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

Hello, and language sketch.

Expand Messages
  • Aodhán Aannestad
    Hello! I m a longtime lurker, finally deciding to get involved a little. Some of you may have met me at the last LCC, I m with UT Austin s conlang club and I
    Message 1 of 11 , Jun 23, 2013
      Hello! I'm a longtime lurker, finally deciding to get involved a little.
      Some of you may have met me at the last LCC, I'm with UT Austin's
      conlang club and I drove some people around. I've been conlanging on and
      off for quite some time now, but I don't have much to show for it - I
      tend to be singlemindedly focused on realism, and as my level of
      linguistics knowledge has increased I've scrapped and restarted on
      several occasions due to realising how horribly unrealistic of a
      language I'd created. This is kind of the latest iteration of that
      cycle, and as I'm done with all the core undergrad ling classes now
      hopefully I've run out of major things to learn and it'll prove to be
      the last.

      Anyway, on to the language. It doesn't have a name yet, sadly - I'm
      using it for a protolang for a large project, and since said large
      project has yet to really begin, it's not really in a state where I can
      really name anything. (I suppose I could come up with an endonym for it
      - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
      though that doesn't sound that great to my ears.)

      Phonologically, it's not that interesting - five vowels (/a i u e o/),
      two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
      and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
      but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
      l/ round out the inventory.
      Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
      disallowed, and maximal CGVGC syllables are very rare. Affixes have no
      shape target - anything from V to a maximal syllable is in theory
      permitted. Length is kind of phonemic on consonants (really, a 'long
      consonant' is just two adjacent identical consonants - /alla/ 'day' is
      VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
      are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
      'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
      /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
      distinguished in codas. There's not much else in the way of phonological
      rules/alternations, at the moment this is the idealised pre-protolang
      stage, and I'll need to send it through some sound changes before I get
      a good protolang by the technical definition.
      Stress is noncontrastive - it occurs on the heaviest of the last three
      syllables, defaulting to the antepenultimate when they're all equal
      (so/émnira, doráyra,////dalésise (do-person-COP-ATT)/, etc.)/./

      Grammatically, it's agglutinative and erg-abs. The two basic word
      classes are noun and verb (all 'adjectives' are just verbs, and at least
      in the protolanguage all 'adverbs' are clearly nouns or nominalised
      verbs marked with a non-core case). Word order is VSO when there's no
      overt complementiser, and SOV when there is (so /fikol le/ 'the man has
      gone', but /le/ /fikolti/ 'the fact that the man has gone').

      Verbs don't care about person, number, or tense, but there are 5 or 7
      aspect markers (perfective/stative (null-marked), progressive, perfect,
      expective, intentive(?), and hortative and imperative if you count them
      - they're mutually exclusive with aspect). It has a number of 'mood'
      markers: potential and permissive; volitive, suggestive, and
      necessitive, and weak and strong expectation (weak is for 'I bet X is
      the case' and strong is for 'X /has/ to be the case, I just know it').
      Relative clauses are formed by using the attributive affix
      /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
      role the modified noun would have in the clause, that's left up to
      context (so typically you can only relativise obliques when the
      subclause has all of its core arguments overtly specified). (This works
      mostly like the Japanese rentaikei.)

      The only obligatory marking on nouns is case, but there's a number of
      other potential affixes. Number is especially complex - specific
      quantities are marked directly on the noun (so/lemofyethon/ 'twenty
      people', it's base-8 so that breaks down as /le-mo-fye-thon/
      'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
      'more than half (of a group)', 'less than half (of a group)', 'part (of
      a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
      'none', and further by 'the next' or 'the previous' (allowing for very
      long sequences such as /lemofyethondawfag /'none of the last twenty
      people'). Demonstratives are also noun suffixes (there's a two-way
      distinction, 'this'/'that'), and an interrogative marker can slot in
      here too (/lewos?/ 'who? / which person?').

      There are a good number of cases (it's kind of Finno-Ugric in this
      regard :P) Beyond the erg and null-marked abs, there's two kinds of
      genitives, possessive and categorical (for things like 'men of that
      village', 'the strength of an ox', 'a sword of bronze' and so on), both
      of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
      -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
      heard as 'the cat is the man's').
      Locative cases are the following: inessive and exessive (both used for
      general locatives, inessive for being within the boundaries of a place,
      exessive for being near but outside the boundaries of a place or
      object), superessive and subessive, proessive ('in front of') and
      postessive ('behind'), comitative, allative (also used as a dative) and
      ablative (also used as the agent of causatives and volitives), illative
      and ellative, superlative ('going over') and sublative ('going under'),
      circumlative/circumessive, and adspective ('facing') and abspective
      ('facing away from').
      Non-locative cases are benefactive (also used for the experiencer with a
      number of perception verbs - 'see' for example has a BEN subject when
      you would expect ERG, and giving it an ERG subject changes the meaning
      to 'look at'), instrumental, causative, and comparative.
      Copular constructions are formed by affixing the copular verbaliser
      /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
      somewhat idiosyncratic way to express motion - while it's perfectly
      grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
      /literally 'I am going to the river'), it's much more native-sounding to
      say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').

      Beyond three generic 'pronouns' (1st person exclusive and inclusive and
      2nd person), there's not much in the way of real pronouns - nouns can be
      used with any person as their referent, and typical non-personal
      pronouny things are done with nouns plus suffixes (so 'who' is just
      /lewos /(/person-INT/), literally 'which person'). There is a set of
      'generic nouns', though, which are basically nouns that refer to quite
      large categories of things - 'person', 'object', 'place', 'point in
      time', 'state of being', 'piece', 'reason/cause', and 'action' make up
      the set - and this allows for fairly conventionalised pronouns ('person'
      can be 'him/her', 'object' can be 'it', 'place' plus the near
      demonstrative can be 'here', and so on). These nouns are further
      distinguished from other nouns by being used as nominaliser suffixes -
      so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
      died' to make /ryukolub/ 'the reason [the subj] died'.
      There are also two generic verbs, meaning something like 'to do (it)'
      and 'to go (there)'. They can also be used as nominalisers (somewhat
      ironically :P), meaning 'method' and 'process', respectively.

      There's a few other small details here and there (I've left out
      valence-change affixes, for example), but that's a basic overview of the
      language. The goal is realism (indeed, all else is secondary), so some
      comments in regards to how realistic these systems are would be
      appreciated! I've got a few ideas on where to go with it from here, but
      if anyone has any ideas I'd be happy to hear them.

      This is my main conlang project, but I'm also working on a set of
      condialects of Japanese (splitting off at various points after around
      1610?), and I'd be happy to describe them if anyone's interested. (It's
      gotten to the point where I'll slip into my primary condialect every
      once in a while while thinking in Japanese, even when I'm not explicitly
      trying to think in it :P)
    • neo gu
      ... Welcome to active participation! I m not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as
      Message 2 of 11 , Jun 24, 2013
        On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:

        >Hello! I'm a longtime lurker, finally deciding to get involved a little.
        >Some of you may have met me at the last LCC, I'm with UT Austin's
        >conlang club and I drove some people around. I've been conlanging on and
        >off for quite some time now, but I don't have much to show for it - I
        >tend to be singlemindedly focused on realism, and as my level of
        >linguistics knowledge has increased I've scrapped and restarted on
        >several occasions due to realising how horribly unrealistic of a
        >language I'd created. This is kind of the latest iteration of that
        >cycle, and as I'm done with all the core undergrad ling classes now
        >hopefully I've run out of major things to learn and it'll prove to be
        >the last.

        Welcome to active participation!

        I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.

        --
        Jeff Jones

        >Anyway, on to the language. It doesn't have a name yet, sadly - I'm
        >using it for a protolang for a large project, and since said large
        >project has yet to really begin, it's not really in a state where I can
        >really name anything. (I suppose I could come up with an endonym for it
        >- it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
        >though that doesn't sound that great to my ears.)
        >
        >Phonologically, it's not that interesting - five vowels (/a i u e o/),
        >two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
        >and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
        >but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
        >l/ round out the inventory.
        >Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
        >disallowed, and maximal CGVGC syllables are very rare. Affixes have no
        >shape target - anything from V to a maximal syllable is in theory
        >permitted. Length is kind of phonemic on consonants (really, a 'long
        >consonant' is just two adjacent identical consonants - /alla/ 'day' is
        >VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
        >are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
        >'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
        >/dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
        >distinguished in codas. There's not much else in the way of phonological
        >rules/alternations, at the moment this is the idealised pre-protolang
        >stage, and I'll need to send it through some sound changes before I get
        >a good protolang by the technical definition.
        >Stress is noncontrastive - it occurs on the heaviest of the last three
        >syllables, defaulting to the antepenultimate when they're all equal
        >(so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
        >
        >Grammatically, it's agglutinative and erg-abs. The two basic word
        >classes are noun and verb (all 'adjectives' are just verbs, and at least
        >in the protolanguage all 'adverbs' are clearly nouns or nominalised
        >verbs marked with a non-core case). Word order is VSO when there's no
        >overt complementiser, and SOV when there is (so /fikol le/ 'the man has
        >gone', but /le/ /fikolti/ 'the fact that the man has gone').
        >
        >Verbs don't care about person, number, or tense, but there are 5 or 7
        >aspect markers (perfective/stative (null-marked), progressive, perfect,
        >expective, intentive(?), and hortative and imperative if you count them
        >- they're mutually exclusive with aspect). It has a number of 'mood'
        >markers: potential and permissive; volitive, suggestive, and
        >necessitive, and weak and strong expectation (weak is for 'I bet X is
        >the case' and strong is for 'X /has/ to be the case, I just know it').
        >Relative clauses are formed by using the attributive affix
        >/-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
        >role the modified noun would have in the clause, that's left up to
        >context (so typically you can only relativise obliques when the
        >subclause has all of its core arguments overtly specified). (This works
        >mostly like the Japanese rentaikei.)
        >
        >The only obligatory marking on nouns is case, but there's a number of
        >other potential affixes. Number is especially complex - specific
        >quantities are marked directly on the noun (so/lemofyethon/ 'twenty
        >people', it's base-8 so that breaks down as /le-mo-fye-thon/
        >'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
        >'more than half (of a group)', 'less than half (of a group)', 'part (of
        >a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
        >'none', and further by 'the next' or 'the previous' (allowing for very
        >long sequences such as /lemofyethondawfag /'none of the last twenty
        >people'). Demonstratives are also noun suffixes (there's a two-way
        >distinction, 'this'/'that'), and an interrogative marker can slot in
        >here too (/lewos?/ 'who? / which person?').
        >
        >There are a good number of cases (it's kind of Finno-Ugric in this
        >regard :P) Beyond the erg and null-marked abs, there's two kinds of
        >genitives, possessive and categorical (for things like 'men of that
        >village', 'the strength of an ox', 'a sword of bronze' and so on), both
        >of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
        >-/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
        >heard as 'the cat is the man's').
        >Locative cases are the following: inessive and exessive (both used for
        >general locatives, inessive for being within the boundaries of a place,
        >exessive for being near but outside the boundaries of a place or
        >object), superessive and subessive, proessive ('in front of') and
        >postessive ('behind'), comitative, allative (also used as a dative) and
        >ablative (also used as the agent of causatives and volitives), illative
        >and ellative, superlative ('going over') and sublative ('going under'),
        >circumlative/circumessive, and adspective ('facing') and abspective
        >('facing away from').
        >Non-locative cases are benefactive (also used for the experiencer with a
        >number of perception verbs - 'see' for example has a BEN subject when
        >you would expect ERG, and giving it an ERG subject changes the meaning
        >to 'look at'), instrumental, causative, and comparative.
        >Copular constructions are formed by affixing the copular verbaliser
        >/-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
        >somewhat idiosyncratic way to express motion - while it's perfectly
        >grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
        >/literally 'I am going to the river'), it's much more native-sounding to
        >say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
        >
        >Beyond three generic 'pronouns' (1st person exclusive and inclusive and
        >2nd person), there's not much in the way of real pronouns - nouns can be
        >used with any person as their referent, and typical non-personal
        >pronouny things are done with nouns plus suffixes (so 'who' is just
        >/lewos /(/person-INT/), literally 'which person'). There is a set of
        >'generic nouns', though, which are basically nouns that refer to quite
        >large categories of things - 'person', 'object', 'place', 'point in
        >time', 'state of being', 'piece', 'reason/cause', and 'action' make up
        >the set - and this allows for fairly conventionalised pronouns ('person'
        >can be 'him/her', 'object' can be 'it', 'place' plus the near
        >demonstrative can be 'here', and so on). These nouns are further
        >distinguished from other nouns by being used as nominaliser suffixes -
        >so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
        >died' to make /ryukolub/ 'the reason [the subj] died'.
        >There are also two generic verbs, meaning something like 'to do (it)'
        >and 'to go (there)'. They can also be used as nominalisers (somewhat
        >ironically :P), meaning 'method' and 'process', respectively.
        >
        >There's a few other small details here and there (I've left out
        >valence-change affixes, for example), but that's a basic overview of the
        >language. The goal is realism (indeed, all else is secondary), so some
        >comments in regards to how realistic these systems are would be
        >appreciated! I've got a few ideas on where to go with it from here, but
        >if anyone has any ideas I'd be happy to hear them.
        >
        >This is my main conlang project, but I'm also working on a set of
        >condialects of Japanese (splitting off at various points after around
        >1610?), and I'd be happy to describe them if anyone's interested. (It's
        >gotten to the point where I'll slip into my primary condialect every
        >once in a while while thinking in Japanese, even when I'm not explicitly
        >trying to think in it :P)
      • Galen Buttitta
        Based on the location of that first /?/ I m going to guess that it s a voiceless dental fricative. Is that correct, and is the second /?/ supposed to be an
        Message 3 of 11 , Jun 24, 2013
          Based on the location of that first /?/ I'm going to guess that it's a voiceless dental fricative. Is that correct, and is the second /?/ supposed to be an alveolar approximant?

          SATOR
          AREPO
          TENET
          OPERA
          ROTAS

          On Jun 24, 2013, at 14:16, neo gu <qiihoskeh@...> wrote:

          > On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
          >
          >> Hello! I'm a longtime lurker, finally deciding to get involved a little.
          >> Some of you may have met me at the last LCC, I'm with UT Austin's
          >> conlang club and I drove some people around. I've been conlanging on and
          >> off for quite some time now, but I don't have much to show for it - I
          >> tend to be singlemindedly focused on realism, and as my level of
          >> linguistics knowledge has increased I've scrapped and restarted on
          >> several occasions due to realising how horribly unrealistic of a
          >> language I'd created. This is kind of the latest iteration of that
          >> cycle, and as I'm done with all the core undergrad ling classes now
          >> hopefully I've run out of major things to learn and it'll prove to be
          >> the last.
          >
          > Welcome to active participation!
          >
          > I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
          >
          > --
          > Jeff Jones
          >
          >> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
          >> using it for a protolang for a large project, and since said large
          >> project has yet to really begin, it's not really in a state where I can
          >> really name anything. (I suppose I could come up with an endonym for it
          >> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
          >> though that doesn't sound that great to my ears.)
          >>
          >> Phonologically, it's not that interesting - five vowels (/a i u e o/),
          >> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
          >> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
          >> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
          >> l/ round out the inventory.
          >> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
          >> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
          >> shape target - anything from V to a maximal syllable is in theory
          >> permitted. Length is kind of phonemic on consonants (really, a 'long
          >> consonant' is just two adjacent identical consonants - /alla/ 'day' is
          >> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
          >> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
          >> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
          >> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
          >> distinguished in codas. There's not much else in the way of phonological
          >> rules/alternations, at the moment this is the idealised pre-protolang
          >> stage, and I'll need to send it through some sound changes before I get
          >> a good protolang by the technical definition.
          >> Stress is noncontrastive - it occurs on the heaviest of the last three
          >> syllables, defaulting to the antepenultimate when they're all equal
          >> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
          >>
          >> Grammatically, it's agglutinative and erg-abs. The two basic word
          >> classes are noun and verb (all 'adjectives' are just verbs, and at least
          >> in the protolanguage all 'adverbs' are clearly nouns or nominalised
          >> verbs marked with a non-core case). Word order is VSO when there's no
          >> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
          >> gone', but /le/ /fikolti/ 'the fact that the man has gone').
          >>
          >> Verbs don't care about person, number, or tense, but there are 5 or 7
          >> aspect markers (perfective/stative (null-marked), progressive, perfect,
          >> expective, intentive(?), and hortative and imperative if you count them
          >> - they're mutually exclusive with aspect). It has a number of 'mood'
          >> markers: potential and permissive; volitive, suggestive, and
          >> necessitive, and weak and strong expectation (weak is for 'I bet X is
          >> the case' and strong is for 'X /has/ to be the case, I just know it').
          >> Relative clauses are formed by using the attributive affix
          >> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
          >> role the modified noun would have in the clause, that's left up to
          >> context (so typically you can only relativise obliques when the
          >> subclause has all of its core arguments overtly specified). (This works
          >> mostly like the Japanese rentaikei.)
          >>
          >> The only obligatory marking on nouns is case, but there's a number of
          >> other potential affixes. Number is especially complex - specific
          >> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
          >> people', it's base-8 so that breaks down as /le-mo-fye-thon/
          >> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
          >> 'more than half (of a group)', 'less than half (of a group)', 'part (of
          >> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
          >> 'none', and further by 'the next' or 'the previous' (allowing for very
          >> long sequences such as /lemofyethondawfag /'none of the last twenty
          >> people'). Demonstratives are also noun suffixes (there's a two-way
          >> distinction, 'this'/'that'), and an interrogative marker can slot in
          >> here too (/lewos?/ 'who? / which person?').
          >>
          >> There are a good number of cases (it's kind of Finno-Ugric in this
          >> regard :P) Beyond the erg and null-marked abs, there's two kinds of
          >> genitives, possessive and categorical (for things like 'men of that
          >> village', 'the strength of an ox', 'a sword of bronze' and so on), both
          >> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
          >> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
          >> heard as 'the cat is the man's').
          >> Locative cases are the following: inessive and exessive (both used for
          >> general locatives, inessive for being within the boundaries of a place,
          >> exessive for being near but outside the boundaries of a place or
          >> object), superessive and subessive, proessive ('in front of') and
          >> postessive ('behind'), comitative, allative (also used as a dative) and
          >> ablative (also used as the agent of causatives and volitives), illative
          >> and ellative, superlative ('going over') and sublative ('going under'),
          >> circumlative/circumessive, and adspective ('facing') and abspective
          >> ('facing away from').
          >> Non-locative cases are benefactive (also used for the experiencer with a
          >> number of perception verbs - 'see' for example has a BEN subject when
          >> you would expect ERG, and giving it an ERG subject changes the meaning
          >> to 'look at'), instrumental, causative, and comparative.
          >> Copular constructions are formed by affixing the copular verbaliser
          >> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
          >> somewhat idiosyncratic way to express motion - while it's perfectly
          >> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
          >> /literally 'I am going to the river'), it's much more native-sounding to
          >> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
          >>
          >> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
          >> 2nd person), there's not much in the way of real pronouns - nouns can be
          >> used with any person as their referent, and typical non-personal
          >> pronouny things are done with nouns plus suffixes (so 'who' is just
          >> /lewos /(/person-INT/), literally 'which person'). There is a set of
          >> 'generic nouns', though, which are basically nouns that refer to quite
          >> large categories of things - 'person', 'object', 'place', 'point in
          >> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
          >> the set - and this allows for fairly conventionalised pronouns ('person'
          >> can be 'him/her', 'object' can be 'it', 'place' plus the near
          >> demonstrative can be 'here', and so on). These nouns are further
          >> distinguished from other nouns by being used as nominaliser suffixes -
          >> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
          >> died' to make /ryukolub/ 'the reason [the subj] died'.
          >> There are also two generic verbs, meaning something like 'to do (it)'
          >> and 'to go (there)'. They can also be used as nominalisers (somewhat
          >> ironically :P), meaning 'method' and 'process', respectively.
          >>
          >> There's a few other small details here and there (I've left out
          >> valence-change affixes, for example), but that's a basic overview of the
          >> language. The goal is realism (indeed, all else is secondary), so some
          >> comments in regards to how realistic these systems are would be
          >> appreciated! I've got a few ideas on where to go with it from here, but
          >> if anyone has any ideas I'd be happy to hear them.
          >>
          >> This is my main conlang project, but I'm also working on a set of
          >> condialects of Japanese (splitting off at various points after around
          >> 1610?), and I'd be happy to describe them if anyone's interested. (It's
          >> gotten to the point where I'll slip into my primary condialect every
          >> once in a while while thinking in Japanese, even when I'm not explicitly
          >> trying to think in it :P)
        • Aodhán Aannestad
          ... Oh, yeah, the email got messed up because for whatever reason my editor is a weird mix of HTML and WYSIWYG. Those are supposed to be what s /T/ and /4/ in
          Message 4 of 11 , Jun 24, 2013
            On 6/24/2013 1:16 PM, neo gu wrote:
            > On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
            >
            >> Hello! I'm a longtime lurker, finally deciding to get involved a little.
            >> Some of you may have met me at the last LCC, I'm with UT Austin's
            >> conlang club and I drove some people around. I've been conlanging on and
            >> off for quite some time now, but I don't have much to show for it - I
            >> tend to be singlemindedly focused on realism, and as my level of
            >> linguistics knowledge has increased I've scrapped and restarted on
            >> several occasions due to realising how horribly unrealistic of a
            >> language I'd created. This is kind of the latest iteration of that
            >> cycle, and as I'm done with all the core undergrad ling classes now
            >> hopefully I've run out of major things to learn and it'll prove to be
            >> the last.
            > Welcome to active participation!
            >
            > I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
            >
            > --
            > Jeff Jones
            Oh, yeah, the email got messed up because for whatever reason my editor
            is a weird mix of HTML and WYSIWYG. Those are supposed to be what's /T/
            and /4/ in X-SAMPA.

            >
            >> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
            >> using it for a protolang for a large project, and since said large
            >> project has yet to really begin, it's not really in a state where I can
            >> really name anything. (I suppose I could come up with an endonym for it
            >> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
            >> though that doesn't sound that great to my ears.)
            >>
            >> Phonologically, it's not that interesting - five vowels (/a i u e o/),
            >> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
            >> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
            >> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
            >> l/ round out the inventory.
            >> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
            >> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
            >> shape target - anything from V to a maximal syllable is in theory
            >> permitted. Length is kind of phonemic on consonants (really, a 'long
            >> consonant' is just two adjacent identical consonants - /alla/ 'day' is
            >> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
            >> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
            >> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
            >> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
            >> distinguished in codas. There's not much else in the way of phonological
            >> rules/alternations, at the moment this is the idealised pre-protolang
            >> stage, and I'll need to send it through some sound changes before I get
            >> a good protolang by the technical definition.
            >> Stress is noncontrastive - it occurs on the heaviest of the last three
            >> syllables, defaulting to the antepenultimate when they're all equal
            >> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
            >>
            >> Grammatically, it's agglutinative and erg-abs. The two basic word
            >> classes are noun and verb (all 'adjectives' are just verbs, and at least
            >> in the protolanguage all 'adverbs' are clearly nouns or nominalised
            >> verbs marked with a non-core case). Word order is VSO when there's no
            >> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
            >> gone', but /le/ /fikolti/ 'the fact that the man has gone').
            >>
            >> Verbs don't care about person, number, or tense, but there are 5 or 7
            >> aspect markers (perfective/stative (null-marked), progressive, perfect,
            >> expective, intentive(?), and hortative and imperative if you count them
            >> - they're mutually exclusive with aspect). It has a number of 'mood'
            >> markers: potential and permissive; volitive, suggestive, and
            >> necessitive, and weak and strong expectation (weak is for 'I bet X is
            >> the case' and strong is for 'X /has/ to be the case, I just know it').
            >> Relative clauses are formed by using the attributive affix
            >> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
            >> role the modified noun would have in the clause, that's left up to
            >> context (so typically you can only relativise obliques when the
            >> subclause has all of its core arguments overtly specified). (This works
            >> mostly like the Japanese rentaikei.)
            >>
            >> The only obligatory marking on nouns is case, but there's a number of
            >> other potential affixes. Number is especially complex - specific
            >> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
            >> people', it's base-8 so that breaks down as /le-mo-fye-thon/
            >> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
            >> 'more than half (of a group)', 'less than half (of a group)', 'part (of
            >> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
            >> 'none', and further by 'the next' or 'the previous' (allowing for very
            >> long sequences such as /lemofyethondawfag /'none of the last twenty
            >> people'). Demonstratives are also noun suffixes (there's a two-way
            >> distinction, 'this'/'that'), and an interrogative marker can slot in
            >> here too (/lewos?/ 'who? / which person?').
            >>
            >> There are a good number of cases (it's kind of Finno-Ugric in this
            >> regard :P) Beyond the erg and null-marked abs, there's two kinds of
            >> genitives, possessive and categorical (for things like 'men of that
            >> village', 'the strength of an ox', 'a sword of bronze' and so on), both
            >> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
            >> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
            >> heard as 'the cat is the man's').
            >> Locative cases are the following: inessive and exessive (both used for
            >> general locatives, inessive for being within the boundaries of a place,
            >> exessive for being near but outside the boundaries of a place or
            >> object), superessive and subessive, proessive ('in front of') and
            >> postessive ('behind'), comitative, allative (also used as a dative) and
            >> ablative (also used as the agent of causatives and volitives), illative
            >> and ellative, superlative ('going over') and sublative ('going under'),
            >> circumlative/circumessive, and adspective ('facing') and abspective
            >> ('facing away from').
            >> Non-locative cases are benefactive (also used for the experiencer with a
            >> number of perception verbs - 'see' for example has a BEN subject when
            >> you would expect ERG, and giving it an ERG subject changes the meaning
            >> to 'look at'), instrumental, causative, and comparative.
            >> Copular constructions are formed by affixing the copular verbaliser
            >> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
            >> somewhat idiosyncratic way to express motion - while it's perfectly
            >> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
            >> /literally 'I am going to the river'), it's much more native-sounding to
            >> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
            >>
            >> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
            >> 2nd person), there's not much in the way of real pronouns - nouns can be
            >> used with any person as their referent, and typical non-personal
            >> pronouny things are done with nouns plus suffixes (so 'who' is just
            >> /lewos /(/person-INT/), literally 'which person'). There is a set of
            >> 'generic nouns', though, which are basically nouns that refer to quite
            >> large categories of things - 'person', 'object', 'place', 'point in
            >> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
            >> the set - and this allows for fairly conventionalised pronouns ('person'
            >> can be 'him/her', 'object' can be 'it', 'place' plus the near
            >> demonstrative can be 'here', and so on). These nouns are further
            >> distinguished from other nouns by being used as nominaliser suffixes -
            >> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
            >> died' to make /ryukolub/ 'the reason [the subj] died'.
            >> There are also two generic verbs, meaning something like 'to do (it)'
            >> and 'to go (there)'. They can also be used as nominalisers (somewhat
            >> ironically :P), meaning 'method' and 'process', respectively.
            >>
            >> There's a few other small details here and there (I've left out
            >> valence-change affixes, for example), but that's a basic overview of the
            >> language. The goal is realism (indeed, all else is secondary), so some
            >> comments in regards to how realistic these systems are would be
            >> appreciated! I've got a few ideas on where to go with it from here, but
            >> if anyone has any ideas I'd be happy to hear them.
            >>
            >> This is my main conlang project, but I'm also working on a set of
            >> condialects of Japanese (splitting off at various points after around
            >> 1610?), and I'd be happy to describe them if anyone's interested. (It's
            >> gotten to the point where I'll slip into my primary condialect every
            >> once in a while while thinking in Japanese, even when I'm not explicitly
            >> trying to think in it :P)
          • James Kane
            So the consonant written is a flap? And /x/ is written ? This seems very cool although there s quite a lot to take in, maybe a few example sentences
            Message 5 of 11 , Jun 25, 2013
              So the consonant written <r> is a flap? And /x/ is written <kh>? This seems very cool although there's quite a lot to take in, maybe a few example sentences with gloss outlining the principle features would help.

