Differences

This shows you the differences between two versions of the page.

--- 2.inventories:2.2.counterparts:2.2.1.alternations [2021-08-09] – László Fejes
+++ 2.inventories:2.2.counterparts:2.2.1.alternations [2021-08-17] (current) – László Fejes
@@ Line 5: / Line 5: @@
 ^ R%%|%%T | — | — | — | — | 1/1 | — | **1,**0/3 | ? | — | — | **5,**3/1| — |
-The formula //a(b)/c(d)// reads: there are //a// alternating pairs (//b// if long/reduced ones are counted separately) and there are //c// potentially, but not practically alternating pairs (//d// if long/reduced ones are counted separately).
+<fs smaller>
+F/B: front/back harmony\\
+R|T: rounding, total harmony\\
+—: not applicable
+The formula //a(b)/c(d)// reads: there are //a// alternating pairs (//b// if long/reduced ones are counted separately) and there are //c// potentially, but not practically alternating pairs (//d// if long/reduced ones are counted separately).</fs>
 Long variants of alternating pairs in parentheses, not potential, but alternating pairs emboldened, potential, but not alternating pairs following the bar, asterix indicates that  the vowel is prohibited or strongly dispreferred in non-initial syllables.
@@ Line 16: / Line 22: @@
   ; Erzya
-  : o : e (There are no vowels differing only in F/B. If we ignore the difference in liability, /i/ and /u/ could be a potential alternating pair, but /u/ is practically prohibited in non-initial syllables, and /i/ occurs most of the time in verbal suffixes. An alternative analysis could be: o : e  *u : (*)i.)
+  : **o : e** (There are no vowels differing only in F/B. If we ignore the difference in liability, /i/ and /u/ could be a potential alternating pair, but /u/ is practically prohibited in non-initial syllables, and /i/ occurs most of the time in verbal suffixes. An alternative analysis could be: o : e  *u : (*)i.)
   ; Hill Mari
@@ Line 28: / Line 34: @@
   ; Hungarian
-  : **F/B harmony:** ɑː : eː, ɒ : ɛ, o : ɛ, u, y, uː, yː, o : ø, oː : øː. (The vowels /ɑː/ : /eː/ and /o/ : /ɛ/ are not potential alternating pairs because of their height difference. Although based on the similarity of the alternations /ɛ/ :  /eː/ and /ɒ/ : /ɑː/ we could argue that /eː/ is phonetically low — or /ɛ/ is mid — it is unpreferable exactly because their different behavior in vowel harmony, that is /eː/ shows much more neutral characteristics — a tendency to avoid alternation, higher transparency — than /ɛ/, which could be related to their different height. The vowels /ɒ/ : /ɛ/ is not a potential alternating pair either, because they differ in roundedness. However, in this case, there seems to be no counterargument against an analysis in which this difference is not phonemic but barely phonetic. In addition, there is also a marginal /i/ : /jɒ/, /i/ : /jɑː/ alternation due to F/B harmony in verbal morphology.) **Rounding harmony:** ɛ : ø | eː : øː, i : y, iː : yː. (The only alternating pair is not a potential one because of the difference in height. None of the potential pairs actually alternate, although the phonemes themselves are not restricted in any way.)
+  : **F/B harmony:** **ɑː : eː, ɒ : ɛ, o : ɛ,** u, y, uː, yː, o : ø, oː : øː. (The vowels /ɑː/ : /eː/ and /o/ : /ɛ/ are not potential alternating pairs because of their height difference. Although based on the similarity of the alternations /ɛ/ :  /eː/ and /ɒ/ : /ɑː/ we could argue that /eː/ is phonetically low — or /ɛ/ is mid — it is unpreferable exactly because their different behavior in vowel harmony, that is /eː/ shows much more neutral characteristics — a tendency to avoid alternation, higher transparency — than /ɛ/, which could be related to their different height. The vowels /ɒ/ : /ɛ/ is not a potential alternating pair either, because they differ in roundedness. However, in this case, there seems to be no counterargument against an analysis in which this difference is not phonemic but barely phonetic. In addition, there is also a marginal /i/ : /jɒ/, /i/ : /jɑː/ alternation due to F/B harmony in verbal morphology.) **Rounding harmony:** **ɛ : ø** | eː : øː, i : y, iː : yː. (The only alternating pair is not a potential one because of the difference in height. None of the potential pairs actually alternate, although the phonemes themselves are not restricted in any way.)
   ; Southern Mansi
-  : Due to the lack of systematic phonematisation, it is difficult to give numbers here. However, based on Kannist (1914: 49–54), high F vowels alternate with mid B vowels ([i] : [ɘ]/[ɜ], [y̠] : [ɵ̞]), a mid F vowel with a low B vowel ([e:] : [ɑ̟:]) and a low F vowel with a low B vowel ([æ] : [ɑ]).
+  : Due to the lack of systematic phonematization, it is difficult to give numbers here. However, based on Kannisto (1914: 49–54), high F vowels alternate with mid B vowels ([i] : [ɘ]/[ɜ], [y̠] : [ɵ̞]), a mid F vowel with a low B vowel ([e:] : [ɑ̟:]) and a low F vowel with a low B vowel ([æ] : [ɑ]).
   ; Southern Khanty
-  : /i/ : [ɨ], /e/ : [ɘ], ɑ : æ, ĕ : ɘ̆ | *u : *y, *o : *ø, *ŏ : *ø̆ (The vowels /i/ and /e/ have no B phonemic counterparts, but they alternate allophonically. Rounded vowels are prohibited in non-initial syllables, but the past tense 3SG ending /ot/ is an exception.)
