Sangi/Sandhi

Phonological Processes at CC morpheme boundaries

	First Sound
Second Sound	p	b	t	d	c	g	m	n	ŋ	s	z	ś	ź	ć	ǵ
p	pp	bb	tp	db	cp	gb	mm	mm	mm	sp	zb	śp	źb	ćp	ǵb
b	pp	bb	tp	db	cp	gb	mp	mp	mp	sp	zb	śp	źb	ćp	ǵb
t	pt	bd	tt	dd	ct	gd	nn	nn	nn	st	zd	śt	źd	ćt	ǵd
d	pt	bd	tt	dd	ct	gd	nt	nt	nt	st	zd	śt	źd	ćt	ǵd
c	pc	bg	tc	dg	cc	gg	ŋŋ	ŋŋ	ŋŋ	sc	zg	śc	źg	ćc	ǵg
g	pc	bg	tc	dg	cc	gg	ŋc	ŋc	ŋc	sc	zg	śc	źg	ćc	ǵg
m	mm	bb	nn	dd	ŋŋ	gg	mm	nn	ŋŋ	sm	zm	śm	źm	nnj	ǵǵ
n	mm	bb	nn	dd	ŋŋ	gg	mm	nn	ŋŋ	sn	zn	śm	źn	nnj	ǵǵ
ŋ	mm	bb	nn	dd	ŋŋ	gg	mm	nn	ŋŋ	sn	zn	śn	źn	nnj	ǵǵ
s	ps	bz	ts	dz	cs	gz	ms	ns	ncs	ss	ts	śś	źź	ćś	ǵź
z	ps	bz	ts	dz	cs	gz	ms	ns	ncs	s	zz	ź	ś	ćś	ǵź
ś	pś	bź	ć	ǵ	cś	gź	mś	nś	ncś	ś	ć	śś	źź	ćś	ǵź
ź	pś	bź	ć	ǵ	cś	gź	mś	nś	ncś	śś	ǵ	śś	źź	ćś	ǵź
ć	pć	bǵ	tć	dǵ	cć	gǵ	mć	nć	nć	ść	źǵ	ść	źǵ	ćć	ǵǵ
ǵ	pć	bǵ	tć	dǵ	cć	gǵ	mć	nć	nć	ść	źǵ	ść	źǵ	ćć	ǵǵ

These changes can occur at word boundaries, also, if a final epenthetic vowel is removed or does not occur. The difference however is that a final voiced consonant will become voiceless before an initial voiceless consonant. The only permittable sound changes at word boundaries are those in which the initial consonant remains intact.

In consonant combinations of -r+dental- the dental consonant is made retroflex while the -r- disappears. If an “r” occurs before a retroflex consonant it also becomes retroflex while remaining intact. This feature is also expanded to suffixes of this structure and is quite common, almost obligatory for most people. There are also a set of sound changes which occur with retroflex consonants and plain alveolars. When a retroflex precedes and alveolar, the alvelor becomes retroflex and the same pattern of changes occurs as with plain alveolar combinations, e.g. t.+n = t.+n. = n.n.. The same patterns also occur in retroflex-retroflex combinations, e.g. t.+z. = t.s.. In these changes, a retroflex "r" is preserved, as no process of retroflex creation can occur. The processes are more complex, however, when alveolar consonants precede retroflex ones. The retroflex consonants tend to become the "palatal" equivalent or the alveolar one plus the sound "i".

	First sound
Second Sound	ț	ḑ	ņ	ŗ	ļ	ș	ẓ
t	țț	ḑḑ	ņņ	ŗț	ļț	șț	ẓḑ
d	țț	ḑḑ	ņț	ŗț	ļț	șț	ẓḑ
n	ņņ	ḑḑ	ņņ	ŗņ	ļņ	șņ	ẓņ
r	țŗ	ḑŗ	ņŗ	ŗŗ	ļŗ	șŗ	ẓŗ
l	țļ	ḑļ	ņļ	ŗļ	ļļ	șļ	ẓļ
s	țș	ḑẓ	ņș	ŗș	ļș	șș	țș
z	țș	ḑẓ	ņș	ŗș	ļș	ș	ẓẓ
ś	țș	ḑẓ	ņș	ŗș	ļș	șș	țș
ź	țș	ḑẓ	ņș	ŗș	ļș	ș	ẓẓ
ć	țț	ḑḑ	ņņ	ŗț	ļț	șț	ẓḑ
ǵ	țț	ḑḑ	ņț	ŗț	ļț	șț	ẓḑ

	First Sound
Second Sound	t	d	n	r	l	s	z	ś	ź	ć	ǵ
ț	tć	dǵ	nć	ŗț	lć	sć	zǵ	ść	źǵ	ćć	ǵǵ
ḑ	tć	dǵ	nć	ŗț	lć	sć	zǵ	ść	źǵ	ćć	ǵǵ
ņ	nni	nni	nni	ŗņ	lni	sni	sni	śni	śni	ćni	ćni
ŗ	țŗ	ḑŗ	n.ŗ	ŗŗ	ļŗ	șŗ	ẓŗ	śri	źri	ćri	ǵri
ļ	łi	dłi	lni	ŗļ	lli	sli	sli	śli	śli	ćli	ćli
ș	tś	tś	nś	rś	lś	sś	sś	śś	śś	ćś	ćś
ẓ	tś	tś	nś	rś	lś	sś	sś	śś	śś	ćś	ćś

A note on combinations with voiceless nasals. If a voiceless nasal occurs before a voiced consonant then the voiced nasal+voiced consonant cluster becomes devoiced. A voiceless nasal following voiced consonant forms the usual voiced+nasal cluster. A voiceless nasal before a voicless consonant forms the normal cluster, but with voiceless nasals in place of the usual voiced ones and finally, a voicless consonant followed by a voiceless nasal will form the usual cluster unless it would normally form a nasal+nasal cluster, in which case both these nasals are devoiced.

Epenthesis

Epenthesis: When a vowel follows another vowel and does not form an allowable diphthong then a consonant is inserted between them. A "j" precedes front vowels and "w" precedes back vowels". When a consonant follows another and cannot be put together in the above to ways, then a vowel is inserted between them. A “e” is placed between like consonants, and “i” is placed after labials, an “a” is placed after dentals and an “e” is placed after velars. The sounds between vowels, however, are not written down, they ar merely pronounced as the speaker sees fit.

Examples of Epethesis:

First Second	e	i, a	V	C
V	Vje	Vja/i	Vj/wV	VC
C	Ce	Ci/a	CV	CeC
P	Pe	Pi/a	PV	PiC
T	Te	Ti/a	TV	TaC
K	Ke	Ki/a	KV	KeC