Culture and Communication
Varieties of English
African American English
Background
This linguistic variety is commonly referred to as Black English (BE), Black English Vernacular (BE), African-American Vernacular English (AAVE), and Inner City English (ICE).
There have been three primary theories regarding the source of African-American English. These three theories can be named the following:
-
Decreolized Creole
-
Variety of Southern States English
-
The "Unified" Theory
Proponents of the decreolized creole theory maintain that African-American English arose from a pidgin that was created among slaves from various linguistic backgrounds, primarily from West Africa. This pidgin included features of both the West African languages and English. Over time, this pidgin developed into a creole, and then more recently, became decreolized, and began to resemble English more closely.
Others state that African-American English is a variety of Southern States English, noting that the two varieties have many features in common, such as the Southern Vowel Shift, vowel lowering, and double modals.
Proponents of the unified theory state that African-American English arose from a number of sources, including West African languages and Southern States English, through a variety of evolutionary tracks.
Features of African-American English
African-American English has a number of phonological features, including:
Consonant Cluster Reduction
Description
In many varieties of African-American English, word-final consonant clusters are reduced. That is, certain members of the cluster, such as stops, are dropped. Therefore, words such as desk, post, and walked are pronounced as [dɛs], [pos], and [wɑk]. This phenomenon is sensitive to morphological information. If the cluster reduction would eliminate a grammatical marker, then it is less common. For example, the reduced form of past tense walked would be , which is phonologically identical to the present tense form walk. Therefore, such reductions are less common than the reductions in words such as desk or post.
Consonant cluster reduction can also be found in varieties of American Indian English, as well. Penfield (1976) provides data from the Colorado River Indian Reservation in Parker, Arizona that show that speakers of English in this area (the data are from members of the Mohave, Hopi, and Navajo tribes) drop the final stop in a cluster, particularly when the cluster involves a nasal and a stop. Therefore, words like equipment (SAE [ikwɪpmɛnt]), understand (SAE [ʌndrstænd]), and student (SAE [studɛnt]) would be pronounced as [ikwɪpmɛn], [ʌndrstæn], [studɛn], respectively.
Leap (1976) also describes a similar phenomenon in Isleta English, a variety spoken among members of the Isletan Tiwa group. In that group, the second consonant of a final consonant cluster is dropped (contest (SAE ) [cantɛst]-> [cantɛst]).
Background
Smitherman (X) takes cluster reduction as evidence for a West African source for African-American English features. She notes that many West African languages do not allow complex clusters.
Realization of /θ/ and /ð/ as [t,f] and [d,v]
Description
In varieties of African-American English, the interdental fricatives are realized as either alveolar stops or labiodental fricatives, depending upon the location of the fricative in the word.
When located at the beginning of a word, the interdental fricative [θ] is realized as [t]. For example, the word thin [θɪn] is pronounced as [tɪn]. When located word-medially or word-finally, the interdental fricative [θ] is realized as [f]. For example, the word bath [bæθ] is pronounced as [bæf].
When located at the beginning of a word, the interdental fricative [ð] is realized as [d]. For example, the word this [θɪs] is pronounced as [dis]. When located word-medially or word-finally, the interdental fricative [ð] is realized as [v]. For example, the word brother [brʌðr] is pronounced as [brʌvə].
It is important to note that these rules are quite consistent. In African-American English, it is ungrammatical to pronounce [θ] as [t] unless it is at the beginning of a word (bath *[bæθ]), and it is ungrammatical to pronounce [θ] as [f] if it is at the beginning of a word (thin *[fin]). The same holds true of the voiced interdental fricative.
Background
Smitherman (X) notes the lack of interdental fricatives in West African languages to account for this rule in African-American English, using this as support for a West African source for some features of African-American English.
Vowel Lowering
In many varieties of African-American English and Southern States English, the vowel /ɪ/ is often realized as /æ/ or /ŋ/ when followed by a velar nasal . Therefore, words like sing (SAE [sɪŋ]), thing (SAE [θɪŋ]), and drink (SAE [dɪrŋk]) are pronounced as [sɛŋ], [sæŋ], and [θɛŋ].
z --> d in Contractions
Description
In certain varieties of Southern States English and African-American English, there is a phonological phenomenon in which a voiced alveolar fricative /z/ is pronounced as a voicedalveolar stop [d] before a nasal consonant. For example, the sound [z] in contracted words such as isn't (SAE [ɪznt]) and wasn't (SAE [wʌznt]) is pronounced as the sound [d]. Thus, the words like isn't and wasn't are pronounced [ɪdnt] and [wʌznt]. Troike (1986) names this phenomenon McDavid's Law, in reference to McDavid (1942), in which McDavid first attempted to state this phenomenon as a rule.
There is evidence that the [z] --> [d] distinction can be found in other lexical items in some varieties of Southern States English. For example, a word like business (SAE [bɪznəs]) is pronounced as [bɪdnəs] in some varieties. Other attested examples include cousin [kʌdn], pleasant [plɛdn], and present [prɛdn] (McMillan 1946 (transcriptions by eds.)). Taylor (1997)theorizes that this phenomenon may be part of a general rule in some varieties. This general rule would include application to the voiced affricate /ʤ/ and the voiced fricative /ʒ/ as well as . He cites as evidence pronunciations of regional, emergency, and register with /d/. For example, he provides the transcription of register in some varieties as [rɛ:dstər]
Phonologically, this sound variation can be accounted for by the assimilation of the feature [-continuant] from the nasal consonant to the /z/, changing the sound from a fricative to a stop.
Background
Different sources of evidence seem to indicate that this phenomenon has its origins in southwestern England (Feagin 1979, Troike 1986, Reynolds 1994). There are references to a rule in dialects of this area in which /z/ became /d/ in negative contracted forms of be; that is, in the forms isn't and wasn't. Reynolds (1994) proposes that the /z/ --> /d/ rule began with the word wasn't, which became [wʌdnt] through analogy with other past tense negative auxiliary contractions. The following table, from Reynolds (1994), illustrates the analogy:
could couldn't
would wouldn't
should shouldn't
might mightn't
had hadn't
did didn't
wasX = wadn't
As the table shows, all other negative contractions of past tense auxiliaries ended in either [dnt] or [ʔnt]. Therefore, by analogy, wasn't came to end in [dnt]. A second analogical change resulted in the pronunciation of isn't as [ɪdnt]. Since isn't is the present tense form of wasn't, isn't took on the [d] pronunciation to match the pronunciation of the past tense form.
Once this phenomenon had migrated to the American South, the rule began to generalize, affecting some lexical items as business and cousin. However, this change has been arrested, with only a few lexical items participating in the rule.
Monophthongization
Description
In a number of varieties of Southern States English and African-American English, the diphthong becomes monophthongized. For example, words such as right (SAE ), time (SAE ), and like (SAE ) are pronounced with a low vowel monophthong, such as , , and . Note that the length of the diphthong is preserved as a long vowel in the SSE and AAE forms.
