French and other Romance languages: hypercorrection with "h" in English

elroy

Imperfect mod
US English, Palestinian Arabic bilingual
It's well known that a common hypercorrection found among French native speakers speaking English is to add a superfluous "h" to words beginning with a vowel, for example by saying "honly" instead of "only." Presumably this is because all initial h's in French are silent, while most initial h's in English are not silent, so a French person may be concerned that if they say "only," they're actually dropping an initial "h" due to an influence of French, so they hypercorrect by adding an "h."

So far, so good.

What I'm wondering about is why Spanish speakers don't do the same thing. Or do they? At least I've never encountered this (or heard about it happening) among Spanish speakers. In Spanish, h's work the same way as in French, so why don't Spanish speakers hypercorrect in the same way? Could it be that h's are easier to pronounce for French speakers?

Italian speakers don't do it either (to my knowledge). This could be because Italian orthography generally omits those silent h's (for example: English 'honor', French 'honneur', Spanish 'honor', Italian 'onore'), so perhaps this is not as much of an issue for Italian speakers.

I don't know enough about Portuguese to comment on it in this context, but I've also never heard (of) Portuguese speakers doing this, and I think Portuguese, like French and Spanish, generally keeps these h's in writing.
 
Last edited:
  • Dymn

    Senior Member
    /h/ is a foreign sound to French speakers and inserting it even if it makes no sense might sound "more English" to their ears.

    Spanish speakers mostly use /x/ from the j sound in their language, so it doesn't sound more or less English either way.
     

    merquiades

    Senior Member
    English (USA Northeast)
    French vowels are pronounced clearly with force and muscular precision. When you do this consistently, it can naturally bring about aspirations before or after vowels which French speakers don't hear because /h/ means nothing to them. There are some interesting threads here about aspirations after the vowels like in houih, Nancyh, icih (another topic).. When a vowel begins a word in French the aspirations happen much less frequently because French words are naturally linked one way or another to the preceding word: elision: l'amie... liaison: les Zamis (les amis)... or enjambement: un jeu Nami (un jeune ami). Actually articles are almost always used in French so few spoken sequences start with vowel sounds.

    When students learn English in France, unfortunately little time is spent on pronunciation/articulation [so they naturally associate each English vowel to the closest French vowel and then substitute it until it becomes their habit..] or on consonant or vocalic linking in English. Actually some teachers tell them to make pauses between each word. When an English speaker says "I understand" or "the orange", there is a slight y /j/ sound that links the words together. With "an idea" "white elephant", the consonants do the connecting. So, if you fail to make these links and pronounce English vowels with great intensity you get I x hunderstande, zee x horange, an x hidee, white x helephante. It can be corrected by practicing sentences like "I eat it" versus "I heat it", "I ate" or "I hate". You can torture them with I hear. I have a big ear.

    Spanish speakers do not do this for two reasons. The vowels are articulated nowhere near as strongly as in French. Actually the French have strong accents in Spanish because of the vowel sounds. Compare lo, l'eau, low.... particularly the rounded o, the i with the lips drawn back, the closed e. I think French developed this habit because there are so many vowels in the language (16) and so many one syllable words that you really do need to articulate strongly and precisely to distinguish them. Anyway, having only 5 vowels (neither fully open or closed) means not as much articulation is needed. When speaking English these 5 weaker vowels don't bring about aspirations although other problems arise (but save that for another thread). The second reason is that Spanish speakers don't associate English h with silent Spanish h, they associate it with Jota /x/. They actually never learn English /h/ at all. They just substitute the jota of José or Gibraltar.
    I've wondered what Spanish learners of German do since this language has both /h/ and /x/. Edit: actually they probably just use /x/ all the time: /h/ starts words in German, /x/ is after a vowel at the end, so they can't be confused.
     
    Last edited:

    AndrasBP

    Senior Member
    Hungarian
    It's well known that a common hypercorrection found among French native speakers speaking English is to add a superfluous "h" to words beginning with a vowel, for example by saying "honly" instead of "only." Presumably this is because all initial h's in French are silent, while most initial h's in English are not silent, so a French person may be concerned that if they say "only," they're actually dropping an initial "h" due to an influence of French, so they hypercorrect by adding an "h."
    I don't know if it helps the discussion, but I've heard French speakers do the same in Hungarian, too.
    Hungarian has lots of words starting with a /h/ sound and also pairs like ős-hős, alom-halom, úr-húr, avas-havas, etc., where the /h/ is the only difference between two words. It's a nightmare for the French.:D
     

    Sprache

    Senior Member
    English/inglés
    French vowels are pronounced clearly with force and muscular precision. When you do this consistently, it can naturally bring about aspirations before or after vowels which French speakers don't hear because /h/ means nothing to them. There are some interesting threads here about aspirations after the vowels like in houih, Nancyh, icih (another topic)..
    This is very interesting to me because I've always noticed it when listening to spoken French, especially when words end in one of the high front vowels /i/ and /y/, but I've never seen or heard it talked about anywhere. The end of the word is aspirated so strongly and it's very noticeable when one isn't used to it. I have a feeling that French-speakers aren't even aware of it, and I never would have imagined that the notorious "hypercorrect H" was caused by the same phenomenon.
     

