5.

The Historical Relation of American English to British English

5.1 Regional Words

A considerable body of lexical regionalisms in America, brought to light by the Linguistic Atlas survey, is presented and discussed by H. Kurath in A Word Geography of the Eastern United States [1949]. Commenting on “ranges of the vocabulary,” the following observations are made [pp. 9-10]:

If this point of view is adopted, we may expect that many of the American regionalisms within this wide semantic range are derived from British regional dialects.

We are relatively well prepared for the task of tracing some American regionalisms to specific sections of the British Isles, and projects now underway will considerably better our chances. For England, J. Wright’s English Dialect Dictionary [1898-1905] provides a large body of roughly localized data, and H. Orton’s Survey of English Dialects [1962-71] is gradually furnishing us with a wealth of evidence recorded directly in the countryside. For North America, the collections of the Linguistic Atlas of the Eastern States present ample localized materials for a limited vocabulary; and the Dictionary of American Regional English, now in preparation under the direction of F. G. Cassidy, will vastly increase our resources.

A few examples chosen from the Word Geography of the Eastern United States will serve to illustrate some of the sources.

The evidence presented by J. Wright makes it clear that cade, cade lamb “pet lamb,” current in southeastern New England [Kurath 1949: Fig. 12], was brought to this country from the Midland counties of England and became established in the Plymouth and the Rhode Island colonies. Orton’s survey does not report a single instance of this expression in the north and the south of England.

The word hap “quilt,” still in common use in the mountains of central Pennsylvania [ibid.: Figs. 25, 80], was brought to America by the Ulster Scots. J. Wright [1905] reports this term only for Scotland and the northern counties of England. The fact that Orton’s survey of half a century later no longer found any trace of hap in northern England does not invalidate this inference, since we must reckon with the recession of regional folk words.

For the noise made by the horse, three different words are current in the Eastern States: (1) whinny ~ whinner in the North and the North Midland, (2) nicker in the Upper South and the South Midland (Virginia and the Appalachians), and (3) whicker in the Carolinas and along Chesapeake Bay [ibid.: Fig. 97].

According to Wright and Orton, whicker is the usual term in the southwestern counties of England. It is the obvious source of the Carolina variant.

The Virginian nicker may have been imported from the southeastern counties of England, although Orton and Wright report it also from Cumberland and Durham.

For whinny, now regarded as the standard or literary term, the regional dissemination in British folk speech is as yet only partially known. According to Orton, it is in regular use in all of the northern counties. Its currency in literary English suggests that it is widely used in the Midlands, where neither whicker nor nicker are reported by Wright. Thus the extensive currency of whinny in the American North and North Midland, though supported by literary English, has its primary source in a regional dialect of England.

The “hand-picked” examples presented above are not intended to suggest that the relationship between American and British regionalism are often that clear and simple. In the course of 250- 300 years many recessions and expansions have taken place on both sides of the Atlantic. As far as America is concerned, we must keep in mind that at the time of settlement usage varied in each of the colonies and that the relative uniformity within the several dialect areas, as in the case of whinny ≠ nicker ≠ whicker, developed several generations after the settlement.

In passing it seems worth pointing out that such regional expressions as cade, hap, and nicker can, with proper precaution, function as “tracers” of population movements or migration routes.

5.2 Phonological Features

5.21 The System of Syllabic Phonemes

Any meaningful discussion of the relationship of a derivative dialect to its parent dialects in phonological matters presupposes a phonemic analysis. Not until the phonemic systems of both dialects are set up, is it possible to distinguish between shared and divergent structural features, to describe differences in the phonic realization of shared phonemes, and to point out divergences in the lexical (etymological) incidence of shared phonemes.

In this discussion of the historical relationships of American English to its British sources I shall rely upon the analysis of the system of stressed vowels presented in H. Kurath and R. I. McDavid, Jr., The Pronunciation of English in the Atlantic States [1961]. For the justification of this analysis see especially pp. 3-9. Its usefulness as a frame of reference has been demonstrated in H. Kurath, A Phonology and Prosody of Modern English [1964].

According to this analysis, the stressed vowels of American English and of Standard British English fall into two classes: checked and free. The checked vowels are restricted to preconsonantal position, whereas the free vowels are unrestricted in their distribution.

The three checked front vowels /1 ≠ ε ≠ æ /, as in bit ≠ bet ≠ bat, bear the same relation to each other in Standard British English (SBE) and in all varieties of American English (AE). Except for /3e /, they differ little in lexical incidence.

The three checked back vowels, on the other hand, differ markedly in lexical incidence: (1) SBE has // ≠ ɒ ≠ a / in foot ≠ hot, fog, loss ≠ hut; most varieties of AE have ///≠ ɑ ≠ /in foot ≠ hut ≠ hot; (3) eastern New England has /// ≠ ɵ ≠ Λ /in foot ≠ boat, stone ≠ hut in old-fashioned speech; but checked / ɵ / has been largely replaced by free / o /.

The subsystem of free vowels presented below is common to SBE and to those dialects of AE in which postvocalic / r / is not preserved as such. Other dialects of American English differ from this system only in that they lack the // of bar.

The lexical (etymological) incidence of the free vowels differs little from dialect to dialect. On the other hand, their phonic character varies markedly. Thus the free vowels /i, u, e, o / of bee, do, bay, no are upgliding diphthongs in SBE and in some regional dialects of AE, while in other dialects of AE they are monophthongal and even ingliding. The / / of bar is low central in SBE and in Metropolitan New York, low front in eastern New England, low back in eastern Virginia. The / ɔ / of law is well rounded in SBE and in Metropolitan New York, less so in eastern New England, western Pennsylvania, and eastern Virginia. The / au / of now ranges all the way from [au] to [æu] and [əu] in AE. The / 3 / of fur is unconstricted in SBE and in most AE dialects spoken along the Atlantic and the Gulf of Mexico, and constricted elsewhere in America.

The phonic variants of shared phonemes in AE are displayed in a series of “synopses” in Kurath-McDavid 1961: 31-100.

The essential agreement of the vowel system of AE with that of SBE may be attributed to three distinct factors.

(1) When the American colonies were planted, SBE was already highly standardized in its phonemic system. Hence the leading families in the several colonies must have used essentially the same vowel system.

(2) The speech of the leading families became the model for the unlettered majority of settlers, who, coming from different parts of the British Isles, spoke a great variety of folk dialects, some of which differed greatly from SBE in the system of syllabics, in the phonic realization of shared phonemes, and in their lexical incidence. Generally speaking, this group of settlers, or their descendants, must have adopted the vowel system of SBE, the prestige dialect of their community, but preserved some of their subphonemic peculiarities.

(3) The leading families in the colonial seaports—officials, merchants, financiers, religious leaders, and intellectuals—maintained close ties with the mother country throughout the colonial period and followed English fashions in speech as well as in other cultural matters. Thus eighteenth century changes of SBE found their way into the colonies on the Atlantic seaboard, notably postvocalic / / from earlier / r /, as in fear, fair, four, poor, and the free vowel / σ / of far, hard. These new prestige features spread inland from the seaports, as they still do to some extent, but did not reach the back country.

Only some of the more striking phenomena can be briefly dealt with here to throw light on the complicated task of accounting for the pervasive influence of SBE in shaping the regional varieties of AE and the survival of certain aspects of regional British folk speech.

We shall first consider some phonemic mergers and splits that had taken place in SBE by 1600—i.e., before the first settlements were established in America, but not in certain regional dialects of England. By doing this we shall see that all varieties of AE largely conform to the phonemic pattern of SBE.

5.22 Mergers and Splits

(1) In SBE and in AE, ME close /ẹ̄/ and open /ę̄/ are merged in / i /, as in beet = beat. Some English dialects show the same development, others not. Thus Suffolk, Lincolnshire, and Leistershire have the merger; but Yorkshire, Lancashire, and Gloucestershire preserve the contrast as [i:] ≠ [ia ~ ea].

(2) In SBE and in AE, ME open /ę̄/ and / ā / remain distinct, as in wheat ≠ late. This contrast is also preserved in Lincoln, Leicester, and Lancashire. But in Suffolk, Yorkshire, and Gloucester they are merged in /eə / or / iə /.

(3) In SBE and in AE, ME / ā / and / ai / are merged in /e /, as in tale = tail. Merging is found also in the dialects of Lincoln and Leicester. But the contrast is preserved in large parts of England: as /eə ≠ æj / in Suffolk and Gloucester, as / iə ≠ aj ~ εj / in York-shire, and as /e: ≠ ej ~ aj / in Lancashire.

(4) In SBE and in AE, ME /ǭ/ and /ou / are merged in /o /, as stone grown. Merging occurs in Suffolk and in Leicester, but in most of the English dialects the contrast survives: in Lincoln as /uə ≠ ow /, in Yorkshire as /iə ≠ aw /, in Lancashire as /ua ≠ o: /, in Gloucester as / oə ~ uə ≠ ow /. It should be noted that in old-fashioned New England speech the contrast is partially preserved as checked / ɵ / ≠ free / o /, as in stone ≠ grown.

(5) In SBE, ME short /o / and the diphthong /au / have remained distinct, contrasting as / ɒ ≠ ɔ / in crop, frost, dog ≠ law, taught. The dialects of England exhibit sharply divergent treatments of these two ME phonemes, which is reflected in the complicated situation obtaining in the regional dialects of AE (see Wetmore 1959).

(6) In SBE, ME short / a / developed positional variants which in the latter part of the eighteenth century were phone micized as /æ ≠ /, as in hat, bag, man ≠ bath, glass, aunt. In AE this split appears to some extent in eastern New England and fragmentarily in Virginia; elsewhere / æ / occurs in all positions. In most of the British dialects, ME / a / remains a low vowel, [a ~ a], in all positions. Structurally, AE largely agrees in this respect with the folk dialects of England, but the phonic realization of the vowel reflects that of SBE / æ / in hat, bag, man.

5.23 Post-Settlement Importations

Changes that took place in the vowel system of Standard British English before the planting of the colonies on the American continent are rather fully established in all regional varieties of American English. Later changes in SBE, on the other hand, are confined to coastal areas: eastern New England, Metropolitan New York, eastern Virginia, South Carolina, and the states along the Gulf of Mexico. The most important of these changes is the development of an additional free vowel owing to the loss of postvocalic / r / as such in the latter part of the eighteenth century.

Orthoepists report “weakening” of postvocalic / r / as early as the seventeenth century and its loss as such shortly before 1800 [Jespersen 1909: 318, 360; Luick 1929: 730]. After high and midvowels, it is now represented by the mid central semivowel / /, as in fear, fair, four, poor, and beard, paired, board, moored, but may be lost after the back vowel / ɔ /. In far, hard it merged with the low vowel to create the new phoneme / /. In AE, these developments of SBE are rather closely matched by the usage current in the coastal areas mentioned above.

In SBE the lengthened and/or lowered allophone of the descendant of ME / a /, as in glass, staff, dance, aunt, merged with the / / derived from earlier / ar /, as in farce, starve, barn. In America this merger is largely confined to eastern New England [Kurath 1964: 109–110].

5.24 Regional Variants in the Phonic Realization of Shared Phonemes

While the regional varieties of American English do not differ greatly in the structure of their vowel system and agree substantially with Standard British English in this respect, the phonic character of many of the shared phonemes varies considerably from region to region, both in folk speech and in the speech of the cultured.

Some of the more striking regional diaphones of shared phonemes in AE are listed below. References to maps in Kurath-McDavid 1961 are offered for the convenience of the reader, who will want to inspect the regional dissemination of the variants, which follows recurring patterns.