In some instances, the diphthong is also monophthongized in certain varieties, particularly before liquids. For example, words like boil (SAE ) and toil (SAE ) are sometimes pronounced and .
In British English, Wells (1982b) believes that it is widely agreed that the "mouth" vowel is a "touchstone for distinguishing between "true Cockney" and popular London" and other more standard accents. Cockney usage would include monophthongization of the word mouth
Example:
mouth = [ma:f] (maaf) rather than [mae:f ] mouth
R-lessness
Description
In varieties of African-American English, Southern States English, and New England English, the consonant /ɹ/, the English "r", is sometimes not pronounced in surface forms. For example, in some varieties of Southern States English and New England English, words such as guard and car are pronounced with a lengthened vowel in place of the "r" (eg. SSE guard [gɑːd]and NEE car [caː]). Other words, such as fear or bored, are pronounced with a glided [ə], in place of the "r" (eg. SSE fear [fiə]).
According to Labov (1972), this results in homophony in these varieties, such as illustrated by the following:
guard = god par = pa
nor = gnaw fort = fought
sore = saw court = caught
In these varieties of Southern States English and New England English, when the is followed by a vowel, it is pronounced. For example, in four o' clock, the "r" surfaces.
In African-American English, the r-lessness pattern is often less restricted, and is not pronounced, even when followed by a vowel. Labov (1972) states that this results in homophony like the following:
Carol = Cal
Paris = pass
terrace = test
In British English R-lessness also exists. Trudgill (1999) cites the fact that in Britain about 250 years ago, the "r" started to be dropped in some dialects if it occurred before a consonant such as in arm or far. Where "r" occurred before a vowel, such as in rack, track, or carry it was retained. The change is thought to have started in the southeast of England and to have spread to other parts of the country and is continuing to spread. He sees the older people retaining the "r" in such words as tart, and the younger people losing it.
The process of ‘r’-loss has probably been accelerated by the fact it has been seen to exist in the BBC accent (standard RP) and thus pronouncing r’s has relatively low status in England, and omitting them relatively high social status Trudgill (1999).
Background
Sample Study (Labov 1972)
American Indian English
Background
The term American Indian English refers to a number of varieties of English that are spoken by indigenous communities throughout North America. As Leap (1982) states, "there are many Indian English-es." Each one is unique in its phonology, syntax and semantic properties. In this area of the site, we will explore some of the features that have been studied in terms of different varieties of American Indian English.
There are two primarily studied sources of the features attributed to American Indian English. In some cases, it has been proposed that the features of American Indian English originate from the same sources as other nonstandard varieties of English, such as Southern States English. In other cases, it has been argued that features of American Indian English are the result of influence from the native language.
Some varieties of English that will be represented on this site are Mojave English, Isletan English, Tsimshian English, Lumbee English, Tohono O'odham English, and Inupiaq English.
Features of African-American English
On the following pages are some features associated with different varieties of American Indian English. Not all features are associated with all varieties of American Indian English. On each page, you will be introduced to representative varieties of American Indian English that are associated with that variety.
-
Deletion of Final Voiced Stops
-
Final [iŋ] --> [in]
-
Vowel Shift
The Central Diphthong [ʌɪ]
Description
One of the distinguishing features of speech in the tidewater area of North Carolina and Virginia is the pronunciation of the Standard English diphthong as , where the initial vowel of the diphthong is raised to a more central position. Varieties of Southern States English found in these areas, such as Ocracoke Island, have this feature. Lumbee English, a variety of American Indian English, also has this feature.
Sample Study (Brewer & Reising 1982)
Brewer & Reising (1982) provide data illustrating some identifying features of Lumbee English, a variety of American Indian English spoken in Coastal North Carolina, particularly around the Robeson County area. This review will focus on the data regarding the central diphthong. One of the major findings of this study is that while both white varieties of English in the tidewater areas of Virginia and Lumbee English have the central diphthong , the distribution of this feature is not identical.
Studies of coastal Southern States English (Tresidder 1943; Shewmake 1943; McDavid 1955; O'Cain 1977) have provided evidence indicating that, in non-Lumbee English varieties, the centralized diphthong occurs predictably before voiceless consonants, but not characteristically before voiced consonants.
In Lumbee English, by contrast, although the central diphthong is a feature of the variety, the findings are that Lumbee English speakers use the centralized diphthong in both the voicedand voiceless environments more than 1/2 of the time. It is also found that speakers in different communities have different percentages in the usage of the central diphthong.
The authors interpret this data as supporting evidence for the hypothesis that the central diphthong is a significant feature that establishes identity as a member of the Lumbee community.
The analysis proposed in this preliminary study is based on a sample of 11 oral history tapes from the Lumbee Regional Development Association. These are taped interviews and include 4 males and 7 females between the ages of 69 and 80 years old. These speakers come from three different Lumbee communities: Pembroke, Prospect, and Magnolia. The authors selected the words night, white, and light as tokens to study the distribution of in a voiceless environment. The following table illustrates their results:
As this table shows, /aɪ/ is pronounced as:
-
[əɪ] ( another representation of [ʌɪ]) 52.7% of the time before a voiceless consonant
-
[aɪ] or [ɑɪ] 36.8% of the time before a voiceless consonant
-
[[ɑi] 10.6% of the time before a voiceless consonant
Thus, one can see that the central diphthong is used over half the time in the environment preceding a voiceless consonant.
The authors also selected the words five, nine, spider, and piazza as tokens to study the distribution of in a voiced environment. The following table illustrates their results:
As this table shows, /aɪ/ is pronounced as:
-
[əɪ] ( another representation of [ʌɪ]) 52.7% of the time before a voiceless consonant
-
[aɪ] or [ɑɪ] 26.6% of the time before a voiceless consonant
-
[[ɑi] 19% of the time before a voiceless consonant
Thus, one can see that the central diphthong is used over half the time in the environment preceding a voiceless consonant.
Thus, one can see from the table that the central diphthong is used over half the time in the environment preceding a voiced sound
As this table shows, /aɪ/ is pronounced as:
-
[əɪ] ( another representation of [ʌɪ]) in 38.6% of the tokens in Prospect, as opposed to 14.0% in Pembroke, and 0% in Magnolia
-
[aɪ] or [ɑɪ] in 12.3% of the tokens in Prospect, as opposed to 24.6% in Pembroke, and 0% in Magnolia
-
[[ɑi] in1.8% of the tokens in Prospect, as opposed to 3.5% in Pembroke, and 5.3% in Magnolia
Thus, one can see from the table that the central diphthong is used more often by speakers in Prospect than in the other two communities.
The authors propose that these facts regarding the usage of the central diphthong in Lumbee English point to this feature being a salient indicator of membership in the Lumbee community. The fact that Lumbee English has a wider distribution of the central diphthong that differs from other dialects in the area, indicates that this feature is salient to the community. Also, the community of Prospect is considered to be one of the more conservative Lumbee areas, and consequently, the greater incidence of the central diphthong in that area, indicates that the central diphthong is a conservative feature of the Lumbee community.
Final Devoicing
Description
In some varieties of American Indian English (particularly in the Southwest United States), final voiced stops, fricatives, and affricates are devoiced. This means that words that have a final voiced obstruents in Standard American English (SAE) are pronounced with less voicing or no voicing in American Indian English. For example, Penfield (1976) shows that in Mojave, Hopi, and Navajo English, words like job (SAE [jab]) would be pronounced with less voicing, which Penfield represents as [japh], where the superscript [p] indicates that there is less voicing of the final stop than in Standard American English. Other words, such as questions (SAE [kwestʃenz]) and has (SAE [hæz]) are sometimes pronounced in Mojave, Hopi, and Navajo English as [kwestʃens] and [hæs], respectively.
Deletion of Final Voiced Stops
Final [iŋ] --> [in]
Vowel Shift
British English
Background
The size of the British Isles often leads people to assume that the language spoken in its countries of England, Scotland, Wales and Ireland is somewhat homogeneous and first time visitors are often surprised to find that they have difficulty in understanding the accents and dialects of certain regions. Even within the country of England alone there is great diversity of dialect both regionally and socially. Trudgill (1999) believes that for the majority of English people "where they are from" is very important to them. Accents are clues to where people were born and where they grew up. Although some people may change the way they speak during their lifetimes, most people "carry at least some trace" of their accent and dialect origins throughout their lives:
In addition to the regional accents of England, there can also be class differences reflected in the different accents. The general sociolinguistic issues section discusses this more fully.
Geography of British English
The term "British English" can occasionally be confusing depending upon the regions included by the term British. (See note 1 for further explanation) For the purpose of this project the current study of British English will concentrate on dialects and accents found within the country of England itself and will not include those found in Scotland, Wales, Northern Ireland and Ireland.
Although there is an abundance of different dialects within England that can be referred to as "northern" or "southern" for example, they do not really follow any sharp boundaries or coincide with any county lines. Dialects form a continuum and as Trudgill (1999) describes, they can be differentiated on a "more-or-less" basis rather than an "either-or" one. It is common in Britain for people who display particularly broad accents to be labeled by terms such as "Geordie", "Cockney", "Jock" or "Scouse." All of these identify a specific regional accent, most of which are recognizable to many of the people in the country. Trudgill (1999) discusses specific regional dialects and vocabulary for many areas of Great Britain.
Sociolinguistic of British English
In Britain, "people are often able to make instant and unconscious judgements about a stranger’s class affiliation on the basis of his or her accent." (Wells 1982a) Both the words and pronunciation of many individuals reflect that person’s social position. It is agreed that in England, the "phonetic factors assume a predominating role which they do not generally have in North America" (Wells 1982a).
Traditionally, it has been acknowledged that in England, the relation between social and regional accents can be diagrammed as follows:
Geographical variation is represented along the broad base of the pyramid while the vertical dimension exhibits social variation. It can be seen that working class accents display a good deal of regional variety, but as the pyramid narrows to its apex, up the social scale, it’s also apparent that upper class accents exhibit no regional variation. (Wells 1982a)
Thus by definition, any regional accent would not be considered upper-class and the more localizable the accent, the more it will described as a "broad" accent. Wells (1982a) purports that broad accents reflect:
-
regionally, the highest degree of local distinctiveness
-
socially, the lowest social class
-
linguistically, the maximal degree of difference from RP.
A 1972 survey carried out by National Opinion Polls in England, provides an example of how significantly speech differences are associated with social class differences. (Wells 1982a)The following question was asked:
"Which of the these [eleven specified factors] would you say are most important in being able to tell which class a person is?" Respondents were randomly chosen from the British public. The factor that scored the highest was "the way they speak" followed by "where they live." At the bottom of the list was "the amount of money they have." All this is evidence that then, and to some degree even now, "speech is regarded as more indicative of social class than occupation, education and income."
(Giles & Sassoon, 1983) also cite consistent findings of listeners evaluating anonymous speakers with standard accents more favorably for such status traits as intelligence, success, confidence. In Britain the middle class is associated with having not only a standard accent, but with also speaking in a more "formal and abstract style than working class."
Accents are often characterized by British speakers themselves as either "posh" or "common" accents. Most speakers of British English would recognize these labels and create a fairly accurate image of the sound of these far ends of the spectrum. Conservative or U-"Received Pronunciation" representing the "posh" end and a less broad version of Cockney representing the "common" accent.
The significance of accents and their cultural and social associations is well represented in films and on television in Britain. The critically acclaimed 1964 file My Fair Lady based on George Bernard Shaw’s 1912 play, Pygmalion is often referenced in linguistic discussions as a wonderful example of how social class and accent were, and are still, inextricably linked in Britain. Over the past years, numerous television series have also provided viewers with a glimpse of the lives and accents of the Cockney population of London. The Cockney English section talks more about the current, very popular long running television series EastEnders.
Accents within England
As language change continues to take place within Britain and within England, there are some who claim that a relatively newly established accent, "Estuary English" (EE) is due to replace the traditional educated accent of England Received Pronunciation" (RP). (Wells, 1998) Estuary English is reported to be used by speakers who constitute the social "middle ground" Rosewarne, 1984) and is discussed in detail under the Estuary English section.
It must be emphasized, however, that there are many features in common among these more prevalent accents that are present in England and that they must be thought of as existing on a continuum rather than having strict, non fuzzy boundaries.
The many regional accents within the British Isles are not currently discussed on this site. However, two of the most commonly known and researched accents (RP and Cockney) will be included as will Estuary English as evidence of recent language change. There will also be a discussion of what Katie Wales (1994) refers to as the "Queen’s" English and where this fits into the continuum of British accents.
Received Pronunciation (RP)
Background
There seems to be some disagreement as to the origins of the term "received" in the phrase, "Received Pronunciation" but both A.J. Ellis’ On early English Pronunciation, 1869-1889 as well as John Walker’s Critical Pronouncing Dictionary and Expositor of the English Language of 1791 are among the possibilities for it’s early appearance. (Wells 1982a), however, cites Daniel Jones, as the "great describer and codifier of the Received Pronunciation of English" in the 1890’s. Regardless of its exact origins, the term "received" originally meant "that which is generally accepted" or "that accepted by the best society."
RP has for many years epitomized the "top end of the scale" of British English and it is what English people have traditionally meant when they’ve said that someone "hasn’t got an accent." It remains that RP is often regarded as a "neutral" and often "correct" accent. It is also referred to under the terms "BBC English," "Public School English" or even "Standard English. (Wells 1982a)
RP is also the accent that Americans and possibly other foreigners would likely refer to as the typical British accent. It should also be noted that there is no single accent whose role and status in the United States correspond to that of RP in England. (Wells 1982a)
A number of distinctions even within RP have, over the years, been proposed by various linguists. These include "Mainsteam" RP, "Upper crust" RP, "Adoptive" RP, "Near" RP, "Conservative" RP, and "General" RP." (Wells 1982a) discusses these in detail along with their individual phonetic and lexical nuances. Wales (1994) also differentiates between them in saying that "conservative" RP is often spoken by the older generation, and "advanced" RP by the younger generation. Her discussion of the accents of the royal family can be found in the "Queen’s English" section.
Regardless of the differences within RP, it is an accent commonly recognized and one that has been taught as the standard English in schools for years.
Geography of Received Pronunciation (RP) English
RP is only really associated with England and not the other countries of Great Britain. However, it is not associated with any particular location within England. Because it is what might be thought of as an "educated accent" it appears characteristically in upper and upper middle class speakers and is more sociologically defined rather than geographically defined.
Sociolinguistic Issue of Received Pronunciation (RP) English
Traditionally, there have been certain occupations most typically associated with an RP accent and they include barristers (attorneys in the superior courts) stockbrokers and diplomats. Up until the 1970’s this was the accent that was required to be considered an announcer on the nationally broadcast BBC (British Broadcasting Corporation) television and radio stations, hence the term "BBC English." As discussed in the general sociolinguistic issues section, RP is an accent that is not localizable but is very recognizable as being the standard, neutral accent of the society. It is not, however, used by a large percentage of the population.
Wells (1991) has identified some characteristics of popular accents that have been resisted by RP and also some changes that have been accepted into RP. He asserts that it is important not to lose sight of the fact that "accents, and more generally varieties of language, are not objective entities so much as mental constructs" and that is preference is for a "sociolinguistic definition of RP, which entails recognizing the possibility of change. Some of those changes can reasonably be attributed to influence from Cockney – often overtly despised, but covertly imitated."
-
Intrusive R
To an objective observer, the intrusive r is very prevalent in RP. It involves the "insertion of an r-sound at the end of a word ending in a non-high vowel where the next word begins with a vowel.
examples:
R pronounced
-
put a comma[r]
-
the idea[r] of
-
I saw[r] it happen
R not pronounced
-
a comma may be added
-
idea for
-
I saw them
Trudgill (1999) sees the development of this intrusive r as one of the consequences of r-lessness that developed in more modern British dialects.
Received Pronunciation (RP)
Canadian English
Introduction
Canadian English, for all its speakers, is an under-described variety of English. In popular dialectological literature it is often given little acknowledgement as a distinct and homogeneous variety, save for a paragraph or two dedicated to oddities of Canadian spelling and the fading use of British-sounding lexical items like chesterfield, serviette, and zed.
There is a small body of scholarly research that suggests that if there is such a thing as a Canadian English, all its unique characteristics are being lost. In fact, Lilles (2000) goes so far as to claim that there is no such thing as a distinct Canadian English, and argues that the notion of Canadian English is a myth, fabricated to reinforce a fragile Canadian identity. As evidence, he cites the lack of phonological and orthographic standardization for Canadian English, the paucity of distinct Canadian vocabulary, and the appearance of regionalisms associated with various parts of the United States.
Sutherland (2000) quickly rebuts by pointing out that Canadian English is more than a "network of regionalisms", and that a variety can be distinct by more than its vocabulary. We can add that orthographic standards tell us little about what makes a spoken variety unique. Further, as you have navigated the page, you will have seen that although few linguistic features are unique to any dialect, the confluence of a particular set of features is what makes a dialect unique. This is certainly true of Canadian English: no other dialect has all the same features.
Other research suggests that the few unique traits of Canadian English are disappearing in favour of American forms. Clarke (1993) and Chambers (1998) point to the loss of certain lexical items, like chesterfield and serviette, and the loss of certain phonological traits, like voiceless wh of which and [yu] in news and student. These are seen as a signal of the impending convergence of Canadian and American English. Indeed, Woods (1993) identifies eight phonological variables as characteristically Canadian, and argues that most of them are disappearing.
To the contrary, this site's discussion of Canadian phonology identifies at least four other characteristics not included in Woods' study, all of which remain robust in Canadian speech. The other sections offer further insight into the character of Canadian English. The Table of Contents below provides an overview of the organization of this area of the page.
Table of Content
1. History
2. Phonetics and Phonology
-
Phonetic properties of Canadian vowels: the low-back merger and the Canadian Shift
-
The vowel space
-
Some regional differences
-
Canadian Phonology
-
Canadian Raising: the Central Diphthong
-
front and back diphthongs
-
Phonological conditioning
-
The history of Canadian Raising
-
-
Rhoticity
-
pre-rhotic [o] in sorrow and borrow
-
pre-rhotic in merry and marry
-
-
Borrowings with low vowels
-
The pasta variable and its exceptions
-
A possible explanation
-
3. Morphology
-
Vocabulary
-
Hockey Nicknames: data and description
4. Syntax and discourse
-
The 'eh' tag
-
functions: tag (actual vs pseudo), imperative, discourse marker
-
constraints: intonation, contrast with huh
2. Phonetics and Phonology
Probably the the most salient aspect of Canadian English is its phonological system. Previous research on Canadian phonology identifies three variables as defining characteristics: the centralized diphthongs and , the voiceless initial glide in which and whether, and the palatal glide in news [nyuz] and tune [tyun]. Since the use of the latter two is falling off in Canadian English, some scholars conclude that there is little to say anymore about Canadian English as a distinct variety. Woods (1993) includes these phonological traits in his list of eight characteristically Canadian features. What is interesting (and misleading) about the study is that most features are considered Canadianisms only because they are Britishisms; hence, the loss of a Britishism is also the loss of a Canadianism.
Though Woods' list was probably not intended to be an exhaustive inventory of Canadianisms, it gives the impression that Canadian and American phonological systems are converging, because (as should be no surprise) Canadian and British English have been diverging for centuries. However, a feature need not have a British origin to be considered a Canadianism; this is certainly true, for example, of the / merger discussed below. We contend that a dialect's characteristic features are those in current use, not those of its predecessors.
Indeed, the Language Samples Project has identified a number of other markers of Canadian English that are missing from Woods' list. Some of these are the subject of the published research of other Canadian dialectologists, and some of them emerge from our own research. This section first offers a description of the phonetic characteristics of Canadian vowels, followed by a discussion of numerous phonological patterns.
Since Canadian English is a North American variety, and it shares much phonology and history with American English, often the best way to illustrate a particular Canadian trait is by comparing it with the American dialects in close geographical proximity with it. As a result, you may notice a recurring method of Canadian-American contrasts throughout this discussion. This is not meant to imply that Canadian English should be characterized simply as a non-American North American dialect. Instead, we hope to show Canadian English to be a homogeneous dialect with its own unique collection of properties.
Content for Phonetics and Phonology
Phonetics
-
Properties of Canadian vowels: the low-back merger and the Canadian Shift
-
The vowel space and monophthongs
-
Some regional differences
Phonology
-
General facts
-
Canadian Raising: the Central Diphthong
-
front and back diphthongs
-
Phonological conditioning
-
The history of Canadian Raising
-
-
Rhoticity
-
pre-rhotic [o] in sorrow and borrow
-
pre-rhotic in merry and marry
-
-
Borrowings with low vowels
-
The pasta variable and its exceptions
-
A possible explanation
-
Phonetics
Properties of Canadian vowels: the low-back merger and the Canadian Shift
Although not the most striking feature of Canadian English, a crucial aspect of the dialect is the complete merger of the vowels [a] and [ɔ]. Since this is a property of the vowel inventory, it is technically an aspect of phonology, but it has a strong effect on the phonetic realization of other Canadian vowels. Though this merger has also occurred in the mid-western and western United States, in Canada it has triggered what Clarke et al (1995) call the Canadian Shift.
A result of the merger is that the low-back region of the vowel space is less dense; allowing for the low-front vowel [æ] to retract to a low-central articulation. This retraction moves [æ] in the opposite direction of where it has shifted in the Northern (US) Cities Shift. Boberg (2000) sums up the result rather succinctly: the word stack is pronounced in Windsor, Ontario (indeed, in the rest of Canada) with the same vowel of the word stock as pronounced in Detroit, just across the river.
The following links provide examples of the retraction of [æ] in Canadian English.
-
"... driving like maniacs ..."
-
" ... half a pack a day ...
There is evidence that the merger of [a] and [ɔ] had taken place at least as early as the 1850s. Chambers (1993) cites the published memoirs of Susanna Moodie, a British woman who emigrated to southern Ontario. In one passage, she mocks an Ontarian's pronunciation of the word sauce, saying it sounded like "sarce." Given what we know about r-lessness in British English of the time, Chambers reasons that Moodie's spelling of "sarce" indicates that the Ontarian pronounced it as [sas], with a merged [a] rather than [ɔ].
The vowel space
The vowel space of Canadian English is further distinct in that the entire system is situated slightly farther back in the oral tract. That is, Canadian back vowels are pronounced with the tongue bunched slightly behind where it woud be in British and American dialects.
In acoustic terms, this difference is reflected in the formant structure of each vowel. Recall that F2 varies inversely with backness; a high value for F2 indicates a vowel with the tongue bunched forward, as for front vowels. Each Canadian vowel has an F2 that is a little lower than the equivalent vowel in other dialects. The image below illustrates this; the graph shows the F1/F2 plot for a Canadian speaker. If you roll your mouse over the button below the chart, you will see the formant values for the corresponding American English vowels.
HOVER YOUR POINTER ABOVE THE HAND TO SEE AMERICAN VOWELS (green) vs. CANADIAN ONES(white)
The above graph's plots are F2 on the x-axis and F1 on the y-axis; formant values were extracted using from spectrograms made with Praat software developed by Paul Boersma.
Vowels that are plotted farther to the right have lower F2 values, and are therefore pronounced with the tongue bunched farther back in the mouth. You should notice that the Canadian vowels (in white) tend to be plotted farther to the right than their American counterparts (in red). The differences between the two speakers here are small, since the word-tokens used to identify their formants were in citation (careful) form, but in fast, natural speech, we would expect a greater distance between the American and Canadian vowel plots.
Monophthongs
Generally, even in quick speech, this difference in F2 values is not enough to make a Canadian incomprehensible, but it still is noticeable to the human ear. The following links provide examples of Canadian back vowels. You will also notice that the perception of extra-backness will be heightened by the monophthongal nature of the Canadian vowels. That is, as shown in the discussion of diphthongs, tense vowels tend to be phonetic diphthongs in English. In these examples, however, the tense vowels are steady-state; there is less of a diphthong contour.
Monophthong Examples
1. Back vowels
-
"Some nineteen-sixties commando show ..."
-
"People that you see rushing home ..."
2. Front vowels
-
"There's not a day I don't try."
For some vowels, the differences in formant structure are enough that a particular vowel in one dialect could sound like a different vowel in another dialect. As discussed above, this appears to be the case for low vowels in Canadian English: the plot for the Canadian front is situated rather close to the American back . An interesting perceptual effect emerges from the proximity of these vowels: an American English speaker might actually perceive the Canadian pronunciation of stack as their stock. The opposite effect is much more salient: Canadians often report the American pronunciation of stock to sound like their own stack. While this effect is strongest for speakers of Northern Cities dialects (like Detroit, Chicago, and Buffalo), Canadians associate this with other varieties, including Southern American and Mid-Atlantic English.
Regional Differences
The vowel space of Canadian English is further distinct in that the entire system is situated slightly farther back in the oral tract. That is, Canadian back vowels are pronounced with the tongue bunched slightly behind where it woud be in British and American dialects.
In acoustic terms, this difference is reflected in the formant structure of each vowel. Recall that F2 varies inversely with backness; a high value for F2 indicates a vowel with the tongue bunched forward, as for front vowels. Each Canadian vowel has an F2 that is a little lower than the equivalent vowel in other dialects. The image below illustrates this; the graph shows the F1/F2 plot for a Canadian speaker. If you roll your mouse over the button below the chart, you will see the formant values for the corresponding American English vowels.
Phonology
General Facts
Canadian English shares a number of phonological properties with Standard American English. Among these are syllable-final rhoticity and alveolar flapping. Canadian English can be called rhotic because, like in Standard American and Irish English, syllable-final r is pronounced in words like car and farm. Interestingly, the English spoken in Lunenburg, Nova Scotia, stands out as a rare r-dropping Canadian dialect.
Flapping is the process of replacing an intervocalic t or d with a quick voiced tap of the tongue against the alveolar ridge. In both Canadian and American English, it can only occur if the t or d is between two vowels, and as long as the second vowel is not stressed. As a result, the alveolar stops in waiting, wading, seated, seeded, and capital are all flapped. Flapping can also occur if there is an r between the first vowel and the alveolar stop, as in words like barter and party. In Canadian English, this feature is age-graded. Woods (1993) shows that older Canadians are less likely than younger ones to replace alveolar stops with flaps.
Canadian Raising
One of the most salient characteristics of Canadian English is the pattern that linguists call "Canadian Raising". In the general discussion of diphthongs, we saw that English has three phonemic diphthongs. Two of these have low vowels as their nucleus: [ay] as in ride and [aw] as in loud. In Canadian English, both these diphthongs have variants whose occurrence is obligatory under certain phonological conditions. Before a voiceless consonant, the diphthong [ʌy] replaces [ay], and [ʌw] replaces [aw]. These variants are called raised or centralizeddiphthongs, because their nucleus is no longer a low vowel.
The table below provides lists of words which do and do not have raised diphthongs in Canadian English. Click on any highlighted word to hear the contrast between raised and unraised diphthongs.
five
fife
lies
lice
rider
writer
powder
pouter
couch
gouge
loud
stout
Canadian Raising was first brought to the attention of linguists by Joos (1942). Parts of it are evident in other dialects: in the northern US Great Lakes cities raises [ay] is raised, but not [aw]. Conversely, Virginia English raises only [aw]. The English of Martha's Vineyard raises both diphthongs, but not in flapped words like writer and pouter. In each of the American raising dialects, the pattern is variable, but subject to the same phonological conditioning of the following consonant. In Canadian English, the pattern is essentially categorical: it can only occur before voiceless consonants, where it is obligatory.
