Morphosyntactic Features in Flux Awareness of “Irishnessness” and “Standard” in Hiberno- English Speakers Tamami Shimada This paper highlights sociolinguistic aspects of the morphosyntactic forms in Hiberno-English (HE), focusing on some well-known features. Major concerns are to illustrate speakers‘ awareness of the forms based on a conducted survey and to discuss the direction of change. The evidence from the survey suggests that morphosyntactic forms in Southwest Hiberno-English (SwHE) are unevenly marked in speakers‘ subjective judgments of ‗use‘, ‗non-use‘, ‗Irishness‘, and ‗bad grammar‘. For example, such forms as Taking three plates she is, I do be taking three plates. (‗I usually take three plates.‘), and I am after taking three plates. (‗I have just taken three plates.‘) are mapped differently onto a sociolinguistic dimension, in which ‗Irishness‘ and ‗Standard‘ are two salient categories of speakers‘ awareness. The survey provides the data that serve for a consideration of change in SwHE. They include: diverging judgements within the same syntactic feature category (PP vs. VP fronting), the regional difference in the use of amn‟t I~? , and the contrast of two tense-aspect forms, do be and be after. In the process, the formation of social meaning and the rise of metalinguistic interaction are also addressed. The paper proposes the concept of morphosyntactic conformity to explain the awareness data with regard to language change. 1. Introduction The purpose of this paper is to elucidate speakers‘ roles in language change with data from Hiberno-English (HE). Speakers‘ sociolinguistic awareness of salient characteristics in HE is described and some changes in progress are discussed. The paper, for example, illustrates how such characteristics as do be V-ing and be after V-ing are differently perceived by speakers in terms of ‗use/ non-use‘, ‗Irishness‘ and ‗bad grammar‘. The question as to what provokes such differences is then addressed. The AAA - Arbeiten aus Anglistik und Amerikanistik Band 40 (2015) · Heft 1-2 Gunter Narr Verlag Tübingen Tamami Shimada 50 present argument is based upon field research, and recognition of the interplay of social meanings has emerged from local interviews. A survey was undertaken in order to describe how linguistic forms, including salient grammatical characteristics and lexical items of HE, are perceived by speakers. The description of salient characteristics in grammar and vocabulary is a major research area on HE (e.g. van Hamel 1912, Bliss 1972, Harris 1993, Kallen 1997, Filppula 1999, Dolan 2000, Hickey 2007). The study of HE has been developed by the examination of HE-specific constructions mainly in contrast with British Standard English. 1 In relation to the earlier publications on HE, this paper examines the sociolinguistic aspect of the features that characterise HE. It concentrates on how speakers of HE perceive various features and other non-specific grammatical characteristics and map them in their awareness. This study further attempts to detect factors underlying the selection of particular forms. It is intended to provide clues for exploring changes in progress in HE, by depicting what speakers‘ awareness is really like and by examining its relevance to the selection of linguistic forms. Among the features described in the literature, some of them seem to have become obsolete while others are maintained in contemporary vernaculars. My question was, therefore, if this is the case, what determines the rise and fall of a particular linguistic construction in the grammatical system. This paper investigates what is relevant to the selection of features and what is the role of speakers as agents of sociolinguistic awareness in language change. 2 It is based on the fundamental working assumption that speakers‘ awareness can be a factor that motivates use or disuse of certain linguistic forms and thus, from a broader perspective, affects the grammatical system, leading ultimately to change in the language. Speakers‘ sociolinguistic awareness, in theory, reacts to social variables in the outer world, including norms and significant community icons, and affects the actual use of linguistic forms. This awareness may thus serve as active molecules directly affecting the use of language, thus constituting an external factor in language change. Awareness is assumed to act upon the use or disuse of a particular grammatical and lexical form in the system via interactions and verbal accommodation among speakers. 