Aspects of the Morphology and Syntax of Ulster-Scots

Michael Montgomery

Elsewhere in this volume can be found an extensive bibliography of published writing on Ulster-Scots, compiled and annotated by this author and John G. W. Erskine.(1) The items noted there fall largely into four categories: 1) studies of pronunciation (represented most prominently and authoritatively by the work of Robert J. Gregg, as a result of which far more is known about the pronunciation of Ulster-Scots than about its other components); 2) works on vocabulary (including glossaries, local word-lists, word studies, and the like dating from the late 19th century); 3) studies of Ulster-Scots literature and culture, which provide larger contexts from the 18th century to the present for understanding traditional Ulster-Scots speech; and 4) items examining or inspired by the recent Ulster-Scots revival movement and the cultural and linguistic politics accompanying it over the past decade and a half.

Notably under-represented, however, is research on the grammar of Ulster-Scots. In his doctoral fieldwork Gregg investigated a cluster of grammatical forms — auxiliary and modal verbs and their negatives (hae, cannae, etc.), principal parts of verbs (break, begin, give, take, etc.), and combinations of modal auxiliary verbs and phrases and their negatives (will can, 'll no can, etc.) — because he anticipated that their patterning, like that of vowels and the consonant /x/ in nicht, richt, etc., would reveal a boundary between Ulster-Scots and Ulster-English speech areas. Gregg used that material with considerable success to achieve his mapping objectives and reproduced his raw data in its entirety in his published Ph.D. thesis.(2) As invaluable as that material is, it includes only three small areas of grammar, with no structural analysis or commentary. Until the appearance in 1997 of Philip Robinson's volume Ulster-Scots: A Grammar of the Traditional Written and Spoken Language,(3) which draws copiously on both excerption of historical sources (especially the work of 18th-/19th-century Rhyming Weaver poets) and observations of modern speech in Counties Antrim and Down, there was virtually no other information on the subject. Robinson's aim was mainly descriptive (to give the grammar of traditional Ulster-Scots long overdue, comprehensive attention), but also prescriptive (to set forth, especially with regard to orthography, a form that was reasoned and historically based, as an exercise in language planning and status building for Ulster-Scots). For reference purposes as well as for many others, Robinson's volume will remain a standard work for the foreseeable future.

Because otherwise grammatical patterns have received little attention, especially in a comparative framework, this writer has sought to expand the descriptive base of knowledge about contemporary Ulster-Scots by an in-depth investigation of morphological and syntactic features. The present survey details a number of these and, where appropriate, compares their patterning to Lowland Scots as well as to Ulster English and Irish English.(4) Although not motivated primarily by theoretical concerns, it proposes rules and generalizations where these are justified.

Found as it is in northern reaches of Ireland, Ulster-Scots descends originally from the speech of Lowland Scots who in the 17th century came to Ulster in several waves, crossing a channel which is barely twelve miles wide at its narrowest stretch.(5) It developed mainly as a 'variant of West-Mid Scots' from Ayrshire and Renfrewshire,(6) with additional elements also from the South Mid Scots of Galloway and Kirkcudbrightshire.(7) Subsequently Ulster-Scots has influenced, and been influenced by, varieties of English brought during the Plantation period originally to more interior parts of Ulster (principally from the northwest Midlands of England), and by the Irish language either directly or through varieties of English in Ulster. The Irish influence, which cannot always be distinguished from that produced by earlier, centuries-long contact in Scotland (e.g. beltane 'May 1' is documented as early as 1424, borrowed into Lowland Scots from Scottish Gaelic),(8) is especially prominent in vocabulary.(9) It is unambiguous as well in pronunciation(10) and in grammar directly or indirectly (as with the second-person plural pronoun yous and the habitual verb be/bes).(11) These later influences from English and Irish and other developments are what distinguish Scots in Ulster from Scots in Scotland today. They are products of the most dynamic language contact zone in the British Isles over the past four hundred years (perhaps most dramatically in County Donegal).(12) Even in its most traditional form, Ulster-Scots has been far from the isolated, conservative variety suggested by the rural communities in which it is found today.(13)

Tens of thousands of people speaking Ulster-Scots or Scots-influenced Ulster English migrated to North America in the 18th century, as a result of which many grammatical patterns in American English can be traced to Ulster and from there ultimately to Scotland and northern England.(14) Historically speaking, Ulster formed an important linguistic bridge between Britain and North America.(15) Because Ulster-Scots has steadily diverged from the mainland variety from which it sprang, especially to converge with Ulster English (and vice versa), the linguistic patterns of its most conservative speakers today cannot be assumed to reflect the emigrant language of more than two centuries ago without the confirmation of earlier historical sources from one or both sides of the water, such as emigrant letters, where this is possible.(16)

The present study is based on a series of elicitations conducted with nine native speakers of traditional Ulster-Scots in County Antrim whom the author met or developed friendship with in the course of research in Northern Ireland between 1997 and 2004. These individuals were Mr. James Fenton (a native of Ballinaloob and Drumdarragh townlands in north Antrim) and eight of his long-standing north Antrim and east Antrim consultants who participated in the project to collect traditional Ulster-Scots material that led to his dictionary, The Hamely Tongue: A Personal Record of Ulster-Scots in County Antrim.(17) Fenton nominated these eight people from among the several dozen he had visited and interviewed for more than twenty years. In many ways ideal for the investigation undertaken here, these individuals were quick, confident, and certain in their judgments, having each responded to hundreds of queries by Fenton about the grammar, vocabulary, pronunciation, semantics, and phraseology of traditional Ulster-Scots.(18) For the researcher they represented the best of both worlds, in that they were little traveled and close to their roots, but thoroughly accustomed to answering questions about their speech. Thus seasoned to the process of recalling, judging and articulating their own usage and that of their families and local communities, they needed no coaching in objectively contrasting their native speech habits with the Standard English used in wider spheres in the British Isles.

