Irregular pronounciation-orthography in French?

[Adam Kim]

Are there any words that are irregular (disregarding the liaisons at the end of the words)? For example words with "-ss-" (between vowels) being pronounced like [z] or "-s-" (between vowels) being pronounced like [s].

I wouldn't consider "monsieur" (masc. sing.) irregular because historically it was "mon sieur" in the past with its pronunciation evolved to today.

Just a curiosity. I hope what I wrote is clear.

[Neil Coffey]

Yes, absolutely, there are irregularities in the spelling>sound correspondence (though I think Monsieur is arguably as much an irregularity as any other case).

Some cases to think about:
- In suspect, the final -ct isn't pronounced; in abject, it is;
- In cognac, the gn is pronounced a bit like the ni in English onion; in cognitif, it's pronounced like gn in English cognitive;
- In aiguiser, the ui is normally pronounced as an i; but in aguille, it's pronounced as in the ui of huit;
- There are cases were not all speakers agree on the pronunciation of s between two vowels (e.g. you'll hear people pronounce abasourdir with either [s] or [z]), and there are complications when the preceding vowel is a nasalised vowel-- e.g. in transatlantique, it tends to be a [z], but in intrinsèque it's a [s];
- It's often not totally predictable whether a final consonant is pronounced or not: e.g. the final -c is pronounced in lac, but usually not in estomac;
- The odd other isolated case, e.g. in the words agenda, appendice, the en is generally pronounced as through written in; the word femme as though written famme etc.

I'm not aware off the top of my head of a case where ss is pronounced [z], other than "acidentally" when followed by a d sound (which is a general process in French -- so e.g. in presse de France, the ss would tend to be pronounced as a z because of the following d; but the same is true of pièce d'identité, where the c would tend to be pronounced as a [z] for the same reason).

[Adam Kim]

Well, let's see.

From http://shtooka.net/collections/fra/en/ abasourdir has either [s] or [z].

"agenda, appendice" are recently learned loanwords from Latin and possibly being read in a Catholic Church convention in France at that time.

[Neil Coffey]

Yes, the fact that agenda and (arguably) appendice are "loanwords" is surely a factor, but they're still irregularities. Neither is particularly recent, incidentally-- they've both hand ample ample time for their spelling to be brought in line with the pronunciation if there was a desire to do so.

[Adam Kim]

What about "au fait"? It says that I should pronounce the "-t" in "fait" according to the dictionary.

[Neil Coffey]

Sorry to be slightly nerdy, but I don't think strictly speaking you can class this as liaison-- it's simply a property of the word fait that in the singular, the final -t is usually pronounced.

Compare: "le lait est froid"-- would you ever proounce the "t"...? If this was a liaison context and not just a property of the word "fait", then you'd probably expect it to work for other words too.

[Neil Coffey]

I actually suspect that most of the time in the singular, the [t] is pronounced. You mention what you call "exceptions", but it seems to me that these are very commom expressions, and that the use of "fait" without it being in one lof these expressions is possibly itself the exception. I confess I don't have raw data at hand just now to back this suspicion up, on the other hand.

I wouldn't overinterpret the transcription chosen by a particular dicitonary, incidentally. (By the way, there's no such thing as "the" dictionary-- just different ones compiled by different human editors, none of whom have any special relationship with any linguistic god to tell them the "right" answer.) The editor/transcriber probably isn't going off any actual data either...