The Loss of the Circumflex


Photograph by Tom Page

by Justin E. H. Smith

An official reform of French spelling was recently announced, causing no small uproar on the Internet, and presumably in real life too (I don’t really talk to people), as to whether this is good or bad.

There were three broad sorts of change. The first are changes to the spelling of words in order to better reflect their pronunciation. The most common example cited has been the replacement ofoignon by ognon. I confess I had always thought the first syllable of this word was supposed to be pronounced as in oie (‘goose’), that is, roughly as in the first syllable of the English water. I noticed people around me were pronouncing it as ognon, but took this for a regionalism or a sort of laziness. I can’t say I care so much about this change, but ognon looks awfully strange to me, too much like a variation on some proto-Slavic root for fire, as in the Russian огонь (‘ogon”), whose genitive is огня (‘ognya‘) and whose Sanskrit cousin is the goddess अग्नि (Agni): in all of which cases the g is pronounced before the n, rather than indicating a softening in the termination of the and providing a faint iotation to the vowel that follows. I expect I will be practicing orthographic disobedience whenever I write that word in the future, not out of firm principle, but only out of soft preference.

The second sort of reform has mostly to do with hyphens, e.g., transforming that most French of words (at least since Godard), week-end, into weekend. This seems to follow a broad trend that is much further along in English (a hundred years ago it was common to see dog-house, out-fit, and so on), and I find I really could not care less. This all has more to do with typography than with historical linguistics. As Barrett Brown has established, there is a certain sort of semiliteracy that is signalled by gross overdependence on the hyphen, yet no one will be accused of not having fully learned to write for daring to do without it. Our prison correspondent doesn’t mention the strange habit of treating an adverb that modifies an adjective as somehow part of the same hyphenated word as the adjective, e.g., quickly-accelerating, beautifully-dressed. To me this is worse than in-mate, or any of those ridiculous hyphenated words beloved of postmodernists who have grown tired of parentheses: thus a conference on (trans)portation might be rebranded as trans-portation, as we move from the 90s to the aughts to the teens, ever on the look-out for new ways to make familiar things ‘problematic’.

But anyhow the third and final sort of change, and the one that concerns me most, has to do with the great reduction in the use of the circumflex accent, the elegant ^. One of this noble crown’s common functions, which has been preserved in the post-reform era, is disambiguation, signalling that a word is not the same as one that would otherwise be identical to it, sans circonflexe. Thusdu as the contraction of de and le can be distinguished from as the past participle of devoir, even if we often make do without the accent, and even if the accent plays no phonetic role and has no real historical sense.

Another more significant role for the circumflex, a role that has been greatly reduced by the reform, is to mark the historical disappearance of a consonant, typically an s, that had been inherited from Latin. Thus magister becomes maistre, and then, soon enough, maître. The circumflex accent does not signal that you should pronounce the word any differently, but then again neither did the s in maistre. It had long been silent, a vestige of a sound from an ancestor language. The circumflex simply marks a further degeneration of this unused part, like the almost undetectable pelvis still shuttled around by the enormous right whale, who has not had any need to bend at the waist since he took to the ocean so long ago. The current shift to maitre not only makes the ancestral connection to magister that much harder to discern, it also hides from view the meaning of more orthographically conservative proper names, like Joseph de Maistre. There are countless examples of the way the circumflex serves as a bridge to the past in French: hôte reminds us of the strange autoantonymy latent in the ideas of both host and hostility. Tête preserves a crucial link to testa, and pâte to pasta: that strange word that gives us in English both spaghetti and the stuff we brush our teeth with. Luckily these instances of the circumflex will be preserved, for now. But how much longer before some bureaucrat announces they too serve no real use?

If you think language is in its essence a facilitator of maximally clear and streamlined communication between living people, then this reform constitutes real progress. But if you think as I do that language is a monument and a link to the ancestors, if you were already more comfortable talking to the dead than to the living, then it is a loss: not a great loss, but a small slipping away of something that once mattered.

Piece crossposted with