Chambers 2004 (1)

Embed Size (px)

Citation preview

  • 7/30/2019 Chambers 2004 (1)

    1/19

    Dynamic typology and vernacular universals

    J.K. Chambers

    Abstract

    Vernacular universals arise in the context of sociolinguistic dialectology as

    generalizations about intralinguistic variation, and their universal status is

    emerging from analyses of putative crosslinguistic counterparts. The external

    factors that underlie them have distinctive social and functional aspects. I

    exemplify them by examining one of them, default singulars, a specific type of

    copula nonconcord. In English, default singulars occur as invariant was (as in They

    was too sick to travel). Socially, default singulars appear to develop naturally in

    the absence of contact models, as dramatically illustrated by Schreiers work

    (2002) on Tristan da Cunha. Functionally, they appear to result from stripping

    away inflectional redundancies, especially when they involve complex look-up

    mechanisms. Vernacular universals, unlike UG-based generalizations, are

    identified partly in terms of their social patterning, in so far as there are regularities

    in the way they are socially embedded, and this added dimension may provide a

    concrete basis for coming to grips with them.

    1. Introduction

    2. Vernacular universals

    3. Dynamic typology

    4. Default singulars5. Linguistic constraints

    6. Finding the baseline

    7. Contact and language change

    8. Worldliness and speech

    9. Primitive and learned features

    10. Nonconcord as a natural

    tendency11. Grammatical constraints

    12. Variationist typology

    1. Introduction

    At the dawn of the Chomskyan era, many people believed that progress in

    linguistics would come from the discovery of principles that allowed

    generalizations across language boundaries. Thus Roman Jakobson said,

    We all seem to agree that linguistics is passing from the bare study of

    variegated languages and language families, through systematic

    TYPOLOGICAL research and gradual INTEGRATION, to become a thoroughly

    universal science of language (Jakobson 1963: 275). That historical thrust

  • 7/30/2019 Chambers 2004 (1)

    2/19

    128 J.K. Chambers

    was waylaid, however, by the rise of Chomskyan syntax into the

    mainstream in the 1960s, and by the parallel rise of sociolinguistics starting

    at the same time and reaching mainstream status soon after.

    Neither theoretical syntax nor sociolinguistic variation studies made

    explicit claims about possible tie-ins with language typology. Chomskys

    goals were implicitly typological, to the extent that they were oriented

    toward universal grammar. The typological void remains, for that school,

    an accidental gap. Sociolinguists, for their part, have been preoccupied with

    discovering the distribution of types and tokens in real communities, and

    only recently have begun to look beyond their own borders. As

    sociolinguistics becomes less restricted to local events, it becomes

    comparative and, as the comparative aspect gains weight, cross-linguistic

    generalizations not only become possible but inevitable. Those

    generalizations are typological, and, as I show below, have universalimplications.

    2. Vernacular universals

    Sociolinguists have amassed copious evidence in the past 35 years for a

    surprising conclusion: a small number of phonological and grammatical

    processes recur in vernaculars wherever they are spoken. This conclusion

    follows from the observation that, no matter where in the world the vernac-

    ulars are spoken Newfoundland, Harlem, Ocracoke, Ballymacarrett,

    Tyneside, Buckie, the Fens, the Falklands, inner-city Sydney these

    features inevitably occur.

    Their ubiquity has one of two possible explanations. Either the features

    were diffused there by the founders of the dialect, or they developed there

    independently as natural structural linguistic developments. As I have

    shown at greater length elsewhere (Chambers 2003: 266270), the

    diffusionist explanation is implausible because of geographic spread. It is

    also implausible linguistically, because these features occur not only in

    working-class and rural vernaculars but also in child language, pidgins,

    creoles and interlanguage varieties. Therefore, they appear to be natural

    outgrowths, so to speak, of the language faculty, that is, the species-specific

    bioprogram that allows (indeed, requires) normal human beings to become

    homo loquens.I have characterized these recurring natural processes as vernacular

    roots (Chambers 2003: 266270). The best candidates, based on their

  • 7/30/2019 Chambers 2004 (1)

    3/19

    Dynamic typology and vernacular universals 129recurrence, are listed below with alternative names and a simple English

    example:

    (ng) or alveolar substitution in final unstressed -ing, as in walkin,

    talkinand runnin.

    (CC) or morpheme-final consonant cluster simplification, as in pos

    office, hanful.

    final obstruent devoicing, as in hundret (for hundred), cubbert (for

    cupboard)

    conjugation regularization, or leveling of irregular verb forms, as in

    Yesterday John seen the eclipse andMary heared the good news.

    default singulars, or subject-verb nonconcord, as in They was the last

    ones.

    multiple negation, or negative concord, as inHe didnt see nothing. copula absence, or copula deletion, as in She smart or We going as

    soon as possible.

    Linguistically, these processes include some phonological ones (the first

    three) and some grammatical ones (the other four). That raises the ultimate

    challenge of bringing them together in a unified theory, though that may be

    a premature concern until the framework is better understood. Elsewhere, I

    have discussed morpheme-final consonant cluster simplification,

    symbolized (CC), and conjugation regularization (Chambers 2003: 258

    265), final obstruent devoicing (Chambers 2000), and multiple negation

    (Chambers 2001) in detail and explored some of their implications asvernacular roots. In this article, I will deal mainly with default singulars.

    3. Dynamic typology

    I have listed the vernacular universals with their English names and

    illustrated them with English examples. This is misleading. In so far as

    these processes arise naturally in pidgins, child language, vernaculars, and

    elsewhere, they are primitive features, not learned. As such, they belong to

    the language faculty, the innate set of rules and representations that are the

    natural inheritance of every human being. They cannot be merely English.

    They must have counterparts in the other languages of the world that aredemonstrably the outgrowths of the same rules and representations in the

    bioprogram.

  • 7/30/2019 Chambers 2004 (1)

    4/19

    130 J.K. Chambers

    Here, then, is where sociolinguistics intersects with typology. Luckily,

    just as variation studies have evolved since Jakobsons time, so have

    typological studies. Vernacular universals provide a potential resource for

    typologists working within the dynamic framework characterized by

    Kortmann (1997: 13): typology does not stop at accounting for (limits of)

    variation across languages, but tries to give a unified account of

    intralinguistic variation, crosslinguistic variation, and variation over time as

    essentially the same external factors are held to underlie all three types of

    variation. Like Kortmann, my goal is to make some progress toward

    Crofts dynamic paradigm, in which the study of all types of linguistic

    variation cross-linguistic (typology), intralinguistic (sociolinguistics and

    language acquisition) and diachronic (historical linguistics) are unified

    (Croft 1990: 258259).

    Vernacular universals arise in the context of sociolinguistic dialectologyas generalizations about intralinguistic variation (so far mainly from

    English dialects) but their universal status is emerging from analyses of

    putative crosslinguistic counterparts. The factors that underlie them have

    distinct cognitive and functional aspects. Socially, the vernacular universals

    appear to fall into well-defined patterns in the acrolect-basilect hierarchy,

    but functionally there appear to be several disparate principles at work

    (from motor economy to cognitive overload). Unifying the functional

    principles into a few empirically defensible cognitive strategies may be too

    much to ask of any branch of linguistics at this time, important though it is

    to try.

    Vernacular universals raise the same challenges for typological analysisas do UG-based generalizations. To establish a claim for universality, it is

    necessary to compare processes that occur in two or more languages. One

    challenge arises in finding (or developing) descriptions at comparable

    analytic depth in the two (or more) languages, and another arises from

    determining equivalence between crosslinguistic categories. Unlike

    grammar-based generalizations, however, vernacular universals are

    identified partly in terms of their social patterning, in so far as there are

    regularities in the way they are socially embedded, and this added

    dimension may provide a concrete basis for coming to grips with them.

    I can make many of these points concrete by looking in some detail at

    the process of default singulars.

  • 7/30/2019 Chambers 2004 (1)

    5/19

    Dynamic typology and vernacular universals 1314. Default singulars

    The grammatical phenomenon I am calling default singulars is

    exemplified in this sentence from Feagins research in Anniston, Alabama

    (1979: 202):

    I seen three rats, butthey was all too far off to shoot.

    Here, the subject of the second clause, they (= three rats), is plural, but the

    verb form was is singular. In traditional grammar terms, the verb fails to

    agree with the subject in number. Constructions like these are not

    acceptable in standard English dialects anywhere in the world except in

    highly restricted grammatical environments (discussed below). All standard

    dialects require number concord between subject and verb be. Thenonconcord pattern in which was occurs with all subjects, though

    nonstandard, occurs globally in vernacular dialects in various parts of the

    world. Britain (2002: 17) identifies this as the first and most common of

    two broad dominant patterns of past be across varieties of English,

    defined as follows:

    Vernacular Pattern 1: WAS occurs variably for standard WERE

    throughout the paradigm, both affirmative and negative.

    Its global distribution is evident from the fact that it is reported in Sydney,

    Australia (Eisikovits 1991), Buckie, Scotland, Nova Scotia, Saman (allTagliamonte and Smith 2000), all varieties of African-American English,

    and many other vernaculars.

    There is a second pattern identified by Britain (2002: 19) as follows:

    Vernacular Pattern 2: WAS occurs variably for standard WERE

    in affirmatives, and WERENT in negatives.

    Whereas in Pattern 1, the negated verb remains singular the boys was

    interested, but the girls wasnt in the second pattern the verb form

    changes when it is negated the boys was interested, but the girls werent.

    It also occurs with singular subjects when the verb is negated, as in Johnny

    werent interested at all. This pattern also has fairly wide distribution inEnglish vernaculars, being reported in North Carolina varieties (Ocracoke,

    Lumbee, etc., in Schilling-Estes and Wolfram 1994; Wolfram and Sellers

  • 7/30/2019 Chambers 2004 (1)

    6/19

    132 J.K. Chambers

    1999), Reading (Cheshire 1989), and the Fens (Britain 2002); for a Britain-

    wide survey, see Anderwald (2002). It remains to be seen whether its

    distribution is global.

    What I call default singular explicitly refers to Vernacular Pattern 1.

    That is, it is purposely intended to specify the co-occurrence of the

    unmarked verb form (the singular in standard dialects) with all subjects,

    regardless of number. There is ample evidence that it is the basic vernacular

    system, with was and wasnt in all persons and numbers. Clearly,

    Vernacular Pattern 2, with was in affirmative and werentin negative, adds

    a grammatical complication by, in effect, requiring the bound suppletive

    form were when the negative clitic occurs. It represents movement in the

    direction of a concord system, and in fact imports the standard plural

    concord form (were) to do so. It appears to be at least one step removed

    from the basilect. Wolfram and Sellers (1999) add evidence from diffusionto arrive at the same conclusion. They note that leveling to was [i.e.,

    Vernacular Pattern 1] may occur as an independent innovation, but

    Vernacular Pattern 2 is learned in the sense that cases of leveling to

    were/nt in the United States appear to be traceable to influence from

    British-based donor dialects (1999: 109).

    Vernacular Pattern 1 is simpler, and, by definition, more basic. If my

    assumption is correct, then eventually we should find evidence in some

    vernacular dialects for Pattern 2 developing out of Pattern 1. As we will see

    below, Schreier (2002) provides evidence for Vernacular Pattern 1 coming

    into being spontaneously in an isolated speech community. This is further

    evidence for its basilectal status.

    5. Linguistic constraints

    We now know that the nonstandard default singulars and the standard

    concord patterns are poles on a continuum. In between, there is a graded

    hierarchy in which concord occurs more frequently with certain types of

    subjects than others. Two of them are so well known as to have their own

    names (Britain 2002: 1920):

    The existential constraint (E below): WAS is most frequent after

    there.

  • 7/30/2019 Chambers 2004 (1)

    7/19

    Dynamic typology and vernacular universals 133Northern Subject Rule (B below): WAS is more frequent after

    nonpronominal plural nouns (NPpl) than after pronouns.

    So robust is the existential constraint (E) that it even intrudes into standard

    dialects, allowing nonconcord variants such as Theres too many

    McDonalds in Helsinki alongside regular constructions like Therere too

    many. Nonconcord after expletive there is the only exception to invariant

    concord in standard dialects (and even such nonconcord variants are often

    inveighed against in usage guides, notwithstanding their frequency in

    conversation). The Northern Subject rule (B) is named for the English

    region where it appears to have become established as a rule-like process.

    According to the Northern Subject Rule, plural nouns (as opposed to

    pronouns) prohibit or at least inhibit concord in some dialects.

    In actual speech communities, other subject types are graded along withplural nouns (NPpl) and existentials (there) in terms of the frequency of

    default singulars that co-occur with them. The hierarchy was first

    established, in my experience, by Feagin (1979), in her study of Anniston,

    Alabama. She showed that the subject-types were graded from (A) to (E),

    where (A) is the most inhibiting context for default singulars and (E) is the

    most permissive, as in the following examples (slightly simplified from

    Feagins originals as in Chambers and Trudgill 1998: 134136):

    (A) they They was all born in Georgia, mama and my daddy

    both.

    (B) NPpl All the student teachers was comin out toWellborn.

    (C) we We was in an ideal place for it.

    (D)you You was a majorette?

    (E) there There was about twenty-somethin boys and just

    four girls.

    The hierarchy is implicational. If a dialect permits default singulars afterwe

    (C), then it permits them in the subject categories below it (D and E), but

    not necessarily vice-versa. While this hierarchy holds for Anniston, there

    is, as we shall see below, some variability in the hierarchy in different

    communities. As always, the most predictable constraints are found at the

    poles of the dialect continuum, the acrolect and basilect. Standard dialects(acrolects) permit default singulars variably in (E), but nowhere else (and

    there are no categories below it). The deepest basilect variety has default

  • 7/30/2019 Chambers 2004 (1)

    8/19

    134 J.K. Chambers

    singulars in all contexts from (A) to (E), apparently invariably; in other

    words, basilects have no concord rules at all. I can say this with greater

    confidence than ever before because of recent studies of the English spoken

    on Tristan da Cunha by Daniel Schreier (2002), for reasons discussed in the

    sections below.

    6. Finding the baseline

    Tristan da Cunha (TdC) can be called the sociolinguists Galapagos. It is a

    South Atlantic island, and perhaps the most isolated inhabited territory in

    the world. It was colonized in 18201840 by British army personnel,

    shipwrecked sailors and American whalers (Schreier 2002: 77). The arrival

    of women from St. Helena Island in 1827 (Schreier 2002: 92) led to thefirst generation of native Tristanians in the years immediately following.

    After 1869, when the Suez Canal was completed, Tristan da Cunha became

    increasingly isolated, because the canal diverted traffic that formerly went

    around the Cape and passed within hailing distance. In the half-century

    from 1850 to 1900, the island received only three new settlers a weaver

    from Yorkshire and two Italian sailors (Schreier 2002: 77).

    The worst was yet to come. Schreier (2002: 77) says, The sociocultural

    isolation of TdC peaked in the early twentieth century. Between 1900 and

    1940, the Tristanians were pretty much on their own. There was no mail at

    all for one ten-year period, and no ships for three years (Schreier 2002: 93).

    Left to their own devices, the islanders developed strong networks and

    extremely dense and multiplex ties (Schreier 2002: 93). There was only

    one village, no formal education (until 1942), almost indistinguishable

    gender roles, and, Schreier says, no social ranks.

    In this closed, egalitarian society, the disparate dialects of the

    immigrants were leveled by their offspring into a more or less uniform or

    focused variety. It is predictable that New-World societies will go through

    an astounding linguistic homogenization in the first generation (Chambers

    2003: 65). In TdC, with virtually no social barriers and everyone in close

    proximity every day, the homogenization may well have been nearly

    complete, with the Tristanians shucking off even subtle variants in favour

    of neighbourly norms.

    One of the more striking developments involved default singulars.Colonists dialects were known to have had variable concord marking,

    Schreier says (2002: 91): that is, both was and were could occur with

  • 7/30/2019 Chambers 2004 (1)

    9/19

    Dynamic typology and vernacular universals 135plural subjects in the English of the immigrant population. However, the

    offspring of the immigrant generations, the native Tristanians, eliminated

    concord from the dialect and adopted invariant was regardless of subject

    number. That is, default singulars became the rule, categorically.

    It is important to recognize that, as Schreier (2002: 92) says,

    Categorical was regularization in early twentieth-century TdCE was

    innovative rather than retentive. In other words, default singulars

    developed as an invariant feature out of diverse source dialects in which it

    had been only one of the variants. Schreier develops the following scenario:

    Once homogeneity [of categorical was] emerged and was in place, it was

    sustained by external factors such as immobility, isolation, and close-knit

    social networks. the interplay of these factors leads to acceleration

    (or, in extremis, to the completion) of language-inherent changes and to the

    thriving of vernacular roots (2002: 93).Significantly, as Schreier makes clear in this quotation, he has

    discovered a community in which default singulars, one of the vernacular

    roots as described above, diffused throughout the community as a natural

    development. The social conditions were decidedly eccentric, with total

    absence of mobility, almost no contact with outside groups and no

    immigration for several generations. Because there was little formal

    education and limited literacy, there was not even a codified standard

    language to measure the vernacular against.

    TdC society seems to be rudimentary even compared to pidgin and

    creole societies, because in the latter there are iron-clad social ranks and

    occupational hierarchies, with the more-or-less standard speaking bosses atone end and the creole-speaking workers, often slaves historically, at the

    other. Before Schreier, sociolinguists perhaps never came upon a

    community in which the basilect developed in the complete absence of an

    acrolectal superstrate. The superimposition of an acrolect has well-known

    consequences. It means that many basilectal speakers will accommodate to

    some processes from the higher styles, resulting in the creole continuum.

    It also means that, by the time linguists come along with their tape

    recorders and notebooks, basilectal forms are liable to be mixed with

    intrusive variants by the process of decreolization. As far as I know, TdC in

    the period from 18501945 is the only society ever studied in which there

    was no stylistic continuum and no accommodation. The vernacular that

    developed there out of the original dialect mixing had a kind of free choicein the absence of varieties that carried social values of prestige, learning or

    power.

  • 7/30/2019 Chambers 2004 (1)

    10/19

    136 J.K. Chambers

    7. Contact and language change

    The extreme isolation in which the native TdC vernacular developed could

    not last. Since 1950, improved transportation, first with speedier water

    travel and then with air transport, made the island more accessible to

    outsiders and gave the islanders access to the outside world. Education

    programs were instituted and British officials were posted there as teachers

    and governors, joining the adventurers and wayfarers who found their way

    to the island.

    The linguistic results are predictable. As Schreier (2002: 94) points out,

    variation began occurring after the recent emergence of the community

    from insularity. The interlopers used non-local forms, and literacy, which

    encodes standard grammar as its norm and prescribes it, became general in

    a couple of generations. Standard forms began seeping into local speech,not with the force of a knockout but variably, at first as minor affectations

    in the speech of a few Tristanians, and then more commonly, until they

    eventually went unnoticed.

    So default singulars were no longer invariable. For the generation born

    and raised after 1950, it was possible to say They were tired out and we

    were too, though in some sense everyone knew (and still knows) that the

    normal, unaffected way of saying it is They was tired out and we was too.

    Schreier (2002: 85, 88) compared the use of those Tristanians born and

    raised in the isolated period with younger ones. Figure 1 shows the

    difference. The traditional Tristanians almost never use the concord form,

    were, as expected; but the younger Tristanians sometimes do. (Schreierprovides results for women only in the younger group, and they are, he

    notes, ahead of men in the same age group, as expected in this

    standardizing change.) The gradualness of the change is evident in the

    infrequency with which the concord form occurs in the speech of the

    younger people never more than half the time and usually much less.

    The highest frequency for standard concord occurs in sentences with

    you subjects (as in You were tired), at 50 percent. This violates the subject

    hierarchy discussed in section 4 above, which is expected to show

    descending frequencies from the left to the right on the X-axis, from they as

    the most frequent subject for concord to there as the least. Apart from the

    unruly behaviour ofyou, however, the hierarchy holds. With that lone

    exception, the introduction of subject-verb concord in TdC English isproceeding according to the subject hierarchy established in faraway places

    like Anniston and inner-city Sydney.

  • 7/30/2019 Chambers 2004 (1)

    11/19

    Dynamic typology and vernacular universals 137

    Figure 1.Percentage of standard were with grammatical subjects in Tristan daCunha in isolated period (18501945) and after (based on Schreier 2002:

    Table 3, 85 and Table 5, 88; TdC after 1950 percentages are calculated

    by combining Schreiers Young and Middle groups in Table 5, p. 88).

    8. Worldliness and speech

    Emergence from extreme isolation in Tristan da Cunha brought numerouschanges, and some of them are linguistic. As in the case of default

    singulars, the local norm is giving way variably to norms imported from

    outside. The local norm remains, as we have seen, but it is now one

    possibility among others. In this respect, Tristan da Cunha is not unique in

    any way. There is a law-like relationship between what might be called

    worldliness and speech. Put simply, the more urban and mobile the social

    setting, the more standard the speech.

    Sociolinguistic research on default singulars by several linguists in

    widely separated communities over two or three decades provides the most

    striking empirical evidence for urbanity and mobility as promulgators of

    standard speech. Schreier (2002: Table 3, 85) tabulates several studies ofnonconcord. By a happy accident, it turns out that the studies have taken

    place in settings distinguishable in terms of size, urbanization and

    they NPpl we you there0

    10

    20

    30

    40

    50

    60

    70

    80

    90

    100

    %standardWERE

    Grammatical subject

    Tristan da Cunha 1850-1945

    Tristan da Cunha after 1950 (womenonly)

  • 7/30/2019 Chambers 2004 (1)

    12/19

    138 J.K. Chambers

    movement. Sydney, Australia (Eisikovits 1991) and York, England

    (Tagliamonte 1998) are large cities in highly urbanized regions; Anniston,

    Alabama (Feagin 1979) is a mid-sized city at the hub of a rural region in

    the American south; the Fens (Britain 2002) are an agricultural region in

    the east Midlands of England; and Tristan da Cunha (Schreier 2002) is the

    closeted Atlantic island. The studies of default singulars in these places

    were undertaken independently, without reference to one another, and

    separated by vast geographical distances. They are comparable because of

    their shared methodologies, as quantitative variationist studies.

    The comparison as shown in Figure 2 makes a stunning demonstration

    of the dialect law of worldliness and speech. The communities show a wide

    range of variation, from nearly pure standard concord (100 percent) to its

    complete absence. The communities are stratified, and the stratification

    correlates robustly it is not hyperbolic to say brilliantly, even exquisitely with urban complexity, from York and Sydney at the standard extreme,

    through Anniston and the Fens in the mid region, to Tristan da Cunha at the

    bottom. Equally brilliant is the discovery that the grammatical subject

    hierarchy holds in all these disparate settings with only a few minor

    discrepancies.

    Figure 2. Percentage of standard were with grammatical subjects in five commun-

    ities (based on data compiled by Schreier 2002: Table 3, 85)

    c

    c

    cc

    c

    they NPpl we you there

    0

    10

    20

    30

    40

    50

    60

    70

    80

    90

    100

    %standardWERE

    Grammatical subject

    Tristan da Cunha (1850-1945)

    Fens

    Anniston

    Sydney

    York

    c

    c

  • 7/30/2019 Chambers 2004 (1)

    13/19

    Dynamic typology and vernacular universals 1399. Primitive and learned features

    The intrusion of standard forms in default singulars in Tristan da Cunha (as

    in Figure 1) provides the best empirical evidence so far for the theory of

    Vernacular Roots. The core of the theory maintains that the standard

    dialect differs from other dialects by resisting certain natural tendencies in

    the grammar and phonology (Chambers 2003: 254). The basilectal form is

    PRIMITIVE, part of the innate bioprogram, and the standard is LEARNED, an

    experiential excrescence on the bioprogram. Finding the basilect in what

    might be called a pure state is rare, because human well-being entails

    outbreeding and social intercourse and also perhaps barter and territorial

    expansion, all of which break down barriers. Socialization beyond the tribe

    requires infrastructure, including ranks and protocols. It also, evidently,

    fosters linguistic norms, superimposed on primitive forms. Kroch (1978:18)was perhaps first to recognize that language standardization is ideological,

    with learned linguistic constraints originating as markers of status and rank.

    Figure 2, with its strata of linguistic complexity layered according to social

    complexity, provides a graphic demonstration.

    Finding the basilectal form in its unadulterated state is uncommon.

    Pidgins make the most likely sites, as Bickerton (1981) has argued, but by

    the time linguists arrive pidgins are creolized at the very least, and the

    purest basilectal form must be inferred from the creole continuum.

    Schreiers discovery on Tristan da Cunha represents a breakthrough

    because the default singulars in their pure basilectal form there are

    palpable. Moreover, he has shown that they came into being as a naturalregression to primitive linguistic instincts, a conclusion that follows from

    an apparent-time inference of very shallow time depth and therefore seems

    virtually incontrovertible.

    10. Nonconcord as a natural tendency

    The theory of vernacular roots rests squarely on the assumption that

    basilectal forms are in some demonstrable sense more natural than

    standard forms. Demonstrating naturalness seems relatively easy for

    phonological processes compared to grammatical ones (and perhaps for that

    reason Kroch 1978 considered only phonology). For example, consonantcluster simplification (CC) represents economies in articulatory (motor)

    gestures that are quantifiable (Chambers 2003: 258259). Principles that

  • 7/30/2019 Chambers 2004 (1)

    14/19

    140 J.K. Chambers

    might underlie natural grammatical processes are never so concrete. In

    previous discussions, for conjugation regularization I have proposed an

    underlying principle of cognitive overload (Chambers 2003: 260, based on

    Pinker and Prince 1999, whereby idiosyncratic retrieval of strong forms

    carries higher cost than rule-governed or inferable generalizations), and for

    multiple negation an underlying principle of compositionality (Chambers

    2001, based on Giannikidou 2000, whereby the morphosyntax carries more

    than one negative marker but the semantic interpretation is a single

    negative; by compositionality Giannikidou means that semantic

    interpretation should be determined by morphosyntax).

    For default singulars, a plausible principle suggests itself readily.

    Nonconcord is structure-independent. Looked at from the top down, so to

    speak, concord in standard dialects exacts a processing cost with no

    information gain. Specifying number (and in some languages person andgender) in the verb replicates information already explicit in the subject.

    Putting the inflectional markers in place requires identifying the subject,

    analyzing its properties, and encoding some of those properties as

    inflectional morphemes in the verb. It is structure-dependent in that it

    requires a look-back mechanism to match number in the verb with number

    in the subject. It gets more complicated when the subject is a dummy, like

    there, because it then requires a look-forward mechanism to find the

    number of the semantic subject that there represents (a point I return to

    below). The information encoded by concord markers in the verb is

    absolutely redundant.

    Avoiding absolute redundancy seems like a sound principle for any kindof processing device, including the language faculty. In that sense, it has a

    kind of intuitive credibility that is about the same as motor economy as an

    explanation for (CC). It is perhaps a bit disquieting, I admit, to realize that

    it adds yet another explanation for the naturalness of primitive processes,

    along with the aforementioned cognitive overload and compositionality.

    Offsetting that, I must say that these principles do not appear to be

    unrelated. It seems to me that notions like these might ultimately be seen as

    belonging to one or perhaps two more general principles of linguistic

    cognition. What those general principles might look like awaits a day when

    they are better understood, perhaps in some unforeseeable time when an

    enlightened psycholinguistics will make real headway on linguistic

    processing. It is important, it seems to me, to seek naturalness explanationsif only to accumulate best-guesses that might eventually stimulate more

    concrete explorations into the workings of the language faculty.

  • 7/30/2019 Chambers 2004 (1)

    15/19

    Dynamic typology and vernacular universals 14111. Grammatical constraints

    Turning to the grammatical constraints, the remarkably regular hierarchy of

    subject-types that promote or inhibit verb concord, we can see that

    hierarchy as a kind of evolutionary scale along which concord replaces

    basilectal nonconcord. Though the hierarchy is remarkably orderly across

    heterogeneous social and geographical boundaries, it will be evident that

    attempts to explain the hierarchy differ in kind from attempts to explain

    primitive processes because they require appeals not to general cognitive

    processes but to specific English grammatical features.