(1) The phoneme /e / of day, bracelet [Maps 18, 19] has the diaphones [εI ~ ei ~ e• ~ e^ə];

(2) the phoneme /o / of ago, coat [Maps 20, 21] is realized as [ou ~ ou ~ O• ~ o^ə];

(3) the /u / of two [Map 17] appears as [uu ~ u^• ~ u•];

(4) the / Ɔ / of law [Map 22] has the variants [ɒ• ~ Ɔ• ~ ɒƆ];

(5) the phoneme /ai / of nine, twice [Maps 26, 27] is realized as [ai ~ ɑ^•e ~ a^•ε ~ əi], partly positionally;

(6) the /au / of mountain, out [Maps 28, 29] has the variants [a[ ~ æææ~ əu], partly in positional distribution;

(7) the phoneme / 3 / of thirty [Map 25] appears as a more or less constricted [a ~ 3] or as unconstricted [3 ~ 3I];

(8) the /Λ / of sun [Map 10] appears as [Λ ~ Y];

(9) the phoneme / æ / of sack, ashes [Maps 12, 13] has the diaphones [æ ~ æ^ə ~ æ^ε], partly by position.

Some connections between regional diaphones of AE with British folk dialects seem to be rather clear, as pointed out below.

(1) The upgliding AE variants [εI — eI] of the phoneme / e /, as in day, bracelet [Maps 18, 19], have their counterpart in the folk speech of eastern England (as shown in Figure 22), which underlie SBE. The monophthongal and ingliding AE variants [e^• ~ e^ə] are matched by phones current in the dialects of the west of England.

(2) In a similar way, the AE upgliding variants [ou ~ ɔu] and the monophthongal and ingliding variants [O• ~ o^ə] of the phoneme / o /, as in ago, coat [Maps 20, 21], are matched by regional variants in the folk speech of England (see Figure 23).

(3) The AE variants [æu] and [ΛU ~ əu] of the phoneme / au /, as in mountain, out [Maps 28, 29] also have their counterparts in English folk speech, as shown in Figure 24. The type of [æu] characterizes the eastern counties, the type of [Λu ~ əu] the west. The variant [au ~ ɑu] of the American North and North Midland matches SBE, its probable source.

Further light on the problems of relating American diaphones of shared phonemes to specific British sources will come largely from the field surveys of the folk dialects of England, Scotland, and Ireland. Spelling evidence can be of little help in tracing subphonemic entities.

The problem of tracing the probable British sources of phonemic, phonic, and incidental differences between the regional and social dialects of AE is especially complicated in the low vowels. In this range a tabulation of the phonic reflexes of ME / a ≠ o ≠ au / in the dialects of England is enough to suggest the complicated British background:

FIGURE 22: England: Diaphones of /e /

The syllabic in ● lane apron O make is articulated as an upgliding diphthong [ei ~ εI ~ æI]. Elsewhere ingliding [eə ~ εə ~ iə] or monophthongal [e ~ ε] are current.

When we ask why the several dialects of AE have so much in common in the organization of their vowel systems and differ so much in the phonic realization of shared vowel phonemes, the answer is fairly obvious. Communication is not seriously hampered by variations in the phonic character of shared phonemes as long as they fall within a certain phonic range, as they actually do. Differences in phonemic structure, on the other hand, would seriously interfere with effective communication. Hence the trend toward eliminating structural differences and tolerating substructural regional features in the history of American English.

FIGURE 23: England: Diaphones of /o /

The syllabic in O road ● coat home is articulated as an upgliding diphthong [ou ~ ou ~ ΛU]. Elsewhere ingliding [oə ~ uə] or monophthongal [o ~ ɔ] are current.

FIGURE 24: England: Diaphones of / au /

The syllabic in ● mouth O pound is articulated as [æu ~ εu]. Elsewhere [ΛU ~ əu] are current.