    Penyafort

    Senior Member
    Catalan (Catalonia), Spanish (Spain)
    In Spanish, h's work the same way as in French, so why don't Spanish speakers hypercorrect in the same way? Could it be that h's are easier to pronounce for French speakers?

    As it's been said, h is mostly silent in the Romance languages, yes, but unlike in French or most other Romance languages, in Spanish there is the jota, a velar fricative, very close to the English glottal fricative -in fact, it is pronounced as in English in some varieties- so Spanish speakers are used to pronounce that English h, even if improperly. There is no need for hypercorrections.

    Probably it'd be better to ask: why don't Italians do it? I wonder if it may be related to the so called aspirated h in French, associated to some words beginning with h- in French that prevent liaison from happening.
     

    merquiades

    Senior Member
    English (USA Northeast)
    As it's been said, h is mostly silent in the Romance languages, yes, but unlike in French or most other Romance languages, in Spanish there is the jota, a velar fricative, very close to the English glottal fricative -in fact, it is pronounced as in English in some varieties- so Spanish speakers are used to pronounce that English h, even if improperly. There is no need for hypercorrections.

    Probably it'd be better to ask: why don't Italians do it? I wonder if it may be related to the so called aspirated h in French, associated to some words beginning with h- in French that prevent liaison from happening.
    Yes, Aspirated h is the only example I can think of where you stop the natural flow of linking every word together in French and make a slight pause before a word beginning with a vowel: Le x hibou, also distinguishing between l'eau and le x haut. They also do this with some words that don't have an h: le x onze, trois x euro.... This is quite similar to the pause they make with I x hunderstand or two x happles.
    Sometimes pauses are made in Catalan too and break the flow. There is l'ungla but la x universitat. The pause is not extended in the plural though, is it? I believe les universitats is pronounced together with closed juncture.
     
    Last edited:

    Dymn

    Senior Member
    As for Portuguese speakers I've noticed the Portuguese don't pronounce the h in Anglicisms whereas the Brazilians do because it's the sound for their r (which is mostly in Brazil /h/, but mostly /ʀ/ in Portugal). I don't know what the Portuguese do when speaking English but I've found a foreira once said this hypercorrection does occur as well.

    Yes, Aspirated h is the only example I can think of where you stop the natural flow of linking every word together in French and make a slight pause before a word beginning with a vowel: Le x hibou, also distinguishing between l'eau and le x haut. They also do this with some words that don't have an h: le x onze, trois x euro.... This is quite similar to the pause they make with I x hunderstand or two x happles.
    Is a fully-fleged /h/ or something close to it what results from these liaison-breaking environments, though? I haven't noticed anything similar.

    Sometimes pauses are made in Catalan too and break the flow. There is l'ungla but la x universitat. The pause is not extended in the plural though, is it? I believe les universitats is pronounced together with closed juncture.
    It's not a break at all, "la u" is pronounced in the same syllable, with a falling diphthong. In fact in colloquial speech it's perfectly acceptable to say "l'universitat", this spelling rule gives nightmares to a lot of students (and note "l'organització" is not *la organització, which is incoherent in Central Catalan).

    As for les universitats any consonant in front of a vowel is pronounced in the same syllable (so long as there is no pause in speech), so the syllable separation would be /ɫə.zu.ni.β̞əɾ.si.tats/.
     

    merquiades

    Senior Member
    English (USA Northeast)
    Is a fully-fleged /h/ or something close to it what results from these liaison-breaking environments, though? I haven't noticed anything similar.
    No, the only real coincidence is it is a pause before a word starting with h, so it gives food for thought. H aspiré is supposed to be a real break between the two words. It feels weird to me to do this because you have to cut yourself off: le x haut, so I usually ignore it. I have heard people pronounce h or weak h, but I don't feel it's right. I heard la hache not long ago. It's especially atrocious with euro because there is no h. Other people just make the links anyway especially with common words like les Z haricots. Sometimes there is no avoiding it with en eau and en haut. The meaning changes drastically.
    My reason for them pronouncing h in English is what I said in the first post.

    It's not a break at all, "la u" is pronounced in the same syllable, with a falling diphthong. In fact in colloquial speech it's perfectly acceptable to say "l'universitat", this spelling rule gives nightmares to a lot of students (and note "l'organització" is not *la organització, which is incoherent in Central Catalan).

    As for les universitats any consonant in front of a vowel is pronounced in the same syllable (so long as there is no pause in speech), so the syllable separation would be /ɫə.zu.ni.β̞əɾ.si.tats/.

    Thanks for the information. Yes, it's hard for me to understand the logic of when to reduce to l' and when to leave it la, especially since the plural and other forms are normal. I believe you that it could be a nightmare.
    As for Portuguese speakers I've noticed the Portuguese don't pronounce the h in Anglicisms whereas the Brazilians do because it's the sound for their r (which is mostly in Brazil /h/, but mostly /ʀ/ in Portugal). I don't know what the Portuguese do when speaking English but I've found a foreira once said this hypercorrection does occur as well.
    In Lisbon it's more like a French r, and not rolled much. I have heard Brazilians pronounce R as /h/ in English and French. It depends on their level. I'll have to listen to see if they add h to the beginning; there are really other pronunciation issuse that draw your attention before that, like nasalization and palatalization.
     
    Last edited:

    Yendred

    Senior Member
    Français - France
    It's especially atrocious with euro because there is no h
    I agree. It's indeed a bad habit, it's wrong and not euphonic.
    People do this because of the word preceding euro - usually a number - because they don't master the tricky rules concerning the 's' at the end of a number:
    vingt T'euros
    cent T'euros

    but:
    quatre-vingts Z'euros
    deux cents Z'euros


    So they simplify the problem and do as if there were a H aspiré in every case.
    The problem did not exist when the French currency was the franc.
     
    Last edited:

    Penyafort

    Senior Member
    Catalan (Catalonia), Spanish (Spain)
    Thanks for the information. Yes, it's hard for me to understand the logic of when to reduce to l' and when to leave it la, especially since the plural and other forms are normal. I believe you that it could be a nightmare.

    What students are told is that it's not apostrophised when the initial i- or u- are unstressed. But there are a few more cases too, mainly having to do with supposed ambiguities.

    If you'd like a parallel with the French, some Catalans tend to do a slight stop in La Haia or la hac. Most do a sort of dragging a, though.
     

    berndf

    Moderator
    German (Germany)
    What I'm wondering about is why Spanish speakers don't do the same thing. Or do they? At least I've never encountered this (or heard about it happening) among Spanish speakers. In Spanish, h's work the same way as in French, so why don't Spanish speakers hypercorrect in the same way? Could it be that h's are easier to pronounce for French speakers?
    Having lived among French speakers for a long time now, it has nothing to do with how easy or difficult a "h" is to pronounce but how easy or difficult it is for speakers to perceive the difference between minimal pairs like at and hat and for French speakers they sound pretty much the same while they sound different for Spanish speakers because they have a similar sound and they perceive the difference more clearly. It is the same phenomenon as German speakers confusing short a and short e or w and v in English: The difference simply does not matter to us.
     

    Sobakus

    Senior Member
    This is very interesting to me because I've always noticed it when listening to spoken French, especially when words end in one of the high front vowels /i/ and /y/, but I've never seen or heard it talked about anywhere. The end of the word is aspirated so strongly and it's very noticeable when one isn't used to it. I have a feeling that French-speakers aren't even aware of it, and I never would have imagined that the notorious "hypercorrect H" was caused by the same phenomenon.
    This is known as vowel devoicing - google for "French vowel devoicing" without the quotes for many many results. I would expect any description beyond the level of "the a is pronounced as in English father" to mention it. Wikipedia adds that a phrase-medial devoicing of high vowels exists, and is especially prominent in Quebecois.

    To put what berndf writes above in fewer words, the difference in aspiration is non-phonemic (=non-distinctive) to French speakers. Other Romance language speakers might or might not have more success depending on whether their speech variety has some phoneme that [h] can be understood as an allophone (a realisation variant) of. For Castilian speakers it's /x/, for Brasilian Portuguese speakers /r/. People don't "hear" and can't "say" sounds that aren't distinctive to them despite being able to identify and pronounce them in isolation.
     

    Penyafort

    Senior Member
    Catalan (Catalonia), Spanish (Spain)
    We Catalans don't have any /x/, but since we also speak Spanish, we tend to say h in the Castilian way. If English had been a thing a hundred years ago, Catalans would have done like the French or Italians, a silent h. Whether there'd have been hypercorrections, it's hard to predict.
     

    Red Arrow

    Senior Member
    Dutch - Belgium
    You can hear "mercih" here (the LuckyLuke sample), but it sounds like [ç] to me. It is too different from English h in most words.
     

    Sobakus

    Senior Member
    It is too different from English h in most words.
    I think this implies that if it was identical to the English /h/, it would make any difference - it wouldn't, because this "sound" is merely a realisation variant of the French /i/ phrase-finally. But if English had word-final /h/, French speakers would then randomly substitute English final /i/ for /h/ and the reverse.
     

    berndf

    Moderator
    German (Germany)
    it wouldn't, because this "sound" is merely a realisation variant of the French /i/
    It is a bit different the airflow is continued after the end of voicing. This phenomenon applies to all final vowels but creates the most audible effect with high front vowels ([ i ] and [ y ]). Needless to say, that this is nonstandard and non-phonemic but very popular.
     
    Last edited:

    Sobakus

    Senior Member
    It is a bit different the airflow is continued after the end of voicing. This phenomenon applies to all final vowels but creates the most audible effect with high front vowels ([ i ] and [ y ]). Needless to say, that this is nonstandard and non-phonemic but very popular.
    Different from [ç]? I didn't mean to say it was identical to it. And there need not even be any voicing at all; I'm not sure it typically applies to non-high vowels - the complete lack of voice doesn't at any rate - and this has to do with the fact that high vowels have inherently the shortest duration. The [ç]-like quality is due to the palatalisation (frontness) of the vowels - with other vowels the result will be different. A [h] should theoretically result from a devoiced [a], but again I don't believe that vowel is devoiced (nor common phrase-finally).
     

    berndf

    Moderator
    German (Germany)
    Different from [ç]?
    The process is different. It is not an alternative realisation of the vowel but the airflow is continued past the end of pronuncing the vowel. The apparent sound similar to [ç] is an artefact of this process happening with high front vowels.

    again I don't believe that vowel is devoiced
    That is your right but doesn't change the fact that this is what happens in this particular accent. The /h/ in most Germanic language is the opposite phenomenon: the airflow starts earlier than voicing starts, i.e. /h/ occurs automatically when you omit the glottal stop in the onset of a syllable that starts with a vowel. This is also why the Germanic /h/ cannot be produced in the syllable coda, contrary to, e.g., the Semitic /h/.
     
