French lessons

[Legion]

Introduction: I though I'd make a thread of French lesson, because I may have an original perspective on the language which might benefit some people, or may not. I have no experience with teaching anything, so if you have questions or claims on the form of this course, do not stay silent.

This thread can also be a good place to ask me any precise question about a point of French grammar.

It goes without saying (but it goes better when saying it) that what I will teach will be mostly based on my dialect (which however is relatively standard) and on my own perspective on the language.

That said…

Lesson I: the sounds of French

1) Consonants

According to my own analysis, French has the following inventory of consonants:

Occlusives: /p b t d k g/
Nasals: /m n ɲ (ŋ)/
Fricatives: /f v s z ʃ ʒ/
Approximants: /l j/
Other: /ʁ/

A recording of the consonants (each is pronounced in onset then in coda, except /ŋ/, which is pronounced only in coda).

—As with other Romance languages, and unlike most Germanic languages, /p t k/ are not aspirated and /b d g/ are fully voiced.
—Many speakers do not have a /ɲ/ phoneme distinct from /n/ + palatal glide. For those speakers, "se magner" /səmaˈɲe/ to hurry up (colloquial) and "se manier" /səmaˈnie/ to be handled are homophonous. I don't have this distinction myself, and it can generally be ignored, as there are few minimal pairs, but I will still mark it in the transcription. When this distinction is absent, what should be sequences of /ɲ/ + /i/ tend to be realised as [ni]: "magnifique" magnificent [maːniˈfik]. This can also happens before non-palatal glides: "baignoire" bathtube > [bɛːˈnuaːʁ̞].
—/ŋ/ is found only in loanwords (eg: "parking" /paʁˈkiŋ/ parking lot), and some speakers substitute it with the sequence /ng/, realised [ŋg].
—The coronal phonemes of French /t d n s z l/ have a dental, rather than alveolar articulation point. /s z/ are still sibilant though. Which means they are pronounced on the same articulation point as English /θ ð/, but with the tongue assuming the same shape than with English /s z/.
—/ʁ/ as a wide variety of allophones, that depends both of its position in the word and of the dialect or social status of the speaker. In onset, it tends to be an uvular trill or an uvular fricative (the fricative is more frequent in practice, but the trill was the standard pronounciation in traditional French songs from before the 80s, and can be heard prominently on recordings from Jacques Brel or Georges Brassens). In coda, it tends to weaken to an approximant, and sometimes it is dropped entirely, leaving only a long vowel behind (but as this sounds uneducated, at least in France, it should not be reproduced). It also has voiceless allophones next to voiceless consonants.
—Traditional inventories additionally include approximants /w ɥ/, but I have yet to see convincing evidence that those are but prevocalic allophones of /u y/.
—French doesn't have /h/, except, curiously, in laughter. "hahaha!" /ˈhaˈhaˈha/
—Gemination is not normally contrastive (except in quite pedantic pronounciation), but can happen over word boundaries: "faites tout !" /ˈfɛt ˈtu/ [ˈfɛˌtːu] do everything!


2) Vowels

French has the following vowels:

Oral: /a ɛ e i ø y ə ɔ o u/
Nasal: /ɛ̃ œ̃ ɑ̃ ɔ̃/

A recording of the vowels.

