Sangi/Consonant Mutations

=Sangi Consonant Mutations=

Sangi has two kinds of consonant mutation. The first is "phonological mutation", a process which occurs at morpheme boundaries, affecting only the phonology of a word and not its meaning. The second is "morphological mutation", which changes the meaning of the word as well as its phonology.

I-Affection
I-affection occurs when the last consonant of the stem is immediately followed by the vowels i and e (long, short and diphthong forms) and also the consonant j.

I-Affection does not affect geminate l, though.

The "pf" form occurs with a singular "p" mutating, "ff" with "pp".

A-Affection
Like I-affection, A-affection occurs when the last consonant of the stem is immediately followed by a vowel, in this case a.

A slight irregularity in this system occurs as “w”, “h” and the velar stops are affected in the above way by a following “o”. A-Affection doesn’t affect consonant clusters, unlike I-Affection, except when the final consonant is an h, in which case the h, unexpectedly, becomes an the a-mutated form of the preceding consonant, or the final consonant of this if it is a cluster itself, e.g. nh>nd, rh>rź, etc.

Plural Mutation
This mutation occurs when forming the plural of the four noun stems (see below).

When the plural stem is weakened or strengthened then the retroflex form of the plural is used as the stem consonant. The plural forms of retroflex consonants are the retroflex equivalents of the plural forms of the alveolar counterparts with 2 exceptions. Retroflex plural forms of alveolar stems are preceded with a retroflex "r" when the stem is retroflex and the plural of retroflex "-tt" is retroflex "d".

Stem Gradation
Stem gradation is a weakening of the final consonant of the stem to form different noun and verb stems which are used to built up the meaning of the noun or verb as well as to derive words from different classes of words.

pp > p > b > v > f > w > h > hi > he

tt > t > d > j > i > e*

cc > c > g > j > i > e

m > mb > mm > mp > mi > me

m̌ > mp > m̌m̌ > mf > m̌i > m̌e

n > nd > nn > ns > nt > ni > ne*

ň > ŋ > ng > nc > nct > ŋi > ŋe

ŋ̌ > nc > ŋ̌i > ŋ̌e

l > ld > ll > ls > lt > li > le*

ł > łd > łł > łs > łt > łi > łe

r > rd > rr > rs > rt > ri > re*

ř > řd > řř > řs > řt > ři > ře*

s > t > d > j > i > e

sp > sw > sl > śi > śe

st > ss > ś > si > se

sc > ś > si > se

h > p > b > w > hi > he

Ø > c > g > j > i > e (if the Ø was originally a c)

Ø > t > d > j > i > e (if the Ø was originally a t)*


 * These gradation series can also be applied to retroflex consonants. So if the sequence was originally VVt# then the t is reinstated as retroflex, removing the ambiguity caused by the process VVt#>V#. The “j” however is pronounced as “hj” and written “ĵ”, while the “i” and the “e” are pronounced as normal.

It should be noted that “s”, “tt”, “cc” and “Ø” all eventually weaken to the same pattern, causing a certain amount of ambiguity if the gradation is carried through enough times. Also, grades of the pattern “Ci” and “Ce” can never be the general stem, being analysed as “C-i/e” where C is graded, but “i” and “e” can be strengthened.