The origins of Canadian Raising are debatable. Chambers and others use it as evidence of the Loyalist roots of Canadian English, assuming it was a feature of the English spoken by Loyalists fleeing the American Revolution, as suggested by its perseverance in some American dialects. Yet the feature is also found in certain "linguistic enclaves" in Canada, notably in Nova Scotia and the Ottawa Valley, where the influence of Loyalist English is thought to have been much weaker. Furthermore, a similar pattern is also seen in Scottish and Irish English, both of which may have had a substantial influence on the Nova Scotia and Ottawa varieties of Canadian English.
So while the question of who brought the pattern to southern Ontario (and ultimately to western Canada) is hard to answer, historical linguists do agree on one thing: that the pattern is a fossil of the Great Vowel Shift that occurred in England in the 15th and 16th centuries. The Great Vowel Shift refers to the rearrangement of the entire English vowel system from Middle to Modern English. Prior to the shift, words like five and house were pronounced [fi:v] and [hu:s], with high vowels. The Great Vowel Shift lowered their vowels to their current low-vowel pronunciation, [fayv] and [haws]. It is believed that the diphthong-raising pattern is inherited from certain middle-English dialects in which the lowering of [i:] and [u:] stopped at the mid-vowel height in some words.
Pre-rhotic Vowels
Another characteristic of Canadian vowels is in the distribution of pre-rhotic (before-r) vowels. A notable aspect of Canadian pre-rhotic vowels is their resistance to the emergent pattern in American English of substituting [a] for [o] before inter-vocalic [r]. In a number of highly frequent words, such as sorry, tomorrow, borrow, sorrow, and Laura, this pattern has become obligatory in American English. The pattern is also variably evident in a few more words, such as Florida, orange, oracle, Norwich, adorable, and thesaurus.
The current variability in the pattern suggests that American English is losing [o] before intervocalic [r], but Canadian English maintains [o] in all of these forms. The result is a contrast between the American pronunciations, like [sari] and [baro], and the Canadian pronunciations, [sori] and [boro]. To our knowledge, this feature is not discussed in any published discussion of Canadian dialectology, but we propose that it is a definitive marker of Canadian English.
While Canadian English maintains [o] before intervocalic [r], it has nearly or completely lost the distinction between and in the same position. A result of this is homophony for the word pairs marry-merry and Barry-berry. In a survey of Ottawa residents, Woods (1993) finds the pattern to be age-related: older speakers tend to pronounce marry as [mæri], while younger speakers are likely to pronounce it so that it rhymes with merry. It is Woods' claim that the retention of [æri] is a Canadianism, since it is also a Britishism. We contend that the loss of [æri] is a property of Canadian English that distinguishes it from other varieties in close proximity, like the English of New York and New England, where [æri] is preserved.
sorry
borrow
Borrowing with Low Vowels
A final, curious property of Canadian English is in its foreign /a/ nativization. This is the way the dialect pronounces borrowed words with low vowels, such as pasta, Mazda, drama, and taco. There is a strong tendency for Canadians to pronounce these words with a front [æ], while Americans tend to use [a].
Though the pattern is somewhat variable, Boberg (2000) finds that for 15 such borrowings, the Canadian pronunciation is always more likely than the American pronunciation to have [æ]. For example, 82% of his American informants pronounce panorama with [æ] in the third syllable, compared to 94% of Canadians. In a parallel manner, 5% of the Americans in his study pronounce pasta with [æ] in the first syllable, compared to 81% of Canadians.
The following table provides a list of items Boberg used. You should note that for nearly each word, at least 70% of his Canadian informants had a pronunciation with , with the exception of macho, taco, and Vietnam. In these cases, however, the Canadian pronunciation was still more likely to use than the American pronunciation was.
The Language Samples project has compiled a list of many more borrowed items that tend to have in Canadian English. These are provided and highlighted in the table below:
Some other words that seem to pattern with macho in Canadian English, by tending to use [a] instead of [æ], include Bach, Guatamala, kamikaze, karate and garbanzo. Why the Canadian pronunciation of these items tends to use [a] is a mystery, but as the two lists above show, the overwhelming pattern is for Canadian English to use [æ] in borrowings with low vowels. Furthermore, the use of [æ] instead of [a] is productive: Canadians, when confronted with novel foreign words, continue to assign [æ] to them.
The reason for this property of Canadian English is a matter of conjecture; Boberg does not speculate as to its source. One account might simply be that Canadian English uses the spelling of these words as a basis for their pronunciation. However, another possible answer lies in the distinct configurations of the Canadian and American vowel systems. In the discussion of the low-back merger, we saw that the Northern Cities Shift and the Canadian Shift have pulled the pronunciation of [æ] in opposite directions. The result is that the American pronunciation of stock is almost identical to the Canadian pronunciation of stack. That is, the American [a] is almost the same vowel as the Canadian [æ].
Oddly, then, the American and Canadian pronunciations of borrowed words like lava and pasta are phonetically very similar, even though they are identifiably different phonemes; only a Canadian could rhyme pasta with Mt. Shasta. Because of the phonetic similarity, we can say that both dialects try to assign a low-central vowel to these borrowings; in the American case, the central vowel happens to be the more back of its two low vowels, while in the Canadian case, the central vowel is the more front of the two. Another way of making this distinction is as follows: the Northern Cities shift has pushed [æ] too far forward and too high for it to be an appropriate vowel for these borrowings. Similarly, the Canadian Shift has pulled [a] too far back for it to be an appropriate vowel in the same items.
1. History
Linguistic historians usually cannot identify a single event or period of time as the beginning of a particular dialect. This is so because there are normally many conditions needed for a speech variety to emerge as a distinct homegeneous form. We can attribute the unique characteristics of a dialect not only to the geographical or social isolation of its speakers, but to the speech of the various populations from which it descends or with whom it has had contact.
One (perhaps oversimplistic) way of characterizing Canadian English is as a hybrid of British and American Englishes. It would certainly seem that way to the initiate: Canadians drive trucks, not lorries, but a Canadian who is pissed is intoxicated, and not (necessarily) angry. Canadians use British spellings like labour, colour, and cheque, but American spellings like plow, draft, and program. Although spelling habits don't really say anything about dialect speech patterns, they may be exploited by the cynic who says there is no Canadian spelling. There are also always the anecdotes of Canadians also 'sounding' halfway between the other two, but the lay-person's means of describing dialect differences rarely can quantify that kind of judgement.
The idea that a dialect is a hybrid often implies that it is not a distinct variety. Coupled with a collective sensitivity about their own identity, English Canadians will bristle at the suggestion that their speech is half-American, half-British, and not at all their own. This need not be the case: a dialect can be seen as a hybrid in terms of its history, but as a distinct form in terms of its current usage. That is, Canadians can claim to speak a distinct variety of English that has the English of both Americans and British as its predecessors.
We can tie the dualistic background of Canadian English directly to the dualistic background of the settlement of English Canada. Following the seizing of the French colony of Quebec in 1761, all of eastern North America was under the control of the British Empire. The thirteen American colonies had already been densely settled, and the dialects of the eastern seaboard had begun to emerge. Maritime Canada had also seen settlement, which is part of why Maritime English remains distinct today.
However, Upper Canada, the region that was to become Ontario, now Canada's most populous province, was at that time sparsely settled. Migration of Europeans to Ontario lagged behind that of the eastern colonies for several reasons, notably among them the harshness of the winter and its distance from ocean ports. Following the American Revolution, however, settlement of Ontario increased in pace, both with the continuing arrival of Europeans, but more significantly with the migration of Loyalists (or "Tories") who fled the United States.
Chambers (1993, 1997) claims it is the speech of the first wave of Loyalists, who arrived in southern Ontario from Pennsylvania and Virginia in the 1780's, that forms the basis of early Canadian English. Later waves of New England loyalists and Scottish and Irish immigrants in the mid 19th century are thought to have had little effect on the dialect, except where their numbers were too large to have simply been absorbed into the settled population.
The social conditions of the time both allowed for Canadian English to diverge from other American dialects, and arguably may have encouraged it to do so. It was able to diverge because it was spoken in an area both geographically and politically separated from the eastern United States. Although the introduction of steam power and railroads (and ultimately, air travel) reduces the effect of geographical distance, the political border between Canada and the US has remained a factor in maintaining the relative isolation of Canadian English from its southern neighbour.
One can also imagine how the sentiments of Loyalists and British Canadians may have encouraged them, perhaps not even consciously, to try to avoid using American-sounding speech. Whether they were successful is debatable, but it is certainly clear that British English was beyond revival in North America. Chambers provides an overview of how writers of the time tended to decry the vernacular of Upper Canada. Amid such testimonials is evidence that the English of Upper Canada had become a homogeneous variety by the 1860's: the historian William Canniff noted in 1852 that the speech of those born as Canadians was quite uniform despite the diversity of accents spoken by their parents.
This homogeneity of Upper Canadian speech provides an explanation of why modern Canadian English is so uniform across the entire country. In the late 19th century, the settlement of western Canada saw a surge in migration of Ontarians, who eventually outnumbered the French-speaking and aboriginal communities of the Prairies to such an extent that later waves of immigration assimilated to the English-speaking population. As a result, the phonological differences among the English spoken today in Vancouver, Calgary, and Toronto are minimal.
Language and Identity
In light of the continuing contact between Canada and Britain in the 19th century, Canadians no doubt had ideas of what sounded British and what sounded American. As a young, emerging nation, they may have had a harder time trying to figure out what sounded Canadian and whether it should resemble either of the other varieties. It is likely, however, that they were more conscious of word-choice than of sound patterns.
Thus, a trait of Canadian speech that survives to this day is the occasional preference for perceived British vocabulary as a means for speakers to mark their speech as non-American. In fact, a survey by Lipski (1976) uncovers a tendency among Canadians to use American lexical choice and spelling in unguarded moments, but to use British forms in situations where nationalism is an issue. Never, however, do Canadians try to adopt British phonology for the same purpose; rather, they adhere strictly to Canadian phonology.
A famous recent example is seen in Molson's Joe Canadian advertisement, a patriotic monologue in which the protagonist spiritedly proclaims his dedication to the words chesterfield for couch and zed for zee. The word chesterfield is an example of a British word whose use, according to Chambers, is disappearing from Canadian English, but Molson's Joe clings to it as a non-American word. In the same spot, Joe avers that he never says about as "aboot" (an exaggeration assigned to cinematic stock Canadians), but his emphatic correction features a very Canadian (and un-British) diphthong .
This should not be taken to imply that Canadian English has evolved into a variety whose only unique trait is its pronunciation. While there is indeed plenty to say about Canadian phonology, you will find in this site a discussion of other aspects of Canadian English as well, including its lexical and syntactic properties.
Northeast US English
Introduction
The northeastern United States has a wide variety of distinct accents and dialects. The diversity that exists in the modern northeast is partially a consequence of its older settlement: communities like Boston, New York, and Philadelphia have been around longer than similar-sized communities in the western U.S. As a result, the speech of each urban community has had more time to diverge from the dialects of other nearby cities. Yet as we will see below, some of these divergent innovations are comparatively recent.
This page is intended as a gateway to discussions of several specific northeastern dialects. The general traits page describes a number of features common to all northeastern dialects. These are features which help distinguish them as a group from other American varieties like Southern American English, African American English, and the English of the Mid-west and West. You will also find links to pages about the English in and around Boston and New York City. In addition, the discussion of the Northern Cities Shift describes some phonological features that are typical of Detroit, Buffalo, and Chicago, among others.
Features of Northeastern Dialects
General Traits
Intro
Phonetics and Phonology
-
low-back distinction
-
the northern-cities shift
-
r-less-ness
Variation
Boston English
History
Phonetics and Phonology
-
The vowel space: no Northern Cities Shift
-
Rhoticity 1
-
r-lessness
-
r-intrusion
-
pre-rhotic [ae] and the merry-marry contrast
-
Morphology
-
Vocabulary and Lexicon
Variation
-
around Boston
-
elsewhere in New England
New York English
Introduction
A caveat to be made when discussing the English spoken in New York City is that it is difficult to isolate a single variety that is prototypical of the region. New York has been a dense, populous urban centre for long enough that different neighborhoods, social classes, and ethnicities can easily have their own variety. Although each of these local varieties is identifiably distinct to other New Yorkers, they all share a set of dialectal features that can distinguish any local variety from other American varieties.
Thus, when we speak of "New York English" as a variety of English, we are referring to a composite of the various forms spoken there. Similarly, if we can identify a New Yorker from his or her speech, we are using at least some of the features which collectively help distinguish New Yorkers from other Americans.
In this area of the LSP website, you will encounter features that are unique to New York English, features that it shares with other dialects of English, and features that vary within New York City.
Northern Cities Shift
Introduction
A caveat to be made when discussing the English spoken in New York City is that it is difficult to isolate a single variety that is prototypical of the region. New York has been a dense, populous urban centre for long enough that different neighborhoods, social classes, and ethnicities can easily have their own variety. Although each of these local varieties is identifiably distinct to other New Yorkers, they all share a set of dialectal features that can distinguish any local variety from other American varieties.
Thus, when we speak of "New York English" as a variety of English, we are referring to a composite of the various forms spoken there. Similarly, if we can identify a New Yorker from his or her speech, we are using at least some of the features which collectively help distinguish New Yorkers from other Americans.
In this area of the LSP website, you will encounter features that are unique to New York English, features that it shares with other dialects of English, and features that vary within New York City.
Southern States English
Background
The term Southern American English (also known as Southern States English) refers to a number of varieties of English spoken in many of the southern States, including Alabama, Florida, Georgia, Kentucky, Louisiana, Tennessee, North and South Carolina, Virginia, and parts of Arkansas, Maryland, Oklahoma, Texas, and West Virginia. Although these varieties are not uniform throughout these states, they share certain common characteristics that differentiate them from other varieties found in the Northern and Western United States.
