LINGUIST List 6.1149

Tue Aug 22 1995

Disc: Sapir-Whorf and what to tell students these days

Editor for this issue: Ann Dizdar <>


  • Dan Alford, Re: Sapir-Whorf and what to tell students these days

    Message 1: Re: Sapir-Whorf and what to tell students these days

    Date: Sun, 20 Aug 1995 11:33:35 Re: Sapir-Whorf and what to tell students these days
    From: Dan Alford <>
    Subject: Re: Sapir-Whorf and what to tell students these days

    Here's what I tell my undergrad and grad students about that same chapter in Language Files. I.e., it's difficult to even judge what's going on with Whorf unless you are simultaneously conversant with linguistics, American Indian languages, and at least the insights of modern physics.

    First, the conclusion that is appropriate is that, as I showed in "The Demise of the Whorf Hypothesis" (Berkeley Linguistics Society, 1978?), what Whorf said has little or generally no relation whatever to the entire body of discussion that comes under the name "(Sapir-)Whorf Hypothesis". He showed decades before the critics came up with their own hypotheses, which they failed to name after themselves, that he would never have agreed with their characterization of his thoughts.

    Simple test: Read the LF chapter and then ask, "Who created the Whorf Hypothesis?" And a quick way to answer this is: What did Whorf himself call it in his two or three references? He called it the "principle of linguistic relativity" or the "linguistic relativity principle". My own reading of Whorf never finds the word "hypothesis" at all. So -- right off the bat, and this is a good way to teach scientific nomenclature, who turned Whorf's 'principle' into a(n) 'hypothesis', and why? It wasn't Whorf, because his designation was clear. So what is the difference between the two? Well, a principle is like an axiom in geometry: a starting point which is theoretically unverifiable -- it's just a starting point. You want something else, you begin from a different starting point, and then you develop your hypotheses from there.

    Next: what does Whorf's "linguistic relativity principle" have to do, if anything, with Einstein's "relativity principle" (which I covered in my also BLS, 1980?, paper, "Is Whorf's Relativity Einstein's Relativity?"). Ah, now we've gotten to the crux of it -- much against Pinker's stand (which was copied and intensified in a Lingua Franca short bio of Suzette Elgin Hayden recently, where a digression found its author saying that Whorf cobbled together his theory from a few ill-translated snatches of Apache -- echoing a Pinker statement and relying on Pinker's quoting Whorf correctly, which he didn't, about a canoe on a beach pointwise: which Pinker identified as an Apache sentence, but which Whorf knew quite well was Nootka in the Pacific Northwest rather than Apache in the beachless desert), Whorf was upping the ante on Einstein, who argued that Euclidian geometry, far from being UNIVERSAL, was applicable only to flat surfaces; that for round surfaces, which most of reality is made up of, you need non-Euclidian geometry. I.e., when the phenomena change significantly, you have to change the tool you're using. Well, that's what Whorf said too (see Heisenberg's Lament below), except he moved its domain from mathematics to natural human language; hence: the truly aptly named "principle of linguistic relativity" as Whorf himself named it.

    Admittedly, this doesn't make any sense until you see it in action, in "An American Indian Model of the Universe," where he posits a worldview without our tense/time (past/present/future diorama or river of time), using Manifested/Manifesting (plus other synonyms) instead. That is: SAE grammars/cultures give that aforementioned notion of "time", which is supported and maintained by their tense systems (though, admittedly, English is weird and gets it through the culture side of the language/culture system); the Hopi language/culture system has no such image of time, working from a different worldview principle that sees only cyclical, not linear, time -- round, not flat. Most of the world's grammars that have broken out of the Latinate mold show that the particular time/tense system of SAE is pretty much peculiar to Western European languages -- hence a linguistic/cultural ontology of "time" as we know and practice it, and not the supposed universal we have so fondly believed it to be.

    Side note: for one who reads Whorf closely, he makes five or ten times more universalist statements than relativity statements in his writings, yet he is seen (and reviled in a Chomskyan universalist attitude) as the relativist par excellent.

    So right from the get-go we see that 1) Whorf didn't write and wouldn't agree with the hypothesis that someone(s) named after him; 2) rather than being some deranged crackpot, he was merely literate: Whorf was one of the few interdisciplinary thinkers between physics and linguistics in this century; 3) Whorf's relativity principle had something important to do with Einstein's; and 4) Whorf was a universalist as well as a relativist -- he just had them in balance, a notable enough rarity in current academe, you must admit.

    If I may be so bold, alluding to your posting that "no one has disproved Whorf's mild version of linguistic relativity (let's leave ling. determinism aside, or the stronger version.)", even the mild version wasn't his! Trace back like I did and you will find that Whorf espoused neither strong nor weak versions of determinism, and relativity has nothing to do with determinism when you see it from the physics viewpoint above, as he did. If you read carefully, writers about the Whorf Hypothesis admit that even Whorf didn't hold a strong version of determinism (so if he didn't, who did? and if nobody did, why bring it up?), and that ALL the critics hold the weak version that they dreamed up (even though Whorf wouldn't hold it because it's at least weakly deterministic and therefore Newtonian). So what's going on? The problem is that Whorf had already, from his acquaintance with physics, moved from Newtonian monocausal determinism as an ideal into systems thinking -- where sometimes the opposite of one profound truth is another profound truth, where everything is INTERdependent, multicausal, interconnected: language shapes culture while culture is shaping language; language shapes thinking while thinking is shaping language. The cumulative effect of The (Humboldt/Boas/Sapir-)Whorf Hypothesis literature has been primarily to throw up a smokescreen around his ideas so that people, including grad students in linguistics, psychology, anthropology and sociology, won't read him in the original (English!). I tell my grad students that if they want to REALLY find out what their discipline is about, go find out who their discipline is beating up on and read them; and if you are so lucky as me to find someone that FOUR major academic disciplines are ganging up on -- you know you've hit a goldmine! What in the world could be so important that four academic disciplines create a combined smokescreen?

    Because so few linguists in this century have availed themselves of the changes in thinking about reality that physics has been broadcasting during this entire century, few linguists are even qualified to step into what they didn't realize was an interdisciplinary debate in the history of ideas which Whorf felt so comfortable in. I'll explain. I'll give you a synopsis of a talk I intend to give at a 100th Birthday Conference for Benjamin Whorf, which I intend to get funding for and hold in the Bay Area in Spring 1997. I call it "Heisenberg's Lament." You see, early in this century, that 'uncertain' Heisenberg was among the first to gain a 'glimpse' into the subatomic world; and, having done so, he rendered his opinion that, regarding the subatomic realm, "we have reached the limits of our language." He said this for two reasons: 1) no matter how glibly Western scientists talk about electrons, protons, neutrons, quarks, etc., when we look into that realm there are no 'things', only processes and relationships; but in order to make sense (i.e., complete sentences) in SAE languages, we need NOUNS -- and there's nothing in the subatomic realm that you can, except willy-nilly, attach nouns to. And 2) given that, our most fundamental scientific terms such as "same" and "different" are useless. He didn't know then, and physics doesn't know now, whether there are gazillions of electrons or just one Electron with gazillions of manifestations. We have reached the limits of our language. Fast-forward a few decades and Whorf hears this in his physics classes at Yale (he has unpublished manuscripts on gravity in the Yale Archives), and ponders: Hm, I wonder if this has anything to do with what Prof. Sapir said the other day about Hopi not needing nouns to express ordinary propositions -- just "rehpi", "flashed", instead of "it" or "the/a light flashed": because when you come right down to it, how is the flashing different from the light? Is "light" just a convenient grammatical fiction foisted upon us by SAE grammar? And as he pondered 'light' being noun or verb, particle or wave, depending on how it's viewed, he saw the universe in the same way, with different cultures taking different positions on the question. In this case, since Hopi didn't seem to take too seriously the absence of nouns, PERHAPS, Whorf surmised, Hopi could be of use for physicists in exploring and reporting back about the quantum world, that realm that didn't have thingy nouns.

