Gradual Loss of Phonemes and Handling the Eventual Confusion

For the discussion of language mechanics, grammar, vocabulary, trends, and other such linguistic topics, in english and other languages.

Moderators: gmalivuk, Moderators General, Prelates

ri.kenji
Posts: 5
Joined: Thu Jun 21, 2012 9:14 pm UTC

Gradual Loss of Phonemes and Handling the Eventual Confusion

Postby ri.kenji » Thu Jun 21, 2012 11:32 pm UTC

I've noticed in many languages, there seems to be a gradual loss of phonemes over long periods of times. Many languages like Chinese and Korean have long shed their rich phonetic inventories, and its descendent dialects have also begun to merge sounds.

Here are some examples:
  • Chinese (Mandarin)
    • Merger of velarized dental affricates and palatalized velar plosives
      This is best exemplified by the terms "Peking duck" and "Peking opera," where "Peking" ([peikiŋ]) was the approximate pronunciation of 北京 before [ki] was palatalized to [tɕi].
  • Chinese (Cantonese)
    • Merger of affricate phonemes [ts] and [tʃ], [tsʰ] and [tʃʰ], and [s] and [ʃ]
      These sounds were separate phonemes during the colonial era, and are still distinguished in Mandarin Chinese.
    • Confusion of [l] and [n]
      In Hong Kong, [n] is frequently pronounced as [l]. Over there, such a pronunciation is termed "lazy tongue."
  • Chinese (Fuzhou Dialect)
    • Merger of rimes [ieu] and [iu] or [eu], and [uai] and [uoi]
      The historical pronunciation is only reflected in the romanization and Qieyin system.
  • Korean (Seoul)
    • Merger of ㅐ ([ai]) and ㅔ ([ei])
      Both ㅐ and ㅔ are pronounced approximately as /e/ by the younger speakers. Older speakers might pronounce them as /ɛ/ and /e/ respectively. This has resulted in creative ways to differentiate between the possessive pronouns 내 ([nai]) and 네 ([nei]) which mean "mine" and "yours" respectively. The former keeps its new pronunciation, and the latter is pronounced 니 ([ni]). Interestingly, this is one of the cases where the orthography becomes non-phonemic as the speaker must know when the syllable represents "yours" and adjust the pronunciation accordingly.
    • Merger of ㅚ ([oi]), ㅙ (oai), and ㅞ (uei)
      All of them are pronounced /we/ by younger speakers. Older speakers might pronounce them as /ø/, /wɛ/, and /we/ respectively.
    • Merger of ㅢ ([ɯi]) and ㅣ ([i]) or ㅔ ([ei])
      ㅢ ([ɯi]) and ㅣ ([i]) are both pronounced as /i/ when preceded by a consonant. The possessive particle 의 ([ɯi]) is specially pronounced as 에 (/e/)—another break in Korean's phonemic orthography.
    • Weakening or loss of ㄴ ([n]) and ㄹ ([l])
      The initial consonant ㄹ ([l]) is frequently weakened to ㄴ ([n]), and both ㄴ ([n]) and ㄹ ([l]) are commonly deleted altogether from the syllable when followed by /i/ or /j/. The Korean surname 李, for example, is romanized as Lee for English speakers, but pronounced as /ji/ in Korean.
  • English (Southern United States)
    • Merger of [ɪ] and [ɛ] before nasal consonants
      Words like "pin" and "pen" sound the same, and must be disambiguated by the terms "stick pin" and "ink pen" respectively. This is similar to the process that created multisyllabic terms in all of the modern Chinese dialects.
Further erosion of phonemes will probably continue into the foreseeable future. The question is… how should this be dealt with? Phonemic orthographies, romanization systems, and lexicons will have to be realigned (lest they decide to follow the example set by English) or they'll have to actively discourage change by younger speakers (Ebonics versus Standard American English).

Derek
Posts: 2156
Joined: Wed Aug 18, 2010 4:15 am UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby Derek » Fri Jun 22, 2012 2:36 am UTC

There is no problem to deal with. Languages change. Mergers happen, but so do splits and borrowings that add new sounds to a language. While I merge "pin" and "pen", I also have /æ/-tensing, in which /æ/ before nasals is raised to something like [ɛə]. This is not yet phonemic for me, but for some speakers it is: They pronounce "can" (to be able to) as /kæn/ and "can" (metal container) as /keən/.