To queries about grammatical patterns respondents showed no overt effect (in the form of hesitation, self-correction, or otherwise) of the stigmatization of their speech by the educational system. Because their experience in the classroom would have taken place a half century or more earlier, this should probably not be a surprise. However, they were quite clearly aware of English as the tongue of power and authority; in Northern Ireland it is identified with Belfast and governmental and other institutions. Awareness of the difference in status was seen in their ability to style shift readily. One verity that researchers on Ulster-Scots (and no doubt other minority language varieties) discover is that speakers are guarded in the presence of, or especially when conversing with, strangers and often shift automatically to English. In conversational interaction before and after questioning, speakers in this study frequently pronounced words differently or used different forms when addressing the investigator than when addressing Fenton (e.g. town and doesn't in the former case, toon and disnae in the latter). The fact that all individuals in this study — among the most conservative one might hope to find — were multi-style speakers presented a classic case of the observer's paradox,(19) which was overcome through the mediation of Fenton. A vernacular native speaker himself, he assisted the investigator at all sessions (which usually took place in the living room of respondents) and introduced the investigator and briefly the object of his work to each respondent, but he intervened thereafter only when the process hit a snag of some kind. His many years of acquaintance with the speakers and his skill as a fieldworker in his own right greatly facilitated this investigation, not least in keeping respondents at their ease. If the researcher, an American, had been on his own, he might have recorded any number of relatively informal interviews with speakers who were Ulster-Scots geographically speaking, but he would have gathered little actual Ulster-Scots.(20)

All nine speakers had been born and raised in rural parts of the county and were over sixty years of age at the time of fieldwork. Most had received only primary-level education. The researcher had two basic goals. Because he was pursuing a long-term project to explore American historical links to local speech (to reconstruct how 18th-century Ulster emigrants spoke), many early questions that were asked concerned whether features of American English grammar were attested in Ulster-Scots. The relative unproductiveness of this line of inquiry and the realization that contemporary Ulster-Scots lacked detailed description for many grammatical features led to a shift to a second emphasis. The same basic approach was employed to gather data: asking speakers to judge the grammaticality/naturalness and in many cases the relative commonness of sentences exemplifying a wide range of morphological and especially syntactic phenomena.

Many constructions were presented in contrasting sets in order to explore specific points of grammar.(21) For example, speakers were asked to evaluate and compare 'the girl wha sa me', 'the girl that sa me' and 'the girl at sa me'(22) to discover which pronoun form(s) in this context (human head-noun as subject of a restrictive relative clause) can be used in Ulster-Scots. When respondents accepted more than one alternative pattern, they were asked which was more common or preferable, in order to gauge variation within Ulster-Scots. Sometimes they identified a difference in meaning. While its methods of data collection were painstaking and a consensus was reached in nearly all cases of what represented the usual patterns of Ulster-Scots, there was somewhat less agreement about what did not conform. This situation reflects not an artefact of the methodology, but rather that some speakers had a mixed grammar for some features that was characterized by variation and on-going change. The minority forms that were thus revealed represent intrusions from either Standard English or from Ulster English into the speech of a seemingly homogenous group of conservative speakers (e.g. some respondents accepted they wur and not they wuz, but others accepted both, almost certainly because of the influence of Ulster English). These forms indicate that Ulster-Scots is, as suggested by the historical background sketched above, not a 'pure' variety today, if it ever has been. In other words, the elicitation procedures, while tapping perhaps the closest thing to traditional Ulster-Scots, was also able to detect considerable variability.(23) This linguistic complexity of Ulster-Scots is presented in what follows as it was discovered by the procedures used, and Ulster-Scots is not presented as a uniform or idealized language, as is often done, even for urban language varieties.(24) One must always remember that few, if any, people speak only Ulster-Scots and thus that for the speakers at hand as well as generally Ulster-Scots and Ulster English form a continuum. Even for older, rural, less-traveled speakers, variation in language may indicate change in progress, suggesting areas for future research to explore; Gregg recognized this fact in his study of the speech of his native Larne two generations ago.

In some cases speakers were asked to rephrase a sentence whose awkwardness they sensed in the course of elicitation, or they offered to rephrase a sentence without prompting, providing very useful insights and suggestions for further inquiry. Sometimes they were asked to translate a sentence from English, which they were always able to do unambiguously. All speakers were asked a core set of questions in the summer of 1997. Subsequently the investigation was expanded in several stages, to well over 300 sentences. Due to this length, the constraints of time and the mortality of three speakers, not all nine were queried with regard to every feature. Prepared sentences were often based on citation sentences from Fenton's dictionary or from other sources,(25) in order to minimize unforeseen lexical, factual, or pragmatic anomalies, to ensure their vernacularity, and to avoid inadvertent mixtures of Ulster-Scots and Ulster English. Initial queries often formed points of departure for the fuller exploration of intuitions. The investigator allowed considerable opportunity for speakers to comment or elaborate on their responses and routinely pursued more detailed lines of inquiry of specific features when this proved productive. Much time was also spent in follow-up sessions comparing responses collected earlier. The data for this study required between twenty-five and thirty hours to collect.

This study concentrates on pronouns, verbal features and a variety of syntactic patterns; data on many other areas of interest (e.g. imperatives, multiple modal verbs) were collected but are not presented here. The consensus judgments of nine individuals, however skillful and uniform they may be as respondents, are not absolutely definitive, and there is no guarantee that a pattern rejected by all might not occasionally occur in Ulster-Scots speech (the findings here usually agree with and build on those presented in Robinson's grammar volume). The limitations of this investigation derive primarily from the smallness of the sample, but these have been counter-balanced by its fine-grained depth. In addition to addressing structural issues, such an exercise goes a long way to set Ulster-Scots in its proper comparative and historical contexts and identify grammatical features for treatment by the historical dictionary of Ulster-Scots being undertaken by the Ulster-Scots Language Society. It complements the data being gathered by the Tape-Recorded Survey of Ulster-Scots and the Ulster-Scots Text-Base, two of the Society's on-going work programmes.

1 Pronouns

1.1 Conjoining and Ordering of Personal Pronouns. In Ulster-Scots the accusative form of a personal pronoun (me, her, him, us, or them) is used as the subject of a clause only when joined with another personal pronoun (1a-4c), with a proper noun (5a-5c), or a common noun (6a-6c). This pattern is shared with many other varieties of Scots and English. In the present tense, the accompanying verb form with conjoined subjects in Ulster-Scots is usually is (often contracted to 's) rather than ir, or it takes the suffix -s; for further commentary on verb forms with these subjects, see §2.1.1. The ordering of pronouns is partially constrained, in that me tends to come first (thus, 3a, 4a, 5a and 6a are preferred to 3b, 4b, 5b and 6b).

1.2 Demonstrative Pronouns and Demonstrative Adjectives. Traditional Ulster-Scots employs this, thon and that as singular demonstratives. Representing a conflation of thon and yon, thon as both a pronoun and an adjective is shared with Lowland Scots and northern English and is first attested in both uses in Jamieson's Dictionary of the Scottish Language (1808). How much older the form may be is unknown. Since it was found throughout Ulster by the end of the 19th century, it quite possibly arose in Scotland in the 18th century and spread to Ulster after the major influx of Scottish settlers in the 1600s.(26) Murray speculates that thon is much older, from its occurrence by the late 19th century from Northumbria to the northernmost Scottish mainland.(27) While both that and thon are equivalent to English that, they only partially overlap in function. That tends to refer to something more concrete and immediately at hand (8a), thon to something more distant from both the speaker and the hearer physically or psychologically in space or time (8b); in this regard the present survey confirms the difference cited by Robinson.(28) Thus, that more often contrasts with this.

Both that and thon function as singular demonstrative adjectives, frequently modifying 'one'.

Thon can function as a plural demonstrative adjective (12a-12c), but not as a plural demonstrative pronoun (12d-12f). On the other hand, them (the plural of that in modern Ulster-Scots) can function as either a pronoun (13a-13b) or an adjective (13c-13d). Those is not used in traditional Ulster-Scots (14a-14b).(29) Thir 'these' and thae 'those' are forms found, for example, in the verse of James Orr (1770-1816) of Ballycarry and other Rhyming Weaver poets, but no longer in Ulster-Scots speech (thae remains in Lowland Scots, however).(30)

In traditional Lowland Scots, according to Murray, thon corresponds to English yon and can participate in a three-way distinction between this/that/yon or this one/that one/yon one: 'thys is used to identify the object nearest the speaker; that is used to identify the object nearest to the person spoken to; thon or yon is used to identify an object remote from both'.(32) Macafee likewise suggests that Lowland Scots today has a similar 'distinction in the demonstrative system: yon [and thon] (singular and plural) expresses a further degree of physical or conceptual distance than that and thae'.(33) While thon may be used in Ulster-Scots to express relative remoteness, the third member of such a series when a three-way contrast is made is usually periphrastic, as by this/that/the other. Likewise, a speaker may use thon yin to single out one instance, item, etc., among others, but not to contrast with both this yin and that yin; in this case, the ither yin is used. Nor does Ulster-Scots make a three-way distinction corresponding in form to English here/there/yonder, expressing this instead as here/thonner/awa thonner.

In traditional Ulster-Scots this and these are both demonstrative pronouns and demonstrative adjectives, usually functioning as singular and plural forms, respectively. Certain constructions permit only this (15a, 16a). This may be followed by a numeral (17a), but not modified by a' 'all' (17b), in which case these is required (17c).

Both that and thon can be used to introduce a statement where English has it (18a-18b), but in some cases it with a post-posed that is an even more common alternative (18c).

1.3 Relative Pronouns and Relative Clauses. The usual relative pronoun in Ulster-Scots, as in Lowland Scots, is that or at, regardless of whether the antecedent head-noun is human or not. At, which Murray calls 'the simple Relative of the Scottish and Northern English dialects',(34) represents either a form inherited from Old Norse, an elision of that (which is always unstressed), or both. Wha, whas, and which (but not wham) occur in Ulster-Scots as interrogative pronouns (which is less common than what as a modifier, as in 'what book ir ye wantin?', and whilk retains marginal usage in Lowland Scots. None of these forms is a relative pronoun in spoken Ulster-Scots. The Dictionary of the Scots Language (DSL) describes wha as 'literary and formal anglicized usage', and Macafee notes at least a marginal use of wha/whas in modern Lowland Scots, apparently in-coming forms from English.(35) As (and its contracted form 's) are found in many parts of England (especially the Midlands), but not among the Ulster-Scots speakers surveyed in this study (21d-21e).(36)

The possessive forms of relative pronouns in Lowland Scots, according to Murray, are phrasal: 'When the Relative is used in the Possessive Case (whose) it is necessary to express it by the conjunction at (that) and the possessive pronoun belonging to the antecedent; thus, "the man at hys weyfe's deid", 'the man whose wife is dead'.(37) Patterns such as 22a-22b, 23a-23b, 24a-24b were rejected by the respondents in this study.(38) Instead, they accepted thats and its variant ats as the usual possessive forms (22c-22d, 23c-23d, 24c-24d). According to DSL (s.v. that), thats derived historically from that + his, but the''s of that (hi)s was later construed as a possessive ending as in its' (and presumably English whose) and came to be used with all head-nouns, whether singular or plural, male, female or neuter.

In Ulster-Scots (as in Lowland Scots)(39), the relative pronoun form is often absent even when representing the subject of the relative clause (omission when the direct object, as the man A sa, or the object of a preposition, as the man A towl ye aboot, is frequent even in formal varieties of English and Scots). Lack of the subject pronoun in Scots occurs particularly when the clause is existential (25a-25d).(40) Henry has identified four overlapping types of sentences in which omission is possible in 'Belfast English'(41): 1) existential sentences; 2) it-cleft sentences; 3) equative sentences with a copula verb; and 4) presentational sentences that introduce a person or thing by such verbs as know, meet, etc. In Ulster-Scots as well, the pronoun is often or normally deleted in these four sentence types.

Existential sentences. In these the verb may be either transitive or intransitive in the simple past tense (25a-25b) or the simple present tense (25c-25d). As in English and Scots generally, that + a form of the auxiliary verb be may be deleted before a present participle (25e).

The verb may also be passive with either an overt auxiliary (26a), or, more commonly, with the relative clause reduced to the past participle (26b); a one-word adjectival complement may also follow the subject (26c). Speakers judged that neither pattern represented in 26a or 26b was as common as the non-existential alternative (26d).

It-cleft sentences. In these a simple sentence is transformed into a complex one by extracting a word or phrase and giving it prominence by placement directly after it's/it is. For example, 'He wants his money' becomes 'It's his money (that) he wants'. This device is well known in Irish English.(42) When involving a personal pronoun, the form of the pronoun shifts from nominative to accusative (27a-27b).

Equative sentences with a copula verb.

Presentational sentences.

Henry concludes that 'it is, in fact, very difficult to characterize syntactically the class of contexts' in which such omissions are possible and that the constraints may not be syntactic at all.(43) Rather, she proposes that what the four sentence types have in common is that they introduce new entities to a discourse. Whether such a function characterizes all possible cases of omission cannot be determined by a study such as hers or the present one, based as they are on eliciting judgments regarding individual sentences having no discourse context. Such a proposal is extremely broad, however, in that the introduction of new items into a discourse is probably the most common, if not the default, function of post-verbal noun phrases (direct objects, subject complements, and so forth), especially indefinite ones. The evidence from Ulster-Scots suggests that subject relative pronouns may be omitted in a wider range of contexts. For example, in presentational sentences head-nouns with omitted relatives may be definite as well (30a, 31a), although less accepted than sentences having an overt relative pronoun (30b, 31b).

Further, unlike in Belfast English as described by Henry, the relative pronoun in Ulster-Scots may be omitted even if the head-noun is the subject of a clause, as in 32a. Crucially it is the intonation contour of the elicitational prompt that determines the acceptability of such a sentence; if 'man' in 32a is given level rather than falling intonation, 'stole my car' is more likely to be to be interpreted as having a relative clause.

2 Verbal Features

2.1 Verbal Concord. Ulster-Scots follows the 'Northern Subject Rule' (NSR),(44) whereby a verb form in the present tense takes an -s suffix unless its subject is an adjacent simple personal pronoun; cf. 33a (with a common noun) and 33b (with an indefinite pronoun) to 33c (with a personal pronoun). This rule operated in Scots and Northern English by the 14th century and is undoubtedly much older.(45) By analogy, is (34a) and has/haes (35a) follow the same constraint, which will henceforth be called 'verbal -s'.(46) In modern Ulster-Scots the NSR is less strict in the past tense, however (36a-36d), and when the subject consists of conjoined personal pronouns, wur is more common than wuz (41a-43b).

2.1.1 Concord with Conjoined Pronouns. Following the NSR, verbal -s operates in the present tense with conjoined-pronoun subjects, whether the pronouns are singular or plural. However, when the verb is stressed, as in expressing emphasis or contrast (37c), this is no longer the case.

By contrast, with conjoined pronoun subjects the verb in the past tense (where only concord with be is possible, as in English) is usually wur. Some respondents judged wuz also to occur, but to be less common or less natural than wur. In sentences with two singular pronouns, as in 43a, wuz was judged more likely than if the pronouns were plural.

The difference in preferred verb forms for the two tenses suggests that, under the influence of English, the NSR is eroding more quickly for the past tense,(47) illustrating very well the mixed nature of Ulster-Scots pointed out earlier, with ir and wur being intruding forms from Standard English, but on different trajectories within Ulster-Scots.

2.1.2 With plural demonstrative pronouns (them, them yins), the choice of verb forms is also variable in Ulster-Scots, but verbal -s is preferred (44a-44c, 45a-45c, 46a-46c, 47a-47b, 48a-48c). However, when the appropriate tag question is added, sentences become ungrammatical because of the co-occurrence of is/-s with they (compare 44e with 44f).

2.1.3 Verbal concord with the second-person plural pronoun yous(e) is variable in Ulster-Scots (49a-49b, 49e-49f, 50a-50b). However, this pattern does not constitute an exception to the NSR (which prohibits a verb taking -s or being is/haes when adjacent to a single simple personal pronoun). You (along with its variant form ye) is the historical member of the paradigm of personal pronouns in Scots. Yous is a fairly recent innovation, developing in Ulster English apparently in the early 19th century, almost certainly as a product of Irish speakers acquiring English, and then spreading to Ulster-Scots and to urban varieties in England (especially Liverpool and Newcastle) and Scotland (Glasgow). That yous has been evolving from a pronominal compound having two morphemes to a personal pronoun is suggested by the existence of both yous is and yous ir.(48) Yous, like other pronouns (e.g. demonstrative them), may take a cliticized form yins, as in yous yins, 'you (plural)', literally 'you ones'. Yous yins has at least one feature of a personal pronoun, in that its meaning is the same as both yous and you (plural). When it is the subject of a clause, yous yins (like yous) often follows the NSR (51a, 52a), but it does not occur in tag questions (51c-51d, 52c-52d).

2.1.4 Verbal Concord in Yes-No Questions. Henry says that in her study of Belfast English verbal -s (which she terms 'singular concord') does not operate in yes-no questions (thus, '*Is the eggs cracked?').(49) This statement is rather surprising, because verbal -s in yes-no questions is reported for Belfast by other researchers(50) and examples of it from transcriptions of Ulster folktales are not hard to find.(51) Inversion is possible in Ulster-Scots, a finding that perhaps shows that the NSR is more vigorous than in local varieties of English. A verb with is/-s is preferred when the subject is a noun phrase (53a, 53c, 53e, 54a, 57a, 57c) or a demonstrative pronoun (57b). Such a verb is also possible if the subject is yous (55b) or conjoined pronouns beginning with ye (56b).

2.1.5 Concord with an intervening adverb. In line with what Henry found for Belfast English,(52) in Ulster-Scots the presence of an intervening adverb can interfere with the Northern Subject Rule, though the constraints on this are not clear. Verbal -s does not occur when an adverb comes between a noun subject and either a copula verb (compare 59a with 59c) or a tensed auxiliary (compare 61a and 61c); however, if the subject is a pronoun, the sentence is ungrammatical regardless of the verb form used (60a, 62a). With a tensed lexical verb, this prohibition appears to operate for only certain adverbs. In 63a aye, mebbe or aply are permissible before cum, but in 64a railly is not; respondents always rephrased the latter as 64c, with post-verbal rail. They accepted 59c, with a pre-verbal adverb (though judging it less common than its alternative, 59b), but they rephrased 60a and 62a, moving the adverb after ir (60b) and hae (62b). That adverbs do not normally precede the verb in Ulster-Scots, regardless of the person or number of the subject, undoubtedly plays a role in the ungrammaticality of sentences like 59a and 61a.

2.1.6 Verbal Concord in Existential Clauses. As in many varieties of English and Scots,(53) an existential sentence having a plural subject often takes is (usually contracted to 's) in Ulster-Scots (65a, 65c). While there has long been employed to introduce existential clauses in English, in Scots by the early-18th century they (distinct from the personal pronoun having the same form) had developed as a variant and is used today in both Lowland Scots(54) and Ulster-Scots. In Ulster-Scots only there is used with is/’s/wuz (compare 65c with 65e, etc.) In contrast, especially in yes-no questions or in answer to such a question, they appears with ir/wur/etc., but not is/’s/wuz, even if the subject is singular (66c, 68a, 69a).

2.2 Negation of be and have. In many varieties of English in England and the United States ain't is the equivalent of negated forms of both be (isn't, aren't, am not) and have (hasn't, haven't). However, it is not used in Ulster-Scots for either (70a, 71a, 71c, 72a, 74a, 75a, 76a), and historically not in Lowland Scots.(55) Haenae (= English haven't) negates both auxiliary and main-verb hae in statements (73a, 73c), but not in questions, where the form is sometimes haen't (74c, 75c, 76c) or haesn't in the third singular (72c), but more commonly hae/haes ... naw (74d, 75d, 76d). These same patterns (with negative particle no rather than naw) also characterize Lowland Scots.(56)

2.3 Habitual Verbs. Like Ulster English, but not Lowland Scots, Ulster-Scots uses be/bes (77b, 78b) and occasionally dae be/daes be (77c, 78c) to express habitual activities or occurrences. It also uses be/bes for conditions (84b, 85b), but not for states (77e). For the habitual these verbs are usually accompanied by an adverb such as whiles or affen. When the verb is negated (79b-79d) or put into a question (80b, 81c), dae or daes is normally inserted. All these usages have almost certainly come into Ulster-Scots through Ulster English, from the influence of the Irish language.(57) However, it remains true that the far most frequent means of expressing the habitual in Ulster-Scots (and also Ulster English) is with wud (77a, 78a, 79a, etc.), although wud is used to express conditions and has other functions as well. One must also note that the simple present-tense (i.e., the non-progressive) form of verbs (81b) frequently expresses the habitual in Ulster-Scots as well as in English. The respondents in this study rejected the suffix -s (possibly of Scottish influence) in (82) as another means of expressing the habitual.(58)

2.4 Complement Shifting. Under certain conditions Ulster-Scots permits the past participle of a have-verb phrase to move rightward to the slot after the direct object, with a consequent slight change in emphasis (compare 86a with 86b). This phenomenon is apparently not documented for Lowland Scots, but is well known in varieties of Irish English (Filppula terms it the 'Medial-Object Perfect'),(59) which is almost certainly the source for the pattern in Ulster-Scots. Ulster-Scots also permits either a past participle (87a) or a present participle of a be-verb phrase ( 87c) to move beyond an adjective phrase.

3 Phrases and Miscellaneous Syntax

3.1 When combined to form a phrase with an adjective, the definite article the is sometimes interpreted to mean 'how' in Ulster-Scots (and a' the much = 'how little').(60)

3.2 When modifying a singular noun or pronoun, the phrase a' the means 'the only' (89a-89c). However, when modifying an adjective it expresses extent and can be translated as 'how' or 'as ... as' (thus, a' the far = 'as far as') in either statements (90a-90d) or questions (91a-91d) in a pattern that is clearly based on the one presented in the previous section. Unlike in some varieties of American English, a' the cannot be used before a comparative adjective to mean the same thing (92a-92b), but the pattern a' the + positive adjective was almost certainly the one from which the American ones with a comparative or superlative form developed.(61) Neither is a' the used before a superlative adjective in Ulster-Scots (93a-93c).

3.3 Placement of a' 'all'. In Ulster-Scots, as in Lowland Scots, the quantifier a' can follow a wh- form (who, what, etc.) in either a statement or a question to specify an exhaustive list rather than a partial one. Normally a' immediately follows (94a, 95a). It may also appear after the first element of the predicate (94b, 95b), but not at the end of it (94c, 95c).

3.4 Pronominal Copy Shift. An indefinite pronoun or pronoun phrase subject such as ivery yin 'every one' (96a), or ivery yin o them (97a) may be extracted from subject position and moved rightward to any of several positions, leaving the appropriate pronominal copy behind. The phrase may move to the slot immediately before a tensed lexical verb (96b, 97b) and thus appear to be in apposition to the subject, or it may come immediately after the tensed verb (96c, 97c).

However, if there is a tensed auxiliary or a tensed copula, the pronominal phrase must appear after the first element in the verb phrase (98c, 99c, 100c, 101c, 102c, 103c) rather than before it (98b, 99b, 100b, 101b, 102b, 103b) or after a second auxiliary (98d, 99d).

For emphasis, an indefinite pronoun phrase may shift to the end of the sentence (104b, 104d-g, 104i) if it has a prepositional phrase, but not if it is a bare indefinite pronoun (104a, 104c). Similarly, a nominal may move if it represents the object of a preposition after an indefinite pronoun (104h, 104j).

3.5 Elliptical Infinitives. In Ulster-Scots a set of verbs (luck, 'look, expect, ask for', want, need, and sometimes lake 'like') can take a past infinitive (105a-106d) or a preposition (107a-108c) without requiring an intervening infinitive phrase (i.e. having implied to be, to go, to come, etc.)

3.6 Infinitives with for tae. The construction for tae (less often for til) is used to introduce infinitive phrases in Ulster-Scots, most often to express purpose (= 'in order to'), as in 109a-109e. However, speakers in this study accepted for tae as a general equivalent to the infinitive marker tae/til (110a-110b).(62) In Ulster-Scots if an overt subject appears for the infinitive, this must immediately precede tae (compare 111a, 112a with 111c, 112c); similarly, for tae + verb is not possible as the subject of a sentence (113b).

3.7 Embedded Questions. In Ulster-Scots, as in Ulster English and Irish English, embedded questions frequently use the inverted word-order found in main clauses in other varieties, either by moving an auxiliary verb (114a, 115b) for wh- questions or by inserting the appropriate form of dae (116a) for yes-no questions.

4 Conclusion

Through exploration and analysis, this survey has sought to increase scholarly knowledge about selected features of Ulster-Scots grammar to the depth exemplified by Robert Gregg's work on pronunciation. It complements Robinson's broad descriptive coverage of traditional Ulster-Scots and provides details about some of the syntactic structures examined by Henry's study of Belfast English. Although it is mainly synchronic in focus, its comparisons to Lowland Scots and Ulster English situate the historical contacts and relationships that Ulster-Scots has had with other language varieties and identify some features that may be innovations in Ulster-Scots. In particular, the many similarities with Lowland Scots, while hardly unexpected, show that Ulster-Scots cannot be regarded simply as one among many varieties of Irish English. Indeed, Ulster English shares many of the same similarities with Lowland Scots, and this fundamental affinity may account for some differences Henry found between Belfast English and Standard English. There remains much to learn about the grammar of Ulster-Scots, of course, but for this study as well as others in this volume, Gregg's insistence on scrupulous, painstaking fieldwork and analysis has shown the way forward.

Notes

(1) The author is profoundly indebted to James Fenton, without whose assistance in innumerable ways this study would not have been possible.

(2) Gregg, Robert J., The Scotch-Irish Dialect Boundaries in the Province of Ulster ([Port Credit], 1985), 257-286.

(3) Robinson, Philip S., Ulster-Scots: A Grammar of the Traditional Written and Spoken Language (Belfast, 1997).

(4) For Lowland Scots these include Murray, James A. H., The Dialect of the Southern Counties of Scotland: Its Pronunciation, Grammar, and Historical Relations (London, 1873); and Macafee, Caroline, 'Characteristics of Non-Standard Grammar in Scotland', unpublished typescript, 1992, at <www.abdn.uk.ac/enl038/grammar.htm>, consulted 14 April 2005. For Irish English the principal sources are Harris, John, 'The Grammar of Irish English', in Milroy, James, and Lesley Milroy (eds.), Real English: The Grammar of English Dialects in the British Isles (London, 1993), 139-186; and Filppula, Markku, The Grammar of Irish English: Language in Hibernian Style (London, 1999).

(5) For migration from Britain to Ulster in the 17th century, see Braidwood, John, 'Ulster and Elizabethan English', in Adams, G. Brendan (ed.), Ulster Dialects: An Introductory Symposium (Holywood, 1964), 5-109; Macafee, W. A., 'The Movement of British Settlers into Ulster in the Seventeenth Century', Familia 2 (1992), 94-111; Fitzgerald, Patrick, 'Black '97': Reconsidering Scottish Migration to Ireland in the Seventeenth Century and the Scotch-Irish in America' in Kelly, William P., and John R. Young (eds.), Ireland and Scotland 1600-2000: History, Language and Identity (Dublin, 2004), 71-84; Montgomery, Michael B., and Robert J. Gregg, 'The Scots Language in Ulster', in Jones, Charles (ed.), The Edinburgh History of the Scots Language (Edinburgh, 1997), 569-622.

(6) Grant, William, 'Ulster Dialects', Scottish National Dictionary, volume 1 (Edinburgh, 1931), xli.

(7) Adams, G. Brendan, 'Introduction', in Adams, G. Brendan (ed.), Ulster Dialects: An Introductory Symposium (Holywood, 1964), 1.

(8) These two possible — and no doubt reinforcing — sources are cited in many etymologies in Macafee, Caroline (ed.), Concise Ulster Dictionary (Oxford, 1996). According to G. Brendan Adams, 'The Emergence of Ulster as a Distinct Dialect Area', Ulster Folklife 4 (1958), 69, 'many Gaelic borrowings at present used in Ulster ... were already embedded in [Scots] before its establishment here, for they show sound changes which either belong to Scottish rather than Irish Gaelic or they belong to an earlier stage of Lowland Scots than that at which it was brought to Ulster'.

(9) For the fullest view of Irish vocabulary borrowed into Ulster-Scots, see Fenton, James, The Hamely Tongue: A Personal Record of Ulster-Scots in County Antrim (Newtownards, 1995; 2nd. edition, 2000; 3rd. edition, 2006).

(10) Adams, G. Brendan, 'The Phonology of the Antrim Dialect', Proceedings of the Royal Irish Academy 57C3 (1956), 69-152.

(11) Kirk, John, and Georgina Millar, 'Verbal Aspect in the Scots and English of Ulster', Scottish Language 17 (1998), 82-107.

(12) For an overview of the historical linguistic landscape, see Montgomery, Michael, 'The Position of Ulster-Scots', Ulster Folklife 45 (1999), 85-104.

(13) For a different assessment, see Tagliamonte, Sali, and Jennifer Smith, '"Either It's not or It isn't": neg/aux Contraction in British Dialects', English World-Wide 23 (2002), 251-281, who describe the Cullybackey community of mid-Antrim as having a 'peripheral geographic location' and 'isolated socio-political circumstances' (p. 256).

(14) Montgomery, Michael, 'The Scotch-Irish Influence on Appalachian English: How Broad? How Deep?' in Blethen, H. Tyler, and Curtis W. Wood Jr. (eds.), Ulster and North America: Trans-Atlantic Perspectives on the Scotch-Irish (Tuscaloosa, 1997), 189-212.

(15) Montgomery, Michael, and Philip Robinson, 'Ulster: A Linguistic Bridge across the North Atlantic', Journal of Scotch-Irish Studies 1 (2000), 40-60.

(16) Montgomery, Michael, 'The Linguistic Value of Ulster Emigrant Letters', Ulster Folklife 41 (1995), 26-41; Montgomery, Michael, 'Making the Trans-Atlantic Link between Varieties of English: the Case of Plural Verbal -s', Journal of English Linguistics 25 (1997), 122-141.

(17) The speakers ranged from Dervock in the north to Ballynure in the east to Doagh in the south, ten miles north of Belfast; see Fenton, op. cit., viii.

(18) For a brief account of his methodology, see Fenton, op.cit., ix.

(19) Labov, William, 'Some Principles of Linguistic Methodology', Language in Society 1 (1972), 97-120.

(20) For example, in Ulster-Scots areas the Tape-Recorded Survey of Hiberno English recorded only one individual who used dinnae, in large part no doubt because the interviewees were strangers and not speakers themselves. See Barry, Michael V. (ed.), Aspects of English Dialects in Ireland (Belfast, 1981); and Kirk, John M., Northern Ireland Transcribed Corpus of Speech (Colchester, 1991).

(21) In this process all examples and queries were presented to respondents orally.

(22) In this paper all patterns are presented in the system of orthography devised over a period of more than two decades by James Fenton in his The Hamely Tongue; see Fenton, op.cit., x-xiii.

(23) For a sophisticated study showing the variability in pronunciation, see Kingsmore, Rona R. K., Ulster-Scots Speech: A Sociolinguistic Study (Tuscaloosa, 1995).

(24) Henry, Alison, Belfast English and Standard English: Dialect Variation and Parameter Setting (Oxford, 1995). Henry's purpose was to present the major syntactic structures of 'Belfast English'; by 'variation' in her title she meant contrasts to 'Standard English', not any differences within Belfast English. Nonetheless, she occasionally alludes to the existence of the latter, stating, for example that 'singular concord is also generally [my emphasis] impossible with inversion' (p. 16), a qualification which she does not explore. For the term 'singular concord', see note 44.

(25) The author had studied the published and archival literature on Ulster-Scots since 1988.

(26) Thon is documented from east county Down in Savage-Armstrong, George Francis, Ballads of Down (London, 1901) to south county Donegal in the west, in Simmons, D. A., A List of Peculiar Words and Phrases Formerly in Common Use in the County Armagh, Together with Expressions at One Time Current in South Donegal (Dublin, 1891).

(27) Murray, op. cit., 187.