              From what I can tell I like the way the -se affix works although I'm not sure I quite understand it, and when the change in word order happens in relative clauses.

              Your condialects project sounds very exciting and I would be interested to have a small peak, although I know almost nothing about Japanese (indeed, I know more about Okinawan)


              James

              On 25/06/2013, at 7:50 AM, Aodhán Aannestad <tolkien_freak@...> wrote:

              > On 6/24/2013 1:16 PM, neo gu wrote:
              >> On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
              >>
              >>> Hello! I'm a longtime lurker, finally deciding to get involved a little.
              >>> Some of you may have met me at the last LCC, I'm with UT Austin's
              >>> conlang club and I drove some people around. I've been conlanging on and
              >>> off for quite some time now, but I don't have much to show for it - I
              >>> tend to be singlemindedly focused on realism, and as my level of
              >>> linguistics knowledge has increased I've scrapped and restarted on
              >>> several occasions due to realising how horribly unrealistic of a
              >>> language I'd created. This is kind of the latest iteration of that
              >>> cycle, and as I'm done with all the core undergrad ling classes now
              >>> hopefully I've run out of major things to learn and it'll prove to be
              >>> the last.
              >> Welcome to active participation!
              >>
              >> I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
              >>
              >> --
              >> Jeff Jones
              > Oh, yeah, the email got messed up because for whatever reason my editor is a weird mix of HTML and WYSIWYG. Those are supposed to be what's /T/ and /4/ in X-SAMPA.
              >
              >>
              >>> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
              >>> using it for a protolang for a large project, and since said large
              >>> project has yet to really begin, it's not really in a state where I can
              >>> really name anything. (I suppose I could come up with an endonym for it
              >>> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
              >>> though that doesn't sound that great to my ears.)
              >>>
              >>> Phonologically, it's not that interesting - five vowels (/a i u e o/),
              >>> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
              >>> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
              >>> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
              >>> l/ round out the inventory.
              >>> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
              >>> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
              >>> shape target - anything from V to a maximal syllable is in theory
              >>> permitted. Length is kind of phonemic on consonants (really, a 'long
              >>> consonant' is just two adjacent identical consonants - /alla/ 'day' is
              >>> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
              >>> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
              >>> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
              >>> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
              >>> distinguished in codas. There's not much else in the way of phonological
              >>> rules/alternations, at the moment this is the idealised pre-protolang
              >>> stage, and I'll need to send it through some sound changes before I get
              >>> a good protolang by the technical definition.
              >>> Stress is noncontrastive - it occurs on the heaviest of the last three
              >>> syllables, defaulting to the antepenultimate when they're all equal
              >>> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
              >>>
              >>> Grammatically, it's agglutinative and erg-abs. The two basic word
              >>> classes are noun and verb (all 'adjectives' are just verbs, and at least
              >>> in the protolanguage all 'adverbs' are clearly nouns or nominalised
              >>> verbs marked with a non-core case). Word order is VSO when there's no
              >>> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
              >>> gone', but /le/ /fikolti/ 'the fact that the man has gone').
              >>>
              >>> Verbs don't care about person, number, or tense, but there are 5 or 7
              >>> aspect markers (perfective/stative (null-marked), progressive, perfect,
              >>> expective, intentive(?), and hortative and imperative if you count them
              >>> - they're mutually exclusive with aspect). It has a number of 'mood'
              >>> markers: potential and permissive; volitive, suggestive, and
              >>> necessitive, and weak and strong expectation (weak is for 'I bet X is
              >>> the case' and strong is for 'X /has/ to be the case, I just know it').
              >>> Relative clauses are formed by using the attributive affix
              >>> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
              >>> role the modified noun would have in the clause, that's left up to
              >>> context (so typically you can only relativise obliques when the
              >>> subclause has all of its core arguments overtly specified). (This works
              >>> mostly like the Japanese rentaikei.)
              >>>
              >>> The only obligatory marking on nouns is case, but there's a number of
              >>> other potential affixes. Number is especially complex - specific
              >>> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
              >>> people', it's base-8 so that breaks down as /le-mo-fye-thon/
              >>> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
              >>> 'more than half (of a group)', 'less than half (of a group)', 'part (of
              >>> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
              >>> 'none', and further by 'the next' or 'the previous' (allowing for very
              >>> long sequences such as /lemofyethondawfag /'none of the last twenty
              >>> people'). Demonstratives are also noun suffixes (there's a two-way
              >>> distinction, 'this'/'that'), and an interrogative marker can slot in
              >>> here too (/lewos?/ 'who? / which person?').
              >>>
              >>> There are a good number of cases (it's kind of Finno-Ugric in this
              >>> regard :P) Beyond the erg and null-marked abs, there's two kinds of
              >>> genitives, possessive and categorical (for things like 'men of that
              >>> village', 'the strength of an ox', 'a sword of bronze' and so on), both
              >>> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
              >>> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
              >>> heard as 'the cat is the man's').
              >>> Locative cases are the following: inessive and exessive (both used for
              >>> general locatives, inessive for being within the boundaries of a place,
              >>> exessive for being near but outside the boundaries of a place or
              >>> object), superessive and subessive, proessive ('in front of') and
              >>> postessive ('behind'), comitative, allative (also used as a dative) and
              >>> ablative (also used as the agent of causatives and volitives), illative
              >>> and ellative, superlative ('going over') and sublative ('going under'),
              >>> circumlative/circumessive, and adspective ('facing') and abspective
              >>> ('facing away from').
              >>> Non-locative cases are benefactive (also used for the experiencer with a
              >>> number of perception verbs - 'see' for example has a BEN subject when
              >>> you would expect ERG, and giving it an ERG subject changes the meaning
              >>> to 'look at'), instrumental, causative, and comparative.
              >>> Copular constructions are formed by affixing the copular verbaliser
              >>> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
              >>> somewhat idiosyncratic way to express motion - while it's perfectly
              >>> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
              >>> /literally 'I am going to the river'), it's much more native-sounding to
              >>> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
              >>>
              >>> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
              >>> 2nd person), there's not much in the way of real pronouns - nouns can be
              >>> used with any person as their referent, and typical non-personal
              >>> pronouny things are done with nouns plus suffixes (so 'who' is just
              >>> /lewos /(/person-INT/), literally 'which person'). There is a set of
              >>> 'generic nouns', though, which are basically nouns that refer to quite
              >>> large categories of things - 'person', 'object', 'place', 'point in
              >>> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
              >>> the set - and this allows for fairly conventionalised pronouns ('person'
              >>> can be 'him/her', 'object' can be 'it', 'place' plus the near
              >>> demonstrative can be 'here', and so on). These nouns are further
              >>> distinguished from other nouns by being used as nominaliser suffixes -
              >>> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
              >>> died' to make /ryukolub/ 'the reason [the subj] died'.
              >>> There are also two generic verbs, meaning something like 'to do (it)'
              >>> and 'to go (there)'. They can also be used as nominalisers (somewhat
              >>> ironically :P), meaning 'method' and 'process', respectively.
              >>>
              >>> There's a few other small details here and there (I've left out
              >>> valence-change affixes, for example), but that's a basic overview of the
              >>> language. The goal is realism (indeed, all else is secondary), so some
              >>> comments in regards to how realistic these systems are would be
              >>> appreciated! I've got a few ideas on where to go with it from here, but
              >>> if anyone has any ideas I'd be happy to hear them.
              >>>
              >>> This is my main conlang project, but I'm also working on a set of
              >>> condialects of Japanese (splitting off at various points after around
              >>> 1610?), and I'd be happy to describe them if anyone's interested. (It's
              >>> gotten to the point where I'll slip into my primary condialect every
              >>> once in a while while thinking in Japanese, even when I'm not explicitly
              >>> trying to think in it :P)
            • Aodhán Aannestad
              ... Yes, and yes. And thank you ^_^ As for glosses, that s probably a good idea. I ll add a few here. /roy-r-kyag sya-Ø ne-twa see-NEG-POTENTIAL 2-ABS 1-BEN
              Message 6 of 11 , Jun 25, 2013
                On 6/25/2013 4:03 PM, James Kane wrote:
                > So the consonant written <r> is a flap? And /x/ is written <kh>? This seems very cool although there's quite a lot to take in, maybe a few example sentences with gloss outlining the principle features would help.

                Yes, and yes. And thank you ^_^ As for glosses, that's probably a good
                idea. I'll add a few here.

                /roy-r-kyag sya-Ø ne-twa
                see-NEG-POTENTIAL 2-ABS 1-BEN
                /'I can't see you.'

                /puyag Yompan-na-se le-saw-Ø/
                /be.acquainted.with Yompan-CATEG-ATT person/-/3/-/ABS/
                '[I] know three people from Yompan.'

                /sayma-soy-si//-rod
                sea-ALL-COP-STREXP
                /'Surely [he] must be on [his] way to the sea!'

                /ma-geg-mig//-si-mul //than/-/wos//(-//mig)/
                /place-NEARDEM-INESS-COP-NECESS/ /time-Q(-INESS)
                /'When do [you] need to be here?'
                (explicit marking of things that are clearly adverbs probably ought to
                be optional, and so it is here.)

                /el-ob-kol //le-mo-fye-dawf-ag-Ø
                good-enough-PERF person-8-2-none-next-ABS
                /'None of the last 16 men have been good enough.'
                (haven't decided on negative concord or whatever - this could end up as
                /elobkol lemofyedawfag, elobirkol lemofyedawfag, /or /elobirkol
                lemofyetwegag/ (dropping the 'none' affix and just using 'not all'))

                /ek sya-y le-Ø //udad-kol-ti-Ø, ek-ir=pew udad-kol-se le-wos-Ø
                know 2-ERG person-ABS kill-PERF-NOM-ABS, know-NEG=but kill-PERF-ATT
                person-Q-ABS
                /'[I] know you killed a man, but I don't know who you killed.'
                (the first clause on its own could also be interpreted as 'I know you
                killed him/her.')

                >
                > From what I can tell I like the way the -se affix works although I'm not sure I quite understand it, and when the change in word order happens in relative clauses.
                Here's an example, I hope it makes things clear:

                /nyag-okh ne-y man-Ø
                eat-PROG 1-ERG food-ABS
                /'I'm eating (some) food.'
                vs.
                /katrari ne-y nyag-okh-se man-Ø
                //burn 1-ERG //eat-PROG-ATT food-ABS
                /'The food I'm eating is on fire.'

                It works basically like in Middle Japanese, though MJ doesn't change any
                word order:
                /
                ware ga tabyemono wo tab-u/
                /1 SUBJ food OBJ eat/
                'I'm eating (some) food.'
                vs.
                /ware no tab-uru tabyemono ga moy-u/
                /1 SUBJ eat-RENTAIKEI //food SUBJ burn/
                'The food I'm eating is on fire.'
                (the rentaikei is Japanese's version of an attributive suffix, and the
                subject particle changes for relative clauses.)