    Last edited:

    Sobakus

    Senior Member
    The process is different. It is not an alternative realisation of the vowel but the airflow is continued past the end of pronuncing the vowel. The apparent sound similar to [ç] is an artefact of this process happening with high front vowels.
    That is one's first thought for sure, but on closer inspection a question arises: what is the exact nature of the difference? If one didn't know that this is not a realisation of a vowel phoneme, would one be able to tell it from a consonant? Again: the vowel is pronounced voiceless, this is not a separate segment "past the end" of the vowel - it's the vowel itself, devoiced, exactly as when you speak in whisper. In fact, say [iiiii] in whisper right now - you'll get the sound in question. But if someone doesn't know you're saying a whispered [iii], what's to stop them thinking that you're saying an [ç]? To put it plainly, I suspect the answer is "nothing", and that the difference is only in the head of the interpreter.
    That is your right but doesn't change the fact that this is what happens in this particular accent.
    That's not how this works. This is not a religious belief, but one based on reading mentions of the phenomenon and briefly familiarising myself with papers like this one, where the frequency for /a/ is given as 2.8% vs. 11.8% for /i/ (p. 184). In order to challenge my conclusion, you need to present evidence pointing to a high frequency of devoicing in the vowel /a/.
    The /h/ in most Germanic language is the opposite phenomenon: the airflow starts earlier than voicing starts, i.e. /h/ occurs automatically when you omit the glottal stop in the onset of a syllable that starts with a vowel. This is also why the Germanic /h/ cannot be produced in the syllable coda, contrary to, e.g., the Semitic /h/.
    I'm sorry but this makes no sense to me. Firstly, the limit on where a phoneme can occur belongs to phonotactics, which are language-specific and not sound-specific. If there is no phonetic difference between the Germanic and Semitic /h/, both can be produced wherever the phonotactics allow them to be produced. In addition, Germanic languages including Icelandic, Faroese and Norwegian do have syllable-coda aspiration before (I suspect only aspirated) geminate stops; granted, it's not phonemic - for that one can turn to Finnish, which even has double glottal stops (!?).

    Finally, I don't understand on the basis of what you came to the conclusion that aspiration in Germanic is automatic if you simply remove the glottal stop. Some (most?) varieties of Dutch, for example, lack vowel-initial glottal stops, as do Icelandic and Faroese. All three have /h/ as a phoneme. And to repeat, if it was indeed automatic, this would be simple phonotactics, while you're making it look like some physical articulatory constraint. This would be like a Russian speaker insisting that you can't pronounce /ti/ without palatalising the /t/ - they'd say that if you don't velarise it (your omission of glottal stop), it automatically palatalises.
     
    Last edited:

    berndf

    Moderator
    German (Germany)
    If one didn't know that this is not a realisation of a vowel phoneme, would one be able to tell it from a consonant?
    I do think so. I feel reminded of the [ç] of my own language but I would not confuse it with that consonant. You could say this because I live in a French speaking area and I very familiar with French. But I don't think so. These are other mismatches between the phonemic structure of German and French or English I still struggle with at times even though I am very familiar with. But this I perceive very clearly as an artefact.
     

    Red Arrow

    Senior Member
    Dutch - Belgium
    This old video (00:12 - 00:32) has some fragments from Flemish TV with hypercorrection of H. This his ighly hinteresting.

    As far as I know, Walloons never do this when speaking Dutch. They either drop every H or pronounce it correctly.
     
    Last edited:

    Red Arrow

    Senior Member
    Dutch - Belgium
    Yes, the H was dropped in the dialects of West-Flanders, East-Flanders, Antwerp (only the West), Flemish Brabant and Brussels. The hypercorrection occured when speaking Standard Dutch.

    The H returned everywhere thanks to Standard Dutch, except in West-Flanders where the g and ch moved backwards to the point where they became indistinguishable (for most people) from h. So hypercorrection is still happening there with /ɦ ɣ x ∅/.

    About the discussion between you and Sobakus: I think you are both somewhat correct. Germanic /h/ occurs automatically between certain vowels if you drop the glottal stop. I can say "be(ʔ)amen" or "be(h)amen", but not "beamen". This is also true at the beginning of an utterance. I can say "Hallo" or "ʔallo", but not "Allo". But this is simply not true at the beginning of all Germanic words. You are clearly confusing German and Germanic! It is quite typically German to mark the beginning of words with glottal stops, not all Germanic languages do that.

    My Brabantian grandfather never says H and a glottal stops doesn't occur at the beginning of every word that starts with a vowel. So "Kom eens hier" sounds like ['kɔ.mə.'zæ.rəs] rather than ['kɔm.'ʔəs.'ɦær.'ʔəs]. (The last syllable is a duplication of eens) All words are merged together, not unlike French.

    I say ['kɔ.'me:ns.'ɦi:r] in Standard Dutch, not ['kɔm.'ʔe:ns.'ɦi:r].

