• Tidak ada hasil yang ditemukan

Nasal harmony pattern in disyllabic words in underived word domain

Chapter 6 Nasal Harmony in Deori

6.3 Nasal harmony pattern in Deori

6.3.1 Nasal harmony pattern in disyllabic words in underived word domain

In this section, the target and the opaque segment in monomorphemic disyllabic words are discussed. However, the occurrence of glide [w] and glottal fricative /ɦ/ in the span of nasal harmony is not discussed in this section for the following reasons: Glide [w] occurs only as a suffix (for example: w “thematic marker”), hence, its role as a target segment is discussed in the derived domain in section 6.3.2. Glottal fricative /ɦ/ occurs word-initially in root words; hence, its role in the context of nasal harmony in monomorphemic disyllabic roots is irrelevant.

However, its role in the span of nasal harmony in the derived domain is illustrated in section 6.3.2 considering the locative suffix marker /ɦɔ/.

Vowels are the triggering segment in Deori which affects glides, and liquids as can be seen in the following examples.

(76) Liquid [ɹ] and glide /j/ are target segments in Deori.

Liquid [ɹ] Glide /j/

a. ɡɑ ɹ ɔ ‘pot’. d. tʃ ɑ ‘fish/wife of younger brother’

b. ɡ ɹ ‘second month of the e. ɑ ɑ ‘daughter in law’

Assamese calendar’

c. dzuɹɔ ‘summer’ f. ɡijɑ ‘planter’

In (76) (a-f) the final vowel is nasalized as a process of nasal effacement as the underlying representation of the words as mentioned in Jacquesson (2005) and Deori (2008) are: <g ro g>

~ /ɡ ɹɔŋ/ → ɡ ɹ ɔ “pot” Jacquesson, 2005; Deori, 2008), <g ɹing> ~/ɡ ɹiŋ/ ɡ ɹ “second month of the Assamese calendar” (Jacquesson, 2005; Deori, 2008), <dzuɹɔng> ~ /dzuɹɔŋ/ dzuɹɔ “summer” Jacquesson, 2005; Deori, 2008), <tʃ g> ~ /tʃ ŋ/ → tʃ “fish/wife of younger brother” Jacquesson, 2005; Deori, 2008), < g> ~ / ŋ/ → “daughter in law”

(Jacquesson, 2005; Deori, 2008) and <gij g> ~ /ɡi ŋ/ ɡi “planter” Jacquesson, 2005;

Deori, 2008) and the initial vowel is assumed to be underlyingly nasalized. Considering examples (76) (a, b, d, and e) it can be speculated that nasalization of the final nasal vowels in Deori spread leftward thereby nasalizing the sonorant segment occurring to the left of the nasal vowel. However, this generalization would result in wrongly interpreted data as in (76) (c)

*d u ɹ ɔ and (76) (f) *ɡ , which is an unattested form in Deori. Thus, we assume that nasalization in (76) (a, b, d, e) spreads from the initial vowel to the adjacent sonorant segment thereby nasalizing liquid [ɹ] and glide /j/ and the final vowel is nasalized through the process of nasal effacement. Effacement is diachronic in Deori, not synchronic. Nasal harmony in Deori is a synchronic process and the deletion of final nasals may have either been (a) the impulse for nasal harmony or (b) led to the presence of nasalized vowel and consonants or (c) both. We do not have any evidence that it is a process in the synchronic grammar and no such claim has been made. An alternative role of liquid [ɹ] is evident in Deori in the derived domain. While in (76) (a- b), liquid [ɹ] undergoes nasalization and changes to nasalized liquid [ɹ ], in derived domain liquid [ɹ] undergoes nasalization and changes to nasal consonant /n/. This instance of the alternating segment in the span of nasal harmony will be discussed in detail in section 6.3.2.1. Liquid /l/ in Deori is restricted to word-initial position, and the only occurrence of liquid /l/ in the span of nasal harmony is lepeduɹ l e peduɹ “my goat”, li l “2nd person singular/plural.necklace”. ere, oral /l/ changes to nasalized /l / when preceded by a nasal vowel.

In lepeduɹ l e pduɹ “my goat”, nasalization from 1st person singular pronoun spreads to liquid /l/ and then to the adjacent vowel /ɛ/, but the intervening voiceless stop /p/ stops nasalization from propagating further, which shows voiceless stop /p/ is opaque to nasal harmony in Deori.

Apart from vowels, nasal consonants also trigger nasal harmony in Deori as discussed below:

(77) Nasal consonants triggering nasal harmony a. m ɹ ‘uncooked rice’

b. n ‘cook’

c. mɔ kɔ ‘rice’

d. m sɑ ‘grass, weed’

e. mɛ bɑ ‘fat’

f. mɔ si ‘man’

g. m tʃɔ ‘platform of the house’

h. timu ‘mango’

i. tʃim ‘tail’

Example (77) shows that the nasals /m/ or /n/ trigger nasalization to the adjacent sonorant segment. In (77) (a) ɹ “uncooked rice” and in (77) (b) “cook”, we assume that the initial vowel is contextually nasalized (in close proximity to a nasal consonant) and the final vowel is nasalized through the process of nasal effacement as the underlying representation of the words are < ɹung> ~ / ɹuŋ/ ɹ (Jacquesson, 2005) and < g> ~ / ŋ/ → (Deori, 2008) respectively. Thus, nasalization from the nasal consonants spread to the adjacent vowel which in turn spreads to the adjacent sonorant segment and nasalizes the [+continuant] [ɹ] and /j/.

Similarly, in (77) (c) mɔ kɔ “rice” and (77) (d) s “grass, weed”, we assume that the initial nasal consonant triggers nasalization on the adjacent vowel, and the nasalization of the vowel in the final syllable is derived through the process of nasal effacement as the underlying representation of the words are <mɔ kɔng> ~ /mɔ kɔŋ/ → mɔ kɔ (Brown, 1895) and < s ŋ ~ / u s ŋ/ u s (Jacquesson, 2005). Furthermore, in (77) (c, d) the intervening [-voice] stop /k/

and [-voice] fricative /s/ are not affected by nasalization. Example (77) (e-g) mɛ ba “fat”, mɔ si

“man” and tʃɔ “platform of the house”, shows that the initial vowel is contextually nasalized and nasalization is prohibited from spreading further by an intervening [+voice] obstruent /b/, [- voice] fricative /s/, and [-voice] affricate /tʃ/ respectively. Hence, the lexical sets mɛ ba, mɔ si, and tʃɔ, though minimally attested, show that nasalization does not spread through [+voice]

obstruent stop /b/, [-voice] fricative /s/, and [-voice] affricate /tʃ/ in Deori. Words such as (77)(h) ti “mango” and (77)(i) tʃi “tail” show that while nasal consonant triggers nasalization on the adjacent vowel (final vowel), the initial vowel remains oral and is not affected by the following nasal consonant. It exemplifies that oral vowels can occur to the left of the nasal consonant but not to the right of the nasal consonant. Although the examples above suggest that nasalization does not spread leftwards in Deori, directionality in Deori will be clear in derived word sequences which will be discussed in section 6.3.2. Next, we discuss words with a final nasalized vowel, unlike example (76) and (77) (a-g).

(78) [+voice] obstruent stops /b/ /d/ /ɡ/ [-voice] obstruent stops /p/ /t/ /k/

a. bibɔ ‘granary’ h. op ‘arm’

b. ibɑ ‘flower’ i. ditɔ ‘pot’

c. tɑd ‘spoon’ j. ʧit ‘rope’

d. ɡɑd ‘pillow’ k. ɑtɑ ‘arrow’

e. ɑɡ ‘knee’ l. ɑk ‘ear/upland’

f. diɡ ‘thread’ m. ʧikԑ ‘rat’

g. uɡɑ ‘field’ n. ek ‘smoke’

(79) Fricative /s/

a. ɑsɔ ‘bamboo cleft for mat’

b. disɔ ‘pot’

c. ɡisɔ ‘lower rack above fireplace’

d. isɑ ‘shawl’

e. us ‘bracelet’

(80) Affricate /ʧ/ Affricate /dz/

a. ɑʧɔ ‘house’ d. udz ‘navel/bamboo tube’

b. ʧɔʧɛ ‘shame’ e. sudzɛ ‘rice beer’

c. m tʃɔ ‘platform of the house’ f. ɡudz ‘spear’

g. pidzɔ ‘raw’

In the examples (78-80), it can be seen that the final vowels are nasalized and the initial vowels are oral. These examples show that nasalization does not spread leftwards in Deori as it is arrested from spreading regressively by [±voice] obstruent stops, [-voice] fricative, and [±voice]

affricates, which are less compatible with nasalization. Had it been not blocked by the [±voice]

obstruent segments, we would have wrongly interpreted data as * ɔ “granary”, *e k “smoke”,

*u ɡ “field”, * s “shawl”, * ɔ ɛ “shame”, *u d “navel/bamboo tube” which are not attested in the language.

After a detailed examination of the underived words, a nasal harmony pattern in derived words is examined in the following section. Derived words are good test words for the directionality of nasal harmony in Deori.