Edmond Edmont.

Cara Giaimo writes in Atlas Obscura about a man I knew nothing about, as well as others in his line of work:

Long before we had viral quizzes to gather our peculiarities, there was only [Edmond] Edmont—a linguistic assistant who spent the end of the 19th century bicycling around France, speaking to locals, and cataloguing their unique words and phrases. Over four years, Edmont journeyed to over 600 towns, gathering material for what would become the Atlas Linguistique de la France: the world’s first great linguistic atlas.

A century later—after technological revolutions and scholarly schisms wholly reshaped the field—Edmont remains, in the words of one linguist, “a mythical figure in the history of dialect surveys.” Whether you’re the kind of surveyor who spends hours speaking to farmers in Georgia, or the kind who dreams up the Buzzfeed Accent Challenge, his work remains both vital and informative.

There follows a riveting history of dialect studies, including a PhD student named Georg Wenker who “drew up 42 sentences that, in his estimation, covered the most changeable aspects of the German language” (“In the wintertime dried leaves fly about in the air”; “I will slap your ears with the cooking spoon, you monkey!”) and Jules Gilliéron, with “his own set of 2,000 common words and phrases, similarly designed to cover a broad swath of French.” Edmont worked for Gilliéron, and their Atlas Linguistique inspired “dialectologists from Switzerland to Japan.” There’s much more at the link, which I urge you to click on.


  1. marie-lucie says

    Edmond Edmont and his sturdy bicycle, out on the country roads in all kinds of weather, are well-known among dialectologists. But in spite of his dedication and honesty his reports are not entirely reliable linguistically. For each location he was instructed to seek out the oldest men who had spent their entire lives there, which he did, but in many cases those men were too old to be useful for the project, whether because of deficient memory, near-deafness, toothlessness, or other impediments. But many of those men were widowers who lived with a son and his family, and since a wife was likely to be at home while the husband was working in the fields or at some job, an old man’s daughter-in-law would often be around to help with Edmont’s work, prompting words that she remembered the old man saying, or repeating his utterances with clearer pronunciation, so that many answers attributed to very old men were actually provided by middle-aged women, who might have been from different villages.

  2. Fascinating, thanks for that added background!

  3. January First-of-May says

    Reminds me of Vladimir Dal, who, IIRC, also went on a bicycle asking old men about local words – though, as far as I understand it, he was more concerned with the general variety than with the specifics of pronunciation. (And I might be mistaken about the “bicycle” part.)

    Definitely an interesting story.

  4. Edmond Edmont was an oddity for another reason: he had no knowledge whatsoever of the history of French or of Romance languages/dialects: Gilliéron had found that such knowledge had a negative effect on a would-be fieldworker’s perception (i.e. a fieldworker who knew the etymon and (official!) history of a given word was far more inclined to misperceive the sounds of said word and transcribe it in a more etymologically faithful way), and thus deliberately selected someone with a keen ear whose transcriptions would be unaffected by any presuppositions as to what the forms would/should be like.

    The linked article is also misleading in saying of his work that it involved variation in French: while he did work on Romance varieties in France, most of these varieties could not in any way be called varieties/dialects of French. And indeed, the degree of linguistic diversity which was found then is difficult to overstate (Did I ever get the chance to mention to the Hattery that “Gallo-Romance” is a linguistically meaningless label? Really. The total number of exclusive shared innovations within so-called Gallo-Romance varieties is somewhere between zero and nil), and even more difficult to imagine today. Language death/obsolescence even back then was an issue fieldworkers had to deal with, and I think it was Gilliéron himself who bemoaned the fact that he had, to his mind, arrived fifty years too late.

  5. “Gallo-Romance” is a linguistically meaningless label

    It doesn’t even seem to be well-defined. Once outside the langues d’oïl core, people seem to include some random mixture of: the Arpitan group, the Occitan group, the Rhaeto-Romance group, and the Gallo-Italian group, according to their particular theoretical orientation. WIkipedia mentions as a shared innovation the loss of final vowels other than /a/ (except where a propaedeutic vowel is added), but I’m quite prepared to be told that’s wrong.

  6. John Cowan: “Gallo-Romance” normally refers to the langues d’oïl, the Occitan and the Arpitan groups. The loss of final vowels other than /a/ is not just shared with “Rhaeto-Romance” (also a genetically non-existent group) and (most) Gallo-Italian varieties, but also with Catalan. Now, on the basis of this innovation you might try to classify all of the above varieties as “Gallo-Romance”, but there are practically no other common innovations, phonological or lexical or morphological, which could be used in support of this enlarged “Gallo-Romance” subfamily.

    What I find revealing about this innovation is that it is found in the Romance varieties with the strongest tendency to turn stressed vowels into diphthongs/triphthongs (French, Arpitan, Romansch) as well as in geographically adjacent varieties whose treatment of stressed vowels is either quite banal (Gallo-Italian) or downright conservative (most Occitan varieties, Catalan). This suggests to my mind that the fall of final vowels other than /a/ is an innovation which first arose in the former varieties, as a result of more dynamic stress (which also turned stressed vowels into diphthongs or triphthongs), and thence spread to other, neighboring Romance varieties whose stress was no more dynamic than it was in other points of the Romance continuum.

  7. I (as well as WP) was reckoning Catalan in the Occitan group, so that’s all right. Technically, a single synapomorphy is sufficient to establish monophyly, but spread is certainly a possibility here.

  8. Somehow Edmond Edmont doesn’t get mentioned in Graham Robb’s The Discovery of France, which is also based on cycling around the French countryside and discusses the (former) linguistic diversity of France.

Speak Your Mind