    (I typed this with my phone. I apologise for any possible typos)
     
    Last edited:

    berndf

    Moderator
    German (Germany)
    You are clearly confusing German and Germanic! It is quite typically German to mark the beginning of words with glottal stops, not all Germanic languages do that.
    No, that is not the issue. The issue is if it is physically possible to start a word with a vowel and I would say no, that is not possible, unless heavy breathing is non-phonemic. In a language that has heavy breathing (voiced or unvoiced) as a consonant, any attempt to pronounce an initial vowel without a glottal stop would result in an artifact that amounts to a consonant. French is not one of these languages and therefore allows pronouncing initial vowels with or without a glottal stop.
    There are languages that have glottal stops as an independent phoneme and these languages generally do not have the concept of an initial vowel at all. Vowels can only occur after consonants. This is, e.g., the case in (most/all?) Semitic languages.

    The peculiarity of German within Germanic languages is that glottal stops play an important role in morpheme separation within words that doesn't exist to such an extend elsewhere and that makes the glottal stop semi-phonemic in German.
     
    Last edited:

    Sobakus

    Senior Member
    No, that is not the issue. The issue is if it is physically possible to start a word with a vowel and I would say no, that is not possible, unless heavy breathing is non-phonemic. In a language that has heavy breathing (voiced or unvoiced) as a consonant, any attempt to pronounce an initial vowel without a glottal stop would result in an artifact that amounts to a consonant. French is not one of these languages and therefore allows pronouncing initial vowels with or without a glottal stop.
    I can't figure out whether I understand you right because the way I understand it you're repeating the same thing I already demonstrated to be incorrect, which makes absolutely no sense to me.

    Is your assertion that all languages with glottal fricatives have prothetic glottal stops? Even if somehow you weren't aware of there being many languages that have the glottal fricative as a phoneme and no prothetic glottal stops, and didn't have access to wikipedia, I even gave you examples of such Germanic languages. How can you continue to repeat that claim? I even illustrated the exact reason you were mislead into believing this with my Russian palatalisation example. What more can one possibly need? Does one need to present Gascon, basically the next closest thing to French?
    There are languages that have glottal stops as an independent phoneme and these languages generally do not have the concept of an initial vowel at all. Vowels can only occur after consonants. This is, e.g., the case in (most/all?) Semitic languages.
    No, there are Semitic languages which have a consonant-centered morphophonology, with generally no vowel-initial roots, and with the glottal stop as a separate consonant. There also exist many languages that are not Semitic, don't have consonant-initial triconsonantal roots, do have vowel-initial words and have the glottal stop as either a separate consonant or an allophone of another. Some of these languages have the glottal fricative as a separate phoneme as well, particularly in Southeast Asia as well as the Americas. Others might even have the fricative and the stop as allophones.

    I mean, making a typological observation on the correlation of prothetic glottal stops and glottal fricatives is fine as could lead to an interesting discussion, but why would you want to insist on such a rash and sweeping statement despite what I've already written? This topic is dealt with by research on linguistic universals and if you really wanted to, you'd look there to see if your German speaker's intuition finds factual confirmation.
     
    Last edited:

    berndf

    Moderator
    German (Germany)
    I can't figure out whether I understand you right because the way I understand it you're repeating the same thing I already demonstrated to be incorrect, which makes absolutely no sense to me.
    You demonstrated nothing. You one made assertions that make absolutely no sense to me. I don't think it makes sense to continue the yes-no ping-pong.
    This topic is dealt with by research on linguistic universals and if you really wanted to, you'd look there to see if your German speaker's intuition finds factual confirmation.
    In German, the glottal stop is not phonemic. That is not the issue. You could realize phonemically initial vowels differently and it would still sound natural but there is always something in front of the vowel that counts as a consonant in other languages. E.g., Germans hear words like Arabic `arabia as starting with a vowel. On the other hand, a German wouldn't hear glottal stops in words like Arabic ra's, even when it is clearly pronounced (which isn't always the case in everyday speech).

    Are we getting any closer?
     

    Red Arrow

    Senior Member
    Dutch - Belgium
    There are languages that have glottal stops as an independent phoneme and these languages generally do not have the concept of an initial vowel at all. Vowels can only occur after consonants. This is, e.g., the case in (most/all?) Semitic languages.
    This is not the case in Hawaiian and some other Polynesian languages.

    Alo, 'alo and halo are 3 separate words.
    Hawaiian Dictionaries
    Hawaiian Dictionaries
     

    Sobakus

    Senior Member
    You demonstrated nothing. You one made assertions that make absolutely no sense to me. I don't think it makes sense to continue the yes-no ping-pong.
    Ask and I will answer. I gave you examples of Germanic languages that have no prothetic glottal stops (in the same way Russian doesn't) and do have /h/, and a Dutch speaker confirmed this. Anyone who searches for "icelandic spoken" on YT can make sure that this assertion of mine makes sense to them. Icelandic even has vowel elision <3!
    In German, the glottal stop is not phonemic. That is not the issue. You could realize phonemically initial vowels differently and it would still sound natural but there is always something in front of the vowel that counts as a consonant in other languages.
    Not in Icelandic, Faroese, Gascon, Ukrainian or Hindi there isn't.
    E.g., Germans hear words like Arabic `arabia as starting with a vowel. On the other hand, a German wouldn't hear glottal stops in words like Arabic ra's, even when it is clearly pronounced (which isn't always the case in everyday speech).
    Which tells us a lot of things about German phonology and tells us nothing about crosslinguistic universals or any correlation between prothetic glottal stops (yes, I understand you're talking about prothesis!) and phonemic glottal fricatives.
    Are we getting any closer?
    I'm confindent we are.
     

