Cobbinators and vallifractors

« previous post | next post »

Craig Russell, one of the commenters on my post "What sounds like a clearing of the throat", asked a question that deserves an answer:

Even though the consonantal *sound* in the middle is singular, is it really a sin (or even a mistake) to use the word "consonant" to refer to certain letters of the alphabet?

Craig went on to suggest that by implying people should use the term "consonant" for a sound type rather than a letter type I was just being a prescriptive pedant of the type I normally condemn.

Well, the short answer to his question is yes, it's a real mistake. But I'll give a longer one.

First, it is quite correct to point out that when David Owen referred to "the two central consonants" in the Scottish Gaelic word machair, he meant "the consonant-representing letters c and h". But that's putting sensible words into his mouth. What he actually says is that these letters are "represented by what sounds like a clearing of the throat", as if the sounds of speech were attempts to represent the clarity of letters by means of noises in the oro-naso-pharyngeal tract, and that is precisely the sort of unutterable dumbness in prevalent conceptions of language that I so often lament in these pages.

David Owen was explicitly trying to describe a sequence of sounds in what he wrote, and apparently the best he could do was to say that "two central consonants" in the spelling were "represented by what sounds like a clearing of the throat". I'm saying no one who wants to think clearly should be regarding speech sounds as attempts to represent letters, and the sound we're talking about here does not sound remotely like throat-clearing anyway. (Try saying Bach, and then simulate clearing your throat. Hear the difference?)

Now, to the point about calling the letters b, c, d, f, etc. "consonants". I say that is not a coherent use of terms by anyone's standards. The letter y represents a consonant in yes but not in eye; the letter w represents a consonant in wed but not in dew; the word unit begins with a consonant (and hence we say a unit rather than *an unit) but does not begin with a consonant letter; the word hour begins with a vowel (and hence we say an hour rather than *a hour) but begins with a consonant letter… Trying to classify the letters of the alphabet as consonants or vowels leads to incoherence, and in addition it doesn't have any application. There's nothing you can do with it. You might as well call half the letters cobbinators and the other half vallifractors. There's nothing you can say that makes those terms useful either.

An additional point is that in the post I was explaining in coherent terms to Language Log readers, who have at least some interest in such matters, what Owen was trying to say. That's not prescriptive: I wasn't primarily trying to lay down rules for Owen's future behavior (though it wouldn't hurt if he paid a little more attention to how to describe a foreign pronunciation to English-speaking readers); I was trying to explain to you what he must have meant, so I used the word "consonant" in a defensible and coherent way, to describe the speech sound that he was trying to identify.

One Language Log reader seemed to find me harder to understand than Owen, and said "I understood what the writer was trying to say with the pronunciation, but I don't fully understand your point." This is a bit depressing. I added to that comment what I hope is an even clearer explanation, which I will repeat with elaboration here. It proceeds by way of an analogy.

Without being a civil engineer, you nonetheless know how to say and understand "suspension bridge". A writer who described a suspension bridge as "a kind of long flat thingy hanging from a couple of massive straight-up pillars by means of a whole lot of huge cable thingies so cars could cross the river by driving on the flat bit" would be taken as a moron. He certainly wouldn't survive New Yorker editing. But when a writer who is seriously trying to describe the pronunciation of a word gives a much less accurate description than that, he gets published in The New Yorker. Something is wrong with this picture.

In general, David Owen writes beautifully and intelligently about golf and history and geology and crofting, yet on a simple point of phonetics he is content (it seems) to sound like a moron. Everybody treats language this way: they feel they don't need to know what anything is called; they just make stuff up. I say this indicates there is something wrong with our educational culture. Language is being treated as the subject nobody needs to actually know anything about. Language Log is here to take a stand against that view. Nuff said?


  1. Jens Fiederer said,

    April 19, 2009 @ 11:56 am

    I sympathize, but this is not a pain felt only by linguists. Once a term of art enters general use, specialists are free to moan about the details of that general use – but they moan impotently.

    Like it or not, there is a real – if meaningless – classification of letters as vowels or consonants – just google ""vowels in the alphabet" and you will get thousands of results.

    Back when I started in computer science, "hacker" used to mean a practitioner somebody steeped in the subtleties of the field. Among the old guard it still does. To everybody else it now means some sort of criminal. I moan in vain, and find your minor complaint amusing at best. Only when "linguist" enters the vernacular as a term for one skilled at oral sex can our agony even BEGIN to compare.

  2. George Amis said,

    April 19, 2009 @ 12:20 pm

    @Jens Fiederer

    About thirty years ago at UCSC, there was a faculty softball team called "The Cunning Linguists". Of course faculty in-jokes aren't exactly the vernacular, but . . .

  3. Chris Kern said,

    April 19, 2009 @ 12:24 pm

    I think Professor Pullum's point is more that when newspapers write about computers, they seem to take some time to check the facts — for instance, a journalist would be ridiculed for writing "Intel is hoping that its new product will capture the market for those little metal things inside the computers" or if a mouse were referred to as a "plastic doohicky with those clicky things" but that's essentially how linguistic-related issues are described in the news media. Technical terms are commonly used in news articles about science, technology, politics, and other such fields, but when it comes to language, they don't even try.