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 25852
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby gmalivuk » Fri Jun 22, 2012 3:58 am UTC

ri.kenji wrote:Further erosion of phonemes will probably continue into the foreseeable future.
Old English had 7 vowels (though admittedly short an long were contrasted), my dialect of English has 15 or 16, and RP has 20, on account of they don't pronounce /r/ properly. Old English had about 18 contrasted consonants, most modern varieties have 24 or so. I really don't see where this "erosion" is happening long term, but even if it is, 25% fewer phonemes didn't seem to hurt the Anglo Saxons all that much, so I'm not terribly worried.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

goofy
Posts: 911
Joined: Thu May 01, 2008 3:32 pm UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby goofy » Fri Jun 22, 2012 1:14 pm UTC

ri.kenji wrote:Further erosion of phonemes will probably continue into the foreseeable future.


Haven't we been using language for thousands of years? Why haven't we lost all our phonemes already?

There's plenty of examples of new phonemes appearing. For example breaking in Old English, where short front vowels changed in certain environments, leading to the creation of a new vowel. Or palatalization, where /k/ becomes /tʃ/ before front vowels - that happens in lots of languages.

Chinchokmataa
Posts: 19
Joined: Thu May 17, 2012 3:30 pm UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby Chinchokmataa » Sat Jun 23, 2012 12:43 am UTC

Even if there was a general trend towards losing phonemes (which is demonstratibly untrue), there is also a problem with thinking that the totally normal process of phoneme merger will lead to such an "eventual confusion" that it needs to be handled by an outside force. Languages already have means of distinguising between homophones, the main one being the magical force that is context! Human beings understand words by the company that they keep, and the physical environment that they are used in. If I'm in class, and the person next to me asks if I have a [pɪn] I'm going to hand them "writy-pointy-object" not "pokey-pointy-object." If someone were to say to me "That [bɪn] is a liar" I'm gonna assume they're talking about "person-Ben" and not "plastic-bin," because "plastic-bin" is not in the lying business. If there is a very serious need to make the two clear, all you have to do is add extra words, like you've already demonstrated with your "ink pen" and "stick pin" examples. There is nothing wrong with that, the same information is being communicated.

Derek
Posts: 2156
Joined: Wed Aug 18, 2010 4:15 am UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby Derek » Sat Jun 23, 2012 1:10 am UTC

I believe English also has a relatively high number of phonemes right now (although I know there are languages with more), so it wouldn't surprise me if it is undergoing a net loss at the moment, a regression towards the mean.

User avatar
Iulus Cofield
WINNING
Posts: 2917
Joined: Wed Apr 07, 2010 9:31 am UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby Iulus Cofield » Sat Jun 23, 2012 8:31 am UTC

You mean the phonemic inventories of all the languages of the world won't eventually be reduced to just /a/?

ri.kenji
Posts: 5
Joined: Thu Jun 21, 2012 9:14 pm UTC

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby ri.kenji » Sun Jun 24, 2012 1:48 pm UTC

goofy wrote:Haven't we been using language for thousands of years? Why haven't we lost all our phonemes already?

You're mistaken in assuming that I said phonemes would eventually converge towards zero. The trend is for a language's phonemic inventory to converge towards a mean. I think this post says it best what I'm trying to express:
Derek wrote:I believe English also has a relatively high number of phonemes right now (although I know there are languages with more), so it wouldn't surprise me if it is undergoing a net loss at the moment, a regression towards the mean.


I'm saying there is a trend towards a certain number of phonemes. For many languages, this is a reduction in phonemes and at the same time an increase in some other quantity to compensate for the confusion. Chinese, for example, had monosyllabic words which eventually became disyllable and even multisyllabic words since the spoken language had lost enough phonemes that combinations of syllables were necessary to maintain distinctions between words. Basically, written communication (which was more common at the time) was very different than spoken communication, and it was only recently that written Chinese was done in the vernacular.

While English may have increased its number of phonemes for a period of time, that's an exception to the trend or an outlier. And I think you're all missing the point of my question: how should languages (in general) deal with changes (such as the Chinese example I described)? We already have an example of what happens when nothing is done about to the written languages such as English—the alphabet becomes much less phonemic and confusing.

Chinchokmataa wrote:Even if there was a general trend towards losing phonemes (which is demonstratibly untrue), there is also a problem with thinking that the totally normal process of phoneme merger will lead to such an "eventual confusion" that it needs to be handled by an outside force. Languages already have means of distinguising between homophones, the main one being the magical force that is context! Human beings understand words by the company that they keep, and the physical environment that they are used in. If I'm in class, and the person next to me asks if I have a [pɪn] I'm going to hand them "writy-pointy-object" not "pokey-pointy-object." If someone were to say to me "That [bɪn] is a liar" I'm gonna assume they're talking about "person-Ben" and not "plastic-bin," because "plastic-bin" is not in the lying business. If there is a very serious need to make the two clear, all you have to do is add extra words, like you've already demonstrated with your "ink pen" and "stick pin" examples. There is nothing wrong with that, the same information is being communicated.