    berndf

    Moderator
    German (Germany)
    I'm confindent we are.
    😀 Ok, then let me restate my thesis:
    yes, I understand you're talking about prothesis!
    Yes, indeed. A weaker form I my thesis would be: Phonemic initial vowels can't be realized without some form of prothesis. A non-phonemic glottal stop is the most frequently used one*.
    __________________________________
    * At least in European languages. But this discussion makes really only sense within European language as other languages at times have quite different ideas of what constitutes an /h/ than European languages, where it essentially means heavy breathing in those in which it is phonemic.
     
    Last edited:

    Sobakus

    Senior Member
    Yes, indeed. A weaker form I my thesis would be: Phonemic initial vowels can't be realized without some form of prothesis. A non-phonemic glottal stop is the most frequently used one*.
    __________________________________
    * At least in European languages. But this discussion makes really only sense within European language as other languages at times have quite different ideas of what constitutes an /h/ than European languages, where it essentially means heavy breathing in those in which it is phonemic.
    In view of the obvious fact that there are piles of European languages (the whole of Romance) that don't have any form of initial-vowel prothesis, I think you've forgotten to link that statement to the presence of /h/. If you do, this will end up the same statement about there being no languages with /h/ but no prothetic glottal stops, which again there are, European ones included. What gives?
    Also, I don't think the definition of /h/ as the glottal fricative is language- or region-dependent. A glottal fricative is going to be a glottal fricative even on another planet :)
     

    berndf

    Moderator
    German (Germany)
    In view of the obvious fact that there are piles of European languages (the whole of Romance) that don't have any form of initial-vowel prothesis
    I guess this is the core of our disagreement.

    My position is that it is an obvious fact is that they do have of initial-vowel prothesis (obvious because I hear it around me on a daily basis). In the case of French it is a glottal stop except in liaison. But this glottal stop is non-phonemic and could be replaced by other forms or prothesis without native speakers caring. And that is the reason for this hyper-correction in Germanic languages.
     
    Last edited:

    Sobakus

    Senior Member
    I guess this is the core of our disagreement.

    My position is that it is an obvious fact is that they do have of initial-vowel prothesis. In the case of French it is a glottal stop except in liaison. But this glottal stop is non-phonemic and could be replaced by other forms or prothesis without native speakers caring. And that is the reason for this hyper-correction in Germanic languages.
    Whoa whoa whoa, now that's a reversal! The whole point is that the French can't even hear, let alone pronounce the prothetic glottal stop, and the Italians and the... I think basically only the Czechs, the Hungarians and the Finns can, of non-Germanic speakers. These are the same languages that lack "liaison". In other languages glottal stops are extremely marginal, in Russian for instance it exists in a few interjections, like this one of refusal: не-а.

    Real, German-like vowel-initial glottal stops may be possible in these languages in emphatic speech, but their speakers struggle greatly in reproducing them even as prothetics in normal speech. To us it's like reproducing click consonants. Nobody can seriously claim that they hear German glottal stops in Italian. The pronunciations by GreenPixie and Phorcys contain vocal fold closure, but not a stop consonant articulation. This is simply how the human vocal tract works and isn't transcribed as a consonant by anybody. Languages that have phonemic glottal stops will still distinguish these from this purely physiological fold closure.

    Which gives me an idea, though. I'm starting to suspect that the French misaspiration is their attempt at marking word-boundaries in languages where these are marked with glottal stops. Languages that do this are typically word languages, while French is a syllable language. Here's a presentation that summarises the issue quite nicely in relation to precisely Dutch, French and German (reading the whole is recommended, but specifically see the table on p.7+). French speakers recognise that typological difference and hear that word boundaries are marked in some way, but being unable to "hear" or "say" the glottal stop they resort to the one glottal consonant they can produce: the fricative. This actually sounds to me like an explanation to the whole thing.

    Seen in this light, final vowel devoicing might just be the native strategy of French to mark the right phrase boundary in precisely the same way German and Dutch (and Old French) do it using consonant devoicing (cf. the presentation). It could be connected with the prestige of accents that totally lack phrase-final schwas, which may be phonologically implemented as vowel devoicing.
     
    Last edited:

    Red Arrow

    Senior Member
    Dutch - Belgium
    Ask and I will answer. I gave you examples of Germanic languages that have no prothetic glottal stops (in the same way Russian doesn't) and do have /h/, and a Dutch speaker confirmed this.
    No, maybe I didn't explain myself well. I think we do have glottal stops, but some people (especially in Flanders) only say them between certain vowels (be-amen) or at the start of an utterance. Some people say we never say glottal stops, but I don't think that's true. It would be more accurate to say that Dutch has much less glottal stops than German, but not zero.

    Here is a video on how to stop sayibg glottal stops at the start of an utterance.
     

    berndf

    Moderator
    German (Germany)
    Whoa whoa whoa, now that's a reversal! The whole point is that the French can't even hear, let alone pronounce the prothetic glottal stop
    No, not on my part. I always said in French the difference between 'at or hat was negligible and that was the reason for the hyper-correction and nothing else. I never said they couldn't product the difference or, with some effort, perceive it. This is a frequent pattern for hyper-corrections in case of phonemic mismatches: If a difference is irrelevant in your own language you use one or the other sound, which are mere alternative realization in your own language, erratically.
     

