrastination has its good points – I’ve posted my thoughts here.

    Like

    andrewperrin

    January 14, 2011 at 5:35 pm

  21. Great — heading into some meetings just now (unfortunately), will try to react later today.

    Like

    teppo

    January 14, 2011 at 5:38 pm

  22. One quick interjection: @Teppo: Latour (and Callon, more importantly here) are founders of the ANT school which emerged as a reaction against the strong *social* constructionism of the Strong Programme (hence Latour changing the subtitle to Lab Life). So I think your conflation of the two is not merited (though, frustratingly, MacKenzie is historically associated with the Strong Programme, and I actually think that MacKenzie and Callon are doing somewhat different things with the umbrella term of “performativity”). But just to reiterate, ANT criticized the Strong Programme on grounds very parallel to your argument here, that the Strong Programme made the “social” too powerful and too separate from “nature” or “reality”, and in their model, anything was possible. ANT rejects that view.

    Like

    Dan Hirschman

    January 15, 2011 at 3:04 pm

  23. […] the favorite bones of contention in the soc blogosphere: the performativity of economics. Here is Teppo Felin on the universality of colors (and thus the unimportance of theories or descriptions of colors) and here is Andrew Perrin’s […]

    Like

  24. Dan: I’m not sure I see much day light, at all, between ANT and performativity (and I read LOTS of the literature related to the Org Sci discussion). But I’m open to seeing things differently — what should I read?

    Like

    teppo

    January 15, 2011 at 6:07 pm

  25. Teppo – I guess I’m suggesting a distinction within performativity between stuff focused on the true/false dichotomy (the Barnesian performativity of MacKenzie and “An Engine Not a Camera”) and the calculative agencies pushed by Callon and the Market Devices folks (which lines up with ANT). Have you read the recent back and forth in Journal of Cultural Economy between Judith Butler and Michel Callon? They are short pieces, and I think Callon’s clarifications do a nice job of illustrating what he’s doing with the term. There’s also Callon’s 2009 & 2010 pieces with Caliskan on “economization” in Economy and Society. Last, to toot my own horn, my take on performativity is on p8-10 (approximately) of this paper (pdf). I don’t get into this debate exactly, but I frame performativity very differently, which might help make sense out of the distinction I’m pushing here.

    Like

    Dan Hirschman

    January 15, 2011 at 6:20 pm

  26. Dan: “For those keeping score, I give this round to Andy!”

    Now wait a second, on what basis?! ;)

    Like

    teppo

    January 15, 2011 at 6:58 pm

  27. Teppo – The next sentence is “Of course, my sympathies are well-known and somewhat elaborated.” So, my giving the round to Andy was a bit overdetermined. But beyond that, I think Andy’s post does a really nice job of laying out the issues in a reasonably neutral way, and, relevant to my earlier comment, does so without making loose connections to “bad” and somewhat vague positions like “strong constructionism”. I hope that’s not too unfair! But as I said, I was with Andy to begin with here, so it’s hard to be fair about scoring. I wonder if there’s anyone neutral left who could take on such a task (and, alternatively, that would be actually worth trying to convince one way or the other!).

    Like

    Dan Hirschman

    January 15, 2011 at 7:27 pm

  28. Very cool. Thanks for the links about the color research!

    This reminds me of an unscientific survey run by the XKCD web comic that had 200,000 folks label 5,000,000 colors. Its basically just for fun (and is fun), but there’s some slightly interesting stuff on gender as well: http://blog.xkcd.com/2010/05/03/color-survey-results/

    Like

    Benjamin Mako Hill

    February 6, 2011 at 12:10 am

  29. for a view of reality ultimately unrelated to the mathematics of logic (or reality) but instead the aesthetics (or art) of reality please see artist David Foox newest work. It is all about universal reality and bases it upon the aesthetic of the long eared rabbit. http://followpropaganda.com/about

    Like

    Roggie

    July 18, 2011 at 6:59 pm


Comments are closed.

%d bloggers like this: