Loading ...
Sorry, an error occurred while loading the content.
 

Re: conditioned vowel breaking

Expand Messages
  • Alex Fink
    ... What is final apostrophe? (Unreleased?) ... The diphthongisation is probably fine, as long as there was no phonemic palatalisation or labialisation or
    Message 1 of 5 , Jul 1, 2013
      On Mon, 1 Jul 2013 15:27:50 -0400, neo gu <qiihoskeh@...> wrote:

      >In the latest version (Jun26), I'm using a scheme where stressed vowels in the protolanguage are diphthongized according to the vowel in the next syllable, e.g. "ku.pA > "kuA.pA (Diphthongs can also arise from suffixing). The diphthongs are then reduced to simple vowels, with an on-glide in some cases. There are a bunch of subsequent sound changes, such as deletion of unstressed originally simple vowels in some circumstances. Suffixes already present in the protolanguage may shift the stress prior to diphthongization.
      >
      >Sample Resulting Verb Forms
      >
      >Active - Passive
      >"ju.bu - li"bor
      >"tak' - te"kar
      >"gaS - ga"Sar
      >"dzi.mi - du"mjor
      >"fes - fa"Sor
      >"bjaSt' - biS"tar
      >"mjun - mi"nor
      >"kwetS - pu"tSar

      What is final apostrophe? (Unreleased?)

      >I'm concerned that the diphthongization is unnatural and/or that the resulting stem differences are too major.

      The diphthongisation is probably fine, as long as there was no phonemic palatalisation or labialisation or whatnot on consonants at that stage. A very similar thing, e.g., happened in Northern Vanuatu (Alexandre François, _Unraveling the History of the Vowels of Seventeen Northern Vanuatu Languages_). Indeed, your balance between retention of final vowels and reduction of the diphthong contrasts also strikes me as reasonable, i.e. so that the resulting system isn't too lossy but also retains a sensible number of minimal pairs.

      It's interesting that final *iu survive but final *EA fall, since normally high vowels are weaker. But I suppose this actually makes perfect sense if one imagines old final *iu left the preceding consonants strongly palatalised / labialised respectively in the diphthongisation process, and then one had something like
      CA CE Ci Cu >
      CA CE C_ji C_wu >
      C@ C@ C_j@ C_w@ >
      C C Ci Cu.

      ... But, wait, that's not the whole story of the development, is it? In particular the difference between /"jubu/ and /"mjun/ doesn't look explainable on the grounds of vowels alone, so the consonants must have some role here...



      As for the stem differences, well, if the change leading to them was realistic, then a fortiori the outcome is realistic ... for a limited time. But I do suspect that analogy would quickly start acting on these and smoothing out a few things in the system.

      At a guess, actives will be more frequent than passives for most verbs, so (in such verbs) one's likely to begin to see remodelling of the passive to make it more predictable from the active. The single pair which screams at me strongestly for this treatment is the last one. If I'm right to think that the active shows [kw] < earlier [pw], so that there would be lots of /"CwV/ ~ /Cu"/ verbs including even /"kwV/ ~ /ku"/ plus a couple weirdoes doing /"kwV/ ~ /pu"/, then I wouldn't expect that /p/ to hold out long, > /ku"tSar/. The consonant alternations in lines 1 and 4 would probably (at a guess without seeing the inventory) be less immediately susceptible to this: if /dz/ is purely a secondary consonant then there's no "unmutated" pattern for it to fall in with, so it's more interpretable; as for line 1 I suppose the prospective development is passive /i"bor/, but how compelling this is depends on how permitted V-initial verb stems are.

      The vowel alternations I suspect would remain tolerated for a longer time, partly because this seems to be the case crosslinguistically. But for instance, in the first syllable, the patterns active /jV/ > passive /i/, active /wV/ > passive /u/ seem to remain readable enough and so may survive awhile, whereas any active nonhigh vowel can turn into either passive /a/ or /e/ and so that's likely to start re-sorting first. In the appended syllable, again, the patterns active /i/ > passive /jor/, active /u/ > passive /or/ are readable, whereas when the active is closed by a consonant predicting is harder again, and so etc.

      Of course, if there are other forms in the paradigm with yet different alternants that suggest different modes of simplification, all my predictions above may be nonsense.

      Alex
    • neo gu
      ... p , t , k are ejective. Although I suppose they could be aspirated instead. ... I m not sure I understand the provision. There _is_ a change of *k, *g, *x
      Message 2 of 5 , Jul 1, 2013
        On Mon, 1 Jul 2013 21:47:18 -0400, Alex Fink <000024@...> wrote:

        >On Mon, 1 Jul 2013 15:27:50 -0400, neo gu <qiihoskeh@...> wrote:
        >
        >>In the latest version (Jun26), I'm using a scheme where stressed vowels in the protolanguage are diphthongized according to the vowel in the next syllable, e.g. "ku.pA > "kuA.pA (Diphthongs can also arise from suffixing). The diphthongs are then reduced to simple vowels, with an on-glide in some cases. There are a bunch of subsequent sound changes, such as deletion of unstressed originally simple vowels in some circumstances. Suffixes already present in the protolanguage may shift the stress prior to diphthongization.
        >>

        >What is final apostrophe? (Unreleased?)

        p', t', k' are ejective. Although I suppose they could be aspirated instead.

        >>I'm concerned that the diphthongization is unnatural and/or that the resulting stem differences are too major.
        >
        >The diphthongisation is probably fine, as long as there was no phonemic palatalisation or labialisation or whatnot on consonants at that stage. A very similar thing, e.g., happened in Northern Vanuatu (Alexandre François, _Unraveling the History of the Vowels of Seventeen Northern Vanuatu Languages_).
        >
        I'm not sure I understand the provision. There _is_ a change of *k, *g, *x > c, J\, C / __ (*i, *E).

        > Indeed, your balance between retention of final vowels and reduction of the diphthong contrasts also strikes me as reasonable, i.e. so that the resulting system isn't too lossy but also retains a sensible number of minimal pairs.
        >
        >It's interesting that final *iu survive but final *EA fall, since normally high vowels are weaker. But I suppose this actually makes perfect sense if one imagines old final *iu left the preceding consonants strongly palatalised / labialised respectively in the diphthongisation process, and then one had something like
        >CA CE Ci Cu >
        >CA CE C_ji C_wu >
        >C@ C@ C_j@ C_w@ >
        >C C Ci Cu.
        >
        >... But, wait, that's not the whole story of the development, is it? In particular the difference between /"jubu/ and /"mjun/ doesn't look explainable on the grounds of vowels alone, so the consonants must have some role here...
        >
        The vowel-deletion depends on the adjacent consonants, although I considered deleting only the high vowels. I'm still not certain about final vowel deletion.

        >As for the stem differences, well, if the change leading to them was realistic, then a fortiori the outcome is realistic ... for a limited time. But I do suspect that analogy would quickly start acting on these and smoothing out a few things in the system.
        >
        >At a guess, actives will be more frequent than passives for most verbs, so (in such verbs) one's likely to begin to see remodelling of the passive to make it more predictable from the active. The single pair which screams at me strongestly for this treatment is the last one. If I'm right to think that the active shows [kw] < earlier [pw], so that there would be lots of /"CwV/ ~ /Cu"/ verbs including even /"kwV/ ~ /ku"/ plus a couple weirdoes doing /"kwV/ ~ /pu"/, then I wouldn't expect that /p/ to hold out long, > /ku"tSar/. The consonant alternations in lines 1 and 4 would probably (at a guess without seeing the inventory) be less immediately susceptible to this: if /dz/ is purely a secondary consonant then there's no "unmutated" pattern for it to fall in with, so it's more interpretable; as for line 1 I suppose the prospective development is passive /i"bor/, but how compelling this is depends on how permitted V-initial verb stems are.
        >
        >The vowel alternations I suspect would remain tolerated for a longer time, partly because this seems to be the case crosslinguistically. But for instance, in the first syllable, the patterns active /jV/ > passive /i/, active /wV/ > passive /u/ seem to remain readable enough and so may survive awhile, whereas any active nonhigh vowel can turn into either passive /a/ or /e/ and so that's likely to start re-sorting first. In the appended syllable, again, the patterns active /i/ > passive /jor/, active /u/ > passive /or/ are readable, whereas when the active is closed by a consonant predicting is harder again, and so etc.
        >
        All those make sense. Some of the consonant changes are very "recent" such as lj > j (the <l> might even be retained by the orthography). And pw > kw follows earlier kw > k, if that's not too far off.