  4. Mike Albaugh said,

    April 19, 2009 @ 12:46 pm

    As a "hacker" in the original sense, I feel that pain. I am baffled by the notion that journalists actually bother to fact-check technology articles. They usually just lightly dust press-releases and ship them out.

    So, yes, "Technical terms are commonly used in news articles about science, technology, …", But they are not often used _correctly_, and that should count for something.

  5. marie-lucie said,

    April 19, 2009 @ 12:49 pm

    Jens, you sound as if "consonant" was a term recently come into general use, with a meaning which keeps changing. It is not, but the original "consonant" and "vowel" letters in Greek or Latin did refer unambiguously to specific sounds in those languages. The problem is that English pronunciation has changed quite a lot since the Latin letters were introduced (with suitable additions which unfortunately were later discarded) to write Old English, and English has also borrowed a large number of words from other languages, so that some "consonant" and especially "vowel" letters no longer correspond naturally to the sounds of the language. Modern English spelling being what it is, when the "consonant" and "vowel" letters do not correspond reliably to sounds, considerable confusion results: the word unit quoted by GKP begins with a vowel letter but a consonant sound, which is why the general rule "a before a consonant, an before a vowel" does not apply here if only letters are meant. If teachers are not aware of the differences between letters and sounds, they will try to come up with their own homemade "rules" for which they have no explanation, and which will differ from one teacher to another as each of them struggles with a possible explanation (or they may give up altogether, perhaps repeating a rule in the textbook without understanding it).

    Many non-linguists on this blog are saying that linguists should not be complaining any more than other scientists about the indifference or ignorance of the public. But no other specialty has more relevance to ordinary life, since all human beings talk, and in modern society they also write and read. No other specialty has such an obvious relevance to what all children are taught in school in the way of "language arts". In no other specialty are so many people convinced that their own intuitions, or whatever teachers taught them, must be right.

    If astronomers discovered that most children were being taught that the earth was flat and had four corners, while some teachers swore instead by the Ptolemaic system, or by their own homemade understanding of the universe, they would be right to complain about the poor training of teachers, but those ideas would not affect the children's daily lives. When linguists (trained to deal with all sorts of languages, not just their own) complain about teachers (and consequently the public) not knowing the difference between a sound and a letter (among other things), and therefore being unable to deal with the difficulties large numbers of children (and adults too) are having with reading and writing, something which does materially affect the lives of those children and adults, they are accused of wanting preferential treatment.

  6. marie-lucie said,

    April 19, 2009 @ 12:52 pm

    @Chris Kern: very well put.

  7. Craig Russell said,

    April 19, 2009 @ 1:18 pm

    Thank you for giving such a thorough answer to my inquiry. I can certainly sympathize with your distress–my field is Classics, which is (if possible) even less understood than Linguistics (for example: I imagine a substantial number of people, perhaps even here on LL, don't know what "Classics" is). Classicists see plenty of total bunk about our discipline in print as well: Greek words are called Latin words all the time, and vice versa, for example. And Classics was once the basis for all Western education! But I digress.

    It's probably going too far to suggest that a complaint such as the one you made is on the same level as other types of prescriptive pedantry. There is, I understand, a difference between elitism for its own sake, and frustration at the public inability or unwillingness to engage in or acquaint themselves with the research that is being done in a field that, nevertheless, they want to talk about. I get it.

    But I still stand by my original statement. (Let me make the caveat that I was only ever talking about the part where you criticized his description of "ch" as "two consonants"—I wasn't trying to defend his "clearing the throat" statement, although it seems to me to be a somewhat harmless bit of "dialect humor" along the lines of Mark Twain. It's not that the sound ACTUALLY sounds like clearing the throat; it's that it's not part of the sound set of most dialects of English. Part of the point of this article was to describe some of the local color of this part of Scotland, and a common way of describing an unfamiliar region is to highlight aspects of the local speech that will seem exotic to the reader. But, I will admit, this type of humor is inherently derogatory and condescending. Much of this in Twain now seems dangerously close to (if not beyond) crossing the line into racism, so maybe it's something we should be moving away from.)

    But anyway, the only part of Owen's statement I was trying to defend was his use of the word "consonant". I certainly accept that there is little or no *scholarly* value in using the word "consonant" to describe letters rather than sounds, and that a linguist who published a peer-reviewed article in which he used the term in this way would be rightly open to correction. The examples that Geoff gives above show how problematic identifying consonants and vowels only in terms of letters would be if it were the height of our understanding of phonology.

    But there is a long historical precedent for using the words "consonant" and "vowel" to describe the letters rather than the sounds. The OED entry on "consonant" shows that this is the older sense in which the word was used (although this is, admittedly, complicated by the fact that there was not originally such a gulf between spelling and pronunciation; to the Roman grammarians who invented the term, there would have been little difference), and this is certainly the way the word is more commonly understood today.

