Linguistic purity?

« previous post | next post »

"Purity" at xkcd:

Addendum by Simon Holloway at Davar Akher: "Of course, you can't see Linguistics because it's waaaay over to the right."

That's funny, but not it's really accurate. The interesting thing about linguistics is that it's spread out all along Randall Munroe's continuum: there's sociolinguistics, and psycholinguistics, and biological linguistics (as in the relevant aspects of neuroscience), and some applied physics (as in acoustics, vocal tract dynamics, and so on), and mathematical linguistics. There are other relevant dimensions as well, on which you could place historical linguistics, philosophy of language, stylistics, language engineering, and so on.

Chemical linguistics is a neglected area, however.

The unserious best that I do, for now, is an antique and untranslated quotation, from Lavoisier's Mémoire sur la nécessité de réformer et de perfectionner la nomenclature de la chimie:

Les langues n'ont pas seulement pour objet, comme on le croit communément, d'exprimer par des signes des idées et des images ; ce sont, de plus, de véritables méthodes analytiques, à l'aide desquelles nous procédons du connu à l'inconnu, et, jusqu'à un certain point, à la manière des mathématiciens ; essayons de développer cette idée.

L'algèbre est la méthode analytique par excellence ; elle a été imaginée pour faciliter les opérations de l'esprit, pour abréger la marche du raisonnement, pour resserrer dans un petit nombre de lignes ce qui aurait exigé un grand nombre de pages de discussions, enfin pour conduire d'une manière plus commode, plus prompte et plus sûre à la solution de questions très compliquées. Mais un instant de réflexion fait aisément apercevoir que l'algèbre est une véritable langue ; comme toutes les langues, elle a ses signes représentatifs, sa méthode, sa grammaire, s'il est permis de se servir de cette expression. Ainsi une méthode analytique est une langue ; une langue est une méthode analytique, et ces deux expressions sont, dans un certain sens, synonymes.

Cette vérité a été développée avec infiniment de justesse et de clarté dans la Logique de l'abbé de Condillac, ouvrage que les jeunes gens qui se destinent aux sciences ne sauraient trop lire et dont nous ne pouvons nous dispenser d'emprunter quelques idées. Il y a fait voir comment on pouvait traduire le langage algébrique en langage vulgaire et réciproquement ; comment la marche de l'esprit était la même dans les deux cas ; comment l'art de raisonner était l'art d'analyser.

Mais si les langues sont de véritables instruments que les hommes se sont formés pour faciliter les opérations de leur esprit, il est important que ces instruments soient les meilleurs qu'il est possible, et c'est travailler véritablement à l'avancement des sciences que de s'attacher à les perfectionner.



23 Comments

  1. Craig Russell said,

    June 13, 2008 @ 9:14 am

    Is my amateur translation better than Babelfish? I suspect we shall soon see:

    Languages do not just have as their purpose, the way we commonly think, to express things using signs and ideas and images; they are also true analytic methods, with whose help we move from the known to the unknown, and, to a certain degree, using the methods of mathematicians. Let us develop this notion.

    Algebra is the analytic method par excellence; it has been created to facilitate the operations of the mind, to shorten the path of reasoning, to reduce would would have required a large number of pages of discussion into a small number of lines, and finally to arrive at the solutions to very complicated questions in a more convenient, more quick and more certain manner. But a moment of thought makes it easy to see that algebra is a true language: like all languages, it has its representative signs, its method, its grammar—if one can use that expression. Therefore an analytic method is a language; a language is an analytic method, and these two terms are, in a certain sense, synonyms.

    This truth has been developed with infinite justice and clarity in the "Logic" of Father de Condillac, a work that young people who are headed toward the sciences could not read too much, and which we could not use too much to borrow ideas from for ourselves. Here, we are shown how we can translate the language of algebra into the common language, and vice versa, how the march of the mind is the same in both cases, how the art of reasoning is the art of language.

    But if languages are truly instruments that men have shaped for themselves in order to facilitate operations of their minds, it is important that these instruments be the best that they can be, which means to work towards the real advancement of the sciences that they have been attached to, and to perfect them.

  2. language hat said,

    June 13, 2008 @ 9:15 am

    That's funny, but not it's really accurate.

    I didn't even find it particularly funny, because it's so far from being accurate it's as if you were to say "Of course, you can't see social studies because it's waaaay over to the right." The cartoon, on the other hand, is funny 'cause it's true.

