Reasons for h-dropping in languages

[dihydrogen monoxide]

What is it in the phoneme/sound h that is often left unpronounced in most languages, at least at the beginning of the word. The languages would be Serbo-Croatian (dialects), English (dialects), French (according to somewhere on this forum the French dropped the sound but it appeared in writing due to Latin influence), Spanish and so on.
In other words what would be the cause of h-dropping, what is so difficult about this phoneme? I'd understand if it was a cluster of some kind, but I don't see any difficulty with phoneme h.

 

[Sepia]

What is it in the phoneme/sound h that is often left unpronounced in most languages, at least at the beginning of the word. The languages would be Serbo-Croatian (dialects), English (dialects), French (according to somewhere on this forum the French dropped the sound but it appeared in writing due to Latin influence), Spanish and so on.
In other words what would be the cause of h-dropping, what is so difficult about this phoneme? I'd understand if it was a cluster of some kind, but I don't see any difficulty with phoneme h.

I think it would be too narrow-minded only to look at the "h" in this connection. If you take a closer look at what you do with your body to pronounce an "h": you build up tension around the diaphragm in a similar way as you do when you pronounce the letters p t s and a few others in many Germanic languages.

As there is not much difference between a "b" and a "p" - actually the only difference is the pressure you build up behind your lips - you'll notice that just as difficult it is to a Francophone to pronounce an "h", many people of non-Romance languages have difficulties to pronounce a "b" in a way that the Francophone don't understand as a "p".

So I figure the right thing would be to consider these phenomena as a complete packet - maybe even including the Andalusian "s" dropping.

 

[berndf]

In the case of French (and maybe also in Cockney) it is not so much an dropping but rather a [?] insertion; at least in the case of the “ h aspiré”. You often hear people saying the “h” in “le hérisson” be mute. I would say it is not mute but it is pronounced [?] or even [?h] (aspirated glottal stop).

 

[Fred_C]

In the case of French (and maybe also in Cockney) it is not so much an dropping but rather a [?] insertion; at least in the case of the “ h aspiré”. You often hear people saying the “h” in “le hérisson” be mute. I would say it is not mute but it is pronounced [?] or even [?h] (aspirated glottal stop).

Hi,
In my opinion,
le hérisson is seldomly pronounced [l@?eRisO~], and almost always [l@eRisO~]. The glottal stop is more a pause, you use it when you speak very slowly.
I have never heard it pronounced [l@?heRisO~]

 

[Kanes]

In Bulgarian it is usually in the front of the word and droping it would not change the meaning.

 

[Athaulf]

I think it would be too narrow-minded only to look at the "h" in this connection. If you take a closer look at what you do with your body to pronounce an "h": you build up tension around the diaphragm in a similar way as you do when you pronounce the letters p t s and a few others in many Germanic languages.

"Tension around the diaphragm"? Do you have any references for this claim? I don't feel any extra effort on the part of my diaphragm when pronouncing any of these sounds you mention. Also, your list includes both plosives and fricatives, so I really don't see what they might have in common in this regard.

As there is not much difference between a "b" and a "p" - actually the only difference is the pressure you build up behind your lips -

No, this is completely inaccurate. The difference between /b/ and /p/ is in voicing, i.e. whether you let the vocal chords in your throat vibrate while doing the same movements with your lungs and mouth.

 

[berndf]

No, this is completely inaccurate. The difference between /b/ and /p/ is in voicing, i.e. whether you let the vocal chords in your throat vibrate while doing the same movements with your lungs and mouth.

I think you two have different notions of /b/ and /p/ and what "voicing" actually means. German voiced plosives are very different from those in Romance or Slavic languages. A German would understand /b/ instead of /p/ already, if the VOT is less than about +15 or /g/ instead of /k/, if the VOT is less than about +30ms while Romance languages, and even more so Slavic languages, require a noticeably negative VOT to regard a plosive as voiced. A Slavic /b/ sounds very bizarre to a German, as if the speaker is going to vomit because he/she starts to make sounds before the plosive release.

 

[berndf]

Hi,
In my opinion,
le hérisson is seldomly pronounced [l@?eRisO~], and almost always [l@eRisO~]. The glottal stop is more a pause, you use it when you speak very slowly.
I have never heard it pronounced [l@?heRisO~]

[?h-] is probably exaggerated. I won't insist on this. But I would still contend that the pronunciation you hear here (http://en.wiktionary.org/wiki/h%C3%A9risson) is the standard one with a very clear [?].

I once noticed that many French speakers, when trying to pronounce an "h" in foreign languages, e.g. if pronouncing "hat" in English, try to say [h?æt] which is of course all but impossible. It usually helps explaining that the "h" will come naturally once you stop saying [?].

 

[Fred_C]

Hi, I am afraid you are not quite right.
The glottal stop is not a phoneme or an allophone in French, and is absolutely irrelevant to standard French.
It is just a feature that may or may not accidentally happen in front of a vowel, if you have just made a pause in speech.

I am insisting on this, because I know that things are very different in German, where the correct and standard pronunciation requires a glottal stop in front of a leading vowel, and that this glottal stop is sometimes dropped in fast speech, although technically, it should not.

I listened to your audio sample, you are right that a glottal stop can be heard, but I maintain that it is as irrelevant to standard pronunciation as the noise of your breath when you stop speaking... (!)
You could also have heard a [?] between "à" and "Orléans" in "à Orléans", although there is no aspirated H there.

As for the pronunciation of by French speakers, it is true that some speakers find it difficult, but in my opinion, they do not represent a majority. Only the poorly language skilled may encounter difficulties.
Things are different for other phonemes, like the voiced or voiceless TH or the velar L, which are difficult for many French speakers.

 

[berndf]

Fred_C, you are absolutely right the glottal stop has non-phonemic status in French. This was very the point of my remark. For this reason it went unnoticed that the h has not simply been lost but has been replaced.

By the way, I don’t thing [?] has phonemic status in German either. There are a few words where it is important, like in “beenden”. But I would dare to bet quite a few beers that most native speakers would describe the glottal stop in “beenden” as a pause and not as a sound, unless they had linguistic training or speak a language where the glottal stop has phonemic status, like Hebrew or Arabic.

We, Germans, have the opposite problems with aspirated plosives. As we wouldn’t hear an unvoiced plosive unless it is aspirated, we sometimes mistake a “t” for a “d” or a “c” or “q” for a “g”. I actually often hear “guatre” when a native speaker says “quatre”.

 

[Athaulf]

I think you two have different notions of /b/ and /p/ and what "voicing" actually means. German voiced plosives are very different from those in Romance or Slavic languages. A German would understand /b/ instead of /p/ already, if the VOT is less than about +15 or /g/ instead of /k/, if the VOT is less than about +30ms while Romance languages, and even more so Slavic languages, require a noticeably negative VOT to regard a plosive as voiced.

A Slavic /b/ sounds very bizarre to a German, as if the speaker is going to vomit because he/she starts to make sounds before the plosive release.

But when does German /b/ have a positive VOT in its actual pronunciation? My (perhaps wrong) impression is that it can't happen too often, because German /b/ and /p/ usually sound quite ordinary to Croatian speakers. It's not just me -- there is a huge number of German loanwords in colloquial Croatian, and in the vast majority of cases, they have /b/ and /p/ in place of the corresponding German phonemes (and even when not, it's often due to Croatian assimilation rules). See e.g. here for a bunch of examples. It's noteworthy that these words have always existed only in the informal language, and nobody ever consciously tried to make them sound closer to the originals, so they have actually gone through any sound changes that came naturally to Croatian speakers. Admittedly, I can think of a few examples where German /b/ became Croatian /p/ even where it wouldn't violate the Croatian assimilation rules, e.g. peglati (from bügeln) or protvan (from Bratpfanne).


In any case, are there any additional differences between the German /b/ and /p/ that could be reasonably described as "the pressure you build up behind your lips"?

 

[berndf]

What I meant is that a zero VOT is sufficient for a plosive to be considered voiced in German. I have read that 15 or 30ms are the limits below which you can't hear a positive VOT any more.

Maybe I misunderstood Sepia, but I interpreted "the pressure you build up behind your lips" to refer to /p/ and not for the /b/. To avoid confusion with /b/ you have to aspirate the /p/ rather strongly in German and for doing so you need to “build up pressure behind your lips” before the plosive release.

 

[Jack Naples]

Hi,

This is my first post. I thought I'd join here before my lingusitics degree ('99) went completely to waste...

H is a very weak (quiet) sound, really just a glide rather than a full consonant. Consonant sound change usually involves weakening along the following scale (which increases in ease of articulation):

stop -> fricative -> glide -> zero.

H almost never goes back up the scale to become a fricative, so it can only disappear.

 

[Ajura]

Hi,

This is my first post. I thought I'd join here before my lingusitics degree ('99) went completely to waste...

H is a very weak (quiet) sound, really just a glide rather than a full consonant. Consonant sound change usually involves weakening along the following scale (which increases in ease of articulation):

stop -> fricative -> glide -> zero.

H almost never goes back up the scale to become a fricative, so it can only disappear.

Palatalization and assimilation reverses that.

 

[Jack Naples]

Ajura,

Please give some examples.

Cheers,
Jack

 

[Outsider]

H-dropping is a form of lenition, as Jack says (welcome to the forum, BTW, Jack ). This does not mean, however, that sound changes in the reverse direction are impossible.

 

[MarX]

What is it in the phoneme/sound h that is often left unpronounced in most languages, at least at the beginning of the word. The languages would be Serbo-Croatian (dialects), English (dialects), French (according to somewhere on this forum the French dropped the sound but it appeared in writing due to Latin influence), Spanish and so on.
In other words what would be the cause of h-dropping, what is so difficult about this phoneme? I'd understand if it was a cluster of some kind, but I don't see any difficulty with phoneme h.

Good question.
In Indonesian there has also always been a tendency to drop the H in pronunciation.
Unfortunately I don't know the reason behind it.

 

[ff_crafter]

Good question.
In Indonesian there has also always been a tendency to drop the H in pronunciation.
Unfortunately I don't know the reason behind it.

The main reason for H-dropping in Indonesian is because of affected by the local dialect/language, especially people that live in Jakarta metropolitan area because Betawi Malay drop the H so they also drop H when speaking in Indonesian.

 

[Penyafort]

For the Romance languages, the silent h is mostly due to etymological preservation (some like Italian and Occitan don't even write them), so we should go back to Latin and see why there was h-dropping already there.

But it's a different case from some aspirated h's in French.

Then there's the Spanish and Gascon silent h coming from aspirated h's which come from Latin f-, which has been commonly attributed to Basque influence, since Basque didn't have initial f's. (FERRU > Gascon hèr [εr], Spanish hierro ['jero] (Medieval, dialectal ['hjero].

And linking to the previous one, Basque itself did h-drop, as Old Basque and some modern dialects aspirated h's.

 

[merquiades]

It takes a lot of effort to pronounce an h. Just try saying "Her hammer hit his head." I'm sure someday it will be lost in English too. It's already quite common for some speakers to leave the h in difficult clusters like "human", "huge", "historical" and in frequent pronouns "Give it to 'im", "Where is 'e?" The letter h itself has lost its h.

 

[OBrasilo]

Admittedly, I can think of a few examples where German /b/ became Croatian /p/ even where it wouldn't violate the Croatian assimilation rules, e.g. peglati (from bügeln) or protvan (from Bratpfanne).

Slovenian has more examples, for example pir or per from Bier (beer), as well as puter from Butter (butter), also in toponymy, such as Pliberk from Bleiburg, which also indicates that intervocalic voiced consonants in German are more voiced than those in other positions.

 

[Penyafort]

It takes a lot of effort to pronounce an h. Just try saying "Her hammer hit his head." I'm sure someday it will be lost in English too. It's already quite common for some speakers to leave the h in difficult clusters like "human", "huge", "historical" and in frequent pronouns "Give it to 'im", "Where is 'e?" The letter h itself has lost its h.