    We even use this distinction in contexts that have nothing to do with using the letters to pronounce things. From the above-mentioned OED entry (and 1823 example which seems to be talking about letters used to label part of a diagram):

    The vowels A E I O, are used to designate the solid angles; some of the consonants, B C D F G H, to designate the primary edges.

    So Owen's description of the "two central consonants" in the word "machair" to refer to the letters of the alphabet, rather than the (single) sound that they produce, still seems to me an acceptable use of the word "consonant" in non-academic English, with both historical precedent and regular usage in his favor. To claim that you *can't* use the word "consonant" to describe a letter, but only a sound, is, well, prescriptive rather than descriptive.

    (This said: I admit that this use of "consonant" to describe letters rather than sounds most likely does go hand in hand with a general lack of understanding that there is even a difference between the two, which was probably shared by, if not the writer, a majority of his readers. There's no question that people's linguistic knowledge is at a very low level compared to their knowledge of other topics. That's why I read Language Log!)

  8. Jens Fiederer said,

    April 19, 2009 @ 1:35 pm

    Marie-lucie: you sound as though you think I am actually disagreeing with our gracious host, or completely misunderstand the issue. This is not the case. You yourself contrast a "vowel letter" to "a consonant sound", effectively conceding the existence of a secondary (which I ADMIT seems vestigial and of dubious usefulness) definition that applies to the character rather than the sound.

    I DO (as I said) sympathize, and if it were up to me the whole concept of "vowel LETTER" would be consigned to the void. I was pointing out this is NOT going to happen.

    Chris Kern: I strongly suspect you feel this way only because you know something about linguistics. One's tendency to believe journalists bother to check their facts in some specific field varies inversely with one's familiarity with the field. This makes Professor Pullum's bias entirely understandable.

    To be fair, it is probably a bit more extreme in this particular field because while most journalists KNOW they are ignorant of technology they fancy themselves rather well versed in the field of at least their own language.

  9. Josh Millard said,

    April 19, 2009 @ 1:46 pm

    and the sound we're talking about here does not sound remotely like throat-clearing anyway.

    Rabbit-holing it just for kicks, but what if we imagine there's an additional layer of taboo-avoidance in all this, where Mr. Owen or his editor felt that "…represented by what sounds like a hocking up of a major loogie…" was a little too declasse?

  10. Mark A. Mandel said,

    April 19, 2009 @ 1:54 pm

    Jens: It's not just journalists, but almost everyone (ISTM) that "fancy themselves rather well versed in the field of at least their own language", because they speak it. I sometimes wonder if they feel that doctors are a waste of money because, after all, they've been living in their bodies all their lives.

    Geoff: Are "cobbinator" and "vallifractor" made up out of whole cloth? Google has no hits for the latter, but > 100 for "cobbinator", most of them apparently in somebody's on-line ID, and some in the name of a horse.

  11. Sridhar Ramesh said,

    April 19, 2009 @ 2:24 pm

    There is one and only one application of the knowledge of which letters have been designated "vowels" and which "consonants": to wit, preparation for play on Wheel of Fortune.

  12. hsknotes said,

    April 19, 2009 @ 2:27 pm

    "Everybody treats language this way: they feel they don't need to know what anything is called; they just make stuff up. I say this indicates there is something wrong with our educational culture."

    Point 1: As has been sated above by others and in numerous posts, the 'linguists' are not the only aggrieved party here. There's many, many things people all use on a daily basis that they can only discuss in a very basic way. (Think about discussion of health and the body.) The assumption that newspaper editors will defer to 'technical experts' in 'technical cases' hardly always holds up. I would think the linguists would be in a much better position, because I imagine many cases for dealing with descriptions of language, pronunciation, spelling, etc, would have an entry in the publication's style manual. In the case of other technical matters, I can only assume if an expert is not consulted, common knowledge reigns supreme.

    Point 2. This is an interesting topic in regards to the failures of educational system. I'd say the modern problem stems from the death of grammar instruction over the course of 60s, 70s and 80s. From what I can tell, the results of some studies led the general consensus to become, teaching grammar, teaching something close to linguistics, didn't produced better speakers or writers. The aftermath left us with a nation of people who can't diagram a sentence, who has to cling to 'bibles' like Strunk and White, and can't talk about language in a serious manner. What's interesting is, no one besides the prescriptivists (the naysayers, really) and the anti-prescriptivisits really care. People have only ever had dictionaries each having a unique 'key' to pronunciation on every page and the new york times 'phonetically spelling out words' or using whatever method to indicate pronunciation. We don't have use the IPA or any other system and I don't think we ever will. Words like 'coordinate conjunction', not to mention 'fricative', are as foreign to many people as the coxal bone, or even the fibula. They may have heard of the names, or have an idea, or maybe once used to know, but they wouldn't want to be tested.