    Sobakus

    Senior Member
    No, not on my part. I always said in French the difference between 'at or hat was negligible and that was the reason for the hyper-correction and nothing else. I never said they couldn't product the difference or, with some effort, perceive it. This is a frequent pattern for hyper-corrections in case of phonemic mismatches: If a difference is irrelevant in your own language you use one or the other sound, which are mere alternative realization in your own language, erratically.
    Uhh, there is no difference in French between 'at or hat (is the former one a glottal stop?) and basically the French can't hear or reproduce either one correctly for the same reason you and I have mentioned repeatedly. This is not in question. The reversal is that you're saying that French actually has glottal stops. If French had prosthetic glottal stops, they'd have no trouble hearing them and transferring them onto German. My latest conclusion is that they replace them with aspiration - see my previous post.
     

    Sobakus

    Senior Member
    No, maybe I didn't explain myself well. I think we do have glottal stops, but some people (especially in Flanders) only say them between certain vowels (be-amen) or at the start of an utterance. Some people say we never say glottal stops, but I don't think that's true. It would be more accurate to say that Dutch has much less glottal stops than German, but not zero.

    Here is a video on how to stop sayibg glottal stops at the start of an utterance.
    I totally agree that Dutch has glottal stops, but not all varieties and not in all the environments that German does. Incidentally, even some High German varieties (Swiss, Luxembourgish) may lack them. The presentation I linked 2 posts above tackles this directly.
     

    berndf

    Moderator
    German (Germany)
    basically the French can't hear or reproduce either one correctly for the same reason
    "Doesn't matter" is not the same as "can't hear or reproduce". My point has consistently been that the difference doesn't matter not that they have any problems producing one or the other.

    If you think "the French can't hear or reproduce either one correctly" we are in disagreement.
     
    Last edited:

    Swatters

    Senior Member
    French - Belgium, some Wallo-Picard
    As far as I know, Walloons never do this when speaking Dutch. They either drop every H or pronounce it correctly.
    I don't know if I would be as categorical as this, especially about early learners. But (at least in my interlanguage), Dutch /h/ often ends up rendered as a voiceless glottal or pharyngeal fricative, which makes it a real consonant rather than an allophone of zero, while the same solution feels completely wrong in English.
    Seen in this light, final vowel devoicing might just be the native strategy of French to mark the right phrase boundary in precisely the same way German and Dutch (and Old French) do it using consonant devoicing (cf. the presentation).
    Final vowel devoicing is generally analysed as a boundary marking strategy, yeah. It only occurs phrase or utterance finally after all.
     

    Sobakus

    Senior Member
    "Doesn't matter" is not the same as "can't hear or reproduce". My point has consistently been that the difference doesn't matter not that they have any problems producing one or the other.

    If you think "the French can't hear or reproduce either one correctly" we are in disagreement.
    You seem to be trying to say the same thing I say in this message:
    To put what berndf writes above in fewer words, the difference in aspiration is non-phonemic (=non-distinctive) to French speakers. Other Romance language speakers might or might not have more success depending on whether their speech variety has some phoneme that [h] can be understood as an allophone (a realisation variant) of. For Castilian speakers it's /x/, for Brasilian Portuguese speakers /r/. People don't "hear" and can't "say" sounds that aren't distinctive to them despite being able to identify and pronounce them in isolation.
    The difference doesn't matter means it's not phonemic (contrastive), means they can't "hear" or "say" it, even though they can perceive and pronounce it as a paralinguistic sound. This whole thread is about the French not being able to produce /h/ correctly, and I don't imagine anyone disagreeing with this, so I'm not sure what the disagreement seems to be here. I was trying to convey that:
    1. In French, the difference between 'at or hat does not exist in the sense above (granted you haven't explained what 'at is supposed to be and I'm assuming it's /ʔat/ with the glottal sop)
    2. In French, there are no vowel-prothetic consonants, neither glottal stops nor glottal fricatives. This reflects its being a syllable language with "liaison"
    3. The human vocal tract is closed at rest. This opening of the vocal tract is a physiological phenomenon, is not an articulation, is not a consonant
    4. As a German, you perceive the release of air during the opening of the vocal tract as a consonant, but a French speaker needs to "stop speaking" in order to simulate a glottal stop. Stopping speaking is not an articulation, not a consonant
    5. There are languages that have initial glottal stops contrasting with zero. This demonstrates the point above
    6. One can release the air while "starting speaking" more or less forcefully, but a glottal stop requires a pressure build-up and a sudden release to be perceived as such
    7. Doing this is perceived as a paralinguistic feature of emphatic speech in a language like Russian. Normal articulation of Russian doesn't involve this
    8. The French hear some glottal constriction marking word boundaries in English. To reiterate point 1, this is foreign to the French language, so it stands out. They're unable to articulate a glottal stop as a prothetic consonant more or less like you can't employ the dental click as a prothetic consonant (even though you can say "tsk-tsk!"), but they are able to produce a whispered phonation (as they do with their vowel devoicing), and they do know that this is used as a consonant in English, so they replace prothetic glottal stops with prothetic glottal fricatives in order to be able to "continue speaking" while marking phonetic word boundaries somehow
     
    Last edited:

    berndf

    Moderator
    German (Germany)
    The difference doesn't matter means it's not phonemic (contrastive), means they can't "hear" or "say" it
    If that is your position, we won't get together.

