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

Celtic-style mutations

Expand Messages
  • Daniel Bowman
    Greetings, and I hope all of your holidays were happy! I was wondering if someone could explain to me exactly what Celtic-style mutations are (I ve been
    Message 1 of 25 , Jan 1, 2010
    • 0 Attachment
      Greetings, and I hope all of your holidays were happy!

      I was wondering if someone could explain to me exactly what "Celtic-style
      mutations" are (I've been following the "A few things I find underused in
      conlangs" thread with interest).

      As for the "A few things..." poll:

      1. Angosey does not have tone. Considered it, but realized that it would
      require a substantial revision of how I understand the language.

      2. There are grammatical prefixes in Angosey: verbs agree with ergative noun
      class (if present) by means of a prefix. Ditto with postpositions and
      postpositional subjects. However, suffixes outnumber prefixes.

      3. Pending description of Celtic style mutations! However, I doubt Angosey has
      them, unless I independently invented them.

      3. No reduplication. I learned Kiswahili while studying abroad in Africa. While I
      liked most of the language, the reduplication aspects were odd and threw me
      off. I can't imagine introducing that into Angosey.

      4. Not sure, but I think Angosey's about average in terms of innate features.
    • Ph.D.
      ... Example: http://en.wikipedia.org/wiki/Irish_initial_mutations --Ph. D.
      Message 2 of 25 , Jan 2, 2010
      • 0 Attachment
        Daniel Bowman:
        >
        > I was wondering if someone could explain to me exactly what "Celtic-style
        > mutations" are (I've been following the "A few things I find underused in
        > conlangs" thread with interest).

        Example:
        http://en.wikipedia.org/wiki/Irish_initial_mutations

        --Ph. D.
      • Jörg Rhiemeier
        Hallo! ... Indeed. There is no shred of evidence for them in Continental Celtic, and the Continental Celtic conlang I have in the making doesn t have them
        Message 3 of 25 , Jan 2, 2010
        • 0 Attachment
          Hallo!

          On Sat, 2 Jan 2010 11:03:07 +0000, R A Brown wrote:

          > > I meant morphophonemic alternations in general, though after reading
          > > through
          > > the replies to my original post, I now agree that they're overused. The
          > > only
          > > conlangs I had seen that had mutations before were Brithenig and
          > > Breathanach
          > > - both of which have an explicit purpose to imitate Celtic.
          >
          > They are also a feature of Sindarin. But whereas Brithenig
          > and Breathanach mutations are very much imitations of those
          > of Welsh and Irish respectively, Sindarin shows some
          > original features.
          >
          > But they an inevitable part of those many conlangs that want
          > to have a 'Celtic flavor' (tho AFAIK it was never a feature
          > of Continental Celtic).

          Indeed. There is no shred of evidence for them in Continental Celtic,
          and the Continental Celtic conlang I have in the making doesn't have
          them (nor does it have VSO word order nor any of the other Insular
          Celtic peculiarities). The Insular Celtic initial mutations are most
          likely an areal or substratal phenomenon that developed only in the
          British Isles.

          This substratum could have been something like Albic. Old Albic has
          subphonemic initial mutations (each consonant has a fortis and a lenis
          allophone, with the latter occuring after vowels and liquids, with this
          rule sometimes operating across word boundaries); in at least some of
          the daughter languages, these subphonemic initial mutations become
          phonemic ones which are later grammaticalized, in similar ways as in
          the Insular Celtic languages.

          BTW: The notion of an Elvish substratum in Insular Celtic actually was
          something Tolkien played with, too (in the unfinished time travel story
          _The Lost Road_).

          --
          ... brought to you by the Weeping Elf
          http://www.joerg-rhiemeier.de/Conlang/index.html
        • Gary Shannon
          ... I wonder, is it possible for a child to learn Irish Celtic without a degree in linguistics? ;-) I, too, was interested in what was meant by Celtic-style
          Message 4 of 25 , Jan 2, 2010
          • 0 Attachment
            On Sat, Jan 2, 2010 at 2:17 AM, Ph.D. <phil@...> wrote:
            > Daniel Bowman:
            >>
            >> I was wondering if someone could explain to me exactly what "Celtic-style
            >> mutations" are (I've been following the "A few things I find underused in
            >> conlangs" thread with interest).
            >
            > Example:
            > http://en.wikipedia.org/wiki/Irish_initial_mutations

            I wonder, is it possible for a child to learn Irish Celtic without a
            degree in linguistics? ;-)

            I, too, was interested in what was meant by "Celtic-style mutations",
            so I read the Wikipedia article. Is there a similar article written in
            English? I found the jargon so completely dense and impenetrable that
            after reading the article twice I still don't have a clue as to what
            Celtic-style mutations are all about.

            Is there a place where people without linguistics degrees can learn
            about these interesting topics? Surely they can be explained without
            the need for such obscure jargon.

            --gary
          • Kenner Gordon
            When I originally said Celtic-style mutations , I meant mutations in general. The Celtic-style was there for clarity - e.g., it could be interpreted as some
            Message 5 of 25 , Jan 2, 2010
            • 0 Attachment
              When I originally said "Celtic-style mutations", I meant mutations in
              general. The "Celtic-style" was there for clarity - e.g., it could be
              interpreted as some other sense of the word.

              On Sat, Jan 2, 2010 at 9:04 AM, Gary Shannon <fiziwig@...> wrote:

              > On Sat, Jan 2, 2010 at 2:17 AM, Ph.D. <phil@...> wrote:
              > > Daniel Bowman:
              > >>
              > >> I was wondering if someone could explain to me exactly what
              > "Celtic-style
              > >> mutations" are (I've been following the "A few things I find underused
              > in
              > >> conlangs" thread with interest).
              > >
              > > Example:
              > > http://en.wikipedia.org/wiki/Irish_initial_mutations
              >
              > I wonder, is it possible for a child to learn Irish Celtic without a
              > degree in linguistics? ;-)
              >
              > I, too, was interested in what was meant by "Celtic-style mutations",
              > so I read the Wikipedia article. Is there a similar article written in
              > English? I found the jargon so completely dense and impenetrable that
              > after reading the article twice I still don't have a clue as to what
              > Celtic-style mutations are all about.
              >
              > Is there a place where people without linguistics degrees can learn
              > about these interesting topics? Surely they can be explained without
              > the need for such obscure jargon.
              >
              > --gary
              >



              --
              I don't really get the point of signatures.
            • Ben Fields
              I ve been reading Basic Irish and learning about celtic style mutations. In Irish it seems they mostly involve adding a consonant into the middle of a word
              Message 6 of 25 , Jan 2, 2010
              • 0 Attachment
                I've been reading Basic Irish and learning about celtic style mutations.

                In Irish it seems they mostly involve adding a consonant into the middle of
                a word instead of a suffix or prefix for a grammatical distinction or to
                affect pronunciation.

                An example would be the mutation called lenition, where you add an h after
                usually the first consonant beginning a word.

                Lenition can denote: Marking verbs for tense, marking negative verbs,
                signaling gender of nouns, agreement of adjectives and more.

                Being a student of English and Romance languages I found this way of
                grammatically altering words to be inspiring.


                On Sat, Jan 2, 2010 at 2:15 AM, Daniel Bowman <danny.c.bowman@...>wrote:

                > Greetings, and I hope all of your holidays were happy!
                >
                > I was wondering if someone could explain to me exactly what "Celtic-style
                > mutations" are (I've been following the "A few things I find underused in
                > conlangs" thread with interest).
                >
                > As for the "A few things..." poll:
                >
                > 1. Angosey does not have tone. Considered it, but realized that it would
                > require a substantial revision of how I understand the language.
                >
                > 2. There are grammatical prefixes in Angosey: verbs agree with ergative
                > noun
                > class (if present) by means of a prefix. Ditto with postpositions and
                > postpositional subjects. However, suffixes outnumber prefixes.
                >
                > 3. Pending description of Celtic style mutations! However, I doubt Angosey
                > has
                > them, unless I independently invented them.
                >
                > 3. No reduplication. I learned Kiswahili while studying abroad in Africa.
                > While I
                > liked most of the language, the reduplication aspects were odd and threw me
                > off. I can't imagine introducing that into Angosey.
                >
                > 4. Not sure, but I think Angosey's about average in terms of innate
                > features.
                >
              • R A Brown
                ... For more general use, cf. http://en.wikipedia.org/wiki/Consonant_mutation -- Ray ================================== http://www.carolandray.plus.com
                Message 7 of 25 , Jan 2, 2010
                • 0 Attachment
                  Kenner Gordon wrote:
                  > When I originally said "Celtic-style mutations", I meant mutations in
                  > general.

                  For more general use, cf.
                  http://en.wikipedia.org/wiki/Consonant_mutation

                  --
                  Ray
                  ==================================
                  http://www.carolandray.plus.com
                  ==================================
                  Nid rhy hen neb i ddysgu.
                  There's none too old to learn.
                  [WELSH PROVERB]
                • Mark J. Reed
                  You don t add a consonant. The is a written form of lenition, but it represents a thorough change of the consonant sound. For instance, is /m/, but
                  Message 8 of 25 , Jan 2, 2010
                  • 0 Attachment
                    You don't add a consonant. The <h> is a written form of lenition,
                    but it represents a thorough change of the consonant sound. For
                    instance, <m> is /m/, but <mh> is /v/ (or /B/, I forget which), not
                    /mh/.

                    On Saturday, January 2, 2010, Ben Fields <fresiatap@...> wrote:
                    > I've been reading Basic Irish and learning about celtic style mutations.
                    >
                    > In Irish it seems they mostly involve adding a consonant into the middle of
                    > a word instead of a suffix or prefix for a grammatical distinction or to
                    > affect pronunciation.
                    >
                    > An example would be the mutation called lenition, where you add an h after
                    > usually the first consonant beginning a word.
                    >
                    > Lenition can denote: Marking verbs for tense, marking negative verbs,
                    > signaling gender of nouns, agreement of adjectives and more.
                    >
                    > Being a student of English and Romance languages I found this way of
                    > grammatically altering words to be inspiring.
                    >
                    >
                    > On Sat, Jan 2, 2010 at 2:15 AM, Daniel Bowman <danny.c.bowman@...>wrote:
                    >
                    >> Greetings, and I hope all of your holidays were happy!
                    >>
                    >> I was wondering if someone could explain to me exactly what "Celtic-style
                    >> mutations" are (I've been following the "A few things I find underused in
                    >> conlangs" thread with interest).
                    >>
                    >> As for the "A few things..." poll:
                    >>
                    >> 1.  Angosey does not have tone.  Considered it, but realized that it would
                    >> require a substantial revision of how I understand the language.
                    >>
                    >> 2.  There are grammatical prefixes in Angosey:  verbs agree with ergative
                    >> noun
                    >> class (if present) by means of a prefix.  Ditto with postpositions and
                    >> postpositional subjects.  However, suffixes outnumber prefixes.
                    >>
                    >> 3. Pending description of Celtic style mutations!  However, I doubt Angosey
                    >> has
                    >> them, unless I independently invented them.
                    >>
                    >> 3.  No reduplication.  I learned Kiswahili while studying abroad in Africa.
                    >>  While I
                    >> liked most of the language, the reduplication aspects were odd and threw me
                    >> off.  I can't imagine introducing that into Angosey.
                    >>
                    >> 4.  Not sure, but I think Angosey's about average in terms of innate
                    >> features.
                    >>
                    >

                    --
                    Mark J. Reed <markjreed@...>
                  • Ben Fields
                    Oh right! sorry about that!
                    Message 9 of 25 , Jan 2, 2010
                    • 0 Attachment
                      Oh right! sorry about that!

                      On Sat, Jan 2, 2010 at 12:55 PM, Mark J. Reed <markjreed@...> wrote:

                      > You don't add a consonant. The <h> is a written form of lenition,
                      > but it represents a thorough change of the consonant sound. For
                      > instance, <m> is /m/, but <mh> is /v/ (or /B/, I forget which), not
                      > /mh/.
                      >
                      > On Saturday, January 2, 2010, Ben Fields <fresiatap@...> wrote:
                      > > I've been reading Basic Irish and learning about celtic style mutations.
                      > >
                      > > In Irish it seems they mostly involve adding a consonant into the middle
                      > of
                      > > a word instead of a suffix or prefix for a grammatical distinction or to
                      > > affect pronunciation.
                      > >
                      > > An example would be the mutation called lenition, where you add an h
                      > after
                      > > usually the first consonant beginning a word.
                      > >
                      > > Lenition can denote: Marking verbs for tense, marking negative verbs,
                      > > signaling gender of nouns, agreement of adjectives and more.
                      > >
                      > > Being a student of English and Romance languages I found this way of
                      > > grammatically altering words to be inspiring.
                      > >
                      > >
                      > > On Sat, Jan 2, 2010 at 2:15 AM, Daniel Bowman <danny.c.bowman@...
                      > >wrote:
                      > >
                      > >> Greetings, and I hope all of your holidays were happy!
                      > >>
                      > >> I was wondering if someone could explain to me exactly what
                      > "Celtic-style
                      > >> mutations" are (I've been following the "A few things I find underused
                      > in
                      > >> conlangs" thread with interest).
                      > >>
                      > >> As for the "A few things..." poll:
                      > >>
                      > >> 1. Angosey does not have tone. Considered it, but realized that it
                      > would
                      > >> require a substantial revision of how I understand the language.
                      > >>
                      > >> 2. There are grammatical prefixes in Angosey: verbs agree with
                      > ergative
                      > >> noun
                      > >> class (if present) by means of a prefix. Ditto with postpositions and
                      > >> postpositional subjects. However, suffixes outnumber prefixes.
                      > >>
                      > >> 3. Pending description of Celtic style mutations! However, I doubt
                      > Angosey
                      > >> has
                      > >> them, unless I independently invented them.
                      > >>
                      > >> 3. No reduplication. I learned Kiswahili while studying abroad in
                      > Africa.
                      > >> While I
                      > >> liked most of the language, the reduplication aspects were odd and threw
                      > me
                      > >> off. I can't imagine introducing that into Angosey.
                      > >>
                      > >> 4. Not sure, but I think Angosey's about average in terms of innate
                      > >> features.
                      > >>
                      > >
                      >
                      > --
                      > Mark J. Reed <markjreed@...>
                      >
                    • Aidan Grey
                      Mark has it right. In addition, h isn t added usually the first consonant in lenition. If you re going by orthography, whenever lenition occurs, it ALWAYS
                      Message 10 of 25 , Jan 2, 2010
                      • 0 Attachment
                        Mark has it right. In addition, h isn't added "usually the first consonant" in lenition. If you're going by orthography, whenever lenition occurs, it ALWAYS adds the h immediately after the first consonant.

                        A quick summary of mutations, and how they arise:

                        Say feminine nouns end in -a, and in the genitive, in -as. Masculine nouns end in consonants, and in -o in the genitive. The same rules apply to the article "is".

                        Now saw that at some point, unvoiced stops between vowels become voiced, while voiced stops became fricativized (eclipsis).
                        Also, stops after s become fricativized (lenition).

                        Given the words:
                        man: par
                        woman: bana
                        fair: kel(a)

                        With a few other changes, like loss of final syllables, you get:

                        a fair man: par kel
                        the fair man: is par kel > i far kel (perhaps written phar)
                        of a fair man: paro kelo > paro gel > par gel
                        of the fair man: iso paro kelo > i bar gel (perhaps written bpar gkel)

                        a fair woman: bana kel > bana gel
                        the fair woman: isa bana kel > i vana gel
                        of a fair woman: banas kelas > bana hel
                        of the fair woman: isas banas kelas > i vana hel

                        So you end up with:

                        Masc Fem
                        sN - eclipsis
                        sG ecl lenition

                        sND len + - ecl + ecl
                        sGD ecl + ecl ecl + len

                        (sN = singular Nominative, sGD = singular Genitive definite)

                        And those mutations are the only way in which case and definiteness is marked.

                        With new words, say tel for lamp (masc), and the adjective pini/pinya for bright, you'd know that the singular genitive definite would have to be "i dtel bpini".

                        Aidan



                        ----- Original Message ----
                        > From: Mark J. Reed <markjreed@...>
                        >
                        > You don't add a consonant. The is a written form of lenition,
                        > but it represents a thorough change of the consonant sound. For
                        > instance, is /m/, but is /v/ (or /B/, I forget which), not
                        > /mh/.
                        >
                        > On Saturday, January 2, 2010, Ben Fields wrote:
                        > > I've been reading Basic Irish and learning about celtic style mutations.
                        > >
                        > > In Irish it seems they mostly involve adding a consonant into the middle of
                        > > a word instead of a suffix or prefix for a grammatical distinction or to
                        > > affect pronunciation.
                        > >
                        > > An example would be the mutation called lenition, where you add an h after
                        > > usually the first consonant beginning a word.
                        > >
                        > > Lenition can denote: Marking verbs for tense, marking negative verbs,
                        > > signaling gender of nouns, agreement of adjectives and more.
                        > >
                        > > Being a student of English and Romance languages I found this way of
                        > > grammatically altering words to be inspiring.
                        > >
                      • Keith Gaughan
                        ... Yes, though I m sure a lot of Irish schoolchildren would disagree! :-) Bit of a bugbear, but you mean Gaelic rather than Celtic there. Saying Irish
                        Message 11 of 25 , Jan 2, 2010
                        • 0 Attachment
                          On Sat, Jan 02, 2010 at 08:04:30AM -0800, Gary Shannon wrote:
                          > On Sat, Jan 2, 2010 at 2:17 AM, Ph.D. <phil@...> wrote:
                          > > Daniel Bowman:
                          > >>
                          > >> I was wondering if someone could explain to me exactly what "Celtic-style
                          > >> mutations" are (I've been following the "A few things I find underused in
                          > >> conlangs" thread with interest).
                          > >
                          > > Example:
                          > > http://en.wikipedia.org/wiki/Irish_initial_mutations
                          >
                          > I wonder, is it possible for a child to learn Irish Celtic without a
                          > degree in linguistics? ;-)

                          Yes, though I'm sure a lot of Irish schoolchildren would disagree! :-)
                          Bit of a bugbear, but you mean 'Gaelic' rather than 'Celtic' there.
                          Saying 'Irish Celtic' is akin to saying 'French Italic', or 'English
                          Germanic'.

                          Ignoring the introduction of the epithentic t- and h- that sometimes
                          occurs with words starting with vowels, you've got two kinds, lenition
                          (aka, 'softening' or 'aspiration'), and eclipsis, which is characterised
                          by voicing and nasalisation. I think one of the things that traps people
                          when it come to getting mutations is that Irish does its damnedest to
                          preserve the underlying word rather than changing the spelling (as Welsh
                          does). The name 'eclipsis' is misleading because it refers to the
                          orthographic process rather than the mutations themselves.

                          Lenition means that stops become fricatives with a couple of exceptions,
                          thus:

                          /p/ -> /f/
                          /b/ -> /B/ (at least where I'm from, but elsewhere it can end up as
                          [w] when /b/ is velar or [v] when palatal, it's the
                          Spanish 'v' sound)
                          /m/ -> /B/ (ditto)
                          /t/ -> /h/ (was [T] in the distant past, but wore down to [h])
                          /d/ -> /j/ (Would have been [D] in the past, but wore down to [j])
                          /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)
                          /S/ -> /h/ (That's a palatalised /s/)
                          /d'/ -> /G/ (Would have been /D'/ in the past, but ended up as [G])
                          /k/ -> /x/ (Velar 'k', the achlaut)
                          /c/ -> /C/ (Palatal 'k', the ichlaut)
                          /g/ -> /G/

                          /d'/ = palatalised /d/, which is generally realised as [Z] or [dZ], also
                          note that /t/ and /d/ are dental when velarised, but alveolar when
                          palatalised.

                          In school, you're taught palatalisation, and the basic grammatical idea
                          behind séimhiú, but the sounds and conditions under which it occurs are
                          learned by rote, and people recognise the sound by the 'h' following the
                          consonant, thus _bh_ represents the same pair of sounds regardless of
                          where it appears in a word. It's not considered much different from
                          recognising that 'c' and 'ch' and 't' and 'th' in English all have
                          different sounds.

                          With eclipsis (urú), similarly to séimhiú, people learn the
                          circumstances in which it occurs, and then learn the following table:

                          p -> bp
                          b -> mb
                          f -> bhf
                          t -> dt
                          d -> nd
                          c -> gc
                          g -> ng

                          And that in circumstances where eclipsis would happen and the word
                          starts with a vowel, you put 'n-' at the beginning, and that when you
                          see those patterns at the start of the word, you pronounce the first
                          consonant ('bh' counts as a single consonant).

                          The way it's taught it gradual - the first time the mutations are
                          introduced, the kids would probably only know the present tense and
                          would have picked up the sounds associated with séimhiú unconsciously
                          (as they can appear anywhere in a word), but when they learn about the
                          conditional mood and past tense, they'd be told they undergo séimhiú,
                          which as far as they're concerned is just putting a 'h' after the
                          initial consonant or overdot/buailte to change the sound, and would have
                          internalised the patterns subconsciously from when they had to go
                          through the chore of learning the noun declension table[1], and how to
                          recognise the gender and declension of a noun. It can at least be said
                          about the language that declension is fairly systematic, once you learn
                          to recognise the patterns and there's very few irregularities.

                          [1] http://en.wikipedia.org/wiki/Irish_declension

                          > I, too, was interested in what was meant by "Celtic-style mutations",
                          > so I read the Wikipedia article. Is there a similar article written in
                          > English? I found the jargon so completely dense and impenetrable that
                          > after reading the article twice I still don't have a clue as to what
                          > Celtic-style mutations are all about.
                          >
                          > Is there a place where people without linguistics degrees can learn
                          > about these interesting topics? Surely they can be explained without
                          > the need for such obscure jargon.

                          This covers the essentials of what happens rather than when:

                          http://en.wikipedia.org/wiki/Irish_language#Mutations

                          Cló Iar-Chonnachta used to publish a very nice and short Irish grammar,
                          which is what I used back when I was in secondary school, and it's what
                          made the language click for me. I can't find mention of it on their
                          site, but this might do:

                          http://www.cic.ie/product.asp?idproduct=349

                          If it's even a patch on the old grammar I had, it'll be excellent. Cló
                          Iar-Chonnachta are a very good publisher.

                          K.

                          --
                          Keith Gaughan - k@... - http://stereochro.me/ - CF9F6473
                          Intelligence has much less practical application than you'd think.
                          -- Scott Adams, from "Dilbert"
                        • Keith Gaughan
                          ... It s /B/ in the north west, /v/ or /w/ elsewhere. In some places /v/ is consistently used, others /w/, and in others still, it depends on palatalisation. I
                          Message 12 of 25 , Jan 2, 2010
                          • 0 Attachment
                            On Sat, Jan 02, 2010 at 12:55:56PM -0500, Mark J. Reed wrote:

                            > You don't add a consonant. The <h> is a written form of lenition,
                            > but it represents a thorough change of the consonant sound. For
                            > instance, <m> is /m/, but <mh> is /v/ (or /B/, I forget which), not
                            > /mh/.

                            It's /B/ in the north west, /v/ or /w/ elsewhere. In some places /v/ is
                            consistently used, others /w/, and in others still, it depends on
                            palatalisation. I use /B/ on the rare occasion I use the language.

                            --
                            Keith Gaughan - k@... - http://stereochro.me/ - CF9F6473
                            How come we play war and not peace?
                            Too few role models.
                            -- Calvin And Hobbes
                          • Alex Fink
                            ... These two seem backwards. ... Seriously? Why would something that s historically intervocalic lenition end up like that? Especially when all of /t t_j
                            Message 13 of 25 , Jan 2, 2010
                            • 0 Attachment
                              On Sat, 2 Jan 2010 18:37:27 +0000, Keith Gaughan <k@...> wrote:

                              > /d/ -> /j/ (Would have been [D] in the past, but wore down to [j])
                              > /d'/ -> /G/ (Would have been /D'/ in the past, but ended up as [G])
                              >/d'/ = palatalised /d/

                              These two seem backwards.

                              > /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)

                              Seriously? Why would something that's historically intervocalic lenition
                              end up like that? Especially when all of /t t_j s_j/ (to use more
                              orthogonal labels) go to /h/, it's quite strange that /s/ wouldn't.

                              Alex
                            • Kelvin Jackson
                              ... /s/ goes to /h/. /S/ is how is pronounced, unleniated, before a slender vowel ( or ).
                              Message 14 of 25 , Jan 2, 2010
                              • 0 Attachment
                                >> /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)

                                /s/ goes to /h/. /S/ is how <s> is pronounced, unleniated, before a
                                slender vowel (<i> or <e>).
                              • Keith Gaughan
                                ... And you d be completely correct. Sorry about that. ... Yep. ... Well, unless all the Irish speakers I ve met or heard have been involved in a conspiracy to
                                Message 15 of 25 , Jan 2, 2010
                                • 0 Attachment
                                  On Sat, Jan 02, 2010 at 01:55:13PM -0500, Alex Fink wrote:
                                  > On Sat, 2 Jan 2010 18:37:27 +0000, Keith Gaughan <k@...> wrote:
                                  >
                                  > > /d/ -> /j/ (Would have been [D] in the past, but wore down to [j])
                                  > > /d'/ -> /G/ (Would have been /D'/ in the past, but ended up as [G])
                                  > > /d'/ = palatalised /d/
                                  >
                                  > These two seem backwards.

                                  And you'd be completely correct. Sorry about that.

                                  > > /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)
                                  >
                                  > Seriously?

                                  Yep.

                                  > Why would something that's historically intervocalic lenition
                                  > end up like that? Especially when all of /t t_j s_j/ (to use more
                                  > orthogonal labels) go to /h/, it's quite strange that /s/ wouldn't.

                                  Well, unless all the Irish speakers I've met or heard have been involved
                                  in a conspiracy to cover up the reality of the situation, /s/ -> /S/.
                                  :-)

                                  The situation with coronals isn't wholly consistent - remember that /d/
                                  went all over the place when /D/ disappeared from the language. It's
                                  plausible if you consider that the mutations /s/ -> /S/ and /S/ -> /h/
                                  could have existed there long before /T/ degraded to /h/ - they may have
                                  arrived at /h/ by wholly separate mechanism. Now, I'm not saying there
                                  aren't dialects where /s/ and /S/ both go to /h/, but in the NW, from
                                  Donegal down to Connemara, 'shúil' ('eye') is pronounced /Suːl'/, which
                                  means a lot of kids end up misspelling it as 'siúl' ('walking').

                                  Here's the closest I can find on short notice that there's academic
                                  evidence of /s/ -> /S/ in Middle Irish at least:

                                  http://www.uni-due.de/DI/Middle_Irish.htm

                                  It could easily be that some dialects preserved this for velar /s/, but
                                  for palatal /s/, which was already [S], it went to /h/ for both.

                                  Of course, this is speculation, but it's driven by daily exposure to the
                                  language.

                                  K.

                                  --
                                  Keith Gaughan - k@... - http://stereochro.me/ - CF9F6473
                                  The map is not the territory; the sign is not the signified.
                                • Keith Gaughan
                                  ... I m aware that /S/ is how /s/ ends up when palatalised, but I m also saying that in the part of Ireland *I m* from, velar /s/, when lenited is /S/, and
                                  Message 16 of 25 , Jan 2, 2010
                                  • 0 Attachment
                                    On Sat, Jan 02, 2010 at 02:12:29PM -0500, Kelvin Jackson wrote:

                                    > >> /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)
                                    >
                                    > /s/ goes to /h/. /S/ is how <s> is pronounced, unleniated, before a
                                    > slender vowel (<i> or <e>).

                                    I'm aware that /S/ is how /s/ ends up when palatalised, but I'm also
                                    saying that in the part of Ireland *I'm* from, velar /s/, when lenited
                                    is /S/, and palatal /s/ when lenited is /h/ (though it generally sound
                                    more like [C] than [h]).

                                    I'm not saying this to be awkward; this is how my Grandmother, who was a
                                    gaeilgeoir, taught me, and that's what I was taught in school.

                                    K.

                                    --
                                    Keith Gaughan - k@... - http://stereochro.me/ - CF9F6473
                                    That man lives best who's fain to live half-mad, half-sane.
                                    -- Jan van Stijevoort
                                  • Aidan Grey
                                    I for one have contrary experience, that /s/ /h/, not /S/. At Oideas Gael in Gleann Colm Cille, it was always /h/. The President, Mrs. Mary MacAleese,
                                    Message 17 of 25 , Jan 2, 2010
                                    • 0 Attachment
                                      I for one have contrary experience, that /s/ > /h/, not /S/. At Oideas Gael in Gleann Colm Cille, it was always /h/. The President, Mrs. Mary MacAleese, pronounced it /h/. Everyone I talked to while I was in Ireland (mostly Donegal) said /h/.

                                      Are you sure you're understanding this right? It's spelled <sh>, put pronounced /h/. Maybe the people pronouncing it /S/ are the kids who don't actually speak Irish? I'm so confused how we can have such different experiences.

                                      <A Sheamais> = /h/ hence "Hamish"

                                      I wouldn't be surprised if there were other ways to pronounce it, but everyone I learned from and talked to pronounced it /h/. <mo shúil> as /hu:l'/ and everything.

                                      Aidan



                                      ----- Original Message ----
                                      > From: Keith Gaughan <k@...>

                                      > > > /s/ -> /S/ (If you're reading Wikipedia, it's wrong here)
                                      > >
                                      > > Seriously?
                                      >
                                      > Yep.
                                      >
                                      > > Why would something that's historically intervocalic lenition
                                      > > end up like that? Especially when all of /t t_j s_j/ (to use more
                                      > > orthogonal labels) go to /h/, it's quite strange that /s/ wouldn't.
                                      >
                                      > Well, unless all the Irish speakers I've met or heard have been involved
                                      > in a conspiracy to cover up the reality of the situation, /s/ -> /S/.
                                      > :-)
                                      >
                                      > The situation with coronals isn't wholly consistent - remember that /d/
                                      > went all over the place when /D/ disappeared from the language. It's
                                      > plausible if you consider that the mutations /s/ -> /S/ and /S/ -> /h/
                                      > could have existed there long before /T/ degraded to /h/ - they may have
                                      > arrived at /h/ by wholly separate mechanism. Now, I'm not saying there
                                      > aren't dialects where /s/ and /S/ both go to /h/, but in the NW, from
                                      > Donegal down to Connemara, 'shúil' ('eye') is pronounced /Suːl'/, which
                                      > means a lot of kids end up misspelling it as 'siúl' ('walking').
                                    • Benct Philip Jonsson
                                      ... The Dholuo data are interesting, as my conlang Sohlob has something similar. When the singulative suffix is added to a substantive or substantivized
                                      Message 18 of 25 , Jan 5, 2010
                                      • 0 Attachment
                                        R A Brown skrev:
                                        > Kenner Gordon wrote:
                                        >> When I originally said "Celtic-style mutations", I meant mutations in
                                        >> general.
                                        >
                                        > For more general use, cf.
                                        > http://en.wikipedia.org/wiki/Consonant_mutation
                                        >

                                        The Dholuo data are interesting, as my conlang Sohlob
                                        has something similar. When the singulative suffix
                                        is added to a substantive or substantivized adjective
                                        a final voiced consonant becomes devoiced, so that e.g.

                                        _trag_ 'turtles' -> _trakah_ 'a turtle'
                                        _sozg_ 'elephants' -> _soskah_ 'an elephant'
                                        _hel_ 'cities' -> _hehlah_ 'a city'
                                        _æçfer_ 'sons-in-law' -> _æçfehlah_ 'a son-in-law'

                                        But some consonants can't be devoiced:

                                        _gony_ 'wasps' -> _gonyah_ 'a wasp'
                                        _pons_ 'spiders' -> _ponsah_ 'a spider'

                                        Historically the reason is that the suffix is derived
                                        from the numeral _hah_ 'one'. However the result of
                                        the sandhi in every case coincides with an existing
                                        unit phoneme, and can't very well be synchronically
                                        analysed as C+h. N.B. That the Kidilib 'dialect'
                                        which unlike Classical Sohlob has phonemic voiceless
                                        nasals doesn't turn final voiced nasals into voiceless
                                        nasals when adding the suffix:

                                        Kdl.: _vony_ -> _vonya_ 'wasp(s)'
                                        Kdl.: _cahn_ -> _cahna_ 'fire(s)'
                                        C.S.: _cand_ -> _cantah_ 'fire(s)'

                                        Note: with definite noun phrases number is expressed
                                        on determiners, and the noun itself remains invariable:

                                        _hel an_ 'the city' -> _hel andar_ 'the cities'

                                        Note 2: As voiced and voiceless stops aren't normally
                                        distinguished in Sohlob script the mutation often is
                                        not expressed in (native) writing.

                                        /BP 8^)>
                                        --
                                        Benct Philip Jonsson -- melroch atte melroch dotte se
                                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                                        "C'est en vain que nos Josués littéraires crient
                                        à la langue de s'arrêter; les langues ni le soleil
                                        ne s'arrêtent plus. Le jour où elles se *fixent*,
                                        c'est qu'elles meurent." (Victor Hugo)
                                      • Elliott Lash
                                        The Dholuo and Sohlob data also have similarities to Breton and Welsh provection. This is found in forms of some adjectives: teg fair , tecach fairer ,
                                        Message 19 of 25 , Jan 5, 2010
                                        • 0 Attachment
                                          The Dholuo and Sohlob data also have similarities to Breton and Welsh provection.

                                          This is found in forms of some adjectives:

                                          teg "fair", tecach "fairer", teca(f) 'fairest' (Welsh)
                                          gleb 'wet', glepoc'h 'wetter', glepañ 'wettest' (Breton)

                                          And in subjunctives (mostly archaic now):
                                          catwo 'may he keep' (Welsh < cadw 'keeping'

                                          Elliott






                                          ----- Original Message ----
                                          From: Benct Philip Jonsson <bpj@...>
                                          To: CONLANG@...
                                          Sent: Tue, January 5, 2010 2:53:06 PM
                                          Subject: Re: Celtic-style mutations

                                          R A Brown skrev:
                                          > Kenner Gordon wrote:
                                          >> When I originally said "Celtic-style mutations", I meant mutations in
                                          >> general.
                                          >
                                          > For more general use, cf.
                                          > http://en.wikipedia.org/wiki/Consonant_mutation
                                          >

                                          The Dholuo data are interesting, as my conlang Sohlob
                                          has something similar. When the singulative suffix
                                          is added to a substantive or substantivized adjective
                                          a final voiced consonant becomes devoiced, so that e.g.

                                          _trag_ 'turtles' -> _trakah_ 'a turtle'
                                          _sozg_ 'elephants' -> _soskah_ 'an elephant'
                                          _hel_ 'cities' -> _hehlah_ 'a city'
                                          _æçfer_ 'sons-in-law' -> _æçfehlah_ 'a son-in-law'

                                          But some consonants can't be devoiced:

                                          _gony_ 'wasps' -> _gonyah_ 'a wasp'
                                          _pons_ 'spiders' -> _ponsah_ 'a spider'

                                          Historically the reason is that the suffix is derived
                                          from the numeral _hah_ 'one'. However the result of
                                          the sandhi in every case coincides with an existing
                                          unit phoneme, and can't very well be synchronically
                                          analysed as C+h. N.B. That the Kidilib 'dialect'
                                          which unlike Classical Sohlob has phonemic voiceless
                                          nasals doesn't turn final voiced nasals into voiceless
                                          nasals when adding the suffix:

                                          Kdl.: _vony_ -> _vonya_ 'wasp(s)'
                                          Kdl.: _cahn_ -> _cahna_ 'fire(s)'
                                          C.S.: _cand_ -> _cantah_ 'fire(s)'

                                          Note: with definite noun phrases number is expressed
                                          on determiners, and the noun itself remains invariable:

                                          _hel an_ 'the city' -> _hel andar_ 'the cities'

                                          Note 2: As voiced and voiceless stops aren't normally
                                          distinguished in Sohlob script the mutation often is
                                          not expressed in (native) writing.

                                          /BP 8^)>
                                          -- Benct Philip Jonsson -- melroch atte melroch dotte se
                                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                                          "C'est en vain que nos Josués littéraires crient
                                          à la langue de s'arrêter; les langues ni le soleil
                                          ne s'arrêtent plus. Le jour où elles se *fixent*,
                                          c'est qu'elles meurent." (Victor Hugo)
                                        • Benct Philip Jonsson
                                          ... It occurred to me that the conhistory of how number came to be expressed in determiners might be interesting. Proto-Sohlob had the possibility of using
                                          Message 20 of 25 , Jan 5, 2010
                                          • 0 Attachment
                                            On 2010-01-05 I wrote:
                                            > _hel an_ 'the city' -> _hel andar_ 'the cities'

                                            It occurred to me that the conhistory of how number
                                            came to be expressed in determiners might be interesting.

                                            Proto-Sohlob had the possibility of using reduplication
                                            to indicate plurality, which for CV pronominal stems like
                                            _na_ 'animate medial deictic' and _ra_ 'inanimate medial
                                            deictic' meant that the whole morpheme was doubled, so
                                            that there were forms like:

                                            ANIMATE INANIMATE

                                            _*a na_ _*a ra_ 'that there'
                                            _*a nana_ _*a rara_ 'those there'

                                            When later pretonic unstressed syllables were
                                            contracted or lost pluralization was lost in most
                                            words but the above forms became:

                                            _*ana_ _*ara_ 'that'
                                            _*anna_ _*arra_ 'those'

                                            which still later became:

                                            _ana_ _ara_ 'that'
                                            _anda_ _adra_ 'those'

                                            These are the forms actually attested in Kijeb.
                                            In Old Kidilib these had became:

                                            _an_ _ar_
                                            _and_ _ad@r_

                                            Now the ancestral form of Heleb and Classical Sohlob
                                            (to the extent the latter is not a conlang within
                                            the conculture) changed final *nd to *n, but preserved
                                            medial *nd (and of course treated *mb etc. analogously)
                                            so that it got:

                                            _*an_ _*ar_ direct singular
                                            _*anar_ _*arar_ instrumental singular
                                            _*an_ _*aðar_ direct plural
                                            _*andar_ _*aðarar_ instrumental plural

                                            To preserve the number distinction in the animate
                                            direct case forms this was analogously restructured to:

                                            _*an_ _*ar_ direct singular
                                            _*anar_ _*arar_ instrumental singular
                                            _*andar_ _*aðar_ direct plural
                                            _*andarar_ _*aðarar_ instrumental plural

                                            (At this time *[D] was (still) an allophone of */d/.)

                                            Except that *[D] merged with */z/ these are the forms
                                            attested and preserved in Classical Sohlob:

                                            _an_ _ar_ direct singular
                                            _anar_ _arar_ instrumental singular
                                            _andar_ _azar_ direct plural
                                            _andarar_ _azarar_ instrumental plural

                                            Pre-Heleb however merged */nd/ with */n/ (and similar)
                                            also medially, so that it got:

                                            _*an_ _*ar_ direct singular
                                            _*anar_ _*arar_ instrumental singular
                                            _*anar_ _*aðar_ direct plural
                                            _*anarar_ _*aðarar_ instrumental plural

                                            Thus the animate instrumental singular and direct plural
                                            merged. An analogical restructuring based on the
                                            inanimate forms similar to the earlier one restored the
                                            distinction:

                                            _*an_ _*ar_ direct singular
                                            _*anar_ _*arar_ instrumental singular
                                            _*aðnar_ _*aðar_ direct plural
                                            _*aðnarar_ _*aðarar_ instrumental plural

                                            The attested Heleb forms are:

                                            _an_ _ar_ direct singular
                                            _anar_ _arar_ instrumental singular
                                            _aznar_ _azar_ direct plural
                                            _aznarar_ _azarar_ instrumental plural

                                            The Sohloçan philologists thought they could tell that
                                            the second restructuring happened before *[D] > */z/, since */nz/
                                            was a possible sequence in the language, but *[nð] wasn't.
                                            Moreover the forms in Yahab are:

                                            [a~] [ar] direct singular
                                            [a~na~r~] [arar] instrumental singular
                                            [a~Ja~r~] [azar] direct plural
                                            [a~Ja~rar~] [azarar] instrumental plural

                                            Which is consitent with _*aðnar_, since in Yahab:

                                            D > j / _ C
                                            jn > J
                                            D > z / V _ V
                                            (and z > z~ if there is a nasal in the word!)

                                            (I havent worked out the Yahab spellings/transliterations
                                            yet. Very likely the Yahçan don't write their language.
                                            The most sensible *Latinizations* are _an, adarn, ayarn,
                                            ayararn_, since nasality is a feature of the *word* in
                                            Yahab. I don't suppose the Sohloçan philologists
                                            used so rational a transcription, though.

                                            /BP 8^)>
                                            --
                                            Benct Philip Jonsson -- melroch atte melroch dotte se
                                            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                                            "C'est en vain que nos Josués littéraires crient
                                            à la langue de s'arrêter; les langues ni le soleil
                                            ne s'arrêtent plus. Le jour où elles se *fixent*,
                                            c'est qu'elles meurent." (Victor Hugo)
                                          • Jörg Rhiemeier
                                            Hallo! ... Indeed, very interesting! How are nouns pluralized? ... brought to you by the Weeping Elf http://www.joerg-rhiemeier.de/Conlang/index.html
                                            Message 21 of 25 , Jan 6, 2010
                                            • 0 Attachment
                                              Hallo!

                                              On Tue, 5 Jan 2010 19:27:20 +0100, Benct Philip Jonsson wrote:

                                              > On 2010-01-05 I wrote:
                                              > > _hel an_ 'the city' -> _hel andar_ 'the cities'
                                              >
                                              > It occurred to me that the conhistory of how number
                                              > came to be expressed in determiners might be interesting.
                                              >
                                              > [stuff snup]

                                              Indeed, very interesting! How are nouns pluralized?

                                              ... brought to you by the Weeping Elf
                                              http://www.joerg-rhiemeier.de/Conlang/index.html
                                            • Benct Philip Jonsson
                                              ... They aren t. Number is only shown on the determiner. Notionally proto-Sohlob had pluralization through reduplication, so that if the proto-form of the
                                              Message 22 of 25 , Jan 6, 2010
                                              • 0 Attachment
                                                Jörg Rhiemeier skrev:
                                                > Hallo!
                                                >
                                                > On Tue, 5 Jan 2010 19:27:20 +0100, Benct Philip Jonsson wrote:
                                                >
                                                >> On 2010-01-05 I wrote:
                                                >> > _hel an_ 'the city' -> _hel andar_ 'the cities'
                                                >>
                                                >> It occurred to me that the conhistory of how number
                                                >> came to be expressed in determiners might be interesting.
                                                >>
                                                >> [stuff snup]
                                                >
                                                > Indeed, very interesting! How are nouns pluralized?

                                                They aren't. Number is only shown on the determiner.

                                                Notionally proto-Sohlob had pluralization through
                                                reduplication, so that if the proto-form of the word
                                                which became 'city' was something like _*síre_ its
                                                plural was _*sisíre_. However pretonic unstressed
                                                vowels were lost, so the plural became _*ssíre_
                                                and then merged with the singular. The deictics
                                                preserved distinct plurals because their free forms
                                                were always preceded by a local adverbial particle:

                                                SINGULAR PLURAL

                                                PROXIMAL *í na *í nana 'this'
                                                MEDIAL *á na *á nana 'that'
                                                DISTAL *ú na *ú nana 'yonder' (Ger. 'jener')

                                                And similar inanimare forms _*í ra, *í rara_ etc.

                                                When these suffered syncope the result was forms like
                                                sg. _*ina, *ira_, pl. _*inna, *irra_. In these plurals
                                                the geminates were preserved and even developed further
                                                _*nn_ > _nd_, _*rr_ > _dr_ as described in my previous
                                                post.

                                                OTOH indefinite nouns can be explicitly *singularized*
                                                with the _-(h)ah_ suffix -- actually a clitic numeral
                                                'one', but speakers perceive it as a suffix since it
                                                fuses phonetically with a final consonant of the noun,
                                                as I showed in my other post (in the Celtic mutations
                                                thread).

                                                /BP 8^)>
                                                --
                                                Benct Philip Jonsson -- melroch atte melroch dotte se
                                                ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                                                "C'est en vain que nos Josués littéraires crient
                                                à la langue de s'arrêter; les langues ni le soleil
                                                ne s'arrêtent plus. Le jour où elles se *fixent*,
                                                c'est qu'elles meurent." (Victor Hugo)
                                              • Jörg Rhiemeier
                                                Hallo! ... I see. But how about introducing a fortis/lenis distinction of some kind, such that singular nouns begin with a lenis consonant, and plural nouns
                                                Message 23 of 25 , Jan 6, 2010
                                                • 0 Attachment
                                                  Hallo!

                                                  On Wed, 6 Jan 2010 20:49:24 +0100, Benct Philip Jonsson wrote:

                                                  > > Indeed, very interesting! How are nouns pluralized?
                                                  >
                                                  > They aren't. Number is only shown on the determiner.
                                                  >
                                                  > Notionally proto-Sohlob had pluralization through
                                                  > reduplication, so that if the proto-form of the word
                                                  > which became 'city' was something like _*síre_ its
                                                  > plural was _*sisíre_. However pretonic unstressed
                                                  > vowels were lost, so the plural became _*ssíre_
                                                  > and then merged with the singular.

                                                  I see. But how about introducing a fortis/lenis distinction
                                                  of some kind, such that singular nouns begin with a lenis
                                                  consonant, and plural nouns with the corresponding fortis
                                                  consonant which developed from the geminate left by the loss
                                                  of the vowel? As in

                                                  singular *síre > *zíre
                                                  plural *sisíre > *ssíre > *síre

                                                  (using the symbols |s| and |z| for the fortis and lenis
                                                  variants of *s here)

                                                  But it could just as well be the other way round, e.g.

                                                  singular *síre > *síre (unchanged)
                                                  plural *sisíre > *sizíre > *szíre > *zíre

                                                  Maybe an idea to pursue in a future conlang of mine.

                                                  ... brought to you by the Weeping Elf
                                                  http://www.joerg-rhiemeier.de/Conlang/index.html
                                                • Benct Philip Jonsson
                                                  ... It s a nice idea but if I adopted it I d have to change Kijeb synchronic morphology and Sohlob diachronic phonology beyond recognition! Assuming that
                                                  Message 24 of 25 , Jan 7, 2010
                                                  • 0 Attachment
                                                    Jörg Rhiemeier skrev:
                                                    > Hallo!
                                                    >
                                                    > On Wed, 6 Jan 2010 20:49:24 +0100, Benct Philip Jonsson wrote:
                                                    >
                                                    >>> Indeed, very interesting! How are nouns pluralized?
                                                    >> They aren't. Number is only shown on the determiner.
                                                    >>
                                                    >> Notionally proto-Sohlob had pluralization through
                                                    >> reduplication, so that if the proto-form of the word
                                                    >> which became 'city' was something like _*síre_ its
                                                    >> plural was _*sisíre_. However pretonic unstressed
                                                    >> vowels were lost, so the plural became _*ssíre_
                                                    >> and then merged with the singular.
                                                    >
                                                    > I see. But how about introducing a fortis/lenis distinction
                                                    > of some kind, such that singular nouns begin with a lenis
                                                    > consonant, and plural nouns with the corresponding fortis
                                                    > consonant which developed from the geminate left by the loss
                                                    > of the vowel? As in
                                                    >
                                                    > singular *síre > *zíre
                                                    > plural *sisíre > *ssíre > *síre
                                                    >
                                                    > (using the symbols |s| and |z| for the fortis and lenis
                                                    > variants of *s here)
                                                    >
                                                    > But it could just as well be the other way round, e.g.
                                                    >
                                                    > singular *síre > *síre (unchanged)
                                                    > plural *sisíre > *sizíre > *szíre > *zíre
                                                    >
                                                    > Maybe an idea to pursue in a future conlang of mine.

                                                    It's a nice idea but if I adopted it I'd have to
                                                    change Kijeb synchronic morphology and Sohlob
                                                    diachronic phonology beyond recognition! Assuming
                                                    that Proto-Sohlob had the following consonant
                                                    phonemes

                                                    p t k
                                                    b d g
                                                    m n N
                                                    s
                                                    w r j

                                                    I could get a fortis-lenis distinction by the
                                                    following changes, mont of which already are
                                                    employed in the change from PS to Kijeb (tho I'm
                                                    actually 'reconstructing PS from Kijeb! :-).
                                                    Currently 'fortition' of geminate consonants does
                                                    take place between PS and Kijeb, but voiced stops
                                                    and *s are not susceptible to it, and nasals and
                                                    sonorants are only susceptible when they stand
                                                    immediately after the stressed vowel:

                                                    *pp > f (via *pf)
                                                    *tt > st
                                                    *kk > x (via *kx)
                                                    *bb > b
                                                    *dd > d/zd
                                                    *gg > g
                                                    *mm > m-/´-mb-
                                                    *nn > n-/´-nd-
                                                    *NN > N-/´-Ng-
                                                    *ww > w-/´-gw-
                                                    *rr > r-/´-dr-
                                                    *yy > y-/´-gy-

                                                    To be sure Tolkien made admirably elegant use of
                                                    similar fortition of nasals and initial
                                                    prenasalized voiced stops, but partly because of
                                                    that I don't want to use them; they have their
                                                    definite place in Quendian, but don't feel right
                                                    for Kijeb.

                                                    Currently Kijeb also has no voiced fricatives,
                                                    except [z] as an allophone of /s/ before /b d g/
                                                    -- e.g. the word 'elephant' which I transliterate
                                                    _tyazgu_ but which really is /t_jasgu/ (having in
                                                    PS been something like _*teesVgoo(CV)/*teesVgVBV_,
                                                    where *V = any short vowel, *C = any consonant and
                                                    *B = any voiced labial). Surely I could get around
                                                    that by positing something like:

                                                    *b > *B > w
                                                    *d > *D > Ø
                                                    *g > *G > Ø
                                                    *s > *h > Ø

                                                    This would however be a very strange system where
                                                    words with some initials would have no overt
                                                    plurals and other words would have distinct
                                                    plurals but identical singulars. My guess is that
                                                    in such a language pluralization in nouns would
                                                    actually be lost, *especially* as pluralization
                                                    could be unambiguously marked in definite
                                                    determiners and indefinite singularity could be
                                                    marked with a clitic numeral 'one'. Clearly there
                                                    may be vestiges in Kijeb, like a word for 'leaf'
                                                    in _t-_ and a word for 'foliage' which is
                                                    identical except for beginning in _st-_, and
                                                    words for 'head of cattle' in _p-_ and for
                                                    'herd' in _f-_.