    I think the question is less about our 'educational culture', but more about the general culture. Does the society care if they can speak about language in more educated terms? No. Do they necessarily think they should consults language experts on issues of language (the way they consult an expert on science to discuss scientific issues?) Let's say no for the sake of argument. Ok, so why should they? The argument has to come from two directions, one, the 'eyesore' argument, "those people in the real world discussing our field are saying things wrong and sound like a bunch of idiots," or two, the 'prescriptivist' argument, "what they are saying is not only wrong, but leads to 'confusion' and 'problems'. Well, good luck with the 'eyesore argument. As for the 'prescriptivist' case, I don't think people have any trouble understanding 'inelegant' pronunciation schemes. They have trouble understanding bad or 'half-joking' lines, like the 'machir' example. Once someone says, 'Bach', the problem disappear, and the calls for IPA and linguist's expert advice with it. The issue in this case seems to be the editor giving leeway in the interest of 'rhetorical' or something.

    consonant, for the non-expert American speaker, means those letters of the alphabet that aren't a, e, i, o, u, and sometimes y. Some people have a vague feeling that this is an issue (note the 'sometimes y' case), but how many times do people find themselves actually having to discuss consonants or linguistic issues in general? It depends on you and your circle's sensitivity. If all you have is a hammer, everything looks like nails. If you're real busy, and you've got a bunch of tools, maybe you don't care about language issues.

  13. Jens Fiederer said,

    April 19, 2009 @ 2:38 pm

    Sridhar Ramesh: Cute, although there ARE other games in which the distinction is important. But because these tend to be of a pedagogical nature, that probably makes them EXTRA irritating.


    which, in my case, started you right out with the task of finding the "odd man" among: N U Y :-)

  14. J. W. Brewer said,

    April 19, 2009 @ 3:22 pm

    Is there a standard dictionary of American English (or alternative descriptive reference source for contemporary American English) which does not give both the letter meaning and the phoneme meaning for the word "consonant"? The one I have closest to hand (American Heritage with the Calvert Watkins stuff about Indo-European that I read obsessively in junior high school, thus probably playing a causal role in my subsequently majoring in linguistics . . .) certainly gives both. So the word "consonant" has multiple meanings, which can lead to less than perfect clarity in particular contexts. So what? Natural languages are like that. If that's not to your taste, become a prescriptivist or an Esperantist or something like that.

    What the sentence could have used was a word that unambiguously meant in context "a sequence of two letters used orthographically to represent a single phoneme, as for example the English ch, sh, th (ok, two different phonemes) or zh." But what word would that be? It took me a moment or two to get "digraph" out of off-site storage up to the tip of my tongue, and even then I felt obligated to double-check with wikipedia to make sure it was actually the right word for that as opposed to just meaning that calligraphic/typographic thing where you have the A and E running into each other when writing "Caesar." I would not expect the median New Yorker reader to have a clue what "digraph" means. (Although I agree with the commenter in the prior thread who thought that some meaningful percentage of New Yorker readers might have understood an allusion to the Yiddishism "macher.")

  15. JLR said,

    April 19, 2009 @ 3:38 pm

    … the original "consonant" and "vowel" letters in Greek or Latin did refer unambiguously to specific sounds in those languages.

    Except in Latin, when they didn't. Curse those foolish Romans for using 'i' and 'v' as both vowels and consonants. They knew nothing of your newfangled 'j' and 'u'.

  16. J. W. Brewer said,

    April 19, 2009 @ 5:14 pm

    I'm more baffled the more I re-read. Describing any of the 26 letters of the alphabet standardly used for writing English into "consonants" and "vowels" is not "coherent" because at least three of the 26 cannot be fit neatly into the schema? Why does the lexicon of a natural language need to meet some abstract criterion of logic, rationality, or coherence? I agree the "throat-clearing" half of the sentence (which may have caused overreaction to the use of one of the standard meanings of "consonant"?) is descriptively inaccurate or useless — unless, as a previous commenter suggested, it was meant humorously. I think it's an old vaudeville-type line among speakers of standard High German to say that Swiss German is [German words meaning] "not a dialect, but a throat condition," which might be a parallel instance that would be maddeningly wrong if taken as serious descriptive phonetics rather than, you know, humor.

    I'm equally puzzled by Prof. Pullum's level of outrage at the notion of "speech sounds as attempts to represent letters." I think he may be importing a perhaps inappropriately technical understanding of "represent" or "representation" into a loose journalistic context. It seems axiomatic (perhaps a bit of an unexamined prejudice?) in modern linguistics to think that the spoken language is the real thing, and that the way in which a particular language is conventionally written is either a mere curiosity or an unhelpful and misleading distraction. But both in our native language and especially in foreign languages we learn as adults (in the modern age of mass literacy), we acquire new vocabulary by eye as well as by ear — if we are reading aloud a text with an unfamiliar word we will pronounce it as best we can using our knowledge of the phonetics/phonology/orthography of the particular language, and saying that this means figuring out what sounds represent the written letters rather than what sounds the letters represent doesn't strike me as particularly outrageous on the scale of journalistic imprecision.

    (I should perhaps note I haven't read the article, because Gaelic interests me, but golf bores me to tears — during a vacation in Scotland about 15 years ago my traveling companion and I were treated as suspicious characters by the front desk staff at a hotel in St. Andrews because we were checking in without any golf clubs in our possession.)