    This whole thread is about the French not being able to produce /h/ correctly
    No it is not. It is about French saying it in the wrong places. This is typical for distinctions that don't matter in your language even though you are able to distinguish them in perception and production. It is like German speakers erratically confusing /v/ and /w/ although they have with a little bit of training no problems producing both. They perceive /w/ as the English "flavour" of the essential the same sound as /v/ which they perceive as the German "flavour". That is why the overuse /w/ in English (adwisor, SUW, wery, ...). French over and underusing /h/ is a similar phenomenon.

    But it is not always like that. E.g., the distinction between live and leave is indeed one where French speakers have difficulties perceiving the difference.

    That is my summary after more than 30 year living in French and French speaking Switzerland of Europe and observing these phenomena.
     
    Last edited:

    Sobakus

    Senior Member
    No it is not. It is about French saying it in the wrong places. This is typical for distinctions that don't matter in your language even though you are able to distinguish them in perception and production. It is like German speakers erratically confusing /v/ and /w/ although they have with a little bit of training no problems producing both. They perceive /w/ as the English "flavour" of the essential the same sound as /v/ which they perceive as the German "flavour". That is why the overuse /w/ in English (adwisor, SUW, wery, ...). French over and underusing /h/ is a similar phenomenon.

    But it is not always like that. E.g., the distinction between live and leave is indeed one where French speakers have difficulties perceiving the difference.
    I think I see what you mean, and this is a good observation. I think what's going on here is that the French don't perceive the glottal fricative as a phoneme at all. Whereas the vowels of live and leave are allophones of the French /i/, [h] is an allophone of nothing. It's a paralinguistic articulation, a whispered voice that the French try to use to mark word boundaries, probably extended from their own phrase-final vowel devoicing. This is a good qualification to what I say above, but it it doesn't affect it in any other way that I see.
     

    berndf

    Moderator
    German (Germany)
    This is a good qualification to what I say above, but it it doesn't affect it in any other way that I see.
    There are some things I agree with and some things I don't agree with. Especially I have to reiterate that the glottal stop is not phonemic in German and that claims to the contrary you frequently read are a myth (see this parallel thread). But in this thread is getting to far discussing the glottal stop issue any further. I think we have made our positions clear and resolved the misunderstandings. :)
     

    gburtonio

    Senior Member
    UK, English
    It is like German speakers erratically confusing /v/ and /w/ although they have with a little bit of training no problems producing both. They perceive /w/ as the English "flavour" of the essential the same sound as /v/ which they perceive as the German "flavour". That is why the overuse /w/ in English (adwisor, SUW, wery, ...).

    I had always thought this was simply down to orthography: the fact that letter 'w' is pronounced /v/ in German. But would you say that a German speaker would perceive a /w/ in an English word as being a /v/, or somewhat similar to it, even if s/he had no knowledge of English orthography and so didn't know that the sound was spelt in English with a 'w'? (I know that speakers of many languages generally perceive English /w/ as some kind of vowel sound similar to u*)

    *I want to put this 'u' in square brackets to refer to the phonetic symbol but when I do, it disappears and gets replaced with an underline …
     

    berndf

    Moderator
    German (Germany)
    But would you say that a German speaker would perceive a /w/ in an English word as being a /v/, or somewhat similar to it, even if s/he had no knowledge of English orthography and so didn't know that the sound was spelt in English with a 'w'?
    That might have been true 80 years ago when most Germans had no exposure to spoken English but not now. And then you would expect errors like vhat instead or what to be the more frequent ones but contrary that Hollywood movie stereotypes, the opposite mistake (adwisor and SUW) is the much more frequent one. The problem is compounded by the fact that in some regions the sound shift in modern High German from <w>=[w] to <w>=[v] is not yet completed and the prevalent realization is somewhere in the middle (<w>=[ʋ], as in Dutch).
     

    merquiades

    Senior Member
    English (USA Northeast)
    I'm not sure that more exposure really changes anything, and I'm not sure why.... Ingrained habits too late to correct? I have heard German-Americans who literally have spent their whole life in the US, having immigrated some time after WWII when I guess they were still young. They still pronounce v like w occasionally. Adwisor is a typical example.
    Back to the intrusive h. I have also met French speakers who have spent a lot of time in English speaking countries and/or spent many years studying. They are still quite likely to talk about "hair quality" and say "Hi Hingred. How hare you?". It's one of the hardest things for them to correct.
     

    berndf

    Moderator
    German (Germany)
    I'm not sure that more exposure really changes anything, and I'm not sure why.... Ingrained habits too late to correct? I have heard German-Americans who literally have spent their whole life in the US, having immigrated some time after WWII when I guess they were still young. They still pronounce v like w occasionally. Adwisor is a typical example.
    I didn't say anything else.

    Said *the opposite* mistake (systematically saying [v] instead of [w]) would be due to lack of exposure. And that is the Hollywood movie stereotype (Vhat votsch is it? It is six votch ~ Casablanca).
     
    Top