Sound shifts in Caribbean Creoles

Next we reconsider the historical sound changes which led to the relative positions of the vowels in Jamaican English Creole, which are quite different from the positions of corresponding vowels in the other English dialects studied in the present work.

JC /e:, o:/ (as in FACE, GOAT) historically descend from Middle English (14th century) /a:, :/, which raised to Early Modern English /e:/ (circa 1650, merging with ME /:/, according to Prins, 1972:122) and /o:/ (perhaps a century earlier). This raising of ME long low vowels /a:, :/ which occurred as a part of the Great Vowel Shift, was followed upon in other English dialects by diphthongization to [e] and to [o ~]. This diphthongization appears to postdate the establishment of Jamaican Creole: Prins states that ``the Mo[dern]E[nglish] diphthongs [ei] and [ou] ...are late (end 18th c.).''(p. 124) In the antecedents of present-day Jamaican Creole, this process apparently did not occur. Rather, the subsequent step was a further raising of the nuclei of these vowels, so that they are now transcribed with /ie, uo/. While this raising is not strictly a part of the Great Vowel Shift, it can be seen as a continuation of the raising of these vowels which occurred in the GVS.^B.1

Where did the glide in /ie, uo/ come from? Why did this change go to this extreme in JC. Notice that the raising of short-A (Wells' lexical set, TRAP) in American English dialects as far as [i] is an example of the same raising process. It is clear that this identical vowel shift has occurred again and again in the history of English; presumably the same linguistic forces operated this century to raise ``short A'' in American dialects that operated in earlier JC to raise ``long A'' in a similar way.

One scenario to explain the further raising of /e:, o:/ relates this change to another special feature of Jamaican phonology. The sounds here labelled /ou, ai/ (as in MOUTH, PRICE) in Jamaican are the modern reflexes of Middle English /i:, u:/, which became the diphthongs /ay, aw/ of Reference American, via the Great Vowel Shift. In Jamaican, as in certain Northern dialects in Britain, the Great Vowel Shift has not gone to completion: the formerly high-back vowel in MOUTH has not shifted all the way down to low position. Certainly its nucleus is not to be identified with that of /ai/, as may be reasonably done in RA, as shown in Figure . Thus, MOUTH is [moUt].

The phonemes /ou/ and /o:/ (MOUTH and GOAT) would be quite difficult to distinguish without a phonetic difference in their respective nuclei: [o:] and [oU] are not phonetically very far apart. The quality of the diphthong /ou/ is mid-back and sometimes monophthongal, thereby putting functional pressure on the back phoneme /o:/, which is historically mid, back, and either monophthongal or up-gliding -- that is, very much like /ou/. /o:/ might therefore have become differentiated from /ou/ by diphthongizing and raising to [uo]. The raising of /e:/ to [ie] could then appear as a parallel shift, a symmetric raising of a long mid vowel to a high, lowering diphthong.

It is useful to consider in this connection the range of realizations of the three relevant lexical sets in several varieties of Caribbean English. Wells presents phonological and phonetic summaries of several of these creoles, from which information in the following table was taken.

	FACE	GOAT	MOUTH
Jamaica	[ie]	[uo]	[u]
Montserrat	[ie]	[uo]	[ou]
Trinidad	[e]	[o]	[u]
Guyana	[e:]	[o:]	[u]
Barbados	[e:]	[o:]	[u]
Bahamas	[e:]	[o:]	[U]

The sound written [ou] is not low or lower-mid, but mid or high-mid. In the above-described process, the falling of ME /u:/ (MOUTH) to [ou] is associated with the raising of /ow/(GOAT) to [uo], and the parallel raising of /e:/ (FACE) to /ie/. The only Caribbean creole dialect discussed by Wells with the nucleus of ME /u:/ in mid-back position (as [ou]) is that of Montserrat, one of the Windward Islands. The striking coincidence here is that Montserrat is also the only other dialect with raising and breaking of /e:, o:/. This is evidence that the mid-back nucleus of the reflex of ME /u:/ and the raising and breaking of /e:, o:/ are systematically associated. The generalization here is functional:

Where ME /u:/ (MOUTH) becomes [ou], encroaching on /o:/, /e:, o:/ raise and break.

This generalization supports the interpretation above of a push chain shift where /ou/ encroached on the phonetic space of /o:/, which then diphthongized and raised to high-back [uo], while /e:/ diphthongized and raised in a parallel and symmetric shift. If on the other hand the raising of /e:, o:/ to [ie, uo] occurred before ME /u:/ became [o] in Montserrat and Jamaica (that is, if the changes occurred as a drag chain shift rather than a push chain shift) then there would be no explanation for the raising of /e:, o:/ to [ie, uo]. The push chain scenario, on the other hand, gives a plausible explanation for the raising of /e:, o:/, namely the functional pressure of the phonetic approximation of /ou/ with /o:/.