  17. Dan T. said,

    April 19, 2009 @ 5:19 pm

    If you're on Wheel of Fortune and want to buy a vowel, what you get is a letter, not a sound.

  18. Ellen said,

    April 19, 2009 @ 5:57 pm

    I think the problem is not so much simply using "consonant" for the letters, but doing so and then saying that the speech sound represents those consonant letters, which is backwards.

  19. Alissa said,

    April 19, 2009 @ 10:17 pm

    I think there are two problems that have the same cause. People think they know things about language because they use language all the time, and linguists think that people need to know (technical) things about language because people use language all the time.

    I admit the ignorance about linguistics is really frustrating. I am a linguistics major, and more than once, after saying something about language, someone has told me "I just don't think so." I always want to ask them if they would feel so knowledgeable if I had been talking about physics instead. I also think that some things from linguistics can be helpful in other contexts, especially English class where nothing I was taught about grammar made much sense at all. I think learning the IPA symbols would be useful, but for most people, knowing what 'fricative' or 'approximant' means would be pretty much pointless.

    I don't think that the author's possible confusion in the context of an article about golf is harmful. It would be different if we were talking about teaching children to read, but I'm sure most people would be as annoyed by a requirement to talk in technical terms about language at all times as I was in a junior high science class when our teacher said it was wrong to talk about our weight and that we should talk about our mass instead.

  20. D.O. said,

    April 20, 2009 @ 12:10 am

    Alissa, I do not know what exactly your science teacher told you, but yes it is wrong to talk about weight if what you mean is actually mass in physics class. Chemists are not interested and i am not sure about generic science class. Part of high school physics education is to explain that the two are different and that you should pay attention. Of course, you are free to ignore the distinction in other contexts. But part of the school education is to change a way you are doing things and thinking about them (another part is to give your parents time to work).

  21. Alex Fink said,

    April 20, 2009 @ 12:30 am

    Yes, there's quite a gulf between an orthography in which each given letter either always denotes a consonant or always denotes a vowel, and "incoherence". English, though it must be among the four least regular phonologically-based orthographies out there, is still much closer to the former than hypothetical orthographies where every word is spelled with a random sequence of letters. Accordingly there are uses for terms like "cobbinator" and "vallifractor", even if their utility might be confined to the writing system and irrelevant to a phonological description (and this is independent of what one happens to think about the primacy of speech in language).

    For example, without recourse to such terms, I don't know how I'd state the (exceptionful, but still useful) rule that, in spelling the past of English (monosyllabic) verbs, verbs whose base form ends in VC double the final C before the suffix "-ed" in their spelling ("stop" -> "stopped", "sin" -> "sinned") whereas ones that end in VCC don't ("jump" -> "jumped", "climb" -> "climbed", "pick" -> "picked"). And there are several other orthographic rules like this.

    For another example, one of more than just descriptive use, B. V. Sukhotin observes that vallifractors tend to occur next to cobbinators more often than to their own kind, and that this is of use in decrypting a simple substitution cipher. A class of cyphertext letters with vallifractorial behaviour in this sense, and in practice quite likely to be the vallifractors, can be picked out via Sukhotin's algorithm from the bigram frequency matrix alone. Sukhotin wrote in Russian, but Jacques Guy observed the same to work in English. And the algorithm is quite a handy tool to have if one comes across an unknown script for an unknown language and suspects it to be an alphabet.

  22. Alissa said,

    April 20, 2009 @ 1:40 am

    I guess that example didn't work very well. I don't think my teacher was actually being serious; he was just illustrating a point. What I meant was that while being aware of things is important (especially being aware that one doesn't know everything), it would be ridiculous to require people to use words in their technical senses all the time and using words in a looser sense isn't necessarily harmful to an understanding of the world.

  23. A Mueller said,

    April 20, 2009 @ 3:10 am

    I'm pretty sure the fate of most linguists, at least in the context of the English language, is to be terminally misunderstood because they (we) are trained to think about language in ways no one else is.
    Modern culture, perhaps due to the volume of written material available for consumption, is a highly literate one. Literacy = education, and letters = literacy. Unfortunately, letters, as has been stated in previous comments, do not map one-to-one to phonemes anymore (and even the vaunted Greek letters didn't always, either – two gammas in a row represented what, in IPA, would be a engma).
    And we don't, as a culture, teach our students to think in phonemes, anyway. I didn't learn a thing about them until my first year of undergrad, though I had, in the past, realized that F and V represented sounds that had rather a lot in common, or that the sound in question – the uvular trill – felt and sounded funny because it didn't exist in my language. Most people are expected to ACQUIRE their spoken language. LEARNING the written language is considered a separate task (which in English makes sense, considering the rift between pronunciation and orthography).
    And while, as a linguist, I have to sigh a bit, myself, I am rather accustomed to people not thinking in the ways I have been trained to. To the general populace, letters are sounds, and The New Yorker is designed for the general populace (okay, maybe a relatively snooty subset of the general populace…). "Consonant" means a letter to most people. We're used to our 26 letters, not the representation of out entire phonemic inventory.
    And the computer scientists can pat themselves on the back and let us poor little linguists complain a little because THEY'LL always have JOBS, unlike us.