  3. Charles said,

    June 13, 2008 @ 10:57 am

    For lack of a defined field of chemical linguistics, I did a double major in Chemistry and French Linguistics.

    Certainly though, the usage of language in chemistry can be subject to significant analysis. Chemistry has some very interesting morphology and etymology, and a few syntactic and semantic quirks. It has comparative stylistics between languages, and barges in occasionally on phonetic territory (periodic acid anybody?).

  4. Sili said,

    June 13, 2008 @ 11:02 am

    Well – I don't know how people started thinking in terms of phonemes but doesn't it kinda make sense to call them "atoms of sound"?

    Larry Trask was an undergraduate chemist by the way.

    Perhaps IUPAC is the long sought after scientific prescriptivism. (I used to have a mean hand at naming molecules.)

  5. K. said,

    June 13, 2008 @ 11:11 am

    Rather than the study of I the terminology used in chemistry (which would really be "Chemistry linguistics") l'd have imagined "chemical linguistics" to be the study of the role chemicals play in language. You know, neurochemistry and such.

  6. Mark Liberman said,

    June 13, 2008 @ 11:20 am

    K: Rather than the study of I the terminology used in chemistry (which would really be "Chemistry linguistics") l'd have imagined "chemical linguistics" to be the study of the role chemicals play in language.

    Exactly. Unfortunately, this is an area of research for which the old expression "a much-needed gap" seems to have been created. Though, legend has it, Don Hindle and Bill Labov once carefully documented the effect on conversational style of CH3CH2OH.

  7. Philip Spaelti said,

    June 13, 2008 @ 12:00 pm

    Well for another (semi-serious) example you can try Gert Webelhuth's dissertation, which has a lengthy discussion of valence in atoms, and then draws analogies to concepts in syntax. I can't really remember the details. The whole thing seemed a stretch.

  8. Arnold Zwicky said,

    June 13, 2008 @ 12:27 pm

    For another discussion of possible conceptual analogies between linguistics and chemistry: "Linguistics as Chemistry: The Substance Theory of Semantic Primes", viewable at

    http://www.stanford.edu/~zwicky/linguistics-as-chemistry.pdf

    Warning: this is a very old paper of mine (35 years old!), so some of its theoretical assumptions are going to seem quaint to current readers.

  9. Arnold Zwicky said,

    June 13, 2008 @ 12:32 pm

    For another potential conceptual analogy between linguistics and chemistry: "Linguistics as Chemistry: The Substance Theory of Semantic Primes", viewable here:

    http://www.stanford.edu/~zwicky/linguistics-as-chemistry.pdf

    This is a very old paper of mine (35 years old!), so some of its theoretical assumptions might seem quaint to a modern reader.

  10. Andrew Dalke said,

    June 13, 2008 @ 1:26 pm

    If "chemical linguistics" is what I think it is, the field goes back to at least Eugene Garfield. His PhD thesis (1961) is titled "An Algorithm for Translating Chemical Names to Molecular Formulas" and is at http://www.garfield.library.upenn.edu/essays/v7p441y1984.pdf . It directly uses principles from linguistics. It's also an enjoyable read.

  11. mgh said,

    June 13, 2008 @ 4:48 pm

    Many organisms communicate by chemicals — whether they do so in "phrases" that can encode novel meanings dependent on context seems unlikely, but the field is young!

    (And I suspect real linguists would still resist calling it language)

  12. john riemann soong said,

    June 13, 2008 @ 5:30 pm

    "I didn't even find it particularly funny, because it's so far from being accurate it's as if you were to say "Of course, you can't see social studies because it's waaaay over to the right."

    I don't think the point is completely untenable. Mark's explanation does seem more attractive, but you can look at it from the perspective of the (possibly fallacious) idea that "math is another form of language, so mathematicians are really just applied linguists,"

    And at times it seems more like a circle, because when you explore our faculties for math and language (and symbolic thought in general), you come back to psychology and biology again.

    I remember reading a proposal somewhere that some of the /initial/ pressures for protohumans to develop abstraction may have been to compute the ballooning politics of human social relationships (as opposed to "make better tools," etc.) The idea may or may not have merit, and I can't remember who came up with the idea.

  13. john riemann soong said,

    June 13, 2008 @ 5:36 pm

    Oh by the way, is the "summative it" ("…it seems more like a circle…") in my third paragraph ambiguous? I'm suffering from author bias, but if it's unclear, I meant to say I'm referring to comic's idea of ordering fields of discipline in such a manner, as opposed to referring to Simon Holloway's remark.

