I removed the entry for "cheshirization". This is clearly a non-established neologism, and there was a previous discussion to eliminate this term [1], so it has no business being here. Please see WP:NEO. Benwing ( talk) 21:05, 16 June 2009 (UTC)
Even if the Cheshirisation article is not removed (I don't actually think that discussion belongs here in the first place), a decision needs to be made on whether 1) to employ the term on this template, as is done now 2) to use an equivalent but more universally accepted term or circumlocution 3) to move the sound changes grouped under it to other categories.
First, the question of an alternative term. The Cheshirisation article says it is "where a trace remains of an otherwise disappeared sound in a word". What is meant by "a trace"? I suppose not a "weaker" version of the sound, or this would be a case of Lenition. So I can only think of two other possibility: a sound in the vicinity of the "cheshirized" sound changes because the latter disappers, or, alternatively, a sound in the vicinity changes (initially creating an allophone) and then, later, the "cheshirized" sound disappear (making the initially allophonic contrast phonemic).
The article seems to support both possibilities in a way: it says "Before disappearing, a sound may trigger or prevent some phonetic change in its vicinity that would not otherwise have occurred" (emphasis mine). But it also gives "For example, in the English word night, the gh sound disappeared, but as it did so it lengthened the vowel i" (emphasis mine again).
I wonder, however: wouldn't the latter be a case of simple assimilation? Everything ( place of articulation, manner of articulation, etc) of the "g" are assimilated to the "i"; the only thing that remains is the original chroneme of the "g", which lengthens the vowel. I admit this might be a bit far-fetched.
In both cases, anyway, it seems clear that "cheshirisation" isn't just any garden-variety generic sound change, but it is, by definition, a phonological change. The article about that mentions phonemic split, and that, to me, seems to be what "cheshirisation" is ultimately about: when an allophonic rule linked to a disappearing phoneme creates a new phonemic differentiation between what were previously just allphones.
Or, in other words, as the Phonemic differentiation article says: "[...] when a phoneme has two allophones appearing in different environments, but sound change eliminates the distinction between the two environments. For example in umlaut in the Germanic languages, the back vowels /u, o/ originally had front rounded allophones [y, ø] before the vowel /i/ in a following syllable. When sound change caused the syllables containing /i/ to be lost, a phonemic split resulted, making /y, ø/ distinct phonemes."
Note that this ( Germanic umlaut) is one of the examples given in the Cheshirisation article.
My conclusion: the concept phonemic split seems to overlap with that of "cheshirisation" in most circumstances, except possibly in the "sound disappears and concurrently a trace of it is left in another sound" case (the "night" case), which might, however, be explained as assimilation (or the assumption of concurrency dropped). So I suggest that the term "phonemic split" be used in this template instead of "cheshirisation" (the latter being possibly a neologism, and apparently debated here), but I do also recommend that someone with more expertise in these topic have a look at the whole template, and consider whether the current categorizations are appropriate, given some of the changes listed are by necessity phonological, and others might not be.
LjL ( talk) 01:22, 17 June 2009 (UTC)
I never understood this template to refer to non-phonological sound change, so I don't understand the objection. Cheshirization is phonological change (though not necessarily introducing new phonemes), as the triggering environment disappears, but many of the topics in this template are or can be phonological change.
As for not all of the linked articles covering sound change, that's a defect with the articles, not the template. Nasalization should cover development. (I just added a stub.) AFAIK, floating tones are always cheshirization. kwami ( talk) 20:18, 17 June 2009 (UTC)
OK, I'm moving this back to the left side. I didn't see the comments here until now, but I made the same point as LjL in Talk:Cheshirisation. Quoting that article:
The basic issue here is that "cheshirization", like "merger" or "split", is describing the result of a phonetic process, while "lenition" or "assimiilation" are describing the nature of the process. I think this is an important distinction and suggests, as I mentioned, either splitting to two templates or two sections of the same template. In addition, the higher-level result-oriented processes are mostly historical, whereas the lower-level nature-oriented ones are clearly either synchronic or diachronic. ("Mostly" because in cases where a historical change leaves an extremely regular alternation, it's arguably best analyzed as a synchronic process. As an example, take a look at the section on vowel processes in Egyptian Arabic -- there are separate, ordered processes of vowel lengthening, stress assignment, vowel deletion, vowel shortening, vowel epenthesis and syllable linking, and are all regular enough to be (arguably at least) considered to be synchronic processes, although they clearly came about due to diachronic changes (e.g. the closely related Levantine dialects in Syria/Lebanon/Israel/Palestine/Jordan have none of these changes but instead a different set of vowel changes, which for the most part are not active synchronically). Benwing ( talk) 06:32, 19 June 2009 (UTC)
BTW I have absolutely no idea why Kwami is making the accusation above that I deleted nasalization, tonogenesis, and floating tones from the template, as it simply isn't true. AFAIK they weren't in the template before, and they're not in there now either -- maybe you intended to put them there but accidentally didn't do it? Look in the history at my one change and you'll see no evidence that I deleted such things. Now, what I did do is delete the text from the "Cheshirization" page and move it to Matisoff's page; on that page were links to nasalization, tonogenesis, and floating tones under "see also", and they got deleted and not replaced on Matisoff's page, since they didn't belong. But that's a far cry from taking them out of the template. Benwing ( talk) 06:43, 19 June 2009 (UTC)
Following the thread is getting a bit confusing for me now, so I'll see if I can make some specific statements that we may agree or disagree about:
(Again, please see and use this test template to check out how the proposed changes look in practice)
I'm saying this regardless of what happens to the "cheshirisation" article itself. I'll discuss that on its talk page.