—/a/ is rather central, closer to [ɐ] than to [æ] or [ɑ].
—Some speakers merge /ɛ/ and /e/, which are mostly contrastive in open syllables only.
—Some speakers merge /ɔ/ and /o/, which are only contrastive in close syllables. Those are not necessarily the same speakers that merge /ɛ/ and /e/ (I merge /ɛ/ and /e/, but not /ɔ/ and /o/).
—The contrast between /ø/ and /ə/ is actually a contrast of archi-phonemes. For all intend and purpose, the realisations of /ə/ completly overlap with those of /ø/, but with the notable difference that /ə/ can be productively elided (in some contexts this elision is mandatory, in others, elision is optional, and often associated with colloquial language). This is why they will still be transcribed distinctly, even though /ə/ had ceased to be realised as a schwa and had merged phonetically with /ø/ already in Middle French.
—I have seen at least one phonetics manual advising English speakers to use /ɪ ʊ/, rather than /eɪ oʊ/, as the best approximation of /e o/. This is not a bad idea.
—To produce fronted rounded vowels, remember than simply rounding the lips is not enough: they must be protruded, and the tongue moved forward in the mouth, pretty much as if you were going to kiss someone.
—Several dialects (including the standard pronounciation in France) merge /ɛ̃ œ̃/ as a single unrounded nasal vowel, which tends to be realised closer to [æ̃] than to [ɛ̃].
—There is a tendency to merge /ɑ̃/ and /ɔ̃/ in fast speech, but they are still distinct in careful pronounciation, and so one should be careful to contrast them (even if they sound quite close to begin with).
—In southern France, nasal vowels are often pronounced with a noticible velar appendage, eg "poisson" fish as [puaˈsɔ̃ŋ] instead of [puaˈsɔ̃], so this isacceptable as an approximation of the sound, though pure nasal vowels are still prefered if one aims at a neutral pronounciation.
—Some conservative dialects distinguish additional vowels /ɑː/ and /ɛː/, but this has long ceased to be productive in the France standard pronounciation.
—Most descriptions add a front rounded vowel /œ/, but the only minal pair that would have this phoneme contrast with /ø/ is dubious: "jeune" */ʒœn/ young and "jeûne" /ʒøn/ fasting are only contrasted by speakers who read them side by side; when reading them in a text, appart from each others, the contrast disappears. And appart from this pair, the distribution of [ø] and [œ] as allophones of /ø/ seems completely predictable, if a bit chaotic: /ø/ is realized as [ø] in open syllables ("feu" [fø] fire) and as [œ] in close syllables ("peur" [pœːʁ̞] fear), except if the coda consonant is a dental other than /l/ or /n/ ("menteuse" [mɑ̃ˈtøːz] liar (f); but "seul" [sœːl] alone) or if the coda consonant is /l/ and the onset is a labial consonant ("meule" [møːl] millstone).
—Vowels are lengthened before voiced consonants.

A word on diphthongues:

Modern descriptions generally say that French doesn't have diphthongues, but these are the same descriptions that add /w ɥ/ to explain how words like "huit" */ɥit/ eight or "loi" */lwa/ law are still monosyllabic. I consider simpler to assume that French does have diphthongues, falling diphthongues in /i-/, /u-/ and /y-/, including /ia iɛ ie iø iy iɔ io iu iɛ̃ iɑ̃ iɔ̃ ua uɛ ue ui uø uɛ̃ uɑ̃ uɔ̃ ya yɛ ye yi yø yo yɛ̃ yɑ̃ yɔ̃/. All these are realised as a single syllabic nucleus. However, they can sometimes become bisyllabic in some environments (particularly the /i-/ diphthongues). Typically, after a group consonant + /l/ or /r/: "trier" to sort > [tχiˈje], rather than *[ˈtχi͡e] or *[ˈtχje].

Other combinations of vowels are hiatus and always count as distinct syllables. It should be noted that /i/ and /j/ are only contrastive in coda position: "pays" /pɛ.ˈi/ country vs "paye" /ˈpɛj/ pay, "abbaye" /abɛ.ˈi/ abbey vs "abeille" /aˈbɛj/ bee. And even there, a case could be made for an even more reductionist analysis, on account of the difference in syllabification.


3) Phonotactics

French has a relatively liberal syllabic structure, notably compared to other Romance languages, and sometimes even compared to English. Both onset and coda can include up to three consonants in a row. Typical clusters include /s/ + occlusive (+ /l/ or /r/), as well as /f/ or /v/ + /r/ or /l/, both in onset and coda. /gz/ is also frequent in onset, but is not found word finally. /ks/, on the other hand, is not found word initially, but can occure elsewhere. /ʁ/ or /l/ + consonant is also frequent in coda. French has generally more tolerance than English for non-native clusters, including Greek combinations like /gn/ ("gnose" /ˈgnoz/ gnosis), /pn/ ("pneu" /ˈpnø/ tire, a short for "pneumatique") or /pt/ ("ptérosaure" /pteʁoˈzɔʁ/ pterosaur), or africates, which french does not possess natively ("jet" /ˈdʒɛt/ jet, homographic with "jet" /ˈʒɛ/ throwing; "tchétchène" /tʃeˈtʃɛn/ chechen).

Important word-final clusters tend to be eased by non-phonological silent and/or ultrashort [ø], even in places where there was no vowel historically ("film" [ˈfiːlmø̆] film). If this final cluster comes into contact with other consonants, either word internally or over word boundaries, two things can happen: the final vowel will return to a fully phonological state (careful pronounciation), or the cluster will be simplified (lax pronounciation); "quatre personnes" [ˈkatχøpɛːʁ̞̊ˈsɔːn] or [ˈkatpɛːʁ̞̊ˈsɔːn] four people.