  24. joseph palmer said,

    April 20, 2009 @ 8:21 am

    ESL teachers are trained in phonetics. The results of this are not very impressive, to my mind. Most of them will respond better to articles with non-technical descriptions based on things like throat-clearing.

    Of course, school children will on the contrary lap such instruction up.

    By the way, does Language Log officially endorse lamentations about the decline of grammar teaching, or does it hold that it was done all wrong by wicked prescriptivists in any case?

  25. Dan T. said,

    April 20, 2009 @ 8:50 am

    People use computers a lot these days, but tend to be abysmally ignorant of technical things regarding them, and misuse computer terms; this is frustrating to geeks like me. Language is hardly the only field that gets mangled in the popular conception.

  26. Franz Bebop said,

    April 20, 2009 @ 10:27 am

    Originally written by the staff of The Onion®

    Vowels to Bosnia

    Cities of Sjlbvdnzv, Grzny to Be First Recipients

    Before an emergency joint session of Congress yesterday, President Clinton announced US plans to deploy over 75,000 vowels to the war-torn region of Bosnia. The deployment, the largest of its kind in American history, will provide the region with the critically needed letters A,E,I,O,U, and Y, and is hoped to render countless Bosnian words more pronounceable.

  27. Jonathan Lundell said,

    April 20, 2009 @ 11:11 am

    [There's an analogy between IPA and Unicode that I don't have the energy to work out, and it's not all that relevant to my question; please consider it made.]

    I'd like to make more use of IPA (for example, in a recent post I did on NPR Names), but I don't for at least three reasons.

    1. I didn't discover IPA until late in life, I need a primer, and I haven't found one. Good descriptions, sure, but not a set of lessons on how to apply IPA.

    2. IPA is a PITA to type. (And ASCII IPA sucks.)

    3. My audience won't understand it, but they will understand my lame attempts at phonetic interpretation.

    Among other things, we have a bit of a chicken and egg problem. Maybe we need universal IPA education….

    Oh, my question: where can I find a good IPA primer?

  28. J. W. Brewer said,

    April 20, 2009 @ 12:20 pm

    For an IPA primer, I've heard good things about a book called Phonetic Symbol Guide, co-authored by one G.K. Pullum, who may be too modest to use his many media and blogospheric appearances to flack for his own wares. It's certainly more popularly priced than the CGEL (although Amazon now has the latter marked down from $195 to a low low $161.42 for those who can't wait for the twenty-dollar trade paperback).

  29. Jonathan Lundell said,

    April 20, 2009 @ 12:58 pm

    @JWB, thanks, and I now have a copy on order (hey, if I could spring for CGEL, how could I pass up this $18.25 bargain?). It's described more as a reference than a primer, but perhaps it'll do dual duty. (Hmm. du-du-du. I think I pronounce do & duty the same, dual differently, or at least with a different tongue position. Can I say that in IPA?)

  30. marie-lucie said,

    April 20, 2009 @ 1:36 pm

    ESL teachers are trained in phonetics. The results of this are not very impressive, to my mind.

    Not all ESL teachers receive the same training. Some universities require two years of study to get a masters, some private outfits have one-week programs.

  31. marie-lucie said,

    April 20, 2009 @ 1:40 pm

    I'm sure most people would be as annoyed by a requirement to talk in technical terms about language at all times as I was in a junior high science class when our teacher said it was wrong to talk about our weight and that we should talk about our mass instead.

    A Canadian "reader" (a textbook used in language arts classes) includes a story about a woman who, among other things, "had a mass of 96 kilograms" – obviously a translation of "weighed 200 pounds" in order to conform with "scientific correctness" pushed to a ridiculous extent.

  32. Bryn LaFollette said,

    April 20, 2009 @ 4:38 pm

    @Jonathan Lundell
    For a great, full explanation of IPA and transcription (from Phonemic to as phonetic as one might reasonably get), I'd recommend A Course In Phonetics by the late great Peter Ladefoged. This was in fact the very text book we used in my introductory Phonetics class which taught by none other than Geoffrey Pullum! You can pick it up from Amazon among other places.

    But, any good introductory text book in Linguistics will include a section Phonetics, Phonology and transcription, if not using IPA, using something close enough to it to get the idea.

    And as for your question about the pronunciation of "do", "duty" and "dual". My pronunciation isn't quite the same as yours in this respect, I think, but yes, you can represent your pronunciation (as I would guess at it) with IPA with the following transcriptions: "do" [du:], "duty" [du:ɾi] and "dual" [dju:l]. The [j] represents a palatal approximant, which is represented by "j" in IPA rather than "y" (as is more common in English orthography otherwise) due to the winning out of the Germanic orthographic tradition. Hope you find this illuminating!