LjL ( talk) 16:43, 19 June 2009 (UTC)
What's the difference between fusion and coalescence? (And why is one under elision and one under assimilation? It seems clear that both are assimilation.)
If the articles are accurate, the only difference is that fusion is merging of features of sounds, but coalescence is merging of the sounds themselves. If so, the terms are close enough to be covered in one article.
So, is this an accurate analysis? And which term is best for the title of a merged article? — Erutuon ( talk) 23:12, 5 December 2009 (UTC)
Added synalepha under sandhi, with subtypes elision (specifically elision of one vowel in a pair), crasis, synaeresis, synizesis (ordered by how much they change the original vowels: most to least).
Synalepha is usually defined as the uniting of two words by elision of the vowel of one. (This is probably because elision is the only sound-change that occurs in English poetry.) But the OED indicates that the definition includes more types of sound change ("coalescence or contraction of two syllables into one"). The chart of "Types of vowel junction" in Vox Graeca defines synalepha as including crasis, synaeresis, synizesis, and elision (thlipsis).
Since this broader definition agrees with the Greek ("smearing together"), I've glossed synalepha as "contraction". I don't know another general Greek term for English "contraction": all other terms are more specific. (Allen glosses crasis as "contraction", since Greek contraction is crasis, but since English contraction is elision, contraction should be defined as including both.)
Any comments on this analysis are welcome, and especially disagreements with it. — Erutuon ( talk) 21:19, 24 December 2009 (UTC)
This box does not seem to adjust its dimensions like other boxes when I use text zoom in Firefox. This results in words being compressed and overlapped when the lines are too long: http://i44.tinypic.com/izahc7.jpg -- Anthonzi ( talk) 04:54, 6 April 2010 (UTC)
Is it worth adding Lambdacism and/or Lallation to this template? Grutness... wha? 00:58, 13 July 2010 (UTC)
The term "rhinoglottophilia" simply doesn't belong in this template. I am reluctantly tolerating "cheshirization" even though it violates WP:NEO, because it refers to a common sound change. But "rhinoglottophilia" is not only a complete neologism with no major currency, but the concept itself is obscure and not generally accepted. Even the page on this putative type of sound change can only cite a small number of examples in obscure languages. There's a reason why Wikipedia has policies regarding neologisms, fringe theories, and similar stuff that are designed to exclude them. Otherwise we end up completely inappropriately promoting every random scientist's pet theories into prominent parts of a discipline just because some Wikipedia editor happens to really like a particular random scientist (as Kwami appears to in the case of James Matisoff). Benwing ( talk) 01:45, 14 July 2011 (UTC)
I am a teacher of phonology and I agree, both terms do not belong to the list at all. — Preceding unsigned comment added by 192.12.88.15 ( talk) 02:35, 1 September 2012 (UTC)
I don't think "synalepha" belongs as a category under sandhi. The term "elision" is outrighted duplicated here and elsewhere, and the other terms are simply more specific versions of fusion, which already appears under the assimilation category. These terms like "crasis" and "syneresis" are old terms that were created largely to describe specific phenomena in Latin and Greek, and seem to have been applied to a small number of primarily Romance languages, evidently by consciously classicizing Romance-speaking linguists. I really don't think the sound change template should end up being a wastebasket of every term that has ever been used to describe sound changes. As it is, there are already too many terms, and it's just going to end up being overly confusing to the type of reader who would benefit the most from this template. This is especially the case when there is ill-defined overlap between major categories and/or too many major categories.
(For example, rhotacism might be better placed under lenition rather than being its own top-level category, since most cases of rhotacism are in fact lenitions: fricative /z/ to approximant /r/, stop sound /n/ to continuant /r/, etc.)