        >Of course, if there are other forms in the paradigm with yet different alternants that suggest different modes of simplification, all my predictions above may be nonsense.
        >
        >Alex

        I didn't expect so thorough an analysis; otherwise, I would have included more information.
        I've added some forms here (the apostrophe after a vowel is a glottal stop):

        Active - Passive - Imperative - Passive of Causative
        "ju.bu - li"bor - "ju.bu' - li.bu"dor
        "tak' - te"kar - "ta.ka' - te.ka"dor
        "gaS - ga"Sar - "ga.Se' - ga.Se"dor
        "dzi.mi - du"mjor - "dzi.mi' - du.mi"dor
        "fes - fa"Sor - "fe.si' - fa.si"dor
        "bjaSt' - biS"tar - "bjaS.ta' - biS.ta"dor
        "mjun - mi"nor - "mju.nu' - min"dor
        "kwetS - pu"tSar - "kwe.tSe' - pu.tSe"dor
      • Alex Fink
        ... Was there a historical final glottal stop there, or do you just have a rule that final voiceless stops become ejective? The latter is at least a little
        Message 3 of 5 , Jul 1, 2013
          On Mon, 1 Jul 2013 23:54:08 -0400, neo gu <qiihoskeh@...> wrote:

          >On Mon, 1 Jul 2013 21:47:18 -0400, Alex Fink <000024@...> wrote:
          >
          >>What is final apostrophe? (Unreleased?)
          >
          >p', t', k' are ejective. Although I suppose they could be aspirated instead.

          Was there a historical final glottal stop there, or do you just have a rule that final voiceless stops become ejective? The latter is at least a little funny, since there's not much reason for lifting the glottis to arise spontaneously at end of utterance, though I guess they could be fairly weak ejectives.

          >>>I'm concerned that the diphthongization is unnatural and/or that the resulting stem differences are too major.
          >>
          >>The diphthongisation is probably fine, as long as there was no phonemic palatalisation or labialisation or whatnot on consonants at that stage. A very similar thing, e.g., happened in Northern Vanuatu (Alexandre François, _Unraveling the History of the Vowels of Seventeen Northern Vanuatu Languages_).
          >>
          >I'm not sure I understand the provision. There _is_ a change of *k, *g, *x > c, J\, C / __ (*i, *E).

          That's fine, indeed likely, if it's totally conditioned like that. All I was thinking of is that if your protolanguage had a contrast of, say, /EkA/ vs. /Ek_jA/, it's implausible that the anticipatory repositioning of the tongue body needed to get this to develop to [EAkA] would interfere with maintaining the contrast between earlier /k/ and /k_j/.

          >All those make sense. Some of the consonant changes are very "recent" such as lj > j (the <l> might even be retained by the orthography). And pw > kw follows earlier kw > k, if that's not too far off.

          Hah, I have little idea what spelling pronunciation effects would do here. My general impression is that HERE

          I'm not sure in what sense you worry that kw > k might be "too far off". It would throw in some /"kV/ ~ /ku"/ alternations, but maybe those aren't completely foreign in light of the /"Co/ ~ /Cu"/ alternations you already get from *uA, and the general pattern (before regularisation) that active /"Ce/ and /"Ca/ and /"Co/ mostly all have the same set of inherited passive options; they'd probably be reformed along the same lines.

          >I didn't expect so thorough an analysis; otherwise, I would have included more information.
          >I've added some forms here (the apostrophe after a vowel is a glottal stop):
          >
          >Active - Passive - Imperative - Passive of Causative
          >"ju.bu - li"bor - "ju.bu' - li.bu"dor
          >"tak' - te"kar - "ta.ka' - te.ka"dor
          >"gaS - ga"Sar - "ga.Se' - ga.Se"dor
          >"dzi.mi - du"mjor - "dzi.mi' - du.mi"dor
          >"fes - fa"Sor - "fe.si' - fa.si"dor
          >"bjaSt' - biS"tar - "bjaS.ta' - biS.ta"dor
          >"mjun - mi"nor - "mju.nu' - min"dor
          >"kwetS - pu"tSar - "kwe.tSe' - pu.tSe"dor