Thus, more than 3 consonant clusters do not normally happen word internally, because vowels are always inserted between compounds.

As in English, clusters of consonant + /l/ or /ʁ/ are treated as single consonants during syllabification: "adresse" address, dexterity is /a.ˈdʁɛs/ not */ad.ˈʁɛs/.

Unlike English, French has no problem contrasting [sj] vs [ʃ] vs [ʃj], compare "passion" /paˈsiɔ̃/ passion vs "marchons" /maʁˈʃɔ̃/ let's walk vs "marchions" /maʁˈʃiɔ̃/ (we) were walking.

Nasal vowels never occur at the first element of a vowel hiatus, at least word internally. They normally appear before a consonant or word-finally. They do not often occur before nasal consonants, but it can happen: "ennui" /ɑ̃ˈnyi/ boredom. Contrast with "année" /aˈne/ year (but some conservative dialects have a nasal vowel in "année" as well, and indeed in many words where standard French doesn't. However, this can be perceived as uneducated, and should therefore be avoided).


4) Morphophonology

French has lost a great deal of morphology compared to other Romance languages, but still has enough so that various morphophonological processes can manifest themselves. These processes are not numerous, but they are extremely frequent.

Élision /eliˈziɔ̃/ elision

Élision is the process by which an /ə/ vowel is supressed. This obligatorily affects monosyllabic words in /-ə/ when they preceed a word that starts with a vowel: "le" /lə/ + "orage" /oˈʁaʒ/ > "l'orage" /loˈʁaʒ/ the thunderstorm. There are however some words which phonemically start with a vowel, but prevent élision: "le" + "héros" /eˈʁo/ > "le héros" /ləeˈʁo/ the hero. For emphasis, such sequence can be realized as [ləʔeˈːʁo], but thus kind of glottal stop insertion is not compulsory, and the hiatus pronounciation is never incorrect.

Futhermore, facultative élision can happen both word-internally ("maintenant" now /mɛ̃.tə.ˈnɑ̃/ > /mɛ̃t.ˈnɑ̃/) and in sequence of words ("je veux" I want /ʒə.ˈvø/ > /ˈʒvø/). A high number of facultative élisions can be perceived as quite informal, though there is more tolerance for élision word-internally than in sequences of words (many internal élisions have in fact become completly lexicalized, cf Old French "chauderon" > Modern "chaudron" /ʃo.ˈdʁɔ̃/ cauldron).

There is a limit to what can be elided even facultatively. "je reviens" I'm coming back /ʒə.ʁə.ˈviɛ̃/ can become /ʒʁə.ˈviɛ̃/ or /ʒəʁ.ˈviɛ̃/ but not */ˈʒʁviɛ̃/. It's generally rare for more than two consonants in a row to be allowed this way. Facultative élision cannot occur in contexts where obligatory élision is prohibited: "dehors" outside is always /də.ˈɔʁ/, never */ˈdɔʁ/.

A consequence of these rules is that it is not often clear if syllables that preceed an /ə/ are to be treated as open or close, and in words where this does affect pronounciation noticibly, both treatments seems to be possible: "médecin" (medical) doctor /me.də.ˈsɛ̃/ > [meˈtsɛ̃] or [mɛtˈsɛ̃].

Liaison /liɛ.ˈzɔ̃/ linking

Liaison is the process by which a consonant is inserted between two words, the second of which starts with a vowel. The consonant is phonologically linked to the second word, but lexically to the first word. The insertion is not automatic, but strongly depends of syntaxic context. Some liaisons are compulsory, some others are facultative, and yet others are forbidden. Unlike with élision, the use of facultative liaisons marks more formal levels of language.

The most frequently inserted consonants are /t/ and /z/, followed by /n/. More rarely, liaisons in /p/, /g/ and /ʁ/ can also happen.

Compare: "les chiens" /le.ˈʃiɛ̃/ the dogs vs "les enfants" /le.zɑ̃.ˈfɑ̃/ the children, "le petit chien" /lə.pə.ti.ˈʃiɛ̃/ the little dog vs "le petit enfant" /lə.pə.ti.tɑ̃.ˈfɑ̃/ the little child.

The same words that prevent élision also prevent liaison: "les héros" /le.e.ˈʁo/ the heroes.

Liaison will be more detailed in the next lesson, about spelling.

Enchaînement /ɑ̃.ʃɛ.nə.ˈmɑ̃/ chaining

Enchaînement is a process related to liaison, which cause a word-final close syllable with a single coda consonant to become open of the next word starts with a vowel. "ancienne" old, ancient (f) /ɑ̃.ˈsiɛn/ > "ancienne école" old school, former school /ɑ̃.ˈsiɛ.ne.ˈkɔl/. Unlike liaison, this is not conditionned by syntax and tends to apply everywhere. This may include unintuitive orthographical contexts ("vers Antibes" /vɛ.ʁɑ̃.ˈtib/ toward Antibes), but not words that do not allow élision and liaison: "brave héros" /ˈbʁav.e.ˈʁo/ brave hero.


5) Stress

If there are few easy things in French, word stress is one of them. French words are consistently stressed on the last syllable (excluding any final non-phonologic [ø̆]). But this stress is rather weak, it is often a clause or sentence stress, and individual words will have a tendency to flow each into another. As a rule /ə/ is never stressed, except in a few case of verb subject inversion: "dis-le !" /diˈlə/ say it!.


6) Text

To practice all this, let's see a little text, a short "fable" (/fabl/ tale, fable) from Jean de La Fontaine /ˈʒɑ̃.də.la.fɔ̃.ˈtɛn/. Do not worry about spelling and grammar, this is purely a pronounciation exercice. This text contains all the French phonemes I have listed except /j/ (but it does contain /i-/ diphthongues, which can be alternatively analyzed as /j/ + vowel) and /ŋ/.


Le Coq et la Perle. /lə.ˈkɔ.ke.la.ˈpɛʁl/

Un jour un Coq détourna /œ̃.ˈʒu.ʁœ̃.ˈkɔk.de.tuʁ.ˈna/
Une perle, qu'il donna /y.n(ə).ˈpɛʁ.l(ə).kil.do.ˈna/
Au beau premier lapidaire. /o.ˈbo.pʁə.ˈmie.la.pi.ˈdɛʁ/
« Je la crois fine, dit-il ; /ʒə.la.ˈkʁua.ˈfi.n(ə).ˈdi.til/
Mais le moindre grain de mil /mɛ.lə.ˈmuɛ̃.dʁə.ˈgʁɛ̃.də.ˈmil/
Serait bien mieux mon affaire. » /sə.ˈʁɛ.biɛ̃.ˈmiø.mɔ̃.na.ˈfɛʁ/
Un ignorant hérita /œ̃.ni.ɲo.ˈʁɑ̃.te.ʁi.ˈta/
D'un manuscrit, qu'il porta /dœ̃.ma.nys.ˈkʁi.kil.pɔʁ.ˈta/
Chez son voisin le libraire. /ʃe.sɔ̃.vua.ˈzɛ̃.lə.li.ˈbʁɛʁ/
« Je crois, dit-il, qu'il est bon ; /ʒə.ˈkʁua.ˈdi.til.ki.ˈlɛ.ˈbɔ̃/
Mais le moindre ducaton /mɛ.lə.ˈmuɛ̃.dʁə.dy.ka.ˈtɔ̃/
Serait bien mieux mon affaire. » /sə.ˈʁɛ.biɛ̃.ˈmiø.mɔ̃.na.ˈfɛʁ/

Note: the phonemes in parenthesis are not normally pronounced, but they are required here for the metric (those are seven feet lines).

A recording of this text (note that in this text and others that will follow, I don't make any attempt to pronounce distinctions that are not phonemic in my dialect).

English translation:

The Cock and the Pearl

A cock scratched up, one day,
A pearl of purest ray,
Which to a jeweller he bore.
"I think it fine," he said,
"But yet a crumb of bread
To me were worth a great deal more."

So did a dunce inherit
A manuscript of merit,
Which to a publisher he bore.
"It's good," said he, "I'm told,
Yet any coin of gold
To me were worth a great deal more."


And I think that will be quite enough for this time. See you in the next lesson.

[Astraios]

Awesome. I was having serious déjà vu reading this post. Then I remembered that I take a course unit called The Sounds of French, which teaches over a semester (I know, right?) basically exactly what you've written here.

[finlay]

Introduction: I though I'd make a thread of French lesson, because I may have an original perspective on the language which might benefit some people, or may not. I have no experience with teaching anything, so if you have questions or claims on the form of this course, do not stay silent.

This thread can also be a good place to ask me any precise question about a point of French grammar.

It goes without saying (but it goes better when saying it) that what I will teach will be mostly based on my dialect (which however is relatively standard) and on my own perspective on the language.

That said…

Lesson I: the sounds of French

1) Consonants

According to my own analysis, French has the following inventory of consonants:

Occlusives: /p b t d k g/
Nasals: /m n ɲ (ŋ)/
Fricatives: /f v s z ʃ ʒ/
Approximants: /l j/
Other: /ʁ/

A recording of the consonants (each is pronounced in onset then in coda, except /ŋ/, which is pronounced only in coda).

—As with other Romance languages, and unlike most Germanic languages, /p t k/ are not aspirated and /b d g/ are fully voiced.
—Many speakers do not have a /ɲ/ phoneme distinct from /n/ + palatal glide. For those speakers, "se magner" /səmaˈɲe/ to hurry up (colloquial) and "se manier" /səmaˈnie/ to be handled are homophonous. I don't have this distinction myself, and it can generally be ignored, as there are few minimal pairs, but I will still mark it in the transcription. When this distinction is absent, what should be sequences of /ɲ/ + /i/ tend to be realised as [ni]: "magnifique" magnificent [maːniˈfik]. This can also happens before non-palatal glides: "baignoire" bathtube > [bɛːˈnuaːʁ̞].
—/ŋ/ is found only in loanwords (eg: "parking" /paʁˈkiŋ/ parking lot), and some speakers substitute it with the sequence /ng/, realised [ŋg].
—The coronal phonemes of French /t d n s z l/ have a dental, rather than alveolar articulation point. /s z/ are still sibilant though. Which means they are pronounced on the same articulation point than English /θ ð/, but with the tongue assuming the same shape than with English /s z/.

I dunno about you, but I'm pretty sure my /s z/ are lamino-dental. I find it hard to do an apical one. Does that mean I'm the only one or something? I mean, everywhere else just blindly asserts that English has exclusively alveolar coronals and only /θ ð/ are dental (and then apparently doing interdentals here is weird for the UK). And then there's that thing where some English accents are meant to have a distinction between [t_d] and [t_a] instead of [θ] and [t]. And that makes no sense to me either.

I mean literally, it can't be the case that French /s/ has the same articulation point as English /θ/, because for me that's interdental...

[Legion]

I dunno about you, but I'm pretty sure my /s z/ are lamino-dental. I find it hard to do an apical one. Does that mean I'm the only one or something? I mean, everywhere else just blindly asserts that English has exclusively alveolar coronals and only /θ ð/ are dental (and then apparently doing interdentals here is weird for the UK). And then there's that thing where some English accents are meant to have a distinction between [t_d] and [t_a] instead of [θ] and [t]. And that makes no sense to me either.

I mean literally, it can't be the case that French /s/ has the same articulation point as English /θ/, because for me that's interdental...

Well, I really don't know much about English articulatory details, so you may well be right; in any case, French /s z/ are clearly dental sibilant. According to Wikipedia, French coronals are laminal denti-alveolar consonants.

[finlay]

Hmmm. Fair enough. Sounds like a lot of making up terms for not a lot of benefit there. Lamino-dental will do fine for me... I'll wager that there is variation in the production of French coronals, anyway. But that probably goes without saying...

[Ulrike Meinhof]

Great read!

Two questions:

Do you have any examples of initial /gz/?
How is enchaînement realized phonetically? I mean, saying there's a syllable border somewhere doesn't mean much unless you specify how it's pronounced.

[Niedokonany]

But what's the articulatory difference between [ˈbʁav.e.ˈʁo] and [ˈbʁa.ve.ˈʁo]? I think it might affect the realization of the non-close front rounded phoneme as [ø] or [œ] in the preceding syllable, but this wouldn't apply in case of /a/.

[Yiuel Raumbesrairc]

Great read!

Two questions:

Do you have any examples of initial /gz/?
How is enchaînement realized phonetically? I mean, saying there's a syllable border somewhere doesn't mean much unless you specify how it's pronounced.

1. Most words beginning with initial "x" are pronounced with initial /gz/, though some people reduce it to [z]. It's highly idiolectal.