                                                    Actually I think I might do a sister language to
                                                    Kijeb which does mark pluralization by initial
                                                    mutations, after sound changes like:

                                                    *b- > v- / unchanged
                                                    *d- > z-/r- / unchanged
                                                    *g- > Ø-/R- / unchanged
                                                    *m- > v-
                                                    *n- > r-/l-
                                                    *N- > Ø-
                                                    *s- > h- / *sVs- > *sVz- > z-
                                                    *r- > R-/r-
                                                    *pp- > f-
                                                    *tt- > st-/ts-
                                                    *kk- > x-
                                                    *bb- > b- / *bVb- > *bVB- > v-
                                                    *dd- > d-/zd-/dz-
                                                    *gg- > g- / *gVg- > *gVG- > gR- (> gr-)
                                                    *mm- > m-
                                                    *nn- > n-
                                                    *NN- > N-
                                                    *ww- > gw-
                                                    *rr- > gR-/dr-
                                                    *yy- > gy-

                                                    The snag here is that PS had unstressed prefixes
                                                    on verbs which changed valency, aktionsart etc.
                                                    and also deverbal nouns from these, and I don't
                                                    know how the latter would behave WRT
                                                    reduplication. I think there would be different
                                                    results depending on whether the prefix or the
                                                    root initial were reduplicated:

                                                    PREFIX REDUPLICATION SYNCOPE KIJEB
                                                    *paká- *papaká- *ppka- kwa-
                                                    -"- *pakaká- *pkka- xwa-
                                                    *patá- *papatá- *ppta- fra-
                                                    -"- *patatá- *ptta- ta-/twa-

                                                    Currently I have the following changes with prefixes,
                                                    not considering reduplication:

                                                    *kk > x
                                                    *kg > g
                                                    *kt > kr
                                                    *kd > gr
                                                    *kp > kw
                                                    *kb > gw
                                                    *dk > k
                                                    *dg > g
                                                    *dt > st
                                                    *dd > d
                                                    *dp > p
                                                    *db > b
                                                    *pk > kw
                                                    *pg > gw
                                                    *pt > pr
                                                    *pd > br
                                                    *pp > f
                                                    *pb > b

                                                    plus *s + *ptkbdg... which remain unchanged

                                                    And I'm considering:

                                                    *dk > sk
                                                    *dg > [z]g
                                                    *dt > st
                                                    *dd > [z]d
                                                    *dp > tw
                                                    *db > dw
                                                    *pk > xw

                                                    Combining this with reduplication would introduce
                                                    considerably more havoc into the historical
                                                    phonology. It may be fun though. WRT Kijeb and
                                                    Sohlob proper it would clearly just contribute to
                                                    the demise of noun pluralization, but in the
                                                    putative sister language some kind of
                                                    regularization would probably happen. Granted
                                                    initial clusters like _px, tx, tf, pst, fx, stx,
                                                    stf, ft_ may well exist in the language! Consider
                                                    changes like:

                                                    *dt > *tt > ts
                                                    *dd > dz
                                                    *ddt > *tst > st
                                                    *ddd > *dzd > zd

                                                    Though I doubt the latter two would really differ
                                                    from the previous two in a natural language.
                                                    Perhaps if syncope and fricativization proceded in
                                                    stages. Suppose a second unstressed prefix
                                                    syllable had a more reduced vowel which was lost
                                                    earlier (V being a non-schwa but undetermined
                                                    vowel):

                                                    *d@dVtá-
                                                    > *ddVtá-
                                                    > *ddtá-
                                                    > *dztá
                                                    > *tsta-
                                                    > sta-

                                                    and the singular

                                                    *dVtá-
                                                    > *dta-
                                                    > *tta-
                                                    > tsa-

                                                    Consider side by side with thisa possible
                                                    singular-plural pair like

                                                    SINGULAR PLURAL

                                                    *pVtá- *p@pVtá-
                                                    > *ppVtá-
                                                    > *pta- > *ppta-
                                                    > *ptsa- > *pfta-
                                                    > psta- > fta-

                                                    This is supposing that intermediate *pt- remains
                                                    distinct from intermediate *ps- which is by no
                                                    means certain. Perhaps some singulars in _ps-_
                                                    have plurals in _fs-_ (*pps-) while others have
                                                    plurals in _ft-_ (*ppt-).

                                                    With the _*s-_ prefix the going could get really
                                                    evil:

                                                    SINGULAR PLURAL

                                                    *sVtá- *s@sVtá-
                                                    > *ssVtá-
                                                    > *hVtá-
                                                    > *sVtá-
                                                    > *hta- > *sta-
                                                    > tha- > sta-

                                                    This might indeed be an interesting language, and
                                                    it's probably the language of the "Invaders", a
                                                    barbaric people who overran Sohlodar at one time.
                                                    Also an incentive to actually explore Proto-Sohlob
                                                    rather than treating Kijeb as the
                                                    quasi-protolanguage! :-)

                                                    /BP 8^)>
                                                    --
                                                    Benct Philip Jonsson -- melroch atte melroch dotte se
                                                    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
                                                    "C'est en vain que nos Josués littéraires crient
                                                    à la langue de s'arrêter; les langues ni le soleil
                                                    ne s'arrêtent plus. Le jour où elles se *fixent*,
                                                    c'est qu'elles meurent." (Victor Hugo)
                                                  • Jörg Rhiemeier
                                                    Hallo! ... I didn t mean to suggest changing the languages you have; if you are content with them, by any means keep them the way they are! It s just an idea
                                                    Message 25 of 25 , Jan 7, 2010
                                                    • 0 Attachment
                                                      Hallo!

                                                      On Thu, 7 Jan 2010 11:20:15 +0100, Benct Philip Jonsson wrote:

                                                      > Jörg Rhiemeier skrev:
                                                      > > Hallo!
                                                      > [...]
                                                      > > Maybe an idea to pursue in a future conlang of mine.
                                                      >
                                                      > It's a nice idea but if I adopted it I'd have to
                                                      > change Kijeb synchronic morphology and Sohlob
                                                      > diachronic phonology beyond recognition!

                                                      I didn't mean to suggest changing the languages you have;
                                                      if you are content with them, by any means keep them the way
                                                      they are! It's just an idea of mine, which you may pursue
                                                      if you feel like adding another language to your family.
                                                      But if you feel that it makes more problems than sense,
                                                      just don't do it.

                                                      But I can use it for my yet unborn Indo-European conlang,
                                                      where it would be used in forming perfects, e.g.

                                                      present *dhembhoh2 > zemvöx 'I get astonished'
                                                      perfect *dhedhombhh2a > *zzomvxa > domva 'I am astonished'

                                                      or something like that.

                                                      ... brought to you by the Weeping Elf
                                                      http://www.joerg-rhiemeier.de/Conlang/index.html
                                                    Your message has been successfully submitted and would be delivered to recipients shortly.