                >
                > Your condialects project sounds very exciting and I would be interested to have a small peak, although I know almost nothing about Japanese (indeed, I know more about Okinawan)
                I'll post a thing about it when I have time, then. Odd to see someone
                who knows more about Okinawan than Japanese :P

                (I know most Ryukyuan languages have an animacy hierarchy distinction on
                subject/genitive particles - does Okinawan do that too?)

                >
                > James
                >
                > On 25/06/2013, at 7:50 AM, Aodhán Aannestad <tolkien_freak@...> wrote:
                >
                >> On 6/24/2013 1:16 PM, neo gu wrote:
                >>> On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
                >>>
                >>>> Hello! I'm a longtime lurker, finally deciding to get involved a little.
                >>>> Some of you may have met me at the last LCC, I'm with UT Austin's
                >>>> conlang club and I drove some people around. I've been conlanging on and
                >>>> off for quite some time now, but I don't have much to show for it - I
                >>>> tend to be singlemindedly focused on realism, and as my level of
                >>>> linguistics knowledge has increased I've scrapped and restarted on
                >>>> several occasions due to realising how horribly unrealistic of a
                >>>> language I'd created. This is kind of the latest iteration of that
                >>>> cycle, and as I'm done with all the core undergrad ling classes now
                >>>> hopefully I've run out of major things to learn and it'll prove to be
                >>>> the last.
                >>> Welcome to active participation!
                >>>
                >>> I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
                >>>
                >>> --
                >>> Jeff Jones
                >> Oh, yeah, the email got messed up because for whatever reason my editor is a weird mix of HTML and WYSIWYG. Those are supposed to be what's /T/ and /4/ in X-SAMPA.
                >>
                >>>> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
                >>>> using it for a protolang for a large project, and since said large
                >>>> project has yet to really begin, it's not really in a state where I can
                >>>> really name anything. (I suppose I could come up with an endonym for it
                >>>> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
                >>>> though that doesn't sound that great to my ears.)
                >>>>
                >>>> Phonologically, it's not that interesting - five vowels (/a i u e o/),
                >>>> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
                >>>> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
                >>>> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
                >>>> l/ round out the inventory.
                >>>> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
                >>>> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
                >>>> shape target - anything from V to a maximal syllable is in theory
                >>>> permitted. Length is kind of phonemic on consonants (really, a 'long
                >>>> consonant' is just two adjacent identical consonants - /alla/ 'day' is
                >>>> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
                >>>> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
                >>>> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
                >>>> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
                >>>> distinguished in codas. There's not much else in the way of phonological
                >>>> rules/alternations, at the moment this is the idealised pre-protolang
                >>>> stage, and I'll need to send it through some sound changes before I get
                >>>> a good protolang by the technical definition.
                >>>> Stress is noncontrastive - it occurs on the heaviest of the last three
                >>>> syllables, defaulting to the antepenultimate when they're all equal
                >>>> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
                >>>>
                >>>> Grammatically, it's agglutinative and erg-abs. The two basic word
                >>>> classes are noun and verb (all 'adjectives' are just verbs, and at least
                >>>> in the protolanguage all 'adverbs' are clearly nouns or nominalised
                >>>> verbs marked with a non-core case). Word order is VSO when there's no
                >>>> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
                >>>> gone', but /le/ /fikolti/ 'the fact that the man has gone').
                >>>>
                >>>> Verbs don't care about person, number, or tense, but there are 5 or 7
                >>>> aspect markers (perfective/stative (null-marked), progressive, perfect,
                >>>> expective, intentive(?), and hortative and imperative if you count them
                >>>> - they're mutually exclusive with aspect). It has a number of 'mood'
                >>>> markers: potential and permissive; volitive, suggestive, and
                >>>> necessitive, and weak and strong expectation (weak is for 'I bet X is
                >>>> the case' and strong is for 'X /has/ to be the case, I just know it').
                >>>> Relative clauses are formed by using the attributive affix
                >>>> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
                >>>> role the modified noun would have in the clause, that's left up to
                >>>> context (so typically you can only relativise obliques when the
                >>>> subclause has all of its core arguments overtly specified). (This works
                >>>> mostly like the Japanese rentaikei.)
                >>>>
                >>>> The only obligatory marking on nouns is case, but there's a number of
                >>>> other potential affixes. Number is especially complex - specific
                >>>> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                >>>> people', it's base-8 so that breaks down as /le-mo-fye-thon/
                >>>> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                >>>> 'more than half (of a group)', 'less than half (of a group)', 'part (of
                >>>> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
                >>>> 'none', and further by 'the next' or 'the previous' (allowing for very
                >>>> long sequences such as /lemofyethondawfag /'none of the last twenty
                >>>> people'). Demonstratives are also noun suffixes (there's a two-way
                >>>> distinction, 'this'/'that'), and an interrogative marker can slot in
                >>>> here too (/lewos?/ 'who? / which person?').
                >>>>
                >>>> There are a good number of cases (it's kind of Finno-Ugric in this
                >>>> regard :P) Beyond the erg and null-marked abs, there's two kinds of
                >>>> genitives, possessive and categorical (for things like 'men of that
                >>>> village', 'the strength of an ox', 'a sword of bronze' and so on), both
                >>>> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
                >>>> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
                >>>> heard as 'the cat is the man's').
                >>>> Locative cases are the following: inessive and exessive (both used for
                >>>> general locatives, inessive for being within the boundaries of a place,
                >>>> exessive for being near but outside the boundaries of a place or
                >>>> object), superessive and subessive, proessive ('in front of') and
                >>>> postessive ('behind'), comitative, allative (also used as a dative) and
                >>>> ablative (also used as the agent of causatives and volitives), illative
                >>>> and ellative, superlative ('going over') and sublative ('going under'),
                >>>> circumlative/circumessive, and adspective ('facing') and abspective
                >>>> ('facing away from').
                >>>> Non-locative cases are benefactive (also used for the experiencer with a
                >>>> number of perception verbs - 'see' for example has a BEN subject when
                >>>> you would expect ERG, and giving it an ERG subject changes the meaning
                >>>> to 'look at'), instrumental, causative, and comparative.
                >>>> Copular constructions are formed by affixing the copular verbaliser
                >>>> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
                >>>> somewhat idiosyncratic way to express motion - while it's perfectly
                >>>> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
                >>>> /literally 'I am going to the river'), it's much more native-sounding to
                >>>> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
                >>>>
                >>>> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
                >>>> 2nd person), there's not much in the way of real pronouns - nouns can be
                >>>> used with any person as their referent, and typical non-personal
                >>>> pronouny things are done with nouns plus suffixes (so 'who' is just
                >>>> /lewos /(/person-INT/), literally 'which person'). There is a set of
                >>>> 'generic nouns', though, which are basically nouns that refer to quite
                >>>> large categories of things - 'person', 'object', 'place', 'point in
                >>>> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
                >>>> the set - and this allows for fairly conventionalised pronouns ('person'
                >>>> can be 'him/her', 'object' can be 'it', 'place' plus the near
                >>>> demonstrative can be 'here', and so on). These nouns are further
                >>>> distinguished from other nouns by being used as nominaliser suffixes -
                >>>> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
                >>>> died' to make /ryukolub/ 'the reason [the subj] died'.
                >>>> There are also two generic verbs, meaning something like 'to do (it)'
                >>>> and 'to go (there)'. They can also be used as nominalisers (somewhat
                >>>> ironically :P), meaning 'method' and 'process', respectively.
                >>>>
                >>>> There's a few other small details here and there (I've left out
                >>>> valence-change affixes, for example), but that's a basic overview of the
                >>>> language. The goal is realism (indeed, all else is secondary), so some
                >>>> comments in regards to how realistic these systems are would be
                >>>> appreciated! I've got a few ideas on where to go with it from here, but
                >>>> if anyone has any ideas I'd be happy to hear them.
                >>>>
                >>>> This is my main conlang project, but I'm also working on a set of
                >>>> condialects of Japanese (splitting off at various points after around
                >>>> 1610?), and I'd be happy to describe them if anyone's interested. (It's
                >>>> gotten to the point where I'll slip into my primary condialect every
                >>>> once in a while while thinking in Japanese, even when I'm not explicitly
                >>>> trying to think in it :P)
              • James Kane
                Okay so looking through these examples, is the -se suffix similar to Chinese 的? Are adjectives similar to Chinese adjectives in that they behave more like
                Message 7 of 11 , Jun 25, 2013
                  Okay so looking through these examples, is the -se suffix similar to Chinese 的? Are adjectives similar to Chinese adjectives in that they behave more like verbs? This seems to be the case with the possessive pronouns, where 'my cat' looks more like 'the.cat_which_is.mine'.

                  Anyway, when I said I knew more Okinawan than Japanese I was comparing a minuscule amount of knowledge of one with an even more minuscule amount of knowledge of the other.

                  There is an animacy distinction but I'm not sure how far it extends. I don't think it's in the case particles. It is however in the verb to have, wun is animate and an is inanimate. In contrast, yan is to be for any subjects.


                  James

                  On 26/06/2013, at 10:31 AM, Aodhán Aannestad <tolkien_freak@...> wrote:

                  > On 6/25/2013 4:03 PM, James Kane wrote:
                  >> So the consonant written <r> is a flap? And /x/ is written <kh>? This seems very cool although there's quite a lot to take in, maybe a few example sentences with gloss outlining the principle features would help.
                  >
                  > Yes, and yes. And thank you ^_^ As for glosses, that's probably a good idea. I'll add a few here.
                  >
                  > /roy-r-kyag sya-Ø ne-twa
                  > see-NEG-POTENTIAL 2-ABS 1-BEN
                  > /'I can't see you.'
                  >
                  > /puyag Yompan-na-se le-saw-Ø/
                  > /be.acquainted.with Yompan-CATEG-ATT person/-/3/-/ABS/
                  > '[I] know three people from Yompan.'
                  >
                  > /sayma-soy-si//-rod
                  > sea-ALL-COP-STREXP
                  > /'Surely [he] must be on [his] way to the sea!'
                  >
                  > /ma-geg-mig//-si-mul //than/-/wos//(-//mig)/
                  > /place-NEARDEM-INESS-COP-NECESS/ /time-Q(-INESS)
                  > /'When do [you] need to be here?'
                  > (explicit marking of things that are clearly adverbs probably ought to be optional, and so it is here.)
                  >
                  > /el-ob-kol //le-mo-fye-dawf-ag-Ø
                  > good-enough-PERF person-8-2-none-next-ABS
                  > /'None of the last 16 men have been good enough.'
                  > (haven't decided on negative concord or whatever - this could end up as /elobkol lemofyedawfag, elobirkol lemofyedawfag, /or /elobirkol lemofyetwegag/ (dropping the 'none' affix and just using 'not all'))
                  >
                  > /ek sya-y le-Ø //udad-kol-ti-Ø, ek-ir=pew udad-kol-se le-wos-Ø
                  > know 2-ERG person-ABS kill-PERF-NOM-ABS, know-NEG=but kill-PERF-ATT person-Q-ABS
                  > /'[I] know you killed a man, but I don't know who you killed.'
                  > (the first clause on its own could also be interpreted as 'I know you killed him/her.')
                  >
                  >>
                  >> From what I can tell I like the way the -se affix works although I'm not sure I quite understand it, and when the change in word order happens in relative clauses.
                  > Here's an example, I hope it makes things clear:
                  >
                  > /nyag-okh ne-y man-Ø
                  > eat-PROG 1-ERG food-ABS
                  > /'I'm eating (some) food.'
                  > vs.
                  > /katrari ne-y nyag-okh-se man-Ø
                  > //burn 1-ERG //eat-PROG-ATT food-ABS
                  > /'The food I'm eating is on fire.'
                  >
                  > It works basically like in Middle Japanese, though MJ doesn't change any word order:
                  > /
                  > ware ga tabyemono wo tab-u/
                  > /1 SUBJ food OBJ eat/
                  > 'I'm eating (some) food.'
                  > vs.
                  > /ware no tab-uru tabyemono ga moy-u/
                  > /1 SUBJ eat-RENTAIKEI //food SUBJ burn/
                  > 'The food I'm eating is on fire.'
                  > (the rentaikei is Japanese's version of an attributive suffix, and the subject particle changes for relative clauses.)
                  >
                  >
                  >>
                  >> Your condialects project sounds very exciting and I would be interested to have a small peak, although I know almost nothing about Japanese (indeed, I know more about Okinawan)
                  > I'll post a thing about it when I have time, then. Odd to see someone who knows more about Okinawan than Japanese :P
                  >
                  > (I know most Ryukyuan languages have an animacy hierarchy distinction on subject/genitive particles - does Okinawan do that too?)
                  >
                  >>
                  >> James
                  >>
                  >> On 25/06/2013, at 7:50 AM, Aodhán Aannestad <tolkien_freak@...> wrote:
                  >>
                  >>> On 6/24/2013 1:16 PM, neo gu wrote:
                  >>>> On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
                  >>>>
                  >>>>> Hello! I'm a longtime lurker, finally deciding to get involved a little.
                  >>>>> Some of you may have met me at the last LCC, I'm with UT Austin's
                  >>>>> conlang club and I drove some people around. I've been conlanging on and
                  >>>>> off for quite some time now, but I don't have much to show for it - I
                  >>>>> tend to be singlemindedly focused on realism, and as my level of
                  >>>>> linguistics knowledge has increased I've scrapped and restarted on
                  >>>>> several occasions due to realising how horribly unrealistic of a
                  >>>>> language I'd created. This is kind of the latest iteration of that
                  >>>>> cycle, and as I'm done with all the core undergrad ling classes now
                  >>>>> hopefully I've run out of major things to learn and it'll prove to be
                  >>>>> the last.
                  >>>> Welcome to active participation!
                  >>>>
                  >>>> I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
                  >>>>
                  >>>> --
                  >>>> Jeff Jones
                  >>> Oh, yeah, the email got messed up because for whatever reason my editor is a weird mix of HTML and WYSIWYG. Those are supposed to be what's /T/ and /4/ in X-SAMPA.
                  >>>
                  >>>>> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
                  >>>>> using it for a protolang for a large project, and since said large
                  >>>>> project has yet to really begin, it's not really in a state where I can
                  >>>>> really name anything. (I suppose I could come up with an endonym for it
                  >>>>> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
                  >>>>> though that doesn't sound that great to my ears.)
                  >>>>>
                  >>>>> Phonologically, it's not that interesting - five vowels (/a i u e o/),
                  >>>>> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
                  >>>>> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
                  >>>>> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
                  >>>>> l/ round out the inventory.
                  >>>>> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
                  >>>>> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
                  >>>>> shape target - anything from V to a maximal syllable is in theory
                  >>>>> permitted. Length is kind of phonemic on consonants (really, a 'long
                  >>>>> consonant' is just two adjacent identical consonants - /alla/ 'day' is
                  >>>>> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
                  >>>>> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
                  >>>>> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
                  >>>>> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
                  >>>>> distinguished in codas. There's not much else in the way of phonological
                  >>>>> rules/alternations, at the moment this is the idealised pre-protolang
                  >>>>> stage, and I'll need to send it through some sound changes before I get
                  >>>>> a good protolang by the technical definition.
                  >>>>> Stress is noncontrastive - it occurs on the heaviest of the last three
                  >>>>> syllables, defaulting to the antepenultimate when they're all equal
                  >>>>> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
                  >>>>>
                  >>>>> Grammatically, it's agglutinative and erg-abs. The two basic word
                  >>>>> classes are noun and verb (all 'adjectives' are just verbs, and at least
                  >>>>> in the protolanguage all 'adverbs' are clearly nouns or nominalised
                  >>>>> verbs marked with a non-core case). Word order is VSO when there's no
                  >>>>> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
                  >>>>> gone', but /le/ /fikolti/ 'the fact that the man has gone').
                  >>>>>
                  >>>>> Verbs don't care about person, number, or tense, but there are 5 or 7
                  >>>>> aspect markers (perfective/stative (null-marked), progressive, perfect,
                  >>>>> expective, intentive(?), and hortative and imperative if you count them
                  >>>>> - they're mutually exclusive with aspect). It has a number of 'mood'
                  >>>>> markers: potential and permissive; volitive, suggestive, and
                  >>>>> necessitive, and weak and strong expectation (weak is for 'I bet X is
                  >>>>> the case' and strong is for 'X /has/ to be the case, I just know it').
                  >>>>> Relative clauses are formed by using the attributive affix
                  >>>>> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
                  >>>>> role the modified noun would have in the clause, that's left up to
                  >>>>> context (so typically you can only relativise obliques when the
                  >>>>> subclause has all of its core arguments overtly specified). (This works
                  >>>>> mostly like the Japanese rentaikei.)
                  >>>>>
                  >>>>> The only obligatory marking on nouns is case, but there's a number of
                  >>>>> other potential affixes. Number is especially complex - specific
                  >>>>> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                  >>>>> people', it's base-8 so that breaks down as /le-mo-fye-thon/
                  >>>>> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                  >>>>> 'more than half (of a group)', 'less than half (of a group)', 'part (of
                  >>>>> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
                  >>>>> 'none', and further by 'the next' or 'the previous' (allowing for very
                  >>>>> long sequences such as /lemofyethondawfag /'none of the last twenty
                  >>>>> people'). Demonstratives are also noun suffixes (there's a two-way
                  >>>>> distinction, 'this'/'that'), and an interrogative marker can slot in
                  >>>>> here too (/lewos?/ 'who? / which person?').
                  >>>>>
                  >>>>> There are a good number of cases (it's kind of Finno-Ugric in this
                  >>>>> regard :P) Beyond the erg and null-marked abs, there's two kinds of
                  >>>>> genitives, possessive and categorical (for things like 'men of that
                  >>>>> village', 'the strength of an ox', 'a sword of bronze' and so on), both
                  >>>>> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
                  >>>>> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
                  >>>>> heard as 'the cat is the man's').
                  >>>>> Locative cases are the following: inessive and exessive (both used for
                  >>>>> general locatives, inessive for being within the boundaries of a place,
                  >>>>> exessive for being near but outside the boundaries of a place or
                  >>>>> object), superessive and subessive, proessive ('in front of') and
                  >>>>> postessive ('behind'), comitative, allative (also used as a dative) and
                  >>>>> ablative (also used as the agent of causatives and volitives), illative
                  >>>>> and ellative, superlative ('going over') and sublative ('going under'),
                  >>>>> circumlative/circumessive, and adspective ('facing') and abspective
                  >>>>> ('facing away from').
                  >>>>> Non-locative cases are benefactive (also used for the experiencer with a
                  >>>>> number of perception verbs - 'see' for example has a BEN subject when
                  >>>>> you would expect ERG, and giving it an ERG subject changes the meaning
                  >>>>> to 'look at'), instrumental, causative, and comparative.
                  >>>>> Copular constructions are formed by affixing the copular verbaliser
                  >>>>> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
                  >>>>> somewhat idiosyncratic way to express motion - while it's perfectly
                  >>>>> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
                  >>>>> /literally 'I am going to the river'), it's much more native-sounding to
                  >>>>> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
                  >>>>>
                  >>>>> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
                  >>>>> 2nd person), there's not much in the way of real pronouns - nouns can be
                  >>>>> used with any person as their referent, and typical non-personal
                  >>>>> pronouny things are done with nouns plus suffixes (so 'who' is just
                  >>>>> /lewos /(/person-INT/), literally 'which person'). There is a set of
                  >>>>> 'generic nouns', though, which are basically nouns that refer to quite
                  >>>>> large categories of things - 'person', 'object', 'place', 'point in
                  >>>>> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
                  >>>>> the set - and this allows for fairly conventionalised pronouns ('person'
                  >>>>> can be 'him/her', 'object' can be 'it', 'place' plus the near
                  >>>>> demonstrative can be 'here', and so on). These nouns are further
                  >>>>> distinguished from other nouns by being used as nominaliser suffixes -
                  >>>>> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
                  >>>>> died' to make /ryukolub/ 'the reason [the subj] died'.
                  >>>>> There are also two generic verbs, meaning something like 'to do (it)'
                  >>>>> and 'to go (there)'. They can also be used as nominalisers (somewhat
                  >>>>> ironically :P), meaning 'method' and 'process', respectively.
                  >>>>>
                  >>>>> There's a few other small details here and there (I've left out
                  >>>>> valence-change affixes, for example), but that's a basic overview of the
                  >>>>> language. The goal is realism (indeed, all else is secondary), so some
                  >>>>> comments in regards to how realistic these systems are would be
                  >>>>> appreciated! I've got a few ideas on where to go with it from here, but
                  >>>>> if anyone has any ideas I'd be happy to hear them.
                  >>>>>
                  >>>>> This is my main conlang project, but I'm also working on a set of
                  >>>>> condialects of Japanese (splitting off at various points after around
                  >>>>> 1610?), and I'd be happy to describe them if anyone's interested. (It's
                  >>>>> gotten to the point where I'll slip into my primary condialect every
                  >>>>> once in a while while thinking in Japanese, even when I'm not explicitly
                  >>>>> trying to think in it :P)
                • Aodhán Aannestad
                  ... I m not that familiar with Chinese, but from my knowledge, that s about how it works. 的 after nouns though corresponds to two morphemes (POSS+ATT), and
                  Message 8 of 11 , Jun 25, 2013
                    On 6/25/2013 9:47 PM, James Kane wrote:
                    > Okay so looking through these examples, is the -se suffix similar to Chinese 的? Are adjectives similar to Chinese adjectives in that they behave more like verbs? This seems to be the case with the possessive pronouns, where 'my cat' looks more like 'the.cat_which_is.mine'.
                    I'm not that familiar with Chinese, but from my knowledge, that's about
                    how it works. 的 after nouns though corresponds to two morphemes
                    (POSS+ATT), and you can use POSS as just a verb (e.g. /ne-na-thas
                    nyawa-Ø, //1-POSS-SUGGEST cat-ABS/ 'the cat should be mine').

                    >
                    > Anyway, when I said I knew more Okinawan than Japanese I was comparing a minuscule amount of knowledge of one with an even more minuscule amount of knowledge of the other.
                    >
                    > There is an animacy distinction but I'm not sure how far it extends. I don't think it's in the case particles. It is however in the verb to have, wun is animate and an is inanimate. In contrast, yan is to be for any subjects.
                    Yeah, Japanese does the same thing (iru/oru for 'to exist (of an animate
                    subject)' and aru for 'to exist (of an inanimate subject)', but that's
                    about as far as it goes.
                    Other Ryukyuan languages have a really weird setup with case particles,
                    e.g. Yuwan (from Amami), where both ga and nu can mark both subject and
                    genitive, but ga is for higher-animacy subjects and nu is for lower, and
                    the threshold is different when they're used as subject vs. genitive -
                    ga is genitive with demonstratives and respected kinship terms, nu is
                    the rest; ga is subject for demonstratives, respected kinship terms and
                    human names are zero-marked, and nu is the rest. Yeah.