Geography
The precise boundaries of Southern American English depend upon the variables being studied, however Carver (1987) provides a map of the major dialect areas of the United States, including Southern American English (the background image is based on this map). This map delineates three major divisions of Southern American English: the Upper South, Lower South, and Delta South. There are also some narrower classifications, such as Virginia Piedmont and Southeastern Louisiana. It should be noted that this classification has been criticized in recent years (Frazer 1997).
Features of Southern American English
There are a number of phonetic/phonological features of Southern American English, including the following:
Morphosyntactic features of Southern American English include the following:
-
Double Modals
-
The fixin' to Construction
There are also quite a few lexical distinctions that distinguish Southern American English from other varieties, including:
-
/z/ vs. /s/ in greasy
Southern Vowel Shift (SVF)
The following diagram illustrates the shifts in vowels that have occured in varieties of Southern States English and African-American English:
As the diagram shows, there are a number of shifts among the vowels:
-
the sound in words such as meet is shifting back and downwards
-
the sound in words such as mitt is shifting upwards and forwards, often becoming diphthongized, as is common among tense vowels in Standard English
-
the sound in words such as mate is shifting back and downwards
-
the sound in words such as met is shifting upwards and forwards, becoming diphthongized, as is common among tense vowels in Standard English
-
the sounds and in words such as moot and mote are shifting forwards
According to Wolfram and Schilling-Estes (1998), this shift is largely confined to rural areas of the areas of the Southern States traditionally defined as the South.
Feagin (1986) did a study of the Southern Vowel shift among speakers of English in Anniston, Alabama.
Feagin (1986) did a study of the Southern Vowel shift among speakers of English in Anniston, Alabama. The samples were recorded and the F1 and F2 formants plotted on graphs such as those below (on the graph below, move the mouse over any vowel to see that area highlighted):
As you can see from the graph above of the measurements taken of an older working-class male, the back vowels /u/ and /uw/ are very far forward, particularly in comparison with other back vowels such as /o/ and ow/. This illustrates the back shift in the Southern Vowel Shift. The front vowels, however, do not seem to be undergoing the vowel shift to a large extent. The /e/ vowel has definitely raised above the /ey/ vowel, indicating the front vowel shift, but the /iy/ and /i/ vowels have not shifted greatly.
Now, for comparison, observe the following graph of a teenage male from the same area:
As you can see from this graph, the back vowels have again shifted forward, and in some tokens of /uw/ are in the extreme front. The front shift is even more pronounced in this speaker, with a few tokens of /iy/ (namely she and keeps) still in the upper front region, but most tokens below and behind the /i/ region. The vowel /ey/ is also back and below the vowel /e/.
Feagin theorizes, based on findings like those above and others, that the back vowel shift is an older shift, since it is found in both older and younger speakers. The front shift, on the other hand, is a more recent shift, occuring more dramatically in the younger speakers.
northern
southern
Merger of /ɪ/ and /ɛ/ Before Nasal
Description
In varieties of Southern States English and African-American English, the vowel sounds [ɪ] and [ɛ] merge when they occur before nasal consonants. As a result, words like pin (SAE [pɛn]) and pen (SAE [pɪn]) both have the pronunciation [pɪn] in Southern States English.
Background
It has been theorized that the merger of [ɪ] and [ɛ] before nasals has its sources in British English (Krapp 1925; Kurath and McDavid 1961; Morgan 1969). Morgan (1969) states that "the substitution of the vowel for as in pin for pen" is derived from the British Isles and is common in the speech of North Carolina. Kurath and McDavid (1961) state the following:
"In English folk speech again as the [ɛ] of pen in the central counties...[and] the of pin in the eastern counties...The variants existed in Middle English, survive in English fold speech of today, and had a social rating in London English of the eighteenth century not unlike that in present-day American English."
Krapp (1925) also maintains that this phenomenon (which he finds in southwestern Virginia) is a relic of seventeenth-century colonial English.
Sample Study (Brown 1991)
Brown (1991) discusses the phenomenon of merger in Tennessee, with a particular focus on the spread of the phenomenon throughout the state since before the Civil War. She uses three primary sources for her data in this study. One source is a collection of questionnaires collected by a Tennessee archivist Gus Dwyer between 1918 and 1922. These questionnaires were collected from Tennessee veterans of the Civil War, and thus represent informants that were born between the years of 1820 and 1850. Brown uses the spellings of words such as friendlyas frindly to indicate instances of an informant that has merger. A second source for this study comes Tennessee surveys from LAGS, or the Linguistic Atlas of the Gulf States, the data in which was collected more recently and represents the speech of informants that were born between the years of 1872 and 1960. The third source was designed to fill in the gap between the questionnaires and the LAGS data. This data was from the North Carolina surveys of LAMSAS, or the Linguistic Atlas of the Middle South Atlantic States. Since the speech of North Carolina is very similar to the speech of Tennesee, and since there is data that is collected from informants born between the years of 1840 and 1900, Brown proposed that the insights into the evolution of merger during that time period can be extrapolated for
Tennessee during the same time period.
From these three sources, Brown has data pertaining to merger from informants whose birthdates range from 1820 to 1960. The study of these data have interesting results that may shed some light on how this phenomenon spread throughout Tennessee. Brown's study of the Civil War veterans questionnaires revealed the following statistics (table)
As the above table illustrates, for informants born between the years of 1820 and 1851, there is a higher incidence of no merger in words such as friendly. Brown takes this as an indication that 87.4% of these informants do not have the variant in such words. The following map shows the locations of the informants for this study (map)
As this graphic illustrates, the data collected regarding Civil War veterans seems to indicate that there was variation across most parts of Tennessee, with a few areas of merger. The LAMSAS data revealed the following statistics for informants in North Carolina (table)
This table seems to indicate that there is an increase in merger among speakers born between 1872 and 1900 (compare 1840-1855 12.5%/87.5% to 1872-1900 31.0%/69.0%).
The third source, the LAGS data, revealed the following statistics for informants in Tennessee born between the years of 1872 and 1960 (table)
This table again seems to indicate that there is an increase of merger between the years 1872 and 1960, where merger has increased from 46.2% to 91.2%. The following graphic is a map illustrating the results of the LAGS survey of the words ten, twenty, and pen (map)
The graphic above shows that the LAGS survey, which is more recent, finds more instances of merger across Tennessee. Also, Brown points out that the areas of merger are centered around waterways, roadways, and urban centers such as Nashville and Memphis. Consequently, the incidences of non-merger are relegated to outlying areas.
Based on the regional data, Brown proposes that the increase in incidence of merger among speakers may be due to the increase in urbanization in post-Civil War Tennessee. The proposal is that before the Civil War, there were a few incidences of merger among English and Irish settlers of little education. Among other English settlers, the non-merger form was predominant. However, after the Civil War, many rural speakers moved to urban centers to find work in the new manufacturing industries. At that point, the merger phenomenon may have been associated with the working class, but then lost its stigma and became predominant.