1 Terminologically, the word ‗characteristics‘ is used here to refer to the HE-specific forms and constructions, while ‗features‘ is also employed in line with other studies of HE. Filppula (1999) uses the term ‗distinctive features‘, with HE contrasted to other dialects of English. The word generally seems to refer to ‗non-standard grammatical characteristics‘ as used by Harris (1993), indicating features which ―don‘t correspond to institutionalized norms‖ (ibid: 139). 2 See James Milroy (2003) for the significance of the speaker in language change and the relevant discussion. Morphosyntactic Features in Flux 51 2. Speakers’ awareness of Irishness and Standard 2.1. Background of awareness We may first consider in which socio-historical contexts HE is spoken and on what basis linguistic communication using HE can be produced. Ireland has achieved its present socio-economic status very rapidly in the past forty years or so. Social change must be taken into consideration for the examination of linguistic phenomena in the current Irish situation. More recently dubbed the ‗Celtic Tiger‘, the Republic of Ireland‘s financial development has roared forward, starting in the mid-1990s. Globalization has had a tangible impact on its economy. Immigration from Europe, Asia and South America to rural as well as urban areas has introduced yet another level of complexity to the language situation. Also, speakers of contemporary varieties of HE are increasingly exposed to different varieties of English as well as English as an international language. This recent complexity is notable, since it lays the foundation for how speakers conceptualize their linguistic uniqueness or local value in language as a cultural resource. A more important issue in understanding the sociolinguistic matrix within which speakers of HE conduct their linguistic interactions is the linguistic history of language shift from Irish to English that Ireland has experienced since the seventeenth century. Irish is designated as the first official language of the Republic of Ireland in its Constitution; the Irish language is not daily spoken by most of the people in Ireland but is regarded as indispensable cultural capital in present-day Irish society. It is noteworthy that Irish people are conscious of the history of the colonisation by the English, as shown in the following comment written in a questionnaire: ―Our English is colonial English so it isn‘t as refined [as] ‗British English‘. Sometimes we Irish don‘t take something that is enforced on us‖ (age group: under 30, college student, Cork, 1999). This consciousness of history is likely to raise awareness in the linguistic arena. By virtue of the socio-historical consciousness, speakers are aware of their linguistic distinctiveness from British English on the one hand and of a normative use of English on the other. The former awareness of distinctiveness in HE is referred to as awareness of ‗Irishness‘, and the latter as awareness of ‗Standard‘. The survey to be introduced in section 3 was designed upon the assumption of these two types of awareness and the result of the survey actually confirms ‗Irishness‘ and ‗Standard‘ as significant categories in speakers‘ awareness. 2.2. Awareness of Standard In interview sessions with speakers of HE, they often describe certain forms and speech patterns as ‗bad grammar‘ in their own words. Needless Tamami Shimada 52 to say from a linguistic perspective there is no such a thing as bad grammar as long as the form makes sense as part of a grammatical system. However, speakers of HE in general, seem to refer to ‗Standard‘ when pointing out bad grammar in particular constructions. The word ‗Standard‘ is here used as a representative term to refer to the collective image in speakers‘ minds as to what are ‗correct‘, ‗right‘, ‗authorized‘ or ‗legitimated‘ speech patterns and grammatical forms. Although the words ‗correct‘, ‗legitimated‘, ‗standard‘ and ‗grammatical‘ express notions that should be separated, they are, in speakers‘ perceptions, almost identical or closely related to each other (Milroy 1999). Milroy and Milroy (1998) provide an important remark on the ideological quality of a standard language as follows: ―…it seems appropriate to speak more abstractly of standardisation as an ideology, and a standard language as an idea in the mind rather than a reality −a set of abstract norms to which actual usage may conform to a greater or lesser extent‖ (Milroy and Milroy 1998: 19). My term awareness of Standard is in this vein, and the ‗Standard‘ in speakers‘ minds is sharply distinguished from Standard Language as it is actually spoken. For example, ‗Standard‘ in English speakers‘ minds is not entirely in accordance with Standard English. Standard English is ―a social dialect which is distinguished from other dialects of the language by its grammatical forms‖ (Trudgill 1999: 125); Standard English is the language of education and official communication and has an established orthography (Greenbaum 1996: 14). Standard, on the other hand, is a conceptual entity, which is a set of the abstraction of norms held to be true by speakers of the language. This Standard has been put in quotation marks to indicate that it is constructed in speakers‘ minds through everyday linguistic practice and interaction with the outer world. A certain awareness of Standard, presumably, arises in any variety of language that has a standard or prestigious form. 3 However, in addition to the general legitimacy of Standard that language as a social medium entails, English in Ireland, being taught and learned in the postcolonial context, has placed itself in the position of being a ‗target language‘ that ‗learners‘ should acquire. This may have forced HE speakers to adopt a learner‘s mentality. This mentality, in turn, legitimates the Standard; 3 The subjective judgements of linguistic forms sometimes have close reference to other socially significant categories. For example, Aikhenvald (2001) illustrates the identification of calqued forms as ‗incorrect‘ in Tariana speakers in Brazil. There is a strong constraint against language mixing in Tariana, and the constraint operates against loan forms and items that contain Tuano-like sounds. In this condition, morphosyntactic constructions calqued from east Tuano languages are identified as ‗incorrect‘ Tariana. Socially significant categories depend on linguistic or speech communities. In the case of HE, I assume that ‗Irishness‘ and ‗Standard‘ can be extracted as sociolinguistically significant categories. This assumption has developed especially through participant observation in my fieldwork. Morphosyntactic Features in Flux 53 speakers of HE are exposed to constant reference to Standard. This maintains awareness. 2.3. Awareness of Irishness Speakers of HE are likely to view their language as distinct from other English varieties, and this view, as is often the case, is based on their understanding of the historical background. In a survey that the present author conducted in 1999 in Cork, 4 two thirds of the 103 respondents described their language as ‗Irish English‘ or provided specified equivalents in their own words (for example, Hiberno-English, Gaelic English, Cork English) rather than describing it as ‗English‘. 5 In addition, almost 90% of the respondents reported differences between their English and other Englishes including British English and American English. The reason given by a respondent who reported that Irish English was more appropriate to refer to their language may evidence his/ her awareness of certain patterns of Irish in their speech. A comment was: ―Irish English [describes our English better than English], because we, as Irish, have developed our own interpretation of the English language‖ (age group: under 30, college student, Cork, 1999). A number of the respondents furthermore recognised the significance of ‗Irishness‘ in their language in terms of identity, as in the following remark: ―I prefer ‗Irish English‘ as it does differentiate and helps keep our national identity even though we don‘t speak our national language.‖ (age group: under 30, college student, Cork, 1999); ―our English is colonial English so it isn‘t as refined with ‗British English‘. Sometimes we Irish don‘t take something that is enforced on us‖ (age group: under 30, college student, Cork, 1999). These speakers are aware of the distinctiveness of their English. Many respondents explained the uniqueness of HE as due to its association with the Irish language: ―I must confess that I am not very proficient at Irish, but am still proud of the amount I know. This also influences, in some ways, how I speak English, hence the term Irish English‖ (age group: 31-50, housemaker/ mother, Cork, 1999). ―The people of Ireland, although for the most part, we do not speak Irish, the Irish lan- 4 The 1999 survey was an open-ended questionnaire (N=103). Ninety percent of the respondents were university students under the age of thirty. The survey was done for the purpose of discussing language and identity in the Irish context. 5 The survey included the question: Which do you think describes your language better, „English‟ or „Irish English‟? (If you find another name, please write it down.) Note that the term ‗Hiberno-English‘ is not given in the questionnaire. The results were: English-34 (33.0%), Irish English-59 (57.3%), others-10 (9.7%). ‗Others‘ included ‗Hiberno English‘, ‗Hibernicised English‘, ‗Irish/ English‘, ‗Gaelic-English‘, ‗Cork English‘, ‗Enriched English‘, ‗Waterford Irish English‘, etc. Significantly, respondents not only connected their English in some ways with ‗Irish English‘ but also even specified it more precisely in their own words. The result thus generally reveals the favoured distinction between ‗Irish English‘ and ‗English‘. Tamami Shimada 54 guage has influenced how we speak English. There are direct translations from an Irish sentence to English in many of our ‗sayings‘ the British lack this‖ (age group: under 30, college student, Cork, 1999). These citations signal the importance of HE in terms of ethnolinguistic identity. HE is described by speakers today with such words as ‗our own interpretation‘, ‗direct translations‘, and ‗colonial English‘. This perception amounts to saying that English spoken in Ireland is not the entire adoption of an enforced language. HE is a form of the use of a newly developed language that has been shaped by their own interpretation of English. There are a number of expressions or speech patterns which many speakers of HE regard as ‗Irish‘, despite factual origins. In the survey, these included such lexical items as ‗gas‘ (fun), ‗fella‘ (man), and ‗amadán‘ (stupid) and greeting expressions such as ‗How‘s she cutting? ‘ (How are things going? How are things with you? ), ‗How‘s the craic? ‘ (What is the fun? ) and grammatical characteristics such as ‗She is inside in Roches Stores.‘ (She is in Roches Stores), ‗I am after my work‘ (I have done my work.) and ‗A thirst came on me‘ (I am thirsty). These examples were given in the respondents‘ answers to the question in the survey: Please list expressions or phrases that you regard as Irish English. Irishness can be defined as what speakers know about their reflection of the Irish language in their HE, and also as what they regard as constituting their uniqueness in terms of ethnic and ethnolinguistic identity. It is important to highlight the property of ‗Irishness‘ as a mental construction. The word craic is one of the most telling examples to illustrate that speakers‘ perception of Irishness may differ from etymological fact. In the survey that will be more closely addressed in the later sections, a dominant comment on the Irishness question was ‗They have Irish words‘, but the words that the respondents suggested were not necessarily of Irish origin. One of such examples is craic. The word craic is not Irish-Gaelic origin in a proper sense while the speakers regarded it as an Irish word in their speech in English. According to Dolan‘s Dictionary of Hiberno-English (1999: 77), ―Irish craic is the ModE loanword crack < ME crak, loud conversation, bragging talk; recently reintroduced into HE (usually in its Ir spelling) in the belief that it means high-spirited entertainment‖. However, craic is recognised as an Irish word by speakers. Interestingly, speakers of HE maintain the Irish spelling in their use of English. What can be an effective force in the actual use of language is not so much what it is actually as what speakers themselves find in the language and in a particular linguistic form or its constructions. For example, the respondents commented: ―‗Craic‘ is a uniquely Irish word which we have incorporated into the English language‖ (2006, Listowel, born in 1980s, female); ―craic is originally Irish. Amadan is originally Irish. Cnambhshealing is originally Irish‖ (2006, Cork, born in 1920s, male). Craic gained the highest points in the answer to ‗use‘ in the survey that will be introduced in the following sections; and was highly marked as ‗Irish‘ almost equal to Morphosyntactic Features in Flux 55 the other two lexical items of actual Irish origin, namely amadan (‗stupid‘) and cnambshealing (‗complaining‘). It is noteworthy that the example containing craic is most popular, or favoured, in the listed sentences in the questionnaire. Many of the respondents reported their use of the word; they recognised ‗Irishness‘ in the example having craic. 3. A survey of speakers’ awareness In 2006, I conducted a survey in Cork and Listowel (County Kerry) 6 primarily for the purpose of completing a grammatical description to substantiate the reference to the social connection of grammatical structures in Southwest Hiberno-English (SwHE), i.e. the southwest varieties of HE spoken in the area of the counties Cork and Kerry. A questionnaire was devised to ascertain speakers‘ sociolinguistic awareness of particular grammatical features and lexical items. Sixty-four respondents were sampled: thirty-eight from Listowel and twenty-six from Cork. From Listowel, there were twenty men and eighteen women, aged fifteen to seventyeight years. From Cork, there were nine men and seventeen women, aged thirteen to eighty years. Seven respondents participated in the interview and the remainder filled out an anonymous form. The sample was structured to reflect the community in terms of variety of social categories so as not to be biased in age group and occupation. The number of respondents was relatively even across the age range. In terms of occupation, they included students, retired people, managers, shop keepers, sales assistants, teachers and housewives. A smaller number of the respondents were factory workers, drivers, self-employed, childminders, caretakers, secretaries, volunteer workers, librarians, actors, train conductors, therapists and painters. All except one person had received secondary school education and approximately half of the respondents aged thirty years and over marked University/ College/ Institute as their most recent academic institution. Survey respondents were asked to choose from twenty-six sentences listed below those that accorded with the following five statements: (i) a sentence that they would use themselves; (ii) a sentence that they would not use on any occasion (including family-talk and friend-talk); (iii) a sentence whose meaning they cannot understand; (iv) a sentence that they think contains ‗bad grammar‘; and (v) a sentence that they think shows ‗Irishness‘. This method was adopted so that the markedness of the 6 The population of Cork County and City total was 447,829 in 2002 and 481,295 in 2006; that of County Kerry was 132,527 in 2002 and 139,835 in 2006 (Census 2006: volume 1). Cork City is located on the southwest coast of Ireland and is the second largest city in the Republic of Ireland; Listowel, described as the Literary Capital of Ireland, has produced world famous writers such as John B. Keane and Bryan MacMahon. See 4.4 for these two places in terms of speech community. Tamami Shimada 56 listed features in the speakers‘ awareness could then be highlighted, as compared with a counter-method where they had to judge each sentence under the given five indices. This method has the secondary benefit of revealing speakers‘ attitudes. For example, some respondents chose more numbers in (i) than in (ii) and other respondents did the reverse. This may reveal their positive and negative attitudes towards characteristics of HE and/ or to responding to the survey questions in general. Furthermore, it is well-known that there is often a disjunction between a speaker‘s self-report and actual speech behaviour; behaviour may be inaccurately reported and self-assessment is usually inconsistent. Selfreport, however, yields valuable sociolinguistic data, especially for the consideration of awareness and identity, since reporting their linguistic behaviour can exhibit a speaker‘s aspect of acting identity. Speakers‘ subjective judgements and self-reports represent a part of the knowledge of the language as a living entity in actual use. They can offer an important clue to the nature of the forces causing change in the linguistic system. The twenty-six sentences listed in the questionnaire were chosen to include a selection of the salient grammatical features of HE. Sentences were presented in random order: see Appendix for the task given; the categorised version is given below. A. Unmarked sentence (a) She takes three plates from the cupboard. B. Non-canonical constituent order (b1) From the cupboard she takes three plates. (b2) Taking three plates she is. C. Cleft-like sentence (c1) It is from the cupboard that I take three plates. (c2) ‘Tis lovely she is. (c3) It is lovely that she is. D. There… sentence (d1) There‘s no one can deny it. (d2) I knew there was good news in you. (d3) There was a great housekeeper lost in you. E. Do be V-ing/ AdjP form (e1) I do be taking three plates from the cupboard. (e2) She does be lovely with her long hair. F. Be after V-ing/ NP sentence (f1) I am after taking three plates from the cupboard. (f2) Tom is after his supper. G. So-called perfect sentences (g1) They are visiting here many years. (g2) My sons have visited there for many years. H. Cliticisation (h1) We‘ll visit here tomorrow. Morphosyntactic Features in Flux 57 (h2) The two of us