But the h in human and huge is almost a /ç/, so no wonder it's difficult for many.

 

[berndf]

It takes a lot of effort to pronounce an h. Just try saying "Her hammer hit his head." I'm sure someday it will be lost in English too. It's already quite common for some speakers to leave the h in difficult clusters like "human", "huge", "historical" and in frequent pronouns "Give it to 'im", "Where is 'e?" The letter h itself has lost its h.

Reducing unstressed him to 'im is as frequent as reducing unstressed them to 'em and this is not a sign that th is disappearing but is rather a peculiarity of unstressed pronouns.
I have never encountered h-dropping in historical except for with speakers of a dialect where h-dropping is a general feature, as, e.g. in Eastern London accents. The h-less variant of herb as it is often heard in North America is the original English pronunciation and the h-restoration in writing is an artificial re-Latinization and the pronunciation variant with h is spelling pronunciation. There is no general tendency in late Modern English to drop h but rather to restore etymological h, as in words like hotel or hospital, which are practically always pronounced with an audible h today. Human and huge start with the consonant cluster /hj-/ which is realized either as [j-] or as [ç-]. Such simplifications of onset clusters are common as, e.g., in knee, write and pneumatic. And these simplifications as not a sign of k, w or p disappearing.

As for the pronunciation for the name of the letter, I believe in Ireland haitch (rather than aitch) is the prevalent pronunciation. I hear the variety haitch quite frequently in the London area as well but there it is probably a hyper-correction. The h-less standard variant of the letter name is most likely derived from Old French, where is was pronounced [at͜ʃ] rather than [aʃ] as in modern French. The initial [a] subsequently underwent the diphthongization and the long a of the GVS, which produced the modern pronunciation aitch.

 

[Welsh_Sion]

"To herr is 'uman, to forgive, divine".

- Traditional Cockney

 

[berndf]

"To herr is 'uman, to forgive, divine".

- Traditional Cockney

Lovely example of hyper-correction.

 

[Welsh_Sion]

Nah. "Luvverley".

(c) Eliza Doolittle
1913
All Rights Reserved

 

[Welsh_Sion]

On a more serious note, a lot of speakers of Southern Welsh drop the 'h' - especially in word initial position. This also applies to 'ch' and 'rh' at the beginning of words. This also has a knock on effect in Nasal Mutation where initial 'ngh', 'mh' and 'nh' (nasalised 'c', 'p' and 't') become indistinguishable from 'ng', 'm' and 'n' - which exist already for nasally mutated forms of 'g', 'b' and 'd'.

This 'h dropping' is said to be spreading from the South, although I have yet to hear a Northerner (like me) say 'wara' for 'chwarae' ('playing'). I think Aitchison (!) is wrong in this respect.

 

[Linnets]

In Romance language Latin h has disappeared everywhere because it was not pronounced in Vulgar Latin, but new h's emerged during the developement of languages such as Spanish /f-/ > /h-/ and French h aspiré. Now the only major Romance language that has a pronounced h- at the beginning of a word is Romanian.

 

[Sobakus]

The letter h itself has lost its h.

The h-less standard variant of the letter name is most likely derived from Old French, where is was pronounced [at͜ʃ] rather than [aʃ] as in modern French.

That ch is actually the outcome of the geminate h evidently taught to the French by British, or at any rate foreign, grammarians but eventually adapted as /k/, which later semi-regularly palatalised - compare the Italian name of the letter, acca. So [ɑxːə] > [akːə] > [acːə] > [a:t͜ʃə] > Middle English. I'm not entirely sure why the /a/ lengthened but it had to be short originally if the lack of the expected [aː > æː] is any indication (cf. hache < [hapça]). Or it could have been [ɑçːə] from the start, which would easier explain the palatalisation.

 

[berndf]

I'm not entirely sure why the /a/ lengthened

I don't think it was. The a of French ach was probably imported directly as long into Middle English. I don't think the letter name is inherited from Old English. At least, I am not aware of any evidence it was.