(28) Robinson, op. cit., 74.

(29) Macafee, op. cit. (1992), §20.4.

(30) Ibid.; Miller, Jim, 'The Grammar of Scottish English', in Milroy, James, and Lesley Milroy (eds.), Real English: The Grammar of English Dialects in the British Isles (London, 1993), 108; Macafee, op. cit., §20.4.

(31) Here and elsewhere sentences are marked by "*" to indicate that speakers judged them not to be possible in Ulster-Scots.

(32) Murray, op. cit., 179.

(33) Macafee, op. cit., §20.4.

(34) Murray, op. cit., 194.

(35) Macafee, op. cit., §9.1.

(36) But cf. Robinson, op. cit., 77-78.

(37) Murray, op. cit., 196; see also Macafee, op. cit., §9.2; Miller, op. cit., 111, and DSL, s.v. that. The first citation of the construction is dated 1456.

(38) But cf. Robinson, op. cit., 78.

(39) DSL, s.v. that.

(40) Macafee, op. cit., §9.4.

(41) Henry, op. cit., 125.

(42) Filppula, op. cit., 243ff.

(43) Henry, op. cit., 126.

(44) For this term, see Ihalainen, Ossi, 'The Dialects of England since 1776', in Burchfield, Robert (ed.), Cambridge History of the English Language, Volume 5: English in Britain and Overseas (Cambridge, 1994), 221-222.; Murray, op. cit., 211-212, was apparently the first to describe this rule, though he did not give it a name. The label 'singular concord' is used by Milroy, James, Regional Accents of English: Belfast (Belfast, 1981); Policansky, Linda, 'Grammatical Variation in Belfast English', Belfast Working Papers in Language and Linguistics 6 (1982), 39-66; and Henry, op. cit. It is rather a misnomer, in that the rule pertains to plural contexts. In English (and Scots) it is only nouns and pronouns that have inherent number of either singular or plural, since verbs acquire or are assigned number when finite depending on the number of their subject. If the form or category of a subject constrains the form of a verb, as is clearly the case with the NSR, then concord does take place. A more appropriate term than 'singular concord' is therefore 'plural verbal concord' or 'differential plural concord'.

(45) For a quantitative view of the history of this rule, see Montgomery, Michael, 'The Evolution of Verb Concord in Scots', in Fenton, Alexander, and Donald A. MacDonald (eds.), Studies in Scots and Gaelic: Proceedings of the Third International Conference on the Languages of Scotland (Edinburgh, 1994), 81-95.

(46) Filppula, op. cit., 150ff. reports that this verbal concord operates to a limited extent in the south of Ireland.

(47) Were was used categorically with pronoun subjects in letters written by Ulster emigrants to America in the 18th century; see Montgomery, Michael, op. cit. (JEL 25 [1997]); and Montgomery, Michael, 'A Tale of Two Georges: The Language of Irish Indian Traders in Colonial North America', in Kallen, Jeffrey (ed.), Focus on Ireland (Amsterdam, 1997), 227-254.

(48) Robinson, op. cit., 127, also has examples like 'Yous is tae cum hame noo'.

(49) Henry, op. cit., 16.

(50) For example, Harris, op. cit., 156, states that yes-no questions represent 'the most favourable context for a verbal -s ending to occur' in Irish English, based on data from three large sociolinguistic studies, two of which were in Belfast; see also Finlay, Cathy, 'Syntactic Variation in the Speech of Belfast Schoolchildren,' unpublished Ph.D. dissertation, University of Ulster at Jordanstown, 263-264.

(51) For example, 'Isn't his clothes and boots all here?', from Murphy, Michael J., Culprit of the Shadows (Belfast, 1955), 24; 'Is there no hares at all here?', from Murphy, Michael J., Slieve Gullion's Foot (Dundalk, 1975), 61.

(52) Henry, op. cit., 26.

(53) Macafee, op. cit., §6.1.

(54) Ibid.

(55) Macafee, op. cit., §1.6, reports that it can now be found in Glasgow, but 'only in reversed polarity tag [questions], as the present tense of be + enclitic negative particle (not as a form of have)', presumably as an influence from English.

(56) Miller, op. cit., 114; Miller, Jim, and Keith Brown, 'Aspects of Scottish English Syntax', English World-Wide 2 (1980), 3-17.

(57) This agrees with data from the Northern Ireland Transcribed Corpus of English, as reported in Montgomery, Michael, and John Kirk, 'The Origin of the Habitual Verb be in American Black English: Irish or English or What?' Belfast Working Papers in Linguistics 11 (1996), 308-333.

(58) Robinson, op. cit., 117.

(59) Filppula, op. cit., 107ff.

(60) Cf. Robinson, op. cit., 66, 92-93.

(61) Montgomery, Michael, 'Solving Kurath's Puzzle: Establishing the Antecedents of the American Midland Dialect Region', in Hickey, Raymond (ed.), The Legacy of Colonial English: The Study of Transported Dialects (Cambridge, 2004), 410-425; Thomas, Erik R., 'The Use of all the + Comparative Structure', in Frazer, Timothy C. (ed.), 'Heartland English': Variation and Transition in the American Midwest (Tuscaloosa, 1993), 257-265.

(62) Henry, Alison, 'The Syntax of Belfast English' in Kallen, Jeffrey (ed.), Focus on Ireland (Amsterdam, 1997), 98, states that this broad use of for to 'is now confined to Belfast and a few other areas such as parts of County Armagh', an assessment that would seem to need revised. The general construction for ti is also found in Lowland Scots, according to Macafee, op. cit., §10.2.