From Jamaican Creole to Jamaican English: The vowel system

The only difference between the vowel inventories of the two language varieties involves the vowel /ɔ/ which exists in JamE but not in JamC. There is, therefore, for most vowels, a one-to-one relationship between JamC and JamE variants in cognate lexical items. However, there are three JamC vowels for which there are two possible JamE reflexes. These all involve the JamE vowel /ɔ/ , once as a simple vowel and twice as part of the complex vowels, /ɔɔ/ and /ɔi/ . Below are presented the vowel variants or reflexes across the two language varieties.

We argue that JamE phonological outputs are based on JamC lexical specifications modified by established conversion rules. These rules, we propose, are based on stereotypical notions of the difference between the phonetic outputs of lexical entries in JamC versus the phonetic outputs of their cognates in JamE. The level of success achieved by speakers operating these rules firstly depends on whether the correspondences between JamC and JamE are one-to-one or one-to-many. In the cases of JamC /ia/ > JamE /ee/ and JamC /ua/ > /oo/, we are dealing with one-to-one correspondences. The application of the conversion rule is, therefore, straightforward. The problem is less a linguistic one than a psychological one. With what consistency are speakers actually able to apply these conversion rules? Bilingual speakers will look for ways to keep the language varieties apart while minimizing the effort they put into doing so, giving rise to what we have called differential convergence between the varieties. We have already seen the evidence which suggests that speakers, in their use of JamE, employ more consistently the JamE variant, [oo], in the /ua/ ~ /oo/ variable than they do the JamE variant, [iε] in the /ia/ > /ee/ one. Here, speakers economize on their efforts to keep JamC and JamE apart, by avoiding JamC features more consistently in the former variable than in the latter. As we have already seen, also, this economy of effort may be most active in the environment immediately preceding /r/.

Where two possible JamE reflexes exist for one JamC vowel, matters are more complex. Usually, one JamE reflex is identical phonetically to that in JamC. The other one, however, represents a phonetic form which does not exist in JamC. For any item, the JamE cognate might have a phonetic output identical to its JamC equivalent. On the other hand, the JamE cognate may take the phonetic form that does not exist in JamC. It is the second possibility which is most likely to attract the attention of a speaker relatively unfamiliar with JamE. This produces naïve conversions. Thus, in the variables involving JamE /a/ and /ɔ/ respectively, a naïve conversion would change all the JamC occurrences of /a/, /aa/ and /ai/ to JamE /ɔ/ , /ɔɔ/ and /ɔi/. This approach presumes a one-to-one correspondence with JamC /a/ > JamE /ɔ/ and retains a feature characteristic of JamC:

(i) the vowels of ‘tap’ and ‘top’ not distinguished, here realized as /tɔp/,

(ii) the vowels of ‘mass’ and ‘moss’ not distinguished, both realized as /mɔɔs/, and

(iii) the vowels of ‘tile’ and ‘toil’ not distinguished, both realized as /tɔil/. This is typically discussed in the literature as hypercorrection and is one of the shibboleths of the speech community. It marks the speaker off as uneducated and unaware that the JamC > JamE conversion involves, based on lexical specification, either the form /a/, approximating phonetically to its JamC equivalent, or the form /ɔ/. For many speakers, the lexical marking is done using as a reference the way the words are spelt in English orthography.