Knoschke

I have made a sketch of the phonology for a Noric language which as yet has no name. In the meantime I'll call it "B-Noric", short for "Benct's Noric language".

History

(To be filled in)

Sound system and transcription

"B-Noric" is transcribed into Latin letters as follows. Unless otherwise indicated letters have their IPA values.

• An h following p, t or k indicates aspiration [ʰ].
• A j following another consonant indicates palatalization [ʲ].
• The vowel y is (still) [ɨ]. In "B-Noric" it may also occur long.
• Grave accent is used to indicate low mid vowels, so that è is [ɛ] and ò is [ɔ].
• Vowels with acute accent are long: á [aː], é [eː], í [iː], ó [oː], ú [uː], ý [yː]. Note especially ê for [ɛː] and ô for [ɔː]!
• The letters š and ž indicate retroflex fricatives [ʂ] and [ʐ], with tš and dž for the corresponding affricates [ʈʂ] and [ɖʐ].
• The letter q is used for the voiced velar fricative [ɣ].
• The letter ñ is used word-finally to indicate nasalization of the preceding vowel. (Should this be replaced with µ?)

NB: If the B-noric sound-change applier is run with the "show the steps of the derivation" option checked a number of forms with partly deviant letter-values are shown. These transcriptions are a pure expedient of easy programming. The values are as follows (with some letters changing their significance during the process):

• áéíóú are internal notation for long vowels
• àèìòùý are internal notation for stressed short vowels
• âêîôû are internal notation for stressed long vowels
• þ is internal notation for [ʔ]
• ð is internal notation for [ɾ]
• c is internal notation for [ɣ]
• ç is internal notation for [ɦ]
• x is [x]!
• æ is long [æ:], ä is short [æ], Æ is stressed long ['æ:], ã is stressed short ['æ]

12:25, 21 Jun 2005 (PDT)

Historical phonology

A Proto-Noric to B-Noric sound change applier program (not yet bug-free) is here.

NB that the description of the phonology below is not yet brought up-to-date. Wherever the output of the program differs from what the text makes you expect the program is more reliable!

Transcriptions preceded by * are the agreed-upon transcription. Other transcriptions (not always in slashes/brackets) are IPA.

*CVLC tends to be realized as CLVC when the first C is a stop.

Obstruents are weakened between vowels or between a vowel and a liquid, except where the except where the preceding vowel is stressed:

*b > [β]	*p > [b]	*ph > [f]	*tc > [ʃ]	*ts > [s]	*s/*c > [h]
*d > [ɾ]	*t > [d]	*th > [h]	*tch > [ʃ]	*tsh > [s]	*x > [j]
*g > [ɣ]	*k > [g]	*kh > [x]	*dx > [ʒ]	*dz > [z]	*h > [ɦ]
	*q > [ʔ]	*qh > [h]

Obviously this leads to some mergers in this position: *tc/*tch, *ts/*tsh, *dz/*z, *th/*qh/*s/*c and *x/*j. OTOH the phonemes *b/*p, *d/*t, *g/*k and *th/*h do not actually merge, although there was possibly some realignment.

Remaining *q and *qh merge with *k and *kh respectively. The relative dating of this merger is impossible to determine.

Front vowels (*i, *î, *e, *ê) develop a prothetic [j] and consonants are palatalized before the same vowels (though not before *y). Consonants before inherited *j of course palatalize as well.

Palatalized dentals and velars merge with each other. They however remain distinct from the old "palatals" *tc, *tch, *dx, *c, *x, the latter perhaps becoming retroflex, except for [ɣʲ] which merges with *j. These two sets will be called primary palatals and secondary palatals respectively. It should be noted that primary palatals in palatalizing environment are reflected as secondary palatals (or should they?)

Vowels before /j/ front/palatalize. For a while there were probably front rounded vowels, but these eventually merged with the unrounded front vowels.

*a(:)j > æ(:)j

*o(:)j > e(:)j

*u(:)j > i(:)j

*yj > ij

All *VjC sequences become V:C, so there arises a new phoneme /æ:/ from *ajC.

Unrounded vowels round/retract before *w. The front vowels probably become rounded central vowels at first, but eventually they become back rounded. There arises no new phonemes, since *aw/*âw merge with *ow/*ôw as [o:].

*a(:)w > o(:)w

*e(:)w > o(:)w

*i(:)w > u(:)w

*yw > uw

• All *VwC sequences become V:C, but there arises no new phoneme since *awC merges with *owC and *ôC as [o:].

As a result of these changes the secondary palatals become distinct phonemes.

The [ɣ] allophone of /g/ merges with /j/ after /æ(:)/, /e(:)/ and /i(:)/ but with /w/ after other vowels.

Palatalized labial stops become labiodental fricatives. The [β] allophone of /b/ merges with the new /v/; /mʲ/ merges with /w/.

Word initially and before vowels /w/ merges with /g/, probably by way of [gw]. Sporadically /j/ becomes /dʲ/ under the same circumstances, suggesting that there existed a dialect where the parallell change of j > gj > dʲ was general.

Unstressed short vowels are weakened and eventually lost.

Unstressed long vowels shorten. Perhaps distinctive stress is lost at the same time.

The consonant clusters that arose through the loss of unstressed short vowels are partly simplified through assimilation, dissimilation or loss.

In any cluster of three or more consonants the middle consonant(s) drop.

Main exceptions:

All /VjC/ become /ViC/ and all /VwC/ become /VuC/ -- there is no phonetic change, but the /w/ phoneme disappears.

Dentals before /j/ merge with the secondary palatals. Also /xʲ/ and /hʲ/ ([ç]?) merge with /sʲ/, and [ɦj] merges with /j/.

After palatals /j/ drops. The transcription of "B-Noric" however masks this fact.

Non-sibilant consonants before a sibilant drop; thus C(C)SC becomes SC. This affects affricates as well, so that they lose their closure part.

A [d] next to an obstruent becomes [ɾ], except if that obstruent is [d], [z] or [ʒ].

Progressive voicing assimilation between obstruents; /ʔ/ counts as a voiceless obstruent.

• Complete assimilation of /ʔ/ to any neighboring consonants, except for /ʔ/ + sibilant which becomes an affricate.

• All clusters of obstruent + obstruent become fricative + stop. Exceptions:

• pp, tt, kk, bb, dd, gg, ff, xx remain.

• bv > bb, gG > gg, pf/fp > ff, kx/xk > xx.

• /h/ is wholly assimilated to an adjacent fricative. After an affricate this means loss of /h/. Non-affricate sibilants become geminate.

• A resonant + sibilant becomes resonant + affricate.

• Velar + labial becomes velar + /v/.

• Any stop + /h/ or /h/ + stop becomes a voiceless aspirated stop.

• Initial unclustered stops or stops followed by r/l/ɾ become voiceless aspirated stops.

• Initial nasal + voiceless stop becomes aspirated voiceless stop.

• Initial nasals before voiced stops are lost.

• /t/ + nasal becomes /tt/. Also /pm/ becomes /pp/ but /pn/ remains.

• Labial, dental or palatal voiced stop + nasal becomes geminate nasal with the point of articulation of the stop, while /g/ + nasal becomes/remains /gn/.

• Initial resonants before stops are lost.

• Initial geminates are simplified.

• nl/ln/rl/lr/ɾl/lɾ > ll.

• mn > nn, nm > mm.

• nr/nɾ > rr.

• Vr > Vrr.

• Vɾ > Vr.

• The distinction of length in stressed vowels is replaced by a tenseness distinction, while all short vowels are laxed. There are some mergers:

ɪ/e > e

ɛ/æ/æː > ɛ

a/a: > a

ʊ/o > o

ɔ > /ɔ/

• /ʔ/ and /ɦ/ disappear, giving rise to a great variety of new diphthongs. High mid vowels become high after other vowels (Ve > Vi, Vo > Vu, except ee > ê and oo > ô giving Vi and Vi diphthongs. Va/Vɛ/Vɔ give new long vowels.

• Vh becomes V:

• Final vowel + simple nasal becomes nasalized vowel.

• Final geminates are simplified.

• Vv becomes Vw, with uv becoming û.

To be updated.

Examples

BPJ 07:21, 22 Jun 2005 (PDT)

(Not yet debugged)

*kûdxî`lî > khozjelji
*khitî`do > khrjedò
*uro`phâ > fa
*tcunhû`nzâ > šondza
*duxi`jû > théju
*`âtcâza > atšaz
*`axurhe > ažrj
*`dzîltshize > dzjisj
*`diwdemo > thjunj
*thu`negî > tjèdje
*`jarilda > jarjd
*wî`kajdxâ > khegedža
*aqhi`zî > tsji
*uba`tsû > sfu
*`inwâtshô > jengasò
*`zûjara > zuir
*î`quthê > jutsjè
*râ`rûsu > rarrus
*tci`dxâqî > tšatje
*dîwû`zê > thjogozje
*â`rnery > arnjèr
*`othohi > òtsj
*tsîse`jnu > tsjesjèno
*sâdi`ba > sarjba
*tsi`dâlnô > thlanò
*`îbadu > jibr
*putcû`rgâ > šporqa
*`bidxîbu > vedzjeu
*tchu`lnarha > šnarh
*tcîlî`je > tsjeljejè
*`isâzâ > jesaza
*bu`môqî > motje
*î`dûkhû > jerukho
*`ôwîmâ > ojema
*`thedxûda > tsjèdžor
*uzi`xû > tsju
*dxu`qhyla > žyl
*thûco`nxa > thošša
*âlku`qi > alije
*upu`gû > ku
*`tchopajthi > šòpètsj
*tshî`nôtcho > sjenoš
*îmbja`zê > jemvzje
*blo`khînu > zjbiñ
*ûtsô`co > osô
*`ûrkharî > urje
*etlî`tu > jljedo
*`dxusûqhi > džososj
*`utelu > otjl
*âtshu`tû > asru
*râ`ltshatû > rassato
*u`nilse > njeltsj
*tû`myru > thomyr
*obla`tche > sjfè
*hinzu`kha > zqa
*`olmadzî > òldzje
*u`thapha > haph
*a`zômba > tsomb
*`rarcadô > rašrò
*`akumu > akv

BPJ 04:07, 21 Jun 2005 (PDT)