A post from Heidi Harley on Language Log mentioned a study in which French teenage native speakers showed a lot of variation in “assigning” genders of French nouns. That is for any given word, some teens would think it was masculine gender and others would think it was feminine gender – even if a French dictionary only assigned it to one gender.

To be honest, I’m not terribly surprised at this. Because of the ways French word-final sounds have evolved, the phonological distinction between genders is very weak in Modern French. Compare this to Spanish where most words ending with -o are masculine and most words ending with -a are feminine. The major cues for determining gender of a word in French might be processes like definite article agreement (le vs la) or pronoun replacement (il vs. elle)…and they may not be salient enough for speakers to make consistent judgments.

To me this is evidence that grammatical gender assignment is often based on phonology. A common historical change in gender assignment is for a noun to be reassigned to another gender because it’s ending is more typical of a different gender. A notable example is the Latin word laurus ‘laurel tree’ which was grammatically feminine in Classical Latin but has changed to masculine gender in Italian il lauro (Italian is a descendant of Latin). The French data here is consistent with this idea that phonology is a factor in determining consistent gender. If there is no regular “rule” or phonological cure, you would expect lots of variation.

On an interesting side note, the research also found that adult speakers were much more consistent in their gender assignments than the teens were. Something has happened between generations. This is very speculative, but I wonder if attitudes towards standard grammar or standard grammar education are changing.

I’m thinking irregular English past tense. Almost all native speakers acquire a set of irregular past tense, but there’s actually a lot of variation. For instance in the U.S. the “correct” past tense of bring is brought, but variations like brang (similar to ring/rang) and brung (as in You Got To Dance with Them What Brung You by Molly Ivins). FYI – Neil Diamond used brang in the song Play Me (“Song she sang to me/Song she brang to me.”)

I definitely recall several 3rd grade grammar lessons which required us to memorize “correct” irregular past tense forms (whereas we never had to memorize Question Formation). I suspect 3rd grade French children get to memorize genders of nouns. In fact, I just found a French Guess the Gender game for children, so it’s probably a “tricky grammar point.” So..if the method of grammar instruction changes, you could have the natural variation surfacing again in a population.

I honestly don’t know what grammar instruction is like in modern France, but it would be worthwhile for a researcher to check (without thinking the apocalypse is coming of course).

If nothing else, it would seem like a fascinating historical linguistic phenomenon is in progress.

P.S. On the difficulty of assigning genders by phonology, a French grammar site notes that “If you study these 40 word endings, it is possible to determine the gender of 75% of French nouns with almost 95% precision.” Hmmm!