                    >
                    >
                    > James
                    >
                    > On 26/06/2013, at 10:31 AM, Aodhán Aannestad <tolkien_freak@...> wrote:
                    >
                    >> On 6/25/2013 4:03 PM, James Kane wrote:
                    >>> So the consonant written <r> is a flap? And /x/ is written <kh>? This seems very cool although there's quite a lot to take in, maybe a few example sentences with gloss outlining the principle features would help.
                    >> Yes, and yes. And thank you ^_^ As for glosses, that's probably a good idea. I'll add a few here.
                    >>
                    >> /roy-r-kyag sya-Ø ne-twa
                    >> see-NEG-POTENTIAL 2-ABS 1-BEN
                    >> /'I can't see you.'
                    >>
                    >> /puyag Yompan-na-se le-saw-Ø/
                    >> /be.acquainted.with Yompan-CATEG-ATT person/-/3/-/ABS/
                    >> '[I] know three people from Yompan.'
                    >>
                    >> /sayma-soy-si//-rod
                    >> sea-ALL-COP-STREXP
                    >> /'Surely [he] must be on [his] way to the sea!'
                    >>
                    >> /ma-geg-mig//-si-mul //than/-/wos//(-//mig)/
                    >> /place-NEARDEM-INESS-COP-NECESS/ /time-Q(-INESS)
                    >> /'When do [you] need to be here?'
                    >> (explicit marking of things that are clearly adverbs probably ought to be optional, and so it is here.)
                    >>
                    >> /el-ob-kol //le-mo-fye-dawf-ag-Ø
                    >> good-enough-PERF person-8-2-none-next-ABS
                    >> /'None of the last 16 men have been good enough.'
                    >> (haven't decided on negative concord or whatever - this could end up as /elobkol lemofyedawfag, elobirkol lemofyedawfag, /or /elobirkol lemofyetwegag/ (dropping the 'none' affix and just using 'not all'))
                    >>
                    >> /ek sya-y le-Ø //udad-kol-ti-Ø, ek-ir=pew udad-kol-se le-wos-Ø
                    >> know 2-ERG person-ABS kill-PERF-NOM-ABS, know-NEG=but kill-PERF-ATT person-Q-ABS
                    >> /'[I] know you killed a man, but I don't know who you killed.'
                    >> (the first clause on its own could also be interpreted as 'I know you killed him/her.')
                    >>
                    >>> From what I can tell I like the way the -se affix works although I'm not sure I quite understand it, and when the change in word order happens in relative clauses.
                    >> Here's an example, I hope it makes things clear:
                    >>
                    >> /nyag-okh ne-y man-Ø
                    >> eat-PROG 1-ERG food-ABS
                    >> /'I'm eating (some) food.'
                    >> vs.
                    >> /katrari ne-y nyag-okh-se man-Ø
                    >> //burn 1-ERG //eat-PROG-ATT food-ABS
                    >> /'The food I'm eating is on fire.'
                    >>
                    >> It works basically like in Middle Japanese, though MJ doesn't change any word order:
                    >> /
                    >> ware ga tabyemono wo tab-u/
                    >> /1 SUBJ food OBJ eat/
                    >> 'I'm eating (some) food.'
                    >> vs.
                    >> /ware no tab-uru tabyemono ga moy-u/
                    >> /1 SUBJ eat-RENTAIKEI //food SUBJ burn/
                    >> 'The food I'm eating is on fire.'
                    >> (the rentaikei is Japanese's version of an attributive suffix, and the subject particle changes for relative clauses.)
                    >>
                    >>
                    >>> Your condialects project sounds very exciting and I would be interested to have a small peak, although I know almost nothing about Japanese (indeed, I know more about Okinawan)
                    >> I'll post a thing about it when I have time, then. Odd to see someone who knows more about Okinawan than Japanese :P
                    >>
                    >> (I know most Ryukyuan languages have an animacy hierarchy distinction on subject/genitive particles - does Okinawan do that too?)
                    >>
                    >>> James
                    >>>
                    >>> On 25/06/2013, at 7:50 AM, Aodhán Aannestad <tolkien_freak@...> wrote:
                    >>>
                    >>>> On 6/24/2013 1:16 PM, neo gu wrote:
                    >>>>> On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
                    >>>>>
                    >>>>>> Hello! I'm a longtime lurker, finally deciding to get involved a little.
                    >>>>>> Some of you may have met me at the last LCC, I'm with UT Austin's
                    >>>>>> conlang club and I drove some people around. I've been conlanging on and
                    >>>>>> off for quite some time now, but I don't have much to show for it - I
                    >>>>>> tend to be singlemindedly focused on realism, and as my level of
                    >>>>>> linguistics knowledge has increased I've scrapped and restarted on
                    >>>>>> several occasions due to realising how horribly unrealistic of a
                    >>>>>> language I'd created. This is kind of the latest iteration of that
                    >>>>>> cycle, and as I'm done with all the core undergrad ling classes now
                    >>>>>> hopefully I've run out of major things to learn and it'll prove to be
                    >>>>>> the last.
                    >>>>> Welcome to active participation!
                    >>>>>
                    >>>>> I'm not one of the experts on naturalistic conlanging, but I do have one comment: one of your fricatives is showing up as a /?/ in the archives; also, /w j ? l/.
                    >>>>>
                    >>>>> --
                    >>>>> Jeff Jones
                    >>>> Oh, yeah, the email got messed up because for whatever reason my editor is a weird mix of HTML and WYSIWYG. Those are supposed to be what's /T/ and /4/ in X-SAMPA.
                    >>>>
                    >>>>>> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
                    >>>>>> using it for a protolang for a large project, and since said large
                    >>>>>> project has yet to really begin, it's not really in a state where I can
                    >>>>>> really name anything. (I suppose I could come up with an endonym for it
                    >>>>>> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
                    >>>>>> though that doesn't sound that great to my ears.)
                    >>>>>>
                    >>>>>> Phonologically, it's not that interesting - five vowels (/a i u e o/),
                    >>>>>> two stop series (/p t k/ and /p^(h) t^(h) k^(h)/, transcribed <b d g>
                    >>>>>> and <p t k>), /m n/ - the usual stuff. It only has one fricative series,
                    >>>>>> but it does distinguish /?/ and /s/ (for four total, /f ? s x/). /w j ?
                    >>>>>> l/ round out the inventory.
                    >>>>>> Syllable structure is (C)(G)V(G)(C), where <G> is a glide. /uw ij/ are
                    >>>>>> disallowed, and maximal CGVGC syllables are very rare. Affixes have no
                    >>>>>> shape target - anything from V to a maximal syllable is in theory
                    >>>>>> permitted. Length is kind of phonemic on consonants (really, a 'long
                    >>>>>> consonant' is just two adjacent identical consonants - /alla/ 'day' is
                    >>>>>> VC.CV, not V.C?V), not on vowels - any sequences of two identical vowels
                    >>>>>> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
                    >>>>>> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
                    >>>>>> /dorayra/ 'boy', from /dora /'man' + /ira/). Aspiration isn't
                    >>>>>> distinguished in codas. There's not much else in the way of phonological
                    >>>>>> rules/alternations, at the moment this is the idealised pre-protolang
                    >>>>>> stage, and I'll need to send it through some sound changes before I get
                    >>>>>> a good protolang by the technical definition.
                    >>>>>> Stress is noncontrastive - it occurs on the heaviest of the last three
                    >>>>>> syllables, defaulting to the antepenultimate when they're all equal
                    >>>>>> (so/魮ira, dorṲa,////dal鳩se (do-person-COP-ATT)/, etc.)/./
                    >>>>>>
                    >>>>>> Grammatically, it's agglutinative and erg-abs. The two basic word
                    >>>>>> classes are noun and verb (all 'adjectives' are just verbs, and at least
                    >>>>>> in the protolanguage all 'adverbs' are clearly nouns or nominalised
                    >>>>>> verbs marked with a non-core case). Word order is VSO when there's no
                    >>>>>> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
                    >>>>>> gone', but /le/ /fikolti/ 'the fact that the man has gone').
                    >>>>>>
                    >>>>>> Verbs don't care about person, number, or tense, but there are 5 or 7
                    >>>>>> aspect markers (perfective/stative (null-marked), progressive, perfect,
                    >>>>>> expective, intentive(?), and hortative and imperative if you count them
                    >>>>>> - they're mutually exclusive with aspect). It has a number of 'mood'
                    >>>>>> markers: potential and permissive; volitive, suggestive, and
                    >>>>>> necessitive, and weak and strong expectation (weak is for 'I bet X is
                    >>>>>> the case' and strong is for 'X /has/ to be the case, I just know it').
                    >>>>>> Relative clauses are formed by using the attributive affix
                    >>>>>> /-se/:/fikolse le/ 'the man who has gone'. /-se/ doesn't specify the
                    >>>>>> role the modified noun would have in the clause, that's left up to
                    >>>>>> context (so typically you can only relativise obliques when the
                    >>>>>> subclause has all of its core arguments overtly specified). (This works
                    >>>>>> mostly like the Japanese rentaikei.)
                    >>>>>>
                    >>>>>> The only obligatory marking on nouns is case, but there's a number of
                    >>>>>> other potential affixes. Number is especially complex - specific
                    >>>>>> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                    >>>>>> people', it's base-8 so that breaks down as /le-mo-fye-thon/
                    >>>>>> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                    >>>>>> 'more than half (of a group)', 'less than half (of a group)', 'part (of
                    >>>>>> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
                    >>>>>> 'none', and further by 'the next' or 'the previous' (allowing for very
                    >>>>>> long sequences such as /lemofyethondawfag /'none of the last twenty
                    >>>>>> people'). Demonstratives are also noun suffixes (there's a two-way
                    >>>>>> distinction, 'this'/'that'), and an interrogative marker can slot in
                    >>>>>> here too (/lewos?/ 'who? / which person?').
                    >>>>>>
                    >>>>>> There are a good number of cases (it's kind of Finno-Ugric in this
                    >>>>>> regard :P) Beyond the erg and null-marked abs, there's two kinds of
                    >>>>>> genitives, possessive and categorical (for things like 'men of that
                    >>>>>> village', 'the strength of an ox', 'a sword of bronze' and so on), both
                    >>>>>> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
                    >>>>>> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
                    >>>>>> heard as 'the cat is the man's').
                    >>>>>> Locative cases are the following: inessive and exessive (both used for
                    >>>>>> general locatives, inessive for being within the boundaries of a place,
                    >>>>>> exessive for being near but outside the boundaries of a place or
                    >>>>>> object), superessive and subessive, proessive ('in front of') and
                    >>>>>> postessive ('behind'), comitative, allative (also used as a dative) and
                    >>>>>> ablative (also used as the agent of causatives and volitives), illative
                    >>>>>> and ellative, superlative ('going over') and sublative ('going under'),
                    >>>>>> circumlative/circumessive, and adspective ('facing') and abspective
                    >>>>>> ('facing away from').
                    >>>>>> Non-locative cases are benefactive (also used for the experiencer with a
                    >>>>>> number of perception verbs - 'see' for example has a BEN subject when
                    >>>>>> you would expect ERG, and giving it an ERG subject changes the meaning
                    >>>>>> to 'look at'), instrumental, causative, and comparative.
                    >>>>>> Copular constructions are formed by affixing the copular verbaliser
                    >>>>>> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
                    >>>>>> somewhat idiosyncratic way to express motion - while it's perfectly
                    >>>>>> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
                    >>>>>> /literally 'I am going to the river'), it's much more native-sounding to
                    >>>>>> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
                    >>>>>>
                    >>>>>> Beyond three generic 'pronouns' (1st person exclusive and inclusive and
                    >>>>>> 2nd person), there's not much in the way of real pronouns - nouns can be
                    >>>>>> used with any person as their referent, and typical non-personal
                    >>>>>> pronouny things are done with nouns plus suffixes (so 'who' is just
                    >>>>>> /lewos /(/person-INT/), literally 'which person'). There is a set of
                    >>>>>> 'generic nouns', though, which are basically nouns that refer to quite
                    >>>>>> large categories of things - 'person', 'object', 'place', 'point in
                    >>>>>> time', 'state of being', 'piece', 'reason/cause', and 'action' make up
                    >>>>>> the set - and this allows for fairly conventionalised pronouns ('person'
                    >>>>>> can be 'him/her', 'object' can be 'it', 'place' plus the near
                    >>>>>> demonstrative can be 'here', and so on). These nouns are further
                    >>>>>> distinguished from other nouns by being used as nominaliser suffixes -
                    >>>>>> so /ub/, the generic for 'reason/cause', combines with /ryukol /'has
                    >>>>>> died' to make /ryukolub/ 'the reason [the subj] died'.
                    >>>>>> There are also two generic verbs, meaning something like 'to do (it)'
                    >>>>>> and 'to go (there)'. They can also be used as nominalisers (somewhat
                    >>>>>> ironically :P), meaning 'method' and 'process', respectively.
                    >>>>>>
                    >>>>>> There's a few other small details here and there (I've left out
                    >>>>>> valence-change affixes, for example), but that's a basic overview of the
                    >>>>>> language. The goal is realism (indeed, all else is secondary), so some
                    >>>>>> comments in regards to how realistic these systems are would be
                    >>>>>> appreciated! I've got a few ideas on where to go with it from here, but
                    >>>>>> if anyone has any ideas I'd be happy to hear them.
                    >>>>>>
                    >>>>>> This is my main conlang project, but I'm also working on a set of
                    >>>>>> condialects of Japanese (splitting off at various points after around
                    >>>>>> 1610?), and I'd be happy to describe them if anyone's interested. (It's
                    >>>>>> gotten to the point where I'll slip into my primary condialect every
                    >>>>>> once in a while while thinking in Japanese, even when I'm not explicitly
                    >>>>>> trying to think in it :P)
                  • Alex Fink
                    ... Twas good to meet / bum rides off you there. ... Eh, I m in no rush to give my projects endonyms; I give them codenames or descriptive English names until
                    Message 9 of 11 , Jun 29, 2013
                      On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:

                      >Hello! I'm a longtime lurker, finally deciding to get involved a little.
                      >Some of you may have met me at the last LCC, I'm with UT Austin's
                      >conlang club and I drove some people around.

                      'Twas good to meet / bum rides off you there.

                      >Anyway, on to the language. It doesn't have a name yet, sadly - I'm
                      >using it for a protolang for a large project, and since said large
                      >project has yet to really begin, it's not really in a state where I can
                      >really name anything. (I suppose I could come up with an endonym for it
                      >- it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
                      >though that doesn't sound that great to my ears.)

                      Eh, I'm in no rush to give my projects endonyms; I give them codenames or descriptive English names until then. And anyway, endonyms are often non-transparently formed, especially in areas with lots of cross-cultural interaction, in deriving from place names or tribal names or calqued or borrowed exonyms or whatnot; so if you don't like /lesuj/ it's hardly forced on you.

                      >Affixes have no
                      >shape target - anything from V to a maximal syllable is in theory
                      >permitted.

                      This is kinda odd, unless the affixing is all really new. (I suppose that what I'm calling "really new" could actually èxternally be "well, I have to start sòmewhere with a pre*-proto-language to have something to run diachronics on").

                      What's the realisation of stress? If unstressed syllables are weakerly articulated in whatever way, then weaker syllables could start to lose some of the complexity of their syllable structure (especially monophthongisation, but also cluster simplification &c), which could help set up a more canonical affix shape down the line, and as a bonus potentially yield fun morphophonology.

                      >any sequences of two identical vowels
                      >are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
                      >'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
                      >/dorayra/ 'boy', from /dora /'man' + /ira/).

                      To ask after pedantic completeness: what happens when two different nonhigh vowels meet? When an /i/ and a /u/ meet, which one glides?

                      >Word order is VSO when there's no
                      >overt complementiser, and SOV when there is (so /fikol le/ 'the man has
                      >gone', but /le/ /fikolti/ 'the fact that the man has gone').

                      The /-ti/ is the overt complementiser in there?

                      >Verbs don't care about person, number, or tense, but there are 5 or 7
                      >aspect markers (perfective/stative (null-marked), progressive, perfect,
                      >expective, intentive(?), and hortative and imperative if you count them
                      >- they're mutually exclusive with aspect).

                      Does whether the null-marked aspect is perfective or stative depend on like Aktionsart or some kind of inherent aspect of the verb, or is there something weirder going on here? And then, if there are verb classes, are there any interesting semantic properties of the marked aspects? e.g. often (afaik) inherently stative verbs automatically become inceptives or whatnot when used in a (marked) perfective.

