I recently wrote about the French roots of most older (pre-Second-World-War) international organizations that have the word international (not world or universal) in their names, and hence I in their initialisms or acronyms. But in none of them is the Frenchness as pronounced as in the IPA.

IPA stands for both the International Phonetic Association and the International Phonetic Alphabet; the former is the body responsible for the latter. (Both are API in French.) It seems strange that an organization would knowingly — and, at least to some, confusingly — use the same abbreviated name for itself and for its product. The Ford Motor Company is, to be sure, known briefly as Ford, and one of its cars (if it comes from the Ford Division) is a Ford; while a specific one may be called the Ford (if, for example, a member of a multi-car family announces “I’m driving the Ford today”). But in French the company would be called la Ford, as would the car if a specific one is meant (both are feminine because both compagnie and voiture are feminine). French-speakers seem to be used to such ambiguities and find various ways to resolve them. For example, la Corse means both ‘Corsica’ and ‘the Corsican woman’; Balzac, in his story La Vendetta, refers to the Corsican woman at its center as l’Italienne.

Here I will  refer to the association (when necessary) as IPAss, and limit the use of IPA to the “alphabet.”  I have often toyed with calling it the FPA, because of its French bias and because it doesn’t really make sense to refer to phonetic characters as international — they have to do with languages, not nations. But it can also be questioned whether it’s really phonetic, or really an alphabet.

Let’s start with the last. To most people, an alphabet is a finite set of characters (called letters) that one learns in a specific order; hence the term alphabetic order. Whether modified letters (like á or ç) or digraphs (like æ or ch) constitute distinct entries is a matter decided by the individual language authorities. The IPA, however, is not finite — phoneticians create new characters as needed — and there is no alphabetic order. If it is anything, it’s a code, not an alphabet.

The IPA was first conceived by a French organization of modern-language teachers (L’Association Phonétique des Professeurs de Langues Vivantes) as a way of teaching pronunciation to children, not as a true representation of native pronunciation. This meant giving French approximations to, say, English sounds. Thus the character c (later replaced by ʃ) was proposed for both the French ch in chaud and the English sh in show. These are actually quite different sounds; in the English one the tongue is considerably farther back than in the French one. But when show is used in French as a loanword, it’s pronounced just like chaud; in fact, in IPA the pronunciations of show (as said by a Scot or a West Indian) and chaud would be written the same way: [ʃo], regardless of how different they actually sound.

The real problem with the IPA, for me, has always been the representation of affricates. as, for example, the English sounds represented by ch and j or ge in such words as judge, charge and change. Note that these words all come from  French, and at the time they were borrowed their pronunciation of these consonants in French (Old French, that is) was not that different from what it is in English. But French lost its affricates as it changed from Old to Middle, and when speakers of modern French borrow words from languages that have them, they pronounce these sounds as stop + fricative. That is, a French-speaker would approximate the English phrase ‘catch it’ as ‘cat shit’; one need only  listen to native francophones pronouncing  such words as tchèque, jazz, pizza or tsar. In IPA these are written as [tʃɛk], [dʒaz], [pidza] and [tsar].

But what happened was that the IPAss decided to keep these representations for all languages, even those in which a fricative is a single consonant (in my native language, Polish, there are six such sounds). It was only on the insistence of some non-French linguists that ligatures (such as ʦ, ʧ, ʤ) or over- or under-bows (such as t͡s or t͜s) were allowed “when necessary,” but they are still usually omitted from IPA charts. Many linguists resort to non-IPA symbols, such as č for the ch sound.

So much, then, for “international,” “phonetic” and “alphabet.”


2 Responses to “IPA”

  1. Patricio Moschcovich Says:

    Hi Coby,

    I always enjoy your blog posts.

    Just wanted to say something which indirectly relates to this post.

    You said that “To most people, an alphabet is a finite set of characters (called letters)”

    In India, they commonly call letters “alphabets”.

    Take a look at this question just as an example as to how confusing this is for people (outside of India) who don’t know that:

    Keep up the good work.


  2. jontxe Says:

    Hi, Coby.

    I agree with you that the IPA can sometimes be frustrating. My prime example is that until recently there was no symbol for a central low vowel. The prime candidate, [a], had been taken for a front low vowel which are very rare in language. All because of the desire to represent vowels in a parallelogram, I guess. Now they have come up with [ä] for this sound. Ridiculous. Most languages of the world have this sound. They should let it have the letter [a] and leave [ä] or something else for the front one.

    The representation of affricates with “double” symbols has always bothered me too, if nothing else because they are hard to write. I don’t think the symbol is the place to indicate that a single sound has properties of both stops and fricatives. At least now we have single symbols for them in unicode which do not look so much like digraphs sincere there is no separation between the parts, e.g. [ʤ ʣ ʧ]. Unfortunately, not everybody uses them. In Wikipedia you still find the double symbols and the ligatures. I absolutely hate the ligatures.

    I am not so sure that I agree with you that the [ʃ] sounds of English and French are so different. Of course there may always be a minute difference between two very similar sounds that the IPA has decided is not significant enough to indicate with a different symbol. For that, the IPA created the diacritics, which can indicate, for example, if the tongue is slightly raised or lowered. That is not a design flaw at all, but rather a design property. If you had to have a different symbol for every minor variation, you would need hundreds more symbols, if not thousands.

    Going back to the [ʃ]’s in chaud vs. show, I think the differences you detect here are due to coarticulation, that is to differences in the height of the following vowel, which affect the actual position of the tongue in the preceding consonant. Linguists often decide not to mark such minor differences in their transcriptions, unless they are relevant to what they are describing. It may be that the the French /o/ is phonetically higher and more back than the English /o/. (Note that I am using the phonemic delimiters // here, rather than the phonetic ones [].)

    Anyway, I just thought I would give a linguist’s perspective.

    I too enjoy your posts. I got here, by the way, after finding your article on the Seipel line, which I find fascinating. I also enjoyed The Story of H, also very well done. Actually, that may have been the first article that I found.

    Take care!

    -Jon Aske

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: