Accents of English

All languages exist in many different varieties. The same thing can be said in many different ways. This means not only that we can use different words and different syntactic structures to say (almost) identical things, but that the same words can also often be said in different ways. Some of these differences are due to the speech community the speaker was raised in, some to their age, gender, education, profession, etc, while some are simply idiosyncratic properties of the speaker.

linguicism

There exist speech communities in which some ways of using language are rejected: deemed improper, bad, unacceptable, etc. This attitude of criticising other people’s language use is called linguicism. Since a person has no control over which variety of a language they acquire within their community (although later this may be deliberately altered), linguicism is not justifiable. When a person speaks in a certain way that is meant to project a particular image of them, similarly to cloting, hair style, or other means of communicating their identity.

It is usually the forms used by the most prestigious groups of society (the rich, the educated, the powerful) that is considered to be better than other forms. Such judgements are never linguistically justified, they are not linguistically justifiable. So while it may be true that to avoid the scorn of uninformed people you should use certain linguistic forms rather than others, no variety is linguistically better, clearer, purer, etc than any other.

accent vs dialect

Varieties of a language that share many of their properties are traditionally identified as accents or dialects of that language. The two terms are sometimes used interchangeably, but most people distinguish them, using accent for differences in pronunciation only, and dialect for all sorts of other differences, in morphology, syntax, and the lexicon. Let’s take a sentence like you must eat it up jʉw məst ɪ́jt ɪt ə́p. In the north of England, it would sound something like jə məst ɪ́jt ɪt ʉ́p, in the West Indies as jʉw mɔ̈s ɪ́jt ɪt ɔ̈́p. These varieties exemplify different accents, the same words, in the same order are pronounced in a different way. A traditional northern dialect which has ða mʊŋ gɛ́ɾ ɪt ɛ́ːtn̩, or Jamaican Creole, which has jʊ mɔ̈s njám ɪ áːf, on the other hand, show dialectal differences, since these have different words from our original sentence. To use an illustrative oversimplification, different accents spell a sentence in the same way, different dialects would not. It must be admitted, however, that the difference between accent and dialect is not always clear.

In this course about the pronunciation of English, we will only look at differences between accents, and will not be concerned with dialectal ones. When talking about differences in vowels we will use Lindsey’s narrow, phonetically precise transcriptions symbols for British English pronunciations.

types of accent differences

irregular Some differences between accents are unique, they are not part of larger patterns, but isolated examples. The lot vowel, which is ɔ in British English, is regularly ɑ in American English. So wash is wɔʃ vs wɑʃ, want is wɔnt vs wɑnt, body bɔdɪj vs bɑdi, top tɔp vs tɑp, etc. However, what wɔt is not wɑt, as expected, but wət in AmE. This is an irregular correspondence, which affects this one lexical item only. Such accent differences can only be listed (and learned) one by one. From a linguist’s point of view regular differences are more interesting. These are what we will now look at.

regular Most differences between accents are regular. But not all regular accent differences are of the same type. Some are realizational, they do not affect the sound system. Others do affect it, they are therefore systemic. Differences may also be distributional. We will examine these types of accent differences one after the other.

realizational accent differences

The two narrow diphthongs of BrE, face (ɛj) and goat (əw), are monophthongal in a number of other varieties of English, like Standard Scottish English or some Americant accents: fes, got. This is a one-to-one correspondence: BrE ɛj is always SSE e, SSE e is always BrE ɛj, therefore this difference does not influence the vowel systems of the two accents. No contrasts are merged. (Note that the vowel systems of BrE and SSE are quite radically different, but not because of this realizational difference.) Ferdinand de Saussure, who draws attention to similarities between language and chess, says that if a chess piece, eg a rook, is replaced by some other object, eg a piece of rubber, the game can continue as if nothing happened, the rook is always the piece of rubber, the piece of rubber is always the rook.

systemic accent differences

As opposed to realizational accent difference, systemic differences merge contrasts present in other accents. Consider the vowels in the chart on the right. We see that BrE distinguishes the vowels of palm and lot, AmE does not. On the other hand, BrE has the same vowel in lot and cloth, while AmE has different vowels in these two lexical sets. And again, BrE has distinct vowels in cloth and thought, while many AmE speakers have the same vowels in these two sets. Using Saussure’s chess analogy, if the rook and the knight are both replaced by two identical pieces of rubber, we can only play the game with different rules, since we cannot distinguish these two roles any more.

We have seen earlier that BrE merges a number of contrasts that more conservative AmE accents still retain. Thus north, force, thought, and cure have the same vowel, oː, for many BrE speakers, while some AmE speakers merge north and force only, others pronounce words belonging to these lexical sets with a different vowel each: north nɔrθ, force fors, thought θɔt, and cure kjʊr. BrE turns out to be more conservative in the case of another pre-R environment: the words marry, merry, and Mary are all different here (marɪj, mɛrɪj, and mɛːrɪj, respectively), but most AmE speakers pronounce these three words as homophones. Some only merge merry and Mary, keeping marry distinct, but even so, we witness a systemic accent difference in this case (very–vary or ferry–fairy are pronounced identically by Americans, but not by Brits).

The well-known nurse merger — the fact that in most accents of English heard, Hird, and Hurd are homophones — is also a systemic difference between these accents and conservative Scottish varieties, where these words are all different: hɛrd, hɪrd, and hərd, respectively.

distributional accent differences

The third type of regular differences between accents is distributional. Examining the distribution of j after consonants, for example, we find a wide range of systems, some of which are shown in the chart below.

We see that there are varieties of English that have a palatal segment (actually not j, but i in this case) after the initial consonant of all of these words. So few in Welsh English is fiw, lewd liwd, and chew tʃiw, etc. In the other accents shown here, the syllabicity of iw has swapped: the first part is not syllabic, it is a consonant, and the second part has become syllabic. Thus, in these accents we have ju (or juw) instead of iw. The distribution of this j is variable: in conservative British accents the palatal j is missing only after palatal consonantsWe have seen that homorganicity is dispreferred in initial clusters. (eg in chew tʃuw, not *tʃjuw), in American accents it is typically missing after all coronal consonants (eg also in new or Tuesday), in East Anglian English it is missing in all environments, so even few is fuw, not *fjuw. BrE 1, BrE 2, and BrE 3 represent increasingly more advanced varieties of BrE, so current BrE is moving towards AmE in this respect, though palatality is not lost but merged with the preceding consonant after plosives: Tuesday tʃuwzdɛj, duke dʒuwk.

This chart shows the distribution of r in four types of accents. We again can see a clear case of distributional difference between these four accents. Accent 1 is a typical rhotic accent, like for example General American. Accent 3 is a typical nonrhotic accent, like Current British English. There also are accents between the two, like #2, in which preconsonantal r is not pronounced, but word finally it is. Such accents are associated with, for example, New England. Finally, in southeastern states of the US, some speakers only pronounce r when it is followed by a stressed vowel (or if it is word initial). Such an accent will not even have r in words like América or véry sórry (#4).

We have noted earlier that while in some varieties of English ŋg occurs word finally, in most it does not. The grammaticality of homorganic nasal + voiced plosive clusters at the end of the word also shows a gradual distributional accent difference: the coronal cluster nd is the most widespread word finally, though it does not occur in Scottish Standard English, the velar ŋg (as we have seen) does not occur in London English either (along with most other Englishes), while the labial mb is absent from most accents. Put alternatively: SSE is the most restrictive accent in this respect and Manchester English is the least restrictive.

Distributional accent differences provide evidence for the fact that different varieties of a language incorporate different phonotactic constraints.

nonnative accents

English is the de facto lingua franca of the world today. It is one of the few languages that have more nonnative than native speakers. Therefore we may examine not only accents of English that are used by native speakers, but also others that are used by nonnatives.

One such accent is that used by Hungarian speakers of English. This variety is often called Hunglish, a term which is used with some derogatory overtones. In fact, there is nothing wrong with this accent, it clearly indicates that it’s user has a Hungarian background. Accents inform us of the speakers background in most other cases. Let us see some properties of this accent of English, abbreviated as HE.

Similarly to Scottish Standard English, HE also has monophthongal face and goat. Names like Katy kɛjtɪj or Moby məwbɪj are keːtiSome speakers pronounce the offglide of ɛj word finally though: Milky Way is either milkivé or milkivéj. and moːbi in HE. The fact that ɛj is rendered as eː in HE is a realizational accent difference. In the case of əw, which is oː, however, we see a systemic difference: the contrast of goat and thought is merged, both low ləw and law loː are loː in HE. On the other hand, north and force are kept different from thought in HE, since it is a rhotic accent: lord lord, nerd nørd, Charlie tʃaːrli.

HE also merges dress and trap: both Calvin and Kelvin would be kɛlvin. Some consonantal contrasts are also merged: θ is usually rendered by s (rarely by t) making thick and sick or tick a minimal pair. The voiced counterpart, ð is d, but this replacement is also common in native accents of English. The differnce of w and v also disappears in HE, so there is no contrast between wet and vet or wow and vow, etc.

show me the questions
let’s go back to the contents page