I removed the entry for "cheshirization". This is clearly a non-established neologism, and there was a previous discussion to eliminate this term [1], so it has no business being here. Please see WP:NEO. Benwing ( talk) 21:05, 16 June 2009 (UTC)
Even if the Cheshirisation article is not removed (I don't actually think that discussion belongs here in the first place), a decision needs to be made on whether 1) to employ the term on this template, as is done now 2) to use an equivalent but more universally accepted term or circumlocution 3) to move the sound changes grouped under it to other categories.
First, the question of an alternative term. The Cheshirisation article says it is "where a trace remains of an otherwise disappeared sound in a word". What is meant by "a trace"? I suppose not a "weaker" version of the sound, or this would be a case of Lenition. So I can only think of two other possibility: a sound in the vicinity of the "cheshirized" sound changes because the latter disappers, or, alternatively, a sound in the vicinity changes (initially creating an allophone) and then, later, the "cheshirized" sound disappear (making the initially allophonic contrast phonemic).
The article seems to support both possibilities in a way: it says "Before disappearing, a sound may trigger or prevent some phonetic change in its vicinity that would not otherwise have occurred" (emphasis mine). But it also gives "For example, in the English word night, the gh sound disappeared, but as it did so it lengthened the vowel i" (emphasis mine again).
I wonder, however: wouldn't the latter be a case of simple assimilation? Everything ( place of articulation, manner of articulation, etc) of the "g" are assimilated to the "i"; the only thing that remains is the original chroneme of the "g", which lengthens the vowel. I admit this might be a bit far-fetched.
In both cases, anyway, it seems clear that "cheshirisation" isn't just any garden-variety generic sound change, but it is, by definition, a phonological change. The article about that mentions phonemic split, and that, to me, seems to be what "cheshirisation" is ultimately about: when an allophonic rule linked to a disappearing phoneme creates a new phonemic differentiation between what were previously just allphones.
Or, in other words, as the Phonemic differentiation article says: "[...] when a phoneme has two allophones appearing in different environments, but sound change eliminates the distinction between the two environments. For example in umlaut in the Germanic languages, the back vowels /u, o/ originally had front rounded allophones [y, ø] before the vowel /i/ in a following syllable. When sound change caused the syllables containing /i/ to be lost, a phonemic split resulted, making /y, ø/ distinct phonemes."
Note that this ( Germanic umlaut) is one of the examples given in the Cheshirisation article.
My conclusion: the concept phonemic split seems to overlap with that of "cheshirisation" in most circumstances, except possibly in the "sound disappears and concurrently a trace of it is left in another sound" case (the "night" case), which might, however, be explained as assimilation (or the assumption of concurrency dropped). So I suggest that the term "phonemic split" be used in this template instead of "cheshirisation" (the latter being possibly a neologism, and apparently debated here), but I do also recommend that someone with more expertise in these topic have a look at the whole template, and consider whether the current categorizations are appropriate, given some of the changes listed are by necessity phonological, and others might not be.
LjL ( talk) 01:22, 17 June 2009 (UTC)
I never understood this template to refer to non-phonological sound change, so I don't understand the objection. Cheshirization is phonological change (though not necessarily introducing new phonemes), as the triggering environment disappears, but many of the topics in this template are or can be phonological change.
As for not all of the linked articles covering sound change, that's a defect with the articles, not the template. Nasalization should cover development. (I just added a stub.) AFAIK, floating tones are always cheshirization. kwami ( talk) 20:18, 17 June 2009 (UTC)
OK, I'm moving this back to the left side. I didn't see the comments here until now, but I made the same point as LjL in Talk:Cheshirisation. Quoting that article:
The basic issue here is that "cheshirization", like "merger" or "split", is describing the result of a phonetic process, while "lenition" or "assimiilation" are describing the nature of the process. I think this is an important distinction and suggests, as I mentioned, either splitting to two templates or two sections of the same template. In addition, the higher-level result-oriented processes are mostly historical, whereas the lower-level nature-oriented ones are clearly either synchronic or diachronic. ("Mostly" because in cases where a historical change leaves an extremely regular alternation, it's arguably best analyzed as a synchronic process. As an example, take a look at the section on vowel processes in Egyptian Arabic -- there are separate, ordered processes of vowel lengthening, stress assignment, vowel deletion, vowel shortening, vowel epenthesis and syllable linking, and are all regular enough to be (arguably at least) considered to be synchronic processes, although they clearly came about due to diachronic changes (e.g. the closely related Levantine dialects in Syria/Lebanon/Israel/Palestine/Jordan have none of these changes but instead a different set of vowel changes, which for the most part are not active synchronically). Benwing ( talk) 06:32, 19 June 2009 (UTC)
BTW I have absolutely no idea why Kwami is making the accusation above that I deleted nasalization, tonogenesis, and floating tones from the template, as it simply isn't true. AFAIK they weren't in the template before, and they're not in there now either -- maybe you intended to put them there but accidentally didn't do it? Look in the history at my one change and you'll see no evidence that I deleted such things. Now, what I did do is delete the text from the "Cheshirization" page and move it to Matisoff's page; on that page were links to nasalization, tonogenesis, and floating tones under "see also", and they got deleted and not replaced on Matisoff's page, since they didn't belong. But that's a far cry from taking them out of the template. Benwing ( talk) 06:43, 19 June 2009 (UTC)
Following the thread is getting a bit confusing for me now, so I'll see if I can make some specific statements that we may agree or disagree about:
(Again, please see and use this test template to check out how the proposed changes look in practice)
I'm saying this regardless of what happens to the "cheshirisation" article itself. I'll discuss that on its talk page.