 

[Welsh_Sion]

My Oxford English Concise Dictionary cites the origin of 'aitch' as mid 16th century, from Old French, ache.

GPC says the Welsh borrowed it similarly after the Norman Conquest (1282-84) as we call the letter today aets. (Though prior to the 13th century what we called the letter, I have no idea!). Haitch is considered to be 'a vulgarism' used in the South West of Wales.

 

[Forero]

Reducing unstressed him to 'im is as frequent as reducing unstressed them to 'em and this is not a sign that th is disappearing but is rather a peculiarity of unstressed pronouns.
I have never encountered h-dropping in historical except for with speakers of a dialect where h-dropping is a general feature, as, e.g. in Eastern London accents. The h-less variant of herb as it is often heard in North America is the original English pronunciation and the h-restoration in writing is an artificial re-Latinization and the pronunciation variant with h is spelling pronunciation. There is no general tendency in late Modern English to drop h but rather to restore etymological h, as in words like hotel or hospital, which are practically always pronounced with an audible h today. Human and huge start with the consonant cluster /hj-/ which is realized either as [j-] or as [ç-]. Such simplifications of onset clusters are common as, e.g., in knee, write and pneumatic. And these simplifications as not a sign of k, w or p disappearing.

The pronoun 'em is actually what is left of the native pronoun hem. They, them, and their are based on Old Norse forms and must have been adapted into English to prevent ambiguity.

Also to prevent ambiguity, 'im is often pronounced with the vowel [i] rather than ([ə] or) [ɪ].

In my dialect, huge and human always start with an unvoiced version of [j] (almost [ç]), because their first syllable is always stressed, and wh- words also start with unvoiced sounds (usually [ʍ], but [h][/I] for who- words).

Curiously, the word weapon sometimes gets that initial wh sound ([ʍ]) too.

Most wh- and h- words tend to lose some or all of their aitchiness whenever they are not stressed. This includes how, he, him, her, here, and even historic and historical. These can all retain the full [ʍ] or [h] sound, or be more or less lenis or voiced, depending on the preceding syllable and the speaker's tempo. The phrase "a historic occasion" automatically becomes "an historic occasion" as soon as the [h]
becomes in any way weak.

"C'm'ere" for "Come here" is common, even though this 'ere is stressed, and a lot of English speakers regularly replace [ʍ] with [w] even when stressed and pronounce huge, human, and Houston as "youge", "youman", and "Youston".

The words root and reindeer lost their hs a long time ago.

 

[berndf]

I am not sure [ʍ]-[w] should be understood as a h loss. [ʍ] is what is left of a once independent phoneme. In many parts of the English speaking world this merger is complete, by the way.

Root has lost a w and not an h (preserved, e.g., in High German Wurzel, Dutch/LowGerman wortel). This happened in late OE, probably under ON influence.

 

[Forero]

I am not sure [ʍ]-[w] should be understood as a h loss. [ʍ] is what is left of a once independent phoneme. In many parts of the English speaking world this merger is complete, by the way.

Before it was [ʍ], it was [hw]~[hʍ]. [ h ] merged with [w] the same way it merged with [j].

Root has lost a w and not an h (preserved, e.g., in High German Wurzel, Dutch/LowGerman wortel). This happened in late OE, probably under ON influence.

That makes sense. I must have been thinking of roof.

 

[berndf]

Before it was [ʍ], it was [hw]~[hʍ]. [ h ] merged with [w] the same way it merged with [j].

hw is a phonemic reinterpretation. It was originally xʷ, the reflex of PIE kʷ (compare English what and Latin quod). ʍ does not originate as a merger of h and w. One might argue the the phonemic re-interpretation as /hw/ was already firmly established by the time in w-ʍ merger happened (in those regions where it did happen) and that the merger can therefore be interpreted as an instance of h-dropping. But I am not convinced.

 

[merquiades]

It may well be a reinterpretation but acoustically it does sound like /hw/ to me. I'm going through a region where they speak this way and I hear /hwait/ and /hwat/.