This "magic" doesn't always work. For Korean (which I described in my original post), the words for first person and second person pronouns can't always be disambiguated by context since they sound the same if pronounced as written. They now differentiate it by pronouncing one of the pronouns differently (even if still written with the ambiguous pronunciation). This goes back to my original question: regulate the spoken language, or realign the orthography to reflect the new pronunciation?

User avatar
gmalivuk
GNU Terry Pratchett
Posts: 25852
Joined: Wed Feb 28, 2007 6:02 pm UTC
Location: Here and There
Contact:

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby gmalivuk » Sun Jun 24, 2012 2:41 pm UTC

ri.kenji wrote:I'm saying there is a trend towards a certain number of phonemes.
I strongly doubt this as an overarching trend. There's more likely an effective minimum and maximum for the amount of phonemes a language can maintain without becoming too confusing or too cumbersome for its speakers, and the actual number tends to fluctuate between these two.
Unless stated otherwise, I do not care whether a statement, by itself, constitutes a persuasive political argument. I care whether it's true.
---
If this post has math that doesn't work for you, use TeX the World for Firefox or Chrome

(he/him/his)

Tropylium
Posts: 127
Joined: Tue Jul 25, 2006 9:40 am UTC
Location: Finland

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby Tropylium » Tue Sep 25, 2012 11:11 pm UTC

/wanders in first time in years, immediately delurks to post an essay on something

ri.kenji wrote:
goofy wrote:Haven't we been using language for thousands of years? Why haven't we lost all our phonemes already?

You're mistaken in assuming that I said phonemes would eventually converge towards zero. The trend is for a language's phonemic inventory to converge towards a mean.

It is a good question regardless. Could be restated as: Given that there are more examples of phoneme loss than phoneme introduction, what then are the actual mechanisms that prevent phonemic inventories from reducing to zero?

The answer would appear to be that while sound changes introducing new phonemes are rarer, they also tend to introduce multiple phonemes at once.
• Loss of a single consonant or some other feature, with compensatory lengthening, will immediately double the vowel inventory. (This is underway in some American English dialects, which realize all vowels as long before voiced consonants and short before voiceless; and has recently occurred in Turkish, where <ğ> has muted.) A general loss of vowel length is rarer, and only tends to occur once individual vowel contrasts have expanded to include distinctions other than length (e.g. in the development of both Vulgar Latin and American English, only /a/ and /aː/ merged, while /i e o u/ first became lowered to /ɪ ɛ ɔ ʊ/, so loss of length from /iː eː ᴏː uː/ did not lead to a merger. (That there was a later merger of /ɪ ʊ/ into /e o/ in VL, or that in AE, /a ɔ/ rotated to /æ a/ before the loss of length does not affect the point.))
• Consonant palatalization, labialization or the like will frequently apply to all consonants (or several POA classes of consonants) at once; if something renders this phonemic (e.g. loss of word-final /i/, as happened in Estonian), the inventory will suddendly grow a whole lot. Again, the reverse is rare — a gradual shrinking of the secondarily articulated inventory via assibilation etc. is more typical (just look at the history of Slavic).
• Anything that affects phonation (voicing, aspiration…) will tend to affect all similar segments. E.g. in Indo-Aryan, voiceless stop + laryngeal clusters fused into voiceless aspirate stops, creating five new segments /pʰ tʰ ʈʰ cʰ kʰ/ that could all be phonemicized by the loss of a single phoneme *H; in Hungarian (and rather similarly in Old Japanese), simplification of nasal+stop clusters introduced four voiced stops /b d ɟ g/. A loss of a contrast entirely is not too unusual here tho (eg. a merger of PIE "voiced aspirate" and "plain voiced" stop series was quite common).

User avatar
steewi
Posts: 873
Joined: Mon Feb 18, 2008 3:38 am UTC
Location: Tropical Nowhere

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby steewi » Thu Sep 27, 2012 5:04 am UTC

Tropylium has it pretty much right as far as I'm concerned.

I'll add that the loss of a distinction doesn't have to result in a different *segmental* contrast - it might head suprasegmental. Tone is a very common one (Vietnamese and Lhasa Tibetan, among others), but conceivably, autosegmental palatalisation or nasalisation, contrastive stress and non-modal phonation are all possibilities.

User avatar
tomandlu
Posts: 1058
Joined: Fri Sep 21, 2007 10:22 am UTC
Location: London, UK
Contact:

Re: Gradual Loss of Phonemes and Handling the Eventual Confu

Postby tomandlu » Fri Feb 15, 2013 3:20 pm UTC

How can I think my way out of the problem when the problem is the way I think?


Return to “Language/Linguistics”

Who is online

Users browsing this forum: No registered users and 3 guests