LjL ( talk) 16:43, 19 June 2009 (UTC)
What's the difference between fusion and coalescence? (And why is one under elision and one under assimilation? It seems clear that both are assimilation.)
If the articles are accurate, the only difference is that fusion is merging of features of sounds, but coalescence is merging of the sounds themselves. If so, the terms are close enough to be covered in one article.
So, is this an accurate analysis? And which term is best for the title of a merged article? — Erutuon ( talk) 23:12, 5 December 2009 (UTC)
Added synalepha under sandhi, with subtypes elision (specifically elision of one vowel in a pair), crasis, synaeresis, synizesis (ordered by how much they change the original vowels: most to least).
Synalepha is usually defined as the uniting of two words by elision of the vowel of one. (This is probably because elision is the only sound-change that occurs in English poetry.) But the OED indicates that the definition includes more types of sound change ("coalescence or contraction of two syllables into one"). The chart of "Types of vowel junction" in Vox Graeca defines synalepha as including crasis, synaeresis, synizesis, and elision (thlipsis).
Since this broader definition agrees with the Greek ("smearing together"), I've glossed synalepha as "contraction". I don't know another general Greek term for English "contraction": all other terms are more specific. (Allen glosses crasis as "contraction", since Greek contraction is crasis, but since English contraction is elision, contraction should be defined as including both.)
Any comments on this analysis are welcome, and especially disagreements with it. — Erutuon ( talk) 21:19, 24 December 2009 (UTC)
This box does not seem to adjust its dimensions like other boxes when I use text zoom in Firefox. This results in words being compressed and overlapped when the lines are too long: http://i44.tinypic.com/izahc7.jpg -- Anthonzi ( talk) 04:54, 6 April 2010 (UTC)
Is it worth adding Lambdacism and/or Lallation to this template? Grutness... wha? 00:58, 13 July 2010 (UTC)
The term "rhinoglottophilia" simply doesn't belong in this template. I am reluctantly tolerating "cheshirization" even though it violates WP:NEO, because it refers to a common sound change. But "rhinoglottophilia" is not only a complete neologism with no major currency, but the concept itself is obscure and not generally accepted. Even the page on this putative type of sound change can only cite a small number of examples in obscure languages. There's a reason why Wikipedia has policies regarding neologisms, fringe theories, and similar stuff that are designed to exclude them. Otherwise we end up completely inappropriately promoting every random scientist's pet theories into prominent parts of a discipline just because some Wikipedia editor happens to really like a particular random scientist (as Kwami appears to in the case of James Matisoff). Benwing ( talk) 01:45, 14 July 2011 (UTC)
I am a teacher of phonology and I agree, both terms do not belong to the list at all. — Preceding unsigned comment added by 192.12.88.15 ( talk) 02:35, 1 September 2012 (UTC)
I don't think "synalepha" belongs as a category under sandhi. The term "elision" is outrighted duplicated here and elsewhere, and the other terms are simply more specific versions of fusion, which already appears under the assimilation category. These terms like "crasis" and "syneresis" are old terms that were created largely to describe specific phenomena in Latin and Greek, and seem to have been applied to a small number of primarily Romance languages, evidently by consciously classicizing Romance-speaking linguists. I really don't think the sound change template should end up being a wastebasket of every term that has ever been used to describe sound changes. As it is, there are already too many terms, and it's just going to end up being overly confusing to the type of reader who would benefit the most from this template. This is especially the case when there is ill-defined overlap between major categories and/or too many major categories.
(For example, rhotacism might be better placed under lenition rather than being its own top-level category, since most cases of rhotacism are in fact lenitions: fricative /z/ to approximant /r/, stop sound /n/ to continuant /r/, etc.)