  33. Mark F. said,

    April 20, 2009 @ 6:01 pm

    Earlier I commented that people are used to "ch" representing a single consonant sound, but my claim was disputed so I tried to check it out. I asked a friend whether he thought the "ch" sound was one consonant sound or two, and he said "It depends on which language. For instance, in Hebrew there's one letter for that sound." I tried to press him that I meant consonant sounds, not letters, but it didn't really help. I asked him how many consonant sounds (as opposed to letters) he heard at the beginning of "scrape," and he thought for a while and said "one". Not what I expected, but then I really gave him no definition of what I meant by "one sound" as opposed to a succession of sounds.

    So this is an experiment with an n of 1, but it tends to refute my confident assumption that David Owen on some level must have recognized "ch" as a pair of letters representing a single phoneme.

    I still don't think the case has decisively been made that several years of exposure to the IPA (which is what it would take to keep people from forgetting it) should be a part of everyone's education. I'm not even sure what the ground rules for such a debate would be. I will say that an ignorance of linguistics doesn't seem to stand in the way of good writing.

  34. Bryn LaFollette said,

    April 20, 2009 @ 9:11 pm

    @Mark F.

    I think the more telling thing in this case, at least to my mind, is the "direction" of representation as described by Mr. Owen. He specifically describes the sounds as be a representation of the letters, not the other way around. This, I think, goes to the heart of what I find to be the most profound disconnect between the general public and the linguistically trained; that the written language and letters and spelling is the "truer" representation of the language and that the spoken form is just an attempt to represent the spelling with sound. I don't say this lightly and I in fact believe that J. W. Brewer's assertion that:

    It seems axiomatic (perhaps a bit of an unexamined prejudice?) in modern linguistics to think that the spoken language is the real thing, and that the way in which a particular language is conventionally written is either a mere curiosity or an unhelpful and misleading distraction.

    Is in fact false on several levels. First of all, among linguists, this is not an axiomatic assumption nor is it unexamined. Any peek into an introductory text on modern linguistics will give the argument for this, as will many prior discussions on Language Log. True, a good deal of vocabulary is acquired from written sources (especially for adults), at least among those with sufficient interest in school or reading, but that is a looooong way off from claiming that its primacy eclipses the spoken language. Learning of vocabulary is not the same thing as the native acquisition which occurs by age two or three, well before most all children learn to read to the point that they're acquiring new vocabulary, at least certainly not to the degree of acquiring more through reading than by spoken communication. It's also for well documented biologically grounded reasons that adult learning is a different beast than pre-critical period language acquisition.

    That said, the fact is that this idea that speech is based on spelling, not vice versa, is widespread and wrong. But most people are not familiar enough with the arguments that demonstrate (conclusively, I think) that no matter how abstract the relation, all orthographies of spoken languages are inherently representative of sounds, and the writing systems of languages such as Egyptian and Classical Mayan went undeciphered for centuries because this fact was not heeded.

    Now, beyond that issue is whether it is reasonable to classify letters into Consonants and Vowels. The fact is, whether reasonable or not, this is done throughout elementary through secondary education (at least, in the United States). While I think it would be a great benefit to have better exposure to actual linguistics at these levels of education, if only in so cursory an exposure as to teach the difference between written sumbols and the sounds represented by those symbols (or combinations thereof), I'm not holding my for it to happen very widely any time soon. In my experience the average person isn't aware of the differentiation of sounds from letters, and certainly unless they've spent some time really thinking about it do they even realize that digraphs in English, like 'th' or 'sh' represent single sounds. But I believe this ignorance is really more benign than the belief that speech represents letters, not the other way around. This, however, is only my personal sticking point, and I believe that Geoff is correct in his addressing of his own beef with this author on both of these points.

  35. Dan T. said,

    April 21, 2009 @ 8:36 am

    The "bookish" subset of the public is likely to have some words in their vocabulary that they have only read and not heard, and may be unsure how to pronounce (and pronounce them wrong on the occasion they happen to speak them).

  36. David L. said,

    April 21, 2009 @ 11:54 am

    It seems quite clear that Owen's uses of both "consonant" and "represent", though non-standard within linguistics, are perfectly coherent and common uses in English. "Consonant" as a term for letters, rather than sounds, has already been well-defended in the comments. I would add to its uses in mathematics by noting Viète used vowel letters as variables and consonant letters as constants in algebraic equations before 1600.

    However, as far as I can tell, no one has brought up that GKP is misinterpreting the nature of the representation in the passage from Owen. GKP reveals one mistake most clearly when he says that Owen "referred to 'the two central consonants' in the Scottish Gaelic word machair". Owen did no such thing. Instead, Owen referred to the two central consonants in the text string "mocker", which Owen had presented as a roadmap to the pronunciation of the Gaelic word. In this specific context, written language clearly is being written as a guide to spoken language.

    Now, in this context, the job of Owen's reader is to reconstruct–to build a representation of–the written form "mocker" in spoken sounds, and he's telling us that the representation of "ck" (not "ch") should sound like gargling or whatever (I agree that his characterization of the sound is at least ignorant, and perhaps borders on racist. But that really is beside the point).

    While it is certainly the case that there is a general misconception outside linguistics (where I live) that spoken sounds are used to represent written letters, I don't think Owen is participating in that misconception in the passage. He's constructing, locally, a written form which really is supposed to be used as a template for spoken sounds. So in the local context of the description of this word, the claim that spoken language is a representation of written language is simply correct.