    Starting with expletive there, the subject form that promotes nonconcord

    all the way up to the acrolect, there itself is inherently numberless, and thus

    dictates nothing at all for agreement marking. Concord in there-sentences

    comes from the following subject NP, which requires a look-aheadprocessor, but other concords require looking back. On both counts

    inherent numberlessness of there, and the unique look-ahead processor to

    the logical subject there promotes nonconcord. Expletive subjects other

    than there, in dialects that have them, promote nonconcord in exactly the

    same way. Tristan da Cunha and many other dialects have expletive it, as in

    At that time it was no gas stoves (Schreier 2002: 84). Unlike there,

    expletive it is inherently singular, but like there, concord requires a look-

    ahead processor. In theory, it should promote nonconcord more strongly

    than there, and it probably does.

    The second most resistant site for concord is afteryou subjects. In

    English,you is syncretically singular and plural, since the 15th century lossof singular thou. Its ambiguity may occasionally confound concord in the

    stream of speech.

    The other subjects, we, they and NPpl, have no semantic or grammatical

    feature that distinguishes among them, or at least none that is discernible so

    far. As such, their order in the hierarchy should be arbitrary, and one of the

    empirical consequences should be that they are ordered differently in

    different communities. In fact, they appear to be re-ordered (or dis-ordered)

    in Buckie Scots and Saman (Tagliamonte and Smith 2000). Another

    intimation of arbitrariness at the high end of the hierarchy is visible in

    Figure 2 above, where Sydney and Anniston show a slight preference for

    NPpl overthey, contrary to the expected order.

    My main point in all this is to show that there is a fundamentaldifference between attempts at explaining the primitiveness of default

    singulars and attempts at explaining the subject hierarchy that promotes

  • 7/30/2019 Chambers 2004 (1)

    16/19

    142 J.K. Chambers

    them. The latter is accidental, couched as it is in language-specific

    idiosyncrasies such as numberless there and syncretic you. Because the

    hierarchy hinges upon English-specific facts, it has little generality, and

    obviously no claim to universality. Other languages may (or may not) have

    hierarchies of subject forms that promote nonconcord, but it would be a

    great coincidence if their hierarchies were the same as in English, or even

    similar. They too, presumably, would be determined by facts about

    language-specific morphemes and syntax.

    Default singulars and the other root vernacular features, by contrast,

    cannot be language-specific. They must be universal. As primitive features

    of the language faculty, they cannot be English-specific. Default singulars

    presumably arise naturally in all languages just as they did in Tristan da

    Cunha, and get overlaid with learned features in the process of

    standardization as they did in Sydney and York. My discussion of thesematters depends on English much more than I would like, but careful,

    detailed, broad-based studies of vernaculars exist for English and are

    virtually nonexistent for other languages. (Even non-English sociolinguists

    tend to work on English vernaculars rather than vernaculars closer to home,

    as Edgar Schneider, Heinrich Ramisch, Terttu Nevalainen, Jean-Marc

    Gachelin, Mieko Ogura, and many other distinguished scholars.) There

    may be a feudal residue behind this. I once asked a German professor how

    multiple negatives worked in his language, and when it became clear to him

    that I was asking him about the speech he might have heard from house

    painters and auto mechanics, he replied that he did not think such people

    would use such language in his presence, and certainly not to him.Mechanics and other workers sometimes do use such language with

    professors in North America and many other English-speaking regions.)

    Stuck with adequate analyses mainly from English, I take some solace in

    Chomskys precedent for using available resources with the aim of finding,

    as he put it (1968: 24), conditions that are not accidentally true of the

    existing human languages, but that are rather rooted in the human

    language capacity, and thus constitute what counts as linguistic

    experience.

    12. Variationist typology

    The variationist typology that is implied by this programmatic outline

    obviously needs an international cadre of variationists if it is going to be

  • 7/30/2019 Chambers 2004 (1)

    17/19

    Dynamic typology and vernacular universals 143developed. It must be cross-linguistic, as are all linguistic typologies, in the

    search for default singulars, multiple negatives, consonant cluster

    simplification, and the other vernacular roots as they occur in different

    languages. It must be comparative, as all typology is, in order to discern

    what the processes share in two or more languages that allows them to be

    equated as the same process. Ultimately, the analysis of vernacular

    processes resolves into the search for basilectal forms. That cannot be

    undertaken directly, except in extraordinary circumstances of the kind that

    Schreier unearthed in Tristan da Cunha. Most often, base forms will be

    obscured by the learned excrescences that accumulate in standardization.

    Understanding the constraints that govern the highly restricted variety

    known as the standard makes an instructive starting point, and vernacular

    gradations should peel them away in structural layers. Those layers might

    be hierarchical, as in English default singulars, and the hierarchy might bedetermined by language-specific properties, also as in English default

    singulars. If they are language-specific, they belong to what Chomsky calls

    the accidental conditions of human languages. The search for vernacular

    roots, by contrast, belongs to the essence of language. When we glimpse

    that essence, the core that underlies default singulars and other vernacular

    processes in all languages, it should be possible to inquire into the biology

    of the language faculty, its cognitive and physiological basis.

