Kat wrote:And a silly question of my own: Why do the French only pronounce the first half of their words and swallow all the endings? Have they "always" done that? Is it the result of some kind of sound shift?
This question is so fun. And now I must answer it with an unnecessarily long post!
As Iversen pointed out, no, they haven't always done that.
From the 9th century to the 13th century (or arguably the 15th), French was actually like modern Spanish or German, with spellings that closely (but not perfectly) corresponded to pronunciation. When Evrat, a canon priest at the city of St. Etienne in 1192 was given support by Marie of France, countess of Champagne, to translate the first books of the Bible into French, he wrote:
Tuit li languages sunt & divers & estrange / Fors que li languages franchois / Cest cil que deus entent anchois / Kil lo fist & bel & legier / Sel puet len croistre & abregier / Mielz que toz les altres languages
[Source]
[Listen to a recording I made of these lines.]
All languages are diverse and strange, / except for the French language. / That's the one God likes to hear the most, / as He made it both beautiful and light, / although He can enlarge it or shorten it / more easily than all other languages. (my translation)
The last line, for example, would have been pronounced [ˈmjɛɫts kə ˈtots ləz ˈaɫtɾəs lãŋˈgwadʒəs].
English approximation: MYELTS kuh TOTE-s luh-z AL-truss lang-GWAH-juss.
I would like to correct Iversen, though, about French spelling calcifying with the French Academy. The scholar Togeby he cites is quite correct about French spelling staying close to speech until basically the 15th century, but the Academy didn't publish its dictionary until the end of the 17th century.
French did start to accumulate archaic spellings very early on, from about the 11th century onwards. For example, Evrat's 12th century verse above misspells
en chois (modern
en choix) as
anchois, because he didn't distinguish the
en and
an nasal vowels. As far as Evrat was concerned,
entent anchois could've been "antant en chois" and it would've sounded the same.
Similarly, by the 13th century, the -t of plural verbs (e.g.
ils aiment) was already silent most of the time, and the vowels of
puet (modern
il peut) and
leur sounded the same, which led to people changing the historically correct spelling
puet to "
peut", and
je muer to "
je meur" which then became "
je meurs", and
cuer to "
cœur", etc., etc., although somehow they didn't change
accueil to "
accœuil" and
orgueil to "
orgueuil".
When French was declared the official language of legal texts in the French Kingdom in the 16th-century Ordinance of Villers-Cotterêts, the declaration contained a large number of silent letters already:
Que les arrestz soient clers & entendibles. Arti.c.x.
Et affin quil ny ait cause de doubter sur lintelligence desdictz arrestz. Nous voulons & ordonons quilz soient faictz & escriptz si clairement quil ny ait ne puisse auoir aucune ambiguite ou incertitude ne lieu a en demander interpretation.
[Source, although I retranscribed the article image, as the source has some mistakes in its transcription.]
Decrees must be clear and understandable. Art. 110.
So that there is no reason to doubt the meaning of said decrees. We desire and order that they must be composed and written so clearly that there shall not be, nor be able to be, any ambiguity or uncertainty, nor grounds to request an interpretation of them. (my translation)
Consider that most of the -z letters at the end of words don't stand for any sound, and same goes for the b of
doubter (contrast medieval
douter), the c of
faictz (contrast medieval
faiz), and the -ent of
soient. Why were b and c added to
douter and
fai(t)z? Because people wanted them to look more like the words in Latin they came from,
dubitare and
facta. There was also no reason to spell
faictz & escriptz with -z instead of -s (as in modern
faits et écrits) except that the spellings with -z were the historically correct ones. This is what the French Academy had to work with when it was founded about a hundred years after this document.