+  : **/i/ : [ɨ], /e/ : [ɘ],** ɑ : æ, ĕ : ɘ̆ | *u : *y, *o : *ø, *ŏ : *ø̆ (The vowels /i/ and /e/ have no B phonemic counterparts, but they alternate allophonically. Rounded vowels are prohibited in non-initial syllables, but the past tense 3SG ending /ot/ is an exception.)
   ; Vakh/Vasyugan Khanty
@@ Line 40: / Line 46: @@
   ; Nganasan
-  : **Quasi-rounding harmony:** i : ɑ,  ɨ : ɑ, i : u̯ɑ,  ɨ : u̯ɑ, i̯ɑ : ɑ, i̯ɑ : u̯ɑ, i : y, ɨ : u |  *o : ə (In general, /ɑ/ is in the harmonic class in which all other vowels are rounded, therefore it must be analysed as phonemically rounded. In addition to that, all of its actual alternations (and almost all of those of /u̯ɑ/) are non-potential, since they alternate with high vowels, and some of these alternations are monophthong : diphthong alternations. In principle, there are three kinds of alternations. The first is the case of the potential and actual alternation of high vowels, where the two cases, the alternation of B and the alternation of F vowels are regulated by F/B harmony. The second type is in which a high unrounded vowel alternates with /ɑ/ or /u̯ɑ/. In this type, the Fness/Bness of the high vowel is regulated by F/B harmony, while /u̯ɑ/ occurs after /h/,((Since /u̯ɑ/ occurs almost exclusively after /h/ (and the rare exclusions seem to be mistakes), there is a good chance that /u̯ɑ/ is not a phoneme but an allomorph of /ɑ/ (Fejes 2021). However, here we follow the traditional analysis.)) /ɑ/ everywhere else. The third type is where /i̯ɑ/ alternates with /ɑ/ – or /u̯ɑ/ after /h/. It could also be debated whether  /i̯ɑ/ : /u̯ɑ/ is a potential alternating pair, since the first element of the diphthong is also different in F/B and the second element, rounded phonemically but unrounded phonetically, is not alternating. In addition,  all the alternating vowels, except for diphthongs, occur in suffixes in which they do not alternate.) **F/B harmony:** i : ɨ, u : y | *e : ə.  Additionally, /i̯ɑ/ : /u̯ɑ/ could be a potential alternating pair, but they alternate due to quasi-rounding harmony. In suffixation, F/B harmony functions only when quasi-rounding harmony is also active. However, in bisyllabic stems with two high vowels, the two vowels tend two share the same F/B value, especially but not exclusively when they also share the same rounding value (Fejes 2019: 110–114).
+  : **Quasi-rounding harmony:** **i : ɑ,  ɨ : ɑ, i : u̯ɑ,  ɨ : u̯ɑ, i̯ɑ : ɑ,** i̯ɑ : u̯ɑ, i : y, ɨ : u |  *o : ə (In general, /ɑ/ is in the harmonic class in which all other vowels are rounded, therefore it must be analysed as phonemically rounded. In addition to that, all of its actual alternations (and almost all of those of /u̯ɑ/) are non-potential, since they alternate with high vowels, and some of these alternations are monophthong : diphthong alternations. In principle, there are three kinds of alternations. The first is the case of the potential and actual alternation of high vowels, where the two cases, the alternation of B and the alternation of F vowels are regulated by F/B harmony. The second type is in which a high unrounded vowel alternates with /ɑ/ or /u̯ɑ/. In this type, the Fness/Bness of the high vowel is regulated by F/B harmony, while /u̯ɑ/ occurs after /h/,((Since /u̯ɑ/ occurs almost exclusively after /h/ (and the rare exclusions seem to be mistakes), there is a good chance that /u̯ɑ/ is not a phoneme but an allomorph of /ɑ/ (Fejes 2021). However, here we follow the traditional analysis.)) /ɑ/ everywhere else. The third type is where /i̯ɑ/ alternates with /ɑ/ – or /u̯ɑ/ after /h/. It could also be debated whether  /i̯ɑ/ : /u̯ɑ/ is a potential alternating pair, since the first element of the diphthong is also different in F/B and the second element, rounded phonemically but unrounded phonetically, is not alternating. In addition,  all the alternating vowels, except for diphthongs, occur in suffixes in which they do not alternate.) **F/B harmony:** i : ɨ, u : y | *e : ə.  Additionally, /i̯ɑ/ : /u̯ɑ/ could be a potential alternating pair, but they alternate due to quasi-rounding harmony. In suffixation, F/B harmony functions only when quasi-rounding harmony is also active. However, in bisyllabic stems with two high vowels, the two vowels tend two share the same F/B value, especially but not exclusively when they also share the same rounding value (Fejes 2019: 110–114).
   ; Kamas
   : **F/B harmony:** ɑ : e | i : ɨ, u : y, o : ø, *ĭ : *ɑ̆ (based on Klumpp 2016: 41: Reduced vowels occur in non-initial syllables, but the do not undergo F/B harmony and the only surface form is not identified with any of the reduced vowels in initial syllable. However, a proper notation should be ĭ : *ɑ̆ or *ĭ : ɑ̆.) **Total (full) harmony:** There is no sense to speak about alternating pairs in full harmony.