                      >The only obligatory marking on nouns is case, but there's a number of
                      >other potential affixes. Number is especially complex -

                      Does the below imply that plain old number marking of the singular vs. plural sort, lacking specificity in the quantity and lacking a larger quantity of which it's a part, is not found?

                      >specific
                      >quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                      >people', it's base-8 so that breaks down as /le-mo-fye-thon/
                      >'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                      >'more than half (of a group)', 'less than half (of a group)',

                      Really so specifically 'more / less than half', as opposed to 'relatively many (of a group)', 'relatively few (of a group)'? That precision seems unlikely to me.

                      >'part (of
                      >a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
                      >'none', and further by 'the next' or 'the previous' (allowing for very
                      >long sequences such as /lemofyethondawfag /'none of the last twenty
                      >people').

                      Very long single words arising from numbers seem to me to be one of those things which conlangers are prone to do but natlangs aren't. I don't know any direct data. But one natlang that hàs done this for two digit numbers is Hindi, and the result is that in the modern form all kinds of sandhi goes on and the system is kinda tricky to learn.
                      https://en.wikibooks.org/w/index.php?title=Hindi/Numbers
                      But not many languages have a Hindi-type system. And effects of analogy notwithstanding, my feeling is that this probably means that systems of two-digit numbers mostly don't behave like single words cross-linguistically.

                      Information density is probably not a very good way to get at questions of synchronic wordhood, but it is at least true that grammaticalisation, including univerbation, is characteristic of regions of relatively lower information density, whereas spelling out the digits of a long number in which precision is important intrinsically has relatively high information density. Maybe a more relevant fact is the way people reading telephone numbers and the like tend to give them structureful and nonhomogeneous prosody.

                      >there's two kinds of
                      >genitives, possessive and categorical (for things like 'men of that
                      >village', 'the strength of an ox', 'a sword of bronze' and so on), both
                      >of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
                      >-/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
                      >heard as 'the cat is the man's').

                      This is funny. Leaving aside the use of the label "case", what we have here are two operations which build verbs on nouns.
                      Now, ttbomk, it is a strong tendency that derivational or compounding operations tend to impose indefinite / nonspecific / nonreferential interpretations on their bases. For example, take Mithun's hierarchy of types of noun incorporation (e.g. <www.stanford.edu/~tylers/notes/morphosyntax/Mithun_1984_notes.pdf‎>): the more straightforward types I (and II?) tend to name "unitary activities", like 'birdwatching'; "as you have a unitary activity, the N loses its salience (Mithun 1984: 849)"; the way the other types III and IV develop on this is by running with this nonsalience.
                      On the other hand, possessives at least are most prototypically used with completely definite referential nouns, like SAP pronouns and proper names and stuff; this is totally the opposite of what derivational etc. operations tend to do. Funny. (For the categorical, it doesn't seem so bad to me.)

                      Of course, if you were to have these verbs, their relativised uses for usual noun-modifying-a-noun relations is perfectly natural.

                      Ah, hm, maybe this isn't a problem if these "cases" are regarded as _clitic_ verbs, which just can't occur phonologically free and have to lean on a noun to their left for whatever reason. The copular /-si/ might then permit the same analysis. I suppose that if you don't go with this cliticky kind of thing, the same nonspecificity motivations might motivate a language like this to keep /-si/ for membership in a class ('Gilbert is a farmer') and use a different mechanism for identity ('the morning star is the evening star', 'the murderer is the butler').

                      >Locative cases are the following: inessive and exessive (both used for
                      >general locatives, inessive for being within the boundaries of a place,
                      >exessive for being near but outside the boundaries of a place or
                      >object), superessive and subessive, proessive ('in front of') and
                      >postessive ('behind'), comitative, allative (also used as a dative) and
                      >ablative (also used as the agent of causatives and volitives), illative
                      >and ellative, superlative ('going over') and sublative ('going under'),
                      >circumlative/circumessive, and adspective ('facing') and abspective
                      >('facing away from').

                      Hm, no "general" locative then. That isn't itself weird, but what would be weird is if the system were actually as tidy as presented here; for whatever reason, systems of oblique cases and adpositions seem to be very fertile grounds for subregularities and not-very-productive metaphors and other slightly odd usages to become collocatively fixed.

                      What happens in your language with metaphorical uses of the locative cases -- how do the speakers apply these distinctions of whether things are inessively or exessively etc. "in" a condition or a situation, or a time, or a language, or so forth?

                      >Non-locative cases are benefactive (also used for the experiencer with a
                      >number of perception verbs - 'see' for example has a BEN subject when
                      >you would expect ERG, and giving it an ERG subject changes the meaning
                      >to 'look at'),

                      Good.

                      >instrumental, causative, and comparative.

                      Comparative for standards of comparison, or something wackier? And what does causative case do?

                      >Copular constructions are formed by affixing the copular verbaliser
                      >/-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
                      >somewhat idiosyncratic way to express motion - while it's perfectly
                      >grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
                      >/literally 'I am going to the river'), it's much more native-sounding to
                      >say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').

                      I wonder what the pragmatic effects of using the 'go' verb are, then. (Hm, maybe it foregrounds manner, and the verb you've glossed 'go' is instead more like 'walk'? https://en.wikipedia.org/wiki/Verb_framing)

                      >There are also two generic verbs, meaning something like 'to do (it)'
                      >and 'to go (there)'. They can also be used as nominalisers (somewhat
                      >ironically :P), meaning 'method' and 'process', respectively.

                      Neat, though I'm not entìrely sure what the difference between 'method' and 'process' to you is. Zero-derivation of verbs to nouns isn't a thing that happens in general, then, I take it?

                      >I've got a few ideas on where to go with it from here, but
                      >if anyone has any ideas I'd be happy to hear them.

                      Well, what are your ideas? Anyway, I find the brainstorming easier if you give us a bit of text to chew on, so we can see what kind of structures are asking for reduction or reinterpretation of some sort. (The interlinearised sentences in your next message are a good start! I'll have to read them later than right now, though, I'm already up too long.)

                      Alex, row row rowing
                    • Aodhán Aannestad
                      ... Yeah, that was my reasoning behind not bothering yet. It ll have one eventually. ... That s about where it stands at the moment :P ... I d only really
                      Message 10 of 11 , Jun 30, 2013
                        On 6/30/2013 12:57 AM, Alex Fink wrote:
                        > On Sun, 23 Jun 2013 23:28:43 -0500, Aodhán Aannestad <tolkien_freak@...> wrote:
                        >
                        >> Hello! I'm a longtime lurker, finally deciding to get involved a little.
                        >> Some of you may have met me at the last LCC, I'm with UT Austin's
                        >> conlang club and I drove some people around.
                        > 'Twas good to meet / bum rides off you there.
                        >
                        >> Anyway, on to the language. It doesn't have a name yet, sadly - I'm
                        >> using it for a protolang for a large project, and since said large
                        >> project has yet to really begin, it's not really in a state where I can
                        >> really name anything. (I suppose I could come up with an endonym for it
                        >> - it'd be something like 'Lesuy' (/lesuj/, /person-speak/) or something,
                        >> though that doesn't sound that great to my ears.)
                        > Eh, I'm in no rush to give my projects endonyms; I give them codenames or descriptive English names until then. And anyway, endonyms are often non-transparently formed, especially in areas with lots of cross-cultural interaction, in deriving from place names or tribal names or calqued or borrowed exonyms or whatnot; so if you don't like /lesuj/ it's hardly forced on you.
                        Yeah, that was my reasoning behind not bothering yet. It'll have one
                        eventually.

                        >> Affixes have no
                        >> shape target - anything from V to a maximal syllable is in theory
                        >> permitted.
                        > This is kinda odd, unless the affixing is all really new. (I suppose that what I'm calling "really new" could actually èxternally be "well, I have to start sòmewhere with a pre*-proto-language to have something to run diachronics on").
                        That's about where it stands at the moment :P

                        >
                        > What's the realisation of stress? If unstressed syllables are weakerly articulated in whatever way, then weaker syllables could start to lose some of the complexity of their syllable structure (especially monophthongisation, but also cluster simplification &c), which could help set up a more canonical affix shape down the line, and as a bonus potentially yield fun morphophonology.
                        I'd only really thought about unstressed syllables having reduced
                        vowels, but I very much like this idea - somehow reducing unstressed
                        syllables (or at least affixes) down to simply CV or CVC max. Stress is
                        right-oriented, though, which could serve to create some weird
                        irregularities.

                        I'm honestly a bit worried about having extremely unpredictable
                        morphophonology, but we'll see what happens. Hopefully I can figure out
                        some nice ways to level things out.

                        >> any sequences of two identical vowels
                        >> are shrunk into one (e.g. /emnira/, 'girl', from /emni/ 'woman' + /ira/
                        >> 'child'). /i/ and /u/ become glides when adjacent to other vowels (e.g.
                        >> /dorayra/ 'boy', from /dora /'man' + /ira/).
                        > To ask after pedantic completeness: what happens when two different nonhigh vowels meet? When an /i/ and a /u/ meet, which one glides?
                        Onglides are preferred over offglides, so whichever one's first (/ui/ >
                        [wi], /iu/ > [ju])

                        >
                        >> Word order is VSO when there's no
                        >> overt complementiser, and SOV when there is (so /fikol le/ 'the man has
                        >> gone', but /le/ /fikolti/ 'the fact that the man has gone').
                        > The /-ti/ is the overt complementiser in there?
                        Yeah.

                        >
                        >> Verbs don't care about person, number, or tense, but there are 5 or 7
                        >> aspect markers (perfective/stative (null-marked), progressive, perfect,
                        >> expective, intentive(?), and hortative and imperative if you count them
                        >> - they're mutually exclusive with aspect).
                        > Does whether the null-marked aspect is perfective or stative depend on like Aktionsart or some kind of inherent aspect of the verb, or is there something weirder going on here? And then, if there are verb classes, are there any interesting semantic properties of the marked aspects? e.g. often (afaik) inherently stative verbs automatically become inceptives or whatnot when used in a (marked) perfective.
                        Verbs are inherently either static or dynamic, and aspects tend to
                        change what they imply with each class - so progressive with dynamics is
                        progressive (i.e. ongoing action), but with statics it means that the
                        state is temporary; perfect with dynamics is perfect (i.e. completed
                        action), but with statics it's basically just past tense (i.e. the state
                        is no longer the case), etc. Null-marked dynamics are
                        perfective/inceptive, but null-marked statics are stative (though either
                        can be gnomic, and null-marked dynamics can be habitual also).

                        >
                        >> The only obligatory marking on nouns is case, but there's a number of
                        >> other potential affixes. Number is especially complex -
                        > Does the below imply that plain old number marking of the singular vs. plural sort, lacking specificity in the quantity and lacking a larger quantity of which it's a part, is not found?
                        Yeah, at least as things stand at the moment - if you don't know or
                        don't care how many, you just leave it unmarked. Multiples of 4 and 8
                        are also often used for estimates (c.f. English 'there's a hundred
                        people here' when there's really 118).

                        >
                        >> specific
                        >> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                        >> people', it's base-8 so that breaks down as /le-mo-fye-thon/
                        >> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                        >> 'more than half (of a group)', 'less than half (of a group)',
                        > Really so specifically 'more / less than half', as opposed to 'relatively many (of a group)', 'relatively few (of a group)'? That precision seems unlikely to me.
                        Yeah, it's a bit more fuzzy than that - the idea is 'most' vs. 'a few'.

                        >
                        >> 'part (of
                        >> a unit)', and 'all (of a unit)'. These can be augmented by 'all' or
                        >> 'none', and further by 'the next' or 'the previous' (allowing for very
                        >> long sequences such as /lemofyethondawfag /'none of the last twenty
                        >> people').
                        > Very long single words arising from numbers seem to me to be one of those things which conlangers are prone to do but natlangs aren't. I don't know any direct data. But one natlang that hàs done this for two digit numbers is Hindi, and the result is that in the modern form all kinds of sandhi goes on and the system is kinda tricky to learn.
                        > https://en.wikibooks.org/w/index.php?title=Hindi/Numbers
                        > But not many languages have a Hindi-type system. And effects of analogy notwithstanding, my feeling is that this probably means that systems of two-digit numbers mostly don't behave like single words cross-linguistically.
                        >
                        > Information density is probably not a very good way to get at questions of synchronic wordhood, but it is at least true that grammaticalisation, including univerbation, is characteristic of regions of relatively lower information density, whereas spelling out the digits of a long number in which precision is important intrinsically has relatively high information density. Maybe a more relevant fact is the way people reading telephone numbers and the like tend to give them structureful and nonhomogeneous prosody.
                        So hm - would a numerical classifiers system like
                        Japanese/Chinese/Korean work better for this language, do you think? I
                        know for Old Japanese's version of the Japanese native counting system
                        you had to jump through hoops to specify values between tens (e.g.
                        mis@ka 'thirty days', mis@ka (a)mari k@k@n@ka 'thirty-nine days', lit.
                        'thirty days remainder nine days') - would you expect something with a
                        similar structure to occur here?

