| 2.2.
PHONOTACTICS |
| Hill
defines the term phonotactics as `the area of phonemics which covers
the structural characteristics of sequences’ (Hill : 1958 : 68). To
him phonotactics is essentially a description of the distribution
of the phonemes after their identification. The sequences of phonemes
can be described in terms of their clustering habits. A cluster is
a sequence of two or more phonemes of the same class without the intervention
of a phoneme of another class. As far as the Sema language is concerned,
clusters of both vowels and of `consonants are possible. At the phonemic
level, Sema language has tones (a supra-segmental feature) but this
supra-segmental feature cannot occur without simultaneous segmental
elements. That is, they always co-occur with segmental phonemes. In
addition, there isn’t any special limitations in the occurrence of
different levels of tones. All types of permutations and commutations
of different levels of tones are possible, though a noun beginning
with the vowels i/o/a/ (pronominal clinics) will always have level
tone. A description of the clustering pattern in Sema follows. |
| A
description of the clustering habits of the consonants in Sema may
begin with the initial clusters of consonants. In this the largest
number that can occur is three. A feature to be noted about the clusters
of C1 C2 C3 consonants is that in all instances of the occurrence
of C1 C2 C3 at the word initial position, h occupies either C2 or
C3 position. When h occupies the C3 position, C1 is invariably a bilabial
nasal and the C2 is either a voiceless stop or a lateral. In other
words, in a C1 C2 C3 sequence, when h occupies the C3 position, no
other consonant exception m and a voiceless stop or a lateral can
occur respectively in C1 and C2 positions. Similarly when h occupies
the C2 position C1 C3 positions will invariably have respectively
a voiceless stop and a palatal approximant. That is, in a C1 C2 C3
sequence if h occupies the C2 position, no consonant expecting a voiceless
stop and a palatal approximant can occur respectively in C1 C3 positions.
At the word-initial position, a sequence of either C1 C1 C2 or C1
C2 C2 cannot occur in this language. Given below are a few words illustrating
the occurrence at the word-initial position of the sequences of C1
C2 C3. |
| 1.
m + (p, t, k, l,) +h |
| mph |
mphi |
`not
yet’ |
| mth |
mtha |
`deny’ |
| mkh |
 mkhemu |
`at
least’ |
| mlh |
mlhu |
`below’ |
|
|
| 2. (p.,
k) + hy |
| phy |
phyu |
`bottom
of a trunk’ |
| khy |
khyumu |
`nothing’ |
|
|
| Of
the consonant clusters occurring at the word-initial position the
number of C1 C2 consonant clusters are much higher than C1 C2 C3 consonant
clusters. There are, however, some similarities in both these types
of clusters, i.e., in a C1 C2 C3 cluster a number of consonants were
found to occur in C2 position with m in C1 position. And all the consonants
in a C1 C2 C3 clusters occurring with m in C1 position also occur
in C2 position of C1 C2 C3 cluster having m in C1 position. The C1
C2 C3 clusters having m in C1 position had h in C3 position and a
voiceless stop or a lateral occurred in C2 position. In C1 C2 clusters
also h occurred in C1 C2 C3 cluster. In addition it occurs with some
other consonants. The occurrence of C1 C2 clusters are sub-grouped
into four viz., (1) those having m in C1 position, (2) those having
h in C2 position, (3) those having l in C2 position, and (4) miscellaneous
ones. Given below are a few words for illustrating the occurrence
of C1 C2 clusters. |
| 1.
m + (p, t, k, b, c, s, h, z, n, l) |
| mp- |
mpe |
`year’ |
| mt- |
mtįkipi |
`moderate’ |
| mk- |
mkas |
`ebony’ |
| mb- |
mboms |
`fool’ |
| mc- |
mcassi |
`level’ |
| ms- |
msa |
`fear’ |
| mh- |
mhó |
`dull
weather’ |
| mz- |
mza |
`graze
cattle (v)’ |
| mn- |
mna |
`stink
(natural smell) |
| ml- |
mloba |
`peas’ |
|
|
| 2.
(p, t, k, q, c, m, n, l) + h |
| ph- |
phi |
`read’ |
| th- |
thi |
`die’ |
| kh- |
khi |
`make’ |
| qh- |
qhi |
`keep’ |
| ch- |
chi |
`cut’ |
| mh- |
mhi |
`clouded
sky’ |
| nh- |
nhi |
`marry
(girl)’ |
| lh- |
lhe |
`relate’ |
|
|
|