Xavier /gzavje/
xylophone /gzilofOn/

2. A final consonant with an initial vowel become a whole syllable, linking both words.

Bonne école -> /bOn.e.kOl/ -> /bO.ne.kOl/

[Legion]

Initial /gz/ is simply how is consistently pronounced initial "x": "xylophone" /gzi.lo.̍fɔn/; "Xavier" /gza.̍vie/; "xénophobie" /gze.no.fo.̍bi/.

Enchaînement simply means that word boundaries are ignored during syllabification, so final consonants of a word can end up as onset consonants of the next word instead. There are no pause between words. And yes this can affect the realisation of some vowels: "seul" [̍sœːl] > "seul au monde" [̍søːloːmɔ̃ːd] alone in the world. When a word that doesn't allow that shows up, this realized by the insertion of a short pause or of a glottal stop: "brave héros" > [̍bʁaːv eː̍ʁo] or [̍bʁaːvʔeː̍ʁo].

Edit: damn you, Yiuel, stop reading my thoughts >:V

[Legion]

Enchaînement simply means that word boundaries are ignored during syllabification, so final consonants of a word can end up as onset consonants of the next word instead. There are no pause between words. And yes this can affect the realisation of some vowels: "seul" [̍sœːl] > "seul au monde" [̍søːloːmɔ̃ːd] alone in the world. When a word that doesn't allow that shows up, this realized by the insertion of a short pause or of a glottal stop: "brave héros" > [̍bʁaːv eː̍ʁo] or [̍bʁaːvʔeː̍ʁo].

I just realized that, alternatively, a final -e can be restored: "pauvre héros" > [̍poːvʁøeː̍ʁo] or [̍poːvʁøʔeː̍ʁo] poor hero.

[Yiuel Raumbesrairc]

Enchaînement simply means that word boundaries are ignored during syllabification, so final consonants of a word can end up as onset consonants of the next word instead. There are no pause between words. And yes this can affect the realisation of some vowels: "seul" [̍sœːl] > "seul au monde" [̍søːloːmɔ̃ːd] alone in the world. When a word that doesn't allow that shows up, this realized by the insertion of a short pause or of a glottal stop: "brave héros" > [̍bʁaːv eː̍ʁo] or [̍bʁaːvʔeː̍ʁo].

I just realized that, alternatively, a final -e can be restored: "pauvre héros" > [̍poːvʁøeː̍ʁo] or [̍poːvʁøʔeː̍ʁo] poor hero.

Dialectal note :

Some dialects may also do Enchaînement even then. That is, Enchaînement goes even over the hiatus rule of liaison. The liaison rules still apply even then.

Pauvre héros /povR.ero/ -> [po.vRe.Ro]
Pauvres héros /povR.ero/ -> **[po.vr2.ze.ro]
Pauvres héros /povR.ero/ -> [po.vre.ro]

[AnTeallach]

—The coronal phonemes of French /t d n s z l/ have a dental, rather than alveolar articulation point. /s z/ are still sibilant though. Which means they are pronounced on the same articulation point than English /θ ð/, but with the tongue assuming the same shape than with English /s z/.

I dunno about you, but I'm pretty sure my /s z/ are lamino-dental. I find it hard to do an apical one. Does that mean I'm the only one or something? I mean, everywhere else just blindly asserts that English has exclusively alveolar coronals and only /θ ð/ are dental (and then apparently doing interdentals here is weird for the UK). And then there's that thing where some English accents are meant to have a distinction between [t_d] and [t_a] instead of [θ] and [t]. And that makes no sense to me either.

I mean literally, it can't be the case that French /s/ has the same articulation point as English /θ/, because for me that's interdental...

Ladefoged and Maddieson (The Sounds of the World's Languages, p23 in the edition I have) quote a study which said that although French /t d l n/ are described as laminal dental and English /t d l n/ as apical alveolar, a minority of French speakers used alveolars and a minority of (Californian) English speakers used dentals. (The samples were pretty small.)

As for the Hiberno-English dental/alveolar distinction, I can usually hear that the Hiberno-English reflex of [θ] sounds different from a normal English [t]. I'd guess it's interdental, but I don't actually know.

[Travis B.]

In North American English, I am used to /t d n s z/ being alveolar and having allophonic variation between being apical and laminal, typically being apical by default but being laminal before /uː ʊ w ər/; my dialect also has these being laminal in many consonant clusters, but this I am not as sure about the distribution of. Also, the same applies with /ʃ ʒ tʃ dʒ/ but the variation here is between palatoalveolar and alveolopalatal rather than between apical alveolar and laminal alveolar. (Before I get any comments, I should say I am speaking about much of NAE as a whole; I notice this variation in General American in media content, for instance.)

