Re: Sapir-Whorf hypothesis ?

From: Damien Broderick (damienb@unimelb.edu.au)
Date: Sun Jul 27 2003 - 20:20:36 MDT

  • Next message: Paul Grant: "RE: Food labels and consumer information (was Re: Protesters swarm Calif. biotech meeting)"

    At 11:22 AM 7/27/03 +0200, Frederick H. Cheeseman wrote:

    >There is a linguistic theory--known as the Sapir-Whorf hypothesis--that
    the structure of a human language sets limits on the thinking of those who
    speak it; hence a language could even place constraints on the development
    of the cultures that use it.

    >Is this true ?

    No.

    Well, except that of course and self-evidently it is, since if your
    language by some bizarre circumstance lacks implicit causality you'll have
    a lot of trouble inventing science and learning not to chew your own
    fingers off more than once. But...

    =============
    Umberto Eco has more than once quoted chapter 26, Book II, of Aulus
    Gellius' Noctes Atticae, a second century encyclopaedia, which seems to
    show that Gellius did not discriminate colours according to our (or the
    spectrum's) ordering. How did Gellius see the world, or at any rate
    `elaborat[e] upon literary texts coming from different centuries' (p. 157)?

    "Diachronically speaking, Aulus Gellius was trying to put together the
    codes of at least two centuries of Latin literature and, synchronically
    speaking, the codes of different non-Latin cultures. Gellius must have been
    considering diverse and possibly contrasting segmentations of the chromatic
    field. ... Determined by his cultural information, Gellius cannot trust to
    his personal perceptions, if any, and appears eager to see gold as red as
    fire, and saffron as yellow as the greenish shade of a blue horse." (p. 171)

            This power of verbal categories over our illusory sense of direct,
    unimpeachable perception of the world is known in linguistics as the
    Sapir-Whorf hypothesis; it fell from favour in the 1970s, but is making a
    qualified comeback.

    ...A Scientific American review by Philip E. Ross (February, 1992) finds
    that its subsequent debunking might not have been exhaustive. True, `Whorf
    had asserted that Eskimos use many distinct words in place of the one
    English word "snow".... But Laura E. Martin, an anthropologist at Cleveland
    State University, traced the story to its sources and found that Whorf had
    exaggerated the number of Eskimo snow roots while understating the number
    in English: slush, powder, blizzard and so forth.' Nevertheless, some
    Australian languages suggest that `spatial conceptualization is not
    universal', although such distinctions `are mastered very early in life'
    (pp. 11-2).

    ...What then of Aulus Gellius and his saffron yellow as the greenish shade
    of a blue horse? Has the eye been primed by nature, or are hues secondary?
            The New Guinea Dani have only two colour words: `mola', for warm, vivid
    hues, and `mili' for cold, gloomy tones. Nearly 20 years ago Eleanor Rosch
    tested Dani response to carefully selected coloured chips, and undid her
    Whorfian beliefs in a single stroke. Individual Dani had difficulty coming
    to agreement on which intermediate hues were mola or mili, but (in a
    subsequent test) had virtually none in identifying which of 40 chips each
    had been shown.

    "The Dani turned out to recognize colors in a manner very similar to the
    Americans - making the same confusions, for example - though their overall
    scores were not as high. Thus, differences in naming structure were not
    paralleled by differences in the way in which colors were stored in memory
    or accessed by recall.... The ways in which individuals from different
    cultures remember colors seemed to reflect the organization of the nervous
    system, not the structure of particular lexicons." (p. 344-5)

            More notable still was cross-cultural work done about the same time by
    Berlin and Kay, who `studied the color-naming practices of individuals
    drawn from twenty genetically diverse languages, ranging from Arabic to Ibo
    (Nigeria) to Thai' (p. 348). The number of basic colours per culture ranged
    from two (the Dani, say) to eleven (English). What shook many cultural
    relativists, though, was the discovery of certain invariant factors:

    "[W]hen confronted with chips spanning the full spectrum of hues, their
    informants selected the same focal areas for colors, irrespective of
    whether they had names for them; that is, [they] agreed about what was a
    `good blue' or a `poor green', even if their culture lacked names for these
    colors." (p. 348)

            Perhaps the least predictable finding was a regular ordering, across 98
    different cultures, such that if only two terms exist these translate or
    code to bright/dark, white/black; if there is a third, it is always red.
    Fourth and fifth are yellow or green, in either order; then blue and brown
    compete for sixth and seventh place; purple, pink, orange and grey complete
    the spectrum (p. 349). In this case at least, there is very little scope
    for philology to interfere with ontology, except to mutilate it.
            Eleanor Rosch has summarized the intellectual reverse which these
    discoveries forced upon her and her colleagues, which to my mind resembles
    the blow that empirical research, such as that collated by Farrell, strikes
    at the heart of Freud's doctrines. To their considerable surprise (for they
    had started with a sense that the Whorfian hypothesis was virtually
    self-evidently true), they

    "discovered that colors appeared to be a domain suited to demonstrate just
    the opposite of linguistic relativity, namely, the effect of the human
    perceptual system in determining linguistic categories. Very similar
    evidence exists in the domains of geometric form and emotion categories.
    Furthermore, psychological principles of categorization may apply to the
    formation of all categories, even in culturally relative domains.

    "At present, the Whorfian hypothesis not only does not appear to be
    empirically true in any major respect, but it no longer even seems
    profoundly and ineffably true."

            Why, Rosch asks, `has it been so difficult to demonstrate effects of
    language on thought?' (Ibid). Because, she concludes, it could only be true
    if major lexical distinctions and the categories of `nature' which they fix
    within language(s) were `almost accidentally formed' (Ibid) - in effect,
    the Saussurean postulate. By contrast, her research suggests forcefully
    that, for the categories she has investigated, the links are strongly
    motivated; language categories are shaped by neuro-empirical givens.

    [THE ARCHITECTURE OF BABEL]
    ===============================

    Damien Broderick



    This archive was generated by hypermail 2.1.5 : Sun Jul 27 2003 - 20:29:23 MDT