                        >
                        >> there's two kinds of
                        >> genitives, possessive and categorical (for things like 'men of that
                        >> village', 'the strength of an ox', 'a sword of bronze' and so on), both
                        >> of which form verbs (so//'the man's cat' has to be /lenase nyawa/ with
                        >> -/se/, not */lena nyawa/ - /lena nyawa /is grammatical, but it would be
                        >> heard as 'the cat is the man's').
                        > This is funny. Leaving aside the use of the label "case", what we have here are two operations which build verbs on nouns.
                        > Now, ttbomk, it is a strong tendency that derivational or compounding operations tend to impose indefinite / nonspecific / nonreferential interpretations on their bases. For example, take Mithun's hierarchy of types of noun incorporation (e.g. <www.stanford.edu/~tylers/notes/morphosyntax/Mithun_1984_notes.pdf‎>): the more straightforward types I (and II?) tend to name "unitary activities", like 'birdwatching'; "as you have a unitary activity, the N loses its salience (Mithun 1984: 849)"; the way the other types III and IV develop on this is by running with this nonsalience.
                        > On the other hand, possessives at least are most prototypically used with completely definite referential nouns, like SAP pronouns and proper names and stuff; this is totally the opposite of what derivational etc. operations tend to do. Funny. (For the categorical, it doesn't seem so bad to me.)
                        >
                        > Of course, if you were to have these verbs, their relativised uses for usual noun-modifying-a-noun relations is perfectly natural.
                        >
                        > Ah, hm, maybe this isn't a problem if these "cases" are regarded as _clitic_ verbs, which just can't occur phonologically free and have to lean on a noun to their left for whatever reason. The copular /-si/ might then permit the same analysis. I suppose that if you don't go with this cliticky kind of thing, the same nonspecificity motivations might motivate a language like this to keep /-si/ for membership in a class ('Gilbert is a farmer') and use a different mechanism for identity ('the morning star is the evening star', 'the murderer is the butler').
                        The clitic analysis seems to make sense (and it seems to work
                        structurally, too, at least if I'm remembering right the syntax trees I
                        drew ages ago). That is a direction for diachronic change, though -
                        innovating a separate copular word for use with identity and merging the
                        clitic copula with the categorical (which is already phonetically
                        similar enough that that could happen through sound change alone).

                        >
                        >
                        >> Locative cases are the following: inessive and exessive (both used for
                        >> general locatives, inessive for being within the boundaries of a place,
                        >> exessive for being near but outside the boundaries of a place or
                        >> object), superessive and subessive, proessive ('in front of') and
                        >> postessive ('behind'), comitative, allative (also used as a dative) and
                        >> ablative (also used as the agent of causatives and volitives), illative
                        >> and ellative, superlative ('going over') and sublative ('going under'),
                        >> circumlative/circumessive, and adspective ('facing') and abspective
                        >> ('facing away from').
                        > Hm, no "general" locative then. That isn't itself weird, but what would be weird is if the system were actually as tidy as presented here; for whatever reason, systems of oblique cases and adpositions seem to be very fertile grounds for subregularities and not-very-productive metaphors and other slightly odd usages to become collocatively fixed.
                        >
                        > What happens in your language with metaphorical uses of the locative cases -- how do the speakers apply these distinctions of whether things are inessively or exessively etc. "in" a condition or a situation, or a time, or a language, or so forth?
                        I'm sure there's a lot of depth here I haven't yet plumbed (especially
                        considering this ís an idealised pre-proto-language) - I probably
                        haven't done enough testing either to come up with good situations for
                        things like this. At the moment, I'm assuming that general descriptions
                        are almost always correct (e.g. something during an event would be
                        inessive, because it's within the boundaries of the event; but
                        something around the time of an event would be exessive, because it's
                        not, etc.).

                        >> Non-locative cases are benefactive (also used for the experiencer with a
                        >> number of perception verbs - 'see' for example has a BEN subject when
                        >> you would expect ERG, and giving it an ERG subject changes the meaning
                        >> to 'look at'),
                        > Good.
                        >
                        >> instrumental, causative, and comparative.
                        > Comparative for standards of comparison, or something wackier? And what does causative case do?
                        I'm still working through what the comparative should look like, and I'm
                        leaning towards a combination of standard-of-comparison and what I
                        understand 'essive' to mean (though I'm probably wrong - my
                        understanding is that essive is 'as' in sentences like 'I'm playing [a
                        strategy game] as France' or whatever'. This does require separate
                        comparative morphology on the verb (since when the verb's marked
                        comparative, the case would be interpreted as English 'than'; but when
                        the verb's not marked, it'd be interpreted as 'as' or maybe 'like'), and
                        it might be easier to do it where comparison isn't marked on the verb
                        and you just have to use the case marker (something like how Japanese
                        yori is used).

                        Causative is for reasons (something like 'because of'), but it may be
                        unnecessary - it looks just the same as the bare verb nominaliser
                        (/adverbialiser) that also marks cause. Either that or the bare verb
                        nominaliser should be dropped and verbs should need a normal nominaliser
                        for this (especially since purpose is NOM+BEN - cause might ought to be
                        NOM+CAUS instead of just CAUS as a nominaliser).

                        >
                        >> Copular constructions are formed by affixing the copular verbaliser
                        >> /-si/ to nouns - /dorasi le/ 'the person is a man'. This allows for a
                        >> somewhat idiosyncratic way to express motion - while it's perfectly
                        >> grammatical to say /fyokh ne sakhtasoy /(/go-PROG 1-ABS river-ALL,
                        >> /literally 'I am going to the river'), it's much more native-sounding to
                        >> say /sakhtasoysi ne/ (/river-ALL-COP 1-ABS, /literally 'I am to the river').
                        > I wonder what the pragmatic effects of using the 'go' verb are, then. (Hm, maybe it foregrounds manner, and the verb you've glossed 'go' is instead more like 'walk'? https://en.wikipedia.org/wiki/Verb_framing)
                        The verb is especially useful when leaving destination and direction
                        unspecified, though it pretty much ends up meaning 'leave' in cases like
                        that. More often than not, though, case+COP constructions refer to
                        someone or something being still in transit, and so the verb is required
                        when referring to a whole motion action.

                        A good example of the contrast might be:

                        fi-sif le-Ø sakhta-soy alla-ki
                        go-WEAKEXPECT person-ABS river-ALL day-FARDEM
                        'I bet he'll go to the river that day' - implying that he will set out
                        and arrive on the same day (so the river must be within a day's journey
                        of his starting location)
                        vs.
                        sahkta-soy-si-sif le-Ø alla-ki
                        river-ALL-COP-WEAKEXPECT person-ABS day-FARDEM
                        'I bet he'll be on his way to the river that day' - implying that his
                        journey is at least more than one day long, probably ending on a
                        different day than the one in question (though possibly still starting
                        on the same day); and pragmatically implying that some other action that
                        is occuring on that day besides simply journeying is the overall topic
                        of conversation - either he's doing something on the way, or someone
                        else is establishing a relative chronology of disparate events in their
                        head.

                        I imagine that if the people who spoke this had cellphones, they'd use
                        the case+COP construction when talking about whatever transit the phone
                        call is being made during.


                        This language is very much more satellite-framing most of the time,
                        since you can productively use case endings on verb roots to make
                        compound verbs (so you can have go+ABL = 'leave', go+SUPERESS = 'ride', etc)

                        >
                        >> There are also two generic verbs, meaning something like 'to do (it)'
                        >> and 'to go (there)'. They can also be used as nominalisers (somewhat
                        >> ironically :P), meaning 'method' and 'process', respectively.
                        > Neat, though I'm not entìrely sure what the difference between 'method' and 'process' to you is. Zero-derivation of verbs to nouns isn't a thing that happens in general, then, I take it?
                        Those probably aren't the best terms, and it is pretty fuzzy. 'Do' is
                        more used with the mental conception of an action (e.g. when there's
                        multiple ways to do an action that all lead to largely the same result,
                        'do' is used to refer to those ways), and 'go' is more used with the
                        actual realisation of an action (e.g. it refers to not how something is
                        done, but the fact that it is being done).

                        Zero-derivation is pretty unusual. One of the few cases where it can is
                        when applying case endings to verbs (as a lexical derivation process),
                        in which case the output can be either a noun or a verb.

                        >> I've got a few ideas on where to go with it from here, but
                        >> if anyone has any ideas I'd be happy to hear them.
                        > Well, what are your ideas? Anyway, I find the brainstorming easier if you give us a bit of text to chew on, so we can see what kind of structures are asking for reduction or reinterpretation of some sort. (The interlinearised sentences in your next message are a good start! I'll have to read them later than right now, though, I'm already up too long.)
                        One of the things I'm thinking about is a process involving
                        demonstrative affixes, where something like this happens:
                        near demonstrative > topic marker > fossilised pronoun marker (e.g.
                        person+NEARDEM becomes 3.ANIM, and the affix ceases to be productive)
                        far demonstrative > (contrastive) focus marker > ??

                        And of course there's always the basic sound change stuff (especially
                        things like final consonant loss and cluster simplification, since one
                        of the daugher languages is planned to be straight-up CV). I may have
                        one branch delete unstressed high vowels and have clicks pop out of any
                        resulting prevelar stop(/nasal?) + velar stop clusters (e.g. [pi'kat] >
                        [pkat] > [O\at]).

                        >
                        > Alex, row row rowing
                        >
                        Thanks for the input! You've given me a lot to think about.
                      • Padraic Brown
                        ... Well, it s axiomatic that, at least for human languages, they are always in a state of flux. You ve already said that what you re working on is a sort of
                        Message 11 of 11 , Jun 30, 2013
                          > From: Aodhán Aannestad <tolkien_freak@...>

                          >
                          >> What's the realisation of stress?  If unstressed syllables are weakerly articulated in whatever way, then weaker
                          >> syllables could start to lose some of  the complexity of their syllable structure (especially monophthongisation, but
                          >> also cluster simplification &c), which could help set up a more canonical affix shape down the line, and as a
                          >> bonus potentially yield fun morphophonology.
                          >
                          > I'd only really thought about unstressed syllables having reduced
                          > vowels, but I very much like this idea - somehow reducing unstressed
                          > syllables (or at least affixes) down to simply CV or CVC max. Stress is
                          > right-oriented, though, which could serve to create some weird
                          > irregularities.
                          >
                          > I'm honestly a bit worried about having extremely unpredictable
                          > morphophonology, but we'll see what happens. Hopefully I can figure out
                          > some nice ways to level things out.

                          Well, it's axiomatic that, at least for human languages, they are always in a state of flux.
                          You've already said that what you're working on is a sort of proto-language; just look
                          on what you're doing now as observing the language at a particularly bumpy period in
                          its evolution. So, sit up, strap yourself in and enjoy all the twists and turns!

                          >>> The only obligatory marking on nouns is case, but there's a number of
                          >>> other potential affixes. Number is especially complex - specific
                          >>> quantities are marked directly on the noun (so/lemofyethon/ 'twenty
                          >>> people', it's base-8 so that breaks down as /le-mo-fye-thon/
                          >>> 'person-8-2-4' for (2*8)+4 people), and there are also suffixes for
                          >>> 'more than half (of a group)', 'less than half (of a group)',
                          >
                          >> Really so specifically 'more / less than half', as opposed to 'relatively many (of a group)',
                          >> 'relatively few (of a group)'?  That precision seems unlikely to me.
                          >
                          > Yeah, it's a bit more fuzzy than that - the idea is 'most' vs. 'a few'.

                          I like the idea of associating specific numbers with nouns. I would ask are these numbers
                          supposed to represent a real / assumed / guessed at quantity, or are we looking at
                          certain numbers that, for whatever social reason, get associated with nouns? I'm put in
                          mind of the Old Irish convention of the "thrice fifty" maidens or lads that always seem to
                          accompany a principal character: "The Dagda meanwhile brought his son to Midir's house
                          in Brí Léith in
                          Tethba, to be fostered. There Aengus was reared for the space of nine
                          years.
                          Midir had a great playing-field in Brí Léith. Thrice fifty lads
                          of the young nobles of Ireland
                          were there and thrice fifty maidens of
                          the land of Ireland..."

                          I have no doubt at all but that "thrice fifty maidens" simply means "some largeish quantity
                          of youngish girls"; and perhaps the "space of nine years" likewise simply means "a few
                          years". In other words, the numbers are poetic conventions rather than actual quantities.
                          I've done this sort of thing a time or two in some conlangs and wondered if a similar thing
                          weren't going on here.

                          I also réally like the idea for specific suffixes meaning "more than half / less than half" of
                          a group. I could easily see this sort of thing arising among a people who probably live
                          in smallish groups and are very aware of the presence or absence or location of the
                          others in the group. I could also see another set of "relatively many  / relatively few"
                          and "many / few" suffixes for when the groups become out of hand. Perhaps, for
                          example, during festivals when many people get together from many different lands
                          around and no one can know who all is present or which side of half the group is
                          really there. Dunno how líkely the scheme is, but I do like it!

                          [resto snippato]

                          Padraic
                        Your message has been successfully submitted and would be delivered to recipients shortly.