Palatals and labial velars

The distribution of palatals and labial velars in JamE is clearly influenced by the JamC-to-JamE conversion processes which many speakers carry out. One problem converting JamC lexical inputs into an acceptable JamE realization is the fact that JamC /a/ may be realized as JamE /a/ or /ɔ/ , depending on the lexical item. There is no way, taking the JamC phonological form, /pat/, of knowing whether the JamE form should be /pat/ ‘pat’ or /pɔt/ ‘pot’. However, when JamC /a/ is part of a syllable with a palatal or labial velar stop onset, these invariably predict the correct JamE output.

Let us first take the palatals. In JamE, /kj/ and /gj/, phonetically palatal stops, [c] and [Ɉ] , have a distribution in which they vary with each other before /a/ and /aa/ but not in other environments. Thus, the item ‘cap’ has two realizations in JamE, /kap/ and /kjap/, whereas the items ‘coo’ /kuu/ and ‘queue’, /kjuu/ show a /k/ versus /kj/ phonemic contrast. The JamE /kap/ ~ /kjap/ ‘cap’ variation reflects the fact that /kj/ is part of the lexical specification of cognate items in JamC, serving to distinguish it from /kap/ ‘cop’. With the JamE pronunciation of ‘cop’ being /kɔp/, the use of /kj/ in /kjap/ has no distinctive functional value. It, however, represents a carry-over from JamC which, we argue, provides the lexical input that lies at the base of JamE phonetic output.

In the examples below, the item with /kj/ or /gj/ in the JamC item has /kj/ or /gj/ as variant forms in JamE, followed by /a/. The items which have /k/ or /g/ in the JamC item, require an invariant /k/ or /g/ in the JamE cognate and /ɔ/ as the following vowel. The weight of the phonemic distinction, transferred from the consonant in JamC to the vowel in JamE, is still expressed redundantly in the form of a residual /kj/ variant in JamE.

A very similar kind of situation applies with the labial velars, where again the presence of a semi-vowel linked feature predicts whether JamC /a/ is realized as JamE /a/ or /ɔ/ . The difference is that there are environments in which palatals occur categorically, i.e. before vowels other than /a/ and /aa/. By contrast, labial velars only occur variably in JamE, before the diphthong /ɔi/. Its JamC reflex, /ai/, is the only environment in which they may occur in JamC. In JamE, it represents a redundant feature, the labialization of /b/ in the environment of an /ai/ which has /ɔi/ as its JamE reflex. This represents independent support for the notion that the conversion process is from a JamC underlying input to JamE and not the other way around. Otherwise, we would have no way of understanding how a variable occurrence of /w/ in JamE can be converted into a categorical appearance of this form in the JamC cognates.