    The general lament regarding the ignorance of linguistics in society at large is spot on: the letter meaning of "vowel" and "consonant" obscures their sound meaning, and indeed many people (plausibly even David Owen) really do think that spoken language really is a representation of 'true' written language. Unfortunately, this particular example just doesn't make any of those points, nor derive from those errors.

    @Jens Fiederer

    By the way, I believe the day when linguists can make this complaint is soon approaching, if not actually arrived. The hip hop group CunninLynguists released their first album way back in 2001.

  37. marie-lucie said,

    April 21, 2009 @ 12:16 pm

    Interpretation of digraphs as two consonants:

    One of my students was sure that the word together was a compound: to-get-her.

  38. Bryn LaFollette said,

    April 21, 2009 @ 3:06 pm

    @David L.

    I definitely appreciate your interpretation, and you make a good point. But, in carefully looking at the crucial line from the article "…but with the two central consonants represented by what sounds like…", I think at best the antecedant for "the two central consonants" is ambiguous. It could refer to either "ck" in his pronunciation guide or "ch" in the Gaelic word he cites. So, I think both yours and Geoff's explanation/critique could be seen as equally valid given the context.

  39. David L. said,

    April 21, 2009 @ 4:51 pm

    @ Bryn LaFollette

    I agree that the original statement is somewhat ambiguous with regard to "the two central consonants". However, I'm not ready to say that the two interpretations are equally good. If Owen is really referring to the "ch" in "machair", then the clause has no effect on how "mocker" is pronounced (the entity the phrase appears to be altering). It would be like saying that something was pronounced like "china", more or less, but with all the "v"'s in "variance" replaced by "d"'s. If that is what Owen means to be saying, then it's certainly true that Owen isn't saying anything coherent.

    That said, you're quite right that the original isn't very clear, and GKP's interpretation is not impossible. Really, I just want to suggest that there is a charitable and reasonable interpretation in which the only error Owen is making is to think that the Gaelic "ch" sounds like throat-clearing.

  40. J. W. Brewer said,

    April 21, 2009 @ 10:55 pm

    @ Br. La F. I'm not sure I fundamentally disagree with the consensus notion of the primacy of the spoken language and thus the "direction" of the representation. Maybe where we really disagree is as to whether the direction of representation matters in context, i.e. whether that's a level of understanding or precision one should expect a journalist who's primarily writing about golf courses to care about. I find it qualitatively different from a journalist who talks about, e.g., the "passive tense." My sense from observing my young daughters' teachers is that both directions are used in teaching reading: "what sound does F make" alternates freely with "what letter sounds like fa-fa-fa-fa-fa." I have refrained from pulling rank and saying "I have a B.A. in Linguistics: you're doing it all wrong."

    All of that said, I find it hard to disentangle literacy and all that it entails from my sense of language and linguistic competence. It's like trying to disentangle language from consciousness more generally and think about whether or not your internal thought process uses your native language or is somehow pre-linguistic. You know how Wittgenstein says that if a lion could speak, we wouldn't know what the heck he was talking about? I think that I really have no frickin' idea what it would be like to be a competent adult speaker of a language (and only of that language) that had no written form — it's just not something that I can imagine. I wonder how many of us outside the subset of linguists who have done substantial fieldwork in non-literate speech communities (but who probably messed things up by scribbling everything down in IPA to the bafflement of their native informants . . .) can really comprehend what it would be like to be illiterate and have no real notion of the concept of literacy. I actually took a class as an undergrad (Dep't of Anthropology, but usable as credit for linguistics major) with a guy who'd done such fieldwork, but I must say the sense of what it would be like to live/think/be like that didn't really rub off on me.

  41. Andy said,

    April 22, 2009 @ 3:37 pm

    "In my experience the average person isn't aware of the differentiation of sounds from letters, and certainly unless they've spent some time really thinking about it do they even realize that digraphs in English, like 'th' or 'sh' represent single sounds. But I believe this ignorance is really more benign than the belief that speech represents letters, not the other way around."

    I believe that this "ignorance" of the relationship between sounds and letters is a primary contributor to the relatively poor literacy rates in the United States, and is therefore not benign at all.

  42. Taylor B said,

    September 4, 2012 @ 3:35 pm

    @J. W. Brewer
    A phonetic symbol guide as a primer to the IPA? That's like having a "Handbook of Statistical Tables" as a primer for a statistics course. ?. A good primer to the IPA would be almost any Phonetics or Phonology chapter from any decent first year linguistics textbook, probably not an encyclopaedic reference.

    Personally, people who think sounds = letters, ?. What do you say to these people? It may be just as atrocious as thinking math = numbers, but at least people will readily believe you when you tell them that math is greater in conceptual scope than just numbers. A good number of people, for some reason, don't like to think of sounds as something independent of letters.

    The "?." above is the grammatically complete sentence or phrase that means that I am giving a wtf look.

RSS feed for comments on this post