In some varieties and idiolects, onset laminal /t/ that is aspirated may also be affricated and may become alveolopalatal. This sticks out to me, in particular, in the case of a rather infamous lawyer commercial back in Wisconsin where the woman speaking kept on repeatedly pronouncing twenty with an obvious [t̠ɕʰ] that I remember frequently seeing on local TV several years ago.

[finlay]

—The coronal phonemes of French /t d n s z l/ have a dental, rather than alveolar articulation point. /s z/ are still sibilant though. Which means they are pronounced on the same articulation point than English /θ ð/, but with the tongue assuming the same shape than with English /s z/.

I dunno about you, but I'm pretty sure my /s z/ are lamino-dental. I find it hard to do an apical one. Does that mean I'm the only one or something? I mean, everywhere else just blindly asserts that English has exclusively alveolar coronals and only /θ ð/ are dental (and then apparently doing interdentals here is weird for the UK). And then there's that thing where some English accents are meant to have a distinction between [t_d] and [t_a] instead of [θ] and [t]. And that makes no sense to me either.

I mean literally, it can't be the case that French /s/ has the same articulation point as English /θ/, because for me that's interdental...

Ladefoged and Maddieson (The Sounds of the World's Languages, p23 in the edition I have) quote a study which said that although French /t d l n/ are described as laminal dental and English /t d l n/ as apical alveolar, a minority of French speakers used alveolars and a minority of (Californian) English speakers used dentals. (The samples were pretty small.)

As for the Hiberno-English dental/alveolar distinction, I can usually hear that the Hiberno-English reflex of [θ] sounds different from a normal English [t]. I'd guess it's interdental, but I don't actually know.

oh yeah that was my source for the interdental thing; apparently a majority of californians and a minority of brits used interdental /θ ð/

[Soap]

Thank you for taking the time to type this all out. I notice it says Lesson I at the top so hopefully you'll be writing up more, and I look forward to that too.

I'm just confirming something I think I read on here once (perhaps even from you), is it true that all instances of words beginning with {h} that breaks liaison are loans? If so, does that mean French lost /h/ and then regained it from loanwords and then lost it again?

[finlay]

It did lose /h/ and then gain it again and then lose it again. Whether it was from loans or not I don't know.

[Yiuel Raumbesrairc]

Since hypothèse is a loan yet causes liaison, we can say that being a loan probably is not the reason.

[finlay]

But the question was not "do all loans cause h-aspiré?", it was "are all h-aspiré words caused by loans?". The opposite, and a question you didn't answer there.

It may also be the case that it was loans between certain dates: after the loss of the first /h/, and after the subsequent gain of the second /h/, but before the loss of the second /h/.

[Legion]

Yes there will be more lessons.

"h-" in Germanic loans generally prevents élision and liaison. However, this has sometimes contaminated native words. We'll see that in the next lesson.

[Lukas Kelly]

One tip, if you are designing lessons, use your knowledge of linguistics to make them still useful (which you did, as opposed to most courses), but put it in laymans terms. I know there are many people that have trouble gaining enough useful information if it's in the format of an essay(If we are talking about languages, or any other topics in which you need knowledge in large lumps to be useful).

[Legion]

Well, I'm assuming most people on this board are familiar with the linguistics terminology and stuff. Of course if I was to have this become a general public course, I would adapt it further (but that would take more time).

[Xephyr]

See this is what happens when you make a "Lessons" thread of a language too many people on the board already know: it ceases being a Lessons thread right away and turns into just another Discussion thread.

[Torco]

This is win, tho' it needs more exercises.

EDIT: and yeah, let the man do the teaching xD [which will result in plenty of let's discuss french threads xD]

[Civil War Bugle]

See this is what happens when you make a "Lessons" thread of a language too many people on the board already know: it ceases being a Lessons thread right away and turns into just another Discussion thread.

My French is incredibly poor, having never studied it formally at length, and I am pleased to see lessons up with IPA. I still can't pronounce it but at least I can eradicate the wrong pronunciations I have in some cases.

[Alces]

Thank you for doing this. I have some knowledge of French, but mostly from school, and you learn practically nothing about pronunciation from that.