          So the imperatives are just the actives with the possibly deleted final vowel added back (plus a constant [?], that part's easy). This is the first place where the old final /a/ : /e/ distinction is showing up (unless it can ever be retained in the active?) but other than that its formation seems straightforward; the correlation between the vowel added here and the one added in the passive would be easy to retain, however else the whole issue of final vowels sorts out.
          Ditto for the passives of the causative, pretty much, except that unless the loss of the vowel in /min"dor/ can be construed as a surface morphophonemic rule it's likely that this vowel will be resupplied (unless the passive of the causative is a highly used category or unless this is a highly used form of this particular verb, tolerant of a little irregularity like that).

          Alex
        • neo gu
          ... It seems natural to me (the stops also become ejective before most consonants, after vowel deletion). The sequence voiceless stop + glottal stop can occur
          Message 4 of 5 , Jul 2, 2013
            On Tue, 2 Jul 2013 00:49:22 -0400, Alex Fink <000024@...> wrote:

            >On Mon, 1 Jul 2013 23:54:08 -0400, neo gu <qiihoskeh@...> wrote:
            >
            >>On Mon, 1 Jul 2013 21:47:18 -0400, Alex Fink <000024@...> wrote:
            >>
            >>>What is final apostrophe? (Unreleased?)
            >>
            >>p', t', k' are ejective. Although I suppose they could be aspirated instead.
            >
            >Was there a historical final glottal stop there, or do you just have a rule that final voiceless stops become ejective? The latter is at least a little funny, since there's not much reason for lifting the glottis to arise spontaneously at end of utterance, though I guess they could be fairly weak ejectives.
            >
            It seems natural to me (the stops also become ejective before most consonants, after vowel deletion). The sequence voiceless stop + glottal stop can occur elsewhere (between vowels).

            >>All those make sense. Some of the consonant changes are very "recent" such as lj > j (the <l> might even be retained by the orthography). And pw > kw follows earlier kw > k, if that's not too far off.
            >
            >I'm not sure in what sense you worry that kw > k might be "too far off". It would throw in some /"kV/ ~ /ku"/ alternations, but maybe those aren't completely foreign in light of the /"Co/ ~ /Cu"/ alternations you already get from *uA, and the general pattern (before regularisation) that active /"Ce/ and /"Ca/ and /"Co/ mostly all have the same set of inherited passive options; they'd probably be reformed along the same lines.
            >

            >>I didn't expect so thorough an analysis; otherwise, I would have included more information.
            >>I've added some forms here (the apostrophe after a vowel is a glottal stop):
            >>
            >>Active - Passive - Imperative - Passive of Causative
            >>"ju.bu - li"bor - "ju.bu' - li.bu"dor
            >>"tak' - te"kar - "ta.ka' - te.ka"dor
            >>"gaS - ga"Sar - "ga.Se' - ga.Se"dor
            >>"dzi.mi - du"mjor - "dzi.mi' - du.mi"dor
            >>"fes - fa"Sor - "fe.si' - fa.si"dor
            >>"bjaSt' - biS"tar - "bjaS.ta' - biS.ta"dor
            >>"mjun - mi"nor - "mju.nu' - min"dor
            >>"kwetS - pu"tSar - "kwe.tSe' - pu.tSe"dor
            >
            >So the imperatives are just the actives with the possibly deleted final vowel added back (plus a constant [?], that part's easy). This is the first place where the old final /a/ : /e/ distinction is showing up (unless it can ever be retained in the active?)
            >
            Yes, after b, d, g, and dZ.

            > but other than that its formation seems straightforward; the correlation between the vowel added here and the one added in the passive would be easy to retain, however else the whole issue of final vowels sorts out.
            >Ditto for the passives of the causative, pretty much, except that unless the loss of the vowel in /min"dor/ can be construed as a surface morphophonemic rule it's likely that this vowel will be resupplied (unless the passive of the causative is a highly used category or unless this is a highly used form of this particular verb, tolerant of a little irregularity like that).
            >
            >Alex

            A surface rule is what I had in mind. I don't think the passive of the causative will be highly used.
          Your message has been successfully submitted and would be delivered to recipients shortly.