  14. John Laviolette said,

    June 13, 2008 @ 9:21 pm

    @Language Hat: "I didn't even find it particularly funny, because it's so far from being accurate it's as if you were to say "Of course, you can't see social studies because it's waaaay over to the right." The cartoon, on the other hand, is funny 'cause it's true."

    The cartoon is funnier as-is than with linguistics or any other field because pure mathematics is truly and undeniably divorced from reality — and, more to the point, pure mathematicians are also portrayed as divorced from reality. Although the mathematician's comment *could* be interpreted as another example of one-upmanship, I think it's intended to be ironic, as if the mathematician just woke up from a math-trance and noticed these people arguing about something and said, "Hey! What's up?" The one person who could "win" the debate isn't even aware there was a debate.

  15. marie-lucie said,

    June 13, 2008 @ 9:24 pm

    @Craig Russell: Thank you for translating the French text. The final two paragraphs of are a little difficult to translate. I suggest the following amendments:

    This truth has been developed with infinite justice and clarity in the "Logic" of Father de Condillac, a work that young people who are headed toward scientific careers could do worse than reading again and again and from which we cannot refrain from borrowing a few ideas. In that work he has shown how the language of algebra can be translated into the common language and vice versa, how the steps followed by the mind are the same in both cases, how the art of reasoning is the art of analyzing.

    But if languages are truly instruments that men have shaped for themselves in order to facilitate the operations of their minds, it is important that these instruments be the best that can be, so that we truly work towards the advancement of science when we apply ourselves to perfect them.

  16. Kenny Easwaran said,

    June 13, 2008 @ 9:45 pm

    The Lavoisier quote looks like he's actually anticipating the ideas of an ideal language for science that people like Frege and Tarski (and of course all the logical positivists) were after between 1880 and 1930 or so. I always used to think that just went back to Leibniz and his idea of a perfect calculus for truth, but I suppose the ideas were wider spread than that.

  17. dawn said,

    June 13, 2008 @ 10:59 pm

    As a PhD candidate in rhetoric, I was thinking the same thing: rhetorics are along the continuum and beyond.

  18. Simon Holloway said,

    June 14, 2008 @ 1:02 am

    For the record, my comment was actually concluded with an "(Ahem)"; I was being deliberately facetious. Nonetheless, the point raised in this post is a good one: unlike many other fields, linguistics incorporates a variety of different sub-disciplines, each of which inhabits different places upon the 'purity spectrum'. While I might like to think of my own research as "pure" (in the XKCD sense), I know that it is not.

    Oh, to be more mathematically-minded…

  19. Faith said,

    June 14, 2008 @ 1:34 am

    And what is to the left of the panel? The sociologist is saying to a teacher, "education is just applied sociology." The teacher is turning to a librarian saying, "librarianship is just applied education." And the librarian (me) is saying to kid on a computer, "Google is just applied librarianship."

  20. Adrian Morgan said,

    June 14, 2008 @ 9:42 am

    As for physicists versus chemists, there's a rebuttal I've heard from chemists which goes something like, "The only element physicists understand is helium, which is so boring that chemists don't even bother with it". I think this is quite an amusing way to make the point that purer is not the same as better.

  21. Ran Ari-Gur said,

    June 14, 2008 @ 6:11 pm

    > The interesting thing about linguistics is that it's spread out all along Randall Munroe's continuum: […]

    That may be, but I don't think your examples demonstrate the point. Certainly chemistry applies math as well as physics, biology applies math and physics as well as chemistry, and so on. By the logic in the comic strip, then, I think linguistics is "just" applied sociology. (True, you can split linguistics into separate subfields, each being "just" applications of different fields mentioned by Munroe; but then, that's also true of most of said fields.)

    What would more definitively place linguistics all along the continuum would be a demonstration that the other fields are "just" applied linguistics; and I think one could make that claim for sociology, and perhaps even for psychology, but I don't think one could make it for any of the others.

  22. David Marjanović said,

    June 15, 2008 @ 10:07 am

    "The only element physicists understand is helium, which is so boring that chemists don't even bother with it"

    Be sure to hover over the original picture and read what appears.

  23. Alixtii said,

    June 22, 2008 @ 8:57 pm

    Most humanities run "along" the spectrum, I think–certainly philosophy does, since there is philosophy of history and philosophy of science and philosophy of math and philosophy of language . . . and at the extreme right epistemology and then, as the queen of the sciences, metaphysics.

RSS feed for comments on this post