    I intend this program as a concrete method of proceeding, but at the

    same time I recognize it as one more attempt at realizing an age-old ideal in

    the study of language. Jakobson (1963: 264) again: We [want to] see

    emerging ever new, unforeseen, but henceforth perfectly discernibleuniformities of universal scope, and we are happy to recognize that the

    languages of the world can actually be approached as manifold variations

    of one world-wide theme human language.

    Acknowledgements

    At the Methods XI symposium where this article originated, I benefited

    from astute comments by David Heap, Peter Trudgill and Aila Mielikainen.

    This written version has been greatly improved by the formal scrutiny of

    Melitta Cocan and the critical scrutiny of Lieselotte Anderwald, Lukas

    Pietsch and Bernd Kortmann.

  • 7/30/2019 Chambers 2004 (1)

    18/19

    144 J.K. Chambers

    References

    Anderwald, Lieselotte

    2002 Negation in Non-Standard British English. (Studies in Germanic

    Linguistics 8.) London/New York: Routledge.

    Bickerton, Derek

    1981 Roots of Language. Ann Arbor: Karoma.

    Britain, David

    2002 Diffusion, leveling, simplification and reallocation in the past tense

    of BE in the English Fens. In: Lesley Milroy (ed.), Investigating

    Change and Variation Through Dialect Contact. Special issue of

    Journal of Sociolinguistics 6: 1643.

    Chambers, J.K.

    2000 Universal sources of the vernacular. In: Ulrich Ammon, Klaus J.

    Mattheier and Peter H. Nelde (eds.), The Future of EuropeanSociolinguistics, 1115. (Special issue of Sociolinguistica:

    International Yearbook of European Sociolinguistics 14.) Tbingen:

    Niemeyer.

    2001 Vernacular universals. In: Josep M. Fontana, Louise McNally, M.

    Teresa Turell and Enric Vallduv (eds.), ICLaVE 1:Proceedings of

    the First International Conference on Language Variation in

    Europe, 5260. Barcelona: Universitat Pompeu Fabra.

    2003 Sociolinguistic Theory: Linguistic Variation and Its Social

    Implications. Oxford, UK/Malden, US: Blackwell.

    Chambers, J.K. and Peter Trudgill

    1998 Dialectology,2nd ed. Cambridge: Cambridge University Press.

    Cheshire, Jenny1989 Urban British dialect grammar: The question of dialect leveling.

    English World-Wide 10: 185225.

    Chomsky, Noam

    1968 Language and Mind. New York: Harcourt, Brace and World.

    Croft, William

    1990 Typology and Universals. Cambridge: Cambridge University Press.

    Eisikovits, Edina

    1991 Variation in subject-verb agreement in Inner Sydney English. In:

    Jenny Cheshire (ed.), English Around the World: Sociolinguistic

    Perspectives, 235256. Cambridge: Cambridge University Press.

    Feagin, Crawford

    1979 Variation and Change in Alabama English: A Sociolinguistic Study

    of the White Community, 186208. Washington, D.C.: GeorgetownUniversity Press.

  • 7/30/2019 Chambers 2004 (1)

    19/19

    Dynamic typology and vernacular universals 145Giannakidou, Anastasia

    2000 Negative concord and the scope of universals. Transactions of the

    Philological Society 98: 87120.Jakobson, Roman

    1963 Implications of language universals for linguistics. In: Joseph H.

    Greenberg (ed.), Universals of Language, 263278. Cambridge,

    MA.: MIT Press.

    Kortmann, Bernd

    1997 Adverbial Subordination: Typology and History Based on European

    Languages. Berlin/New York: Mouton de Gruyter.

    Kroch, Anthony

    1978 Toward a theory of social dialect variation.Language in Society 7:

    1736.

    Pinker, Steven and Alan Prince

    1999 The nature of human concepts: Evidence from an unusual source.In: Ray Jackendoff, Paul Bloom and Karen Wynn (eds.), Language,

    Logic and Concepts: Essays in Memory of John Macnamara, 221

    261. Cambridge, MA: MIT Press.

    Schilling-Estes, Natalie and Walt Wolfram

    1994 Convergent explanation and alternative regularization patterns:

    Were/werent levelling in a vernacular English variety. Language

    Variation and Change 6: 273302.

    Schreier, Daniel

    2002 Past be in Tristan da Cunha: The rise and fall of categoricality in

    language change.American Speech 77: 7099.

    Tagliamonte, Sali

    1998 Was/were variation across the generations: View from the city of

    York.Language Variation and Change 10: 153191.

    Tagliamonte, Sali and Jennifer Smith

    2000 Old was, new ecology: Viewing English through the sociolinguistic

    filter. In: Shana Poplack (ed.), The English History of African

    American Speech, 141171. Oxford, UK/Malden, US: Blackwell.

    Wolfram, Walt, and Jason Sellers

    1999 Ethnolinguistic marking of past be in Lumbee Vernacular English.

    Journal of English Linguistics 27: 94114.