    Fast-forward another few decades and physicist David Bohm reads Whorf (which I confirmed personally in talking to him), and then, *in response* (my attribution), writes _Wholeness and the Implicate Order_, in which he, among other things, tries to make English more verby and performative in the "rheomode" -- a brilliant flop; and then launches on the scientific community a view of the universe which does not contain our familiar notions of past/present/future time, but instead an implicate and explicate order of reality -- an "inny/outty" notion where the future is inside us working outward instead of some vague distant goal we are headed toward. An email acquaintance pointedly asked me what the difference was between Bohm's terminology and Whorf's terminology; it took me 6 months to finally answer that there was none, except the Hopis had had theirs for millennia longer. And then it hit me! Bohm, in his own maverick way, appropriated Whorf's answer to Heisenberg's Lament in "An American Indian Model of the Universe" and substituted more scientifically acceptable terminology (implicate/explicate rather than manifesting/manifested) to see how the notion of a universe without linear time would fly in the modern physics and academic community -- and it had qualified success. BUT -- Bohm was no closer to knowing whether Whorf had been accurate in his description of Hopi than he had been before writing the book. And there had been so much bad press on the guy! How was one to know, ultimately? And here's the part that almost no-one knows so far. In 1991, in the last few months of his life, David Bohm launched his most ambitious Thought Experiment to date: with some other physicists and a few psychologists and linguists, and sponsored by the Fetzer Institute, he enticed recognized American Indian intellectual leaders (and some of their elders) to join in Dialogue together in what I can only describe as a roundabout way of asking American Indians whether Whorf was accurate in his description of the 'timeless' Hopi worldview. But it became so much more! The American Indian leaders there had previously read Bohm's book and others, but the physicists knew nothing about Native American worldviews and Native American Science methodologies, so the Indians had to build a bridge over to them in private meetings before the three public days on the themes of time, space and language. During the day on time, Whorf's description of Hopi came up, was read out loud, and discussed, though I don't remember any Hopis being present; nevertheless, the other American Indians present, mostly of Algonquian tribes, gave what can only be called 'independent verification' in scientific terminology by saying essentially: well, I can't speak for the Hopi people, but that's pretty much the way WE do it. In fact, of the many Whorf passages read or discussed in these Dialogues, the physicists and the American Indians present were usually willing to give Whorf his points (proving again, perhaps, the difficulty of being a prophet in one's own country/discipline!).

    But we're not done yet -- the best is yet to come -- the actual conclusion of Heisenberg's Lament! At the beginning of the first Dialogue it was clear that the quantum physicists had their favorite realm to explore and talk about, and the American Indians also had their own favorite realm. As we Dialogued, it began becoming clear that those favorite realms had some fundamental principles in common: the only constant is flux; everything that exists vibrates; everything is interconnected such that the part implicates the whole. In fact, it became crystal clear that the last major obstacle to these realms being the same realm was really only terminological: the physicists are used to calling it "the subatomic realm" whereas the American Indians for millennia have been calling it "the spirit realm". Now that's a big enough surprise -- that modern physics is knocking on the door of spirit without really meaning to -- but not big enough, so now let's take it home! It puzzled the physicists just HOW the American Indians should have foreknowledge of a realm they shouldn't know about, that the Western scientific infrastructure had just recently led us to -- and the Indians had no such scientific infrastructure! And as the physicists gradually understood that, like Hopi, the Algonquian grammatical structures do not demand nouns, do not demand fictitious actors to embody actions (that, as my Mikmaq and Blackfoot friends tell me, they can talk all day long in those languages and never utter a single 'noun'!), they finally had to admit that such languages were indeed much better suited to exploring that realm and reporting back than SAE languages -- Whorf's reply to Heisenberg's Lament was verified and agreed upon. When the phenomena of reality change in a dramatic way, you need to change the tool you're using. Now, of course, the physicists were left with an even larger puzzle, to wit: how is it that these American Indians have a language much better suited than SAE languages to investigate and describe the inner workings of the subatomic realm -- a realm they aren't even supposed to know about!?!

    As you can see by now, Pinker -- like all other facile critics and unindicted co-creators of the so-called hypothesis -- is out of his league altogether in attempting to characterize a major player in one of the most important interdisciplinary discussions ever in the history of ideas. Pinker, like Chomsky, loves logic (which grows out of the grammar of SAE languages just as the philosophy of 'karma' grew out of the grammar of ancient Sanskrit, where it was used earliest as the linguistics term for 'direct object'!!), but has never really gotten with the program this century to replace binary/dualistic thinking with multivariable/ multicausal/ interdependent systems thinking.

    Whorf heard the call, way back then, and may yet prove to have been an entire century ahead of his time in linguistics. Even though we can think in systems for phonology and grammar, we have a tough time doing it for "language and thought"; we feel we have to make them bipolar opposites such that they are distinct and one CAUSES the other invariantly; that's why I so admired how Slobin finally lost those monolithic terms and framed the question instead in terms of "thinking for speaking" -- the at least ONE kind of thinking where your thinking is very much at the mercy of the forms and categories of your language, per Whorf.

    So tell your students, as I do, that the only way to get to the bottom of what Whorf did or did not say is to read his essays in _Language, Thought & Reality_ for themselves, perhaps with the above thoughts as a guideline, and then figure out for themselves whether the Sapir-Whorf Hypothesis smokescreen makes any sense.