Re: Richard's Romanian sound changer

From: Richard Wordingham
Message: 30731
Date: 2004-02-05

--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:
> I've tested Richard's Romanian Sound Change Applier against (most
of) the
> examples given in Bourciez.
>
> The remaining problems that I found are (some belong in several
catgories
> at once):
>
> 1) unstressed vowels
>
> pedica > piedicã not piedecã
> pectinem > pieptine not pieptene
> ra:dici:na > rãdãcinã not rãdecina
> blasphe:ma:re > blestema not blãsfemare
> veci:num > vecin not vãcin
> feli:cem > ferice not fãrice
> petiolus > picior not pãcior
> berbica:rius > berbecar not bãrbecar
> lendina > lindinã not lindenã
> titio:nem > tãciune not teciune
> anellum > inel not îniel
> fenestra > fereastrã not fãniastrã
>
> [This is going to be hard to fix, because the rules are not easy to
> extract. In verbal forms, the quality of the stressed vowel is
also
> important:
> *pe:ti:re > petzi not pãtzire
> vena:re > vâna(re) not vãnare
> vide:re > vedea/vedere not vãdere
> ]
>
> 2) v ~ b ~ 0
> pulverem > pulbere not pulvere
> hi:berna > iarnã not jbiarnã
> caballum > cal not cãbal
> cervum > cerb not cerv
> clavem > cheie not cheve
> vi:va > vie not vivã
> scri:bo: > scriu not scrib
> deus > zeu not ze
> reus > rãu not re
> ri:vus > râu not râv
> cavito: > caut not cavet
> grevum > greu not grev
> civita:tem > cetate not cevetate
> habe:re > avea/avere not abere
> novem > nouã not noave
> lava:re > la(re) nor lãvare
>
> 3) l^
> léporem > iepure not liepure
> folia > foaie not foale
> galli:na > gãinã not gãlinã
> fi:lium > fiu not fili
> mulie:rem > muiere not mulere
> li:num > in not lin
> li:cia > itze not litzã
> li:berta:re > ierta not libertare
>
> 4) -ia(-)
> cere:sia > cirea$ã not cerease
> area > arie not are
> auricla > urechie not aureache
> vi:nea > vie not viie
> urceolum > ulciór not urcéur
> fascia > fa$e not fa$ã
>
> 5) by > yb
> habeat > aibã not abe
> robeum > roib not rob
> cubium > cuib not cub
>
> 6) final -u after Cl/Cr/V
> nigrum > negru not negr
> duplum > duplu not dupl
> oclum > ochiu not ochi
> coliculum > curechiu not curechi
>
> 7) -br- > -wr-
> fabrum > faur not fabr
>
> 8) x > s/$
> maxilla > mãsea not mãp$ea
> lixi:va > le$ie not lep$ivã
>
> 10) st
> sci:re > $ti not $tzi
> ste:lla > stea not $tea
> investi:re > înve$ti(re) not înve$tzire
>
> x) others
> me:a > mea not ma

Should this be, 'mea > mea not *mia'?

> gu:- > gu- not b-
> conventum > cuvînt not cunvînt
> no:s > noi not no
> vo:s > voi not vo
> occi:dere > ucide not uccide(re) [kc not simplified]
> matteu:ca > mãciucã not mãtciucã
> ia:cit > zace not jace
> deorsum > jos not giors
> squama > scamã not spamã
> mergere > merge not mierge [?]
> mer(u)la > mierlã not miarlã [?]

Here's my list of problems (last updated Sunday 1st February):

Evolution of 'baiula' (Cybalist 17527) showed the following
omissions:

(i) Unstressed /j/ did not affricate - cf. Sp. cuyo 'whose'.
(ii) Rhotacism /l/ > /r/ is not yet addressed. Yes it is – but
environment omits stress flag! - Fixed 18/1/04.

ps > s before stressed vowels (Cybalist 17544).

Sort out intro: > 1ntru (Cybalist 17760 ). May have to ask Miguel!

Don't seem to have implemented Cybalist 18207.

Dropping of intervocalic n – Cybalist 18319
Also, fre:num > frâu

bi > ib | V_V – Cybalist 18412

n^ > y – Cybalist 18412 Implemented 18/1/04

hospitem > oaspete, not ospete. Why? Cybalist 18582.

Dropping of final -r Cybalist 18557 and account of sora.

Syncopation = batra:nu, genuchi . See Cybalist 18642.

Stress and yodicisation – Cybalist 18594 (problem with _urceolum_
above). Also look at 18642 for -li-.

Looking at 18614 – fonta:na presents problems. - i.e.
fonta:na > fântânã, not *funtânã

Need to investigate l^ > y. femeie (18642). Also rule ordering
issues, to get ea > e, not a.

Note, unstressed /a/ rule change – Cybalist 25317.

Dropping of intervocalic -b- e.g. Cybalist 18657

Treatment of -bi- e.g. Cybalist 28465.

anellus > inel Cybalist 18732

Check Cybalist 18733 has been incorporated.

ll > w Cybalist 18780, 18784, 18799

Cybalist: 19299:

viTea has the stress on /a/ and viTel on /e/
vitã has the accent on /i/.
The suffix change the stress of the root, in this case the stress
will be on the vowel of the suffix

vitzea Cybalist 19300

-(e)lla > -(e)a Cybalist 1300

stinge Cybalist 19399 – Why not $tinge?

Unstressed e > & - not always – Cybalist 28849

pinnna (or penna?) > panã Cybalist 19441 – investigate rule or=
dering.

n > r Cybalist 19551

VrVn > VnVn Cybalist 19553

novem > nouã Cybalist 29130

el pronounced as though iel Cybalist 29823

â/î selection rule: Conlang
http://groups.yahoo.com/group/conlang/message/104952
The most recent one,
in 1990, made it so that the two are pure allographs: _î_ is used as
the first
or last character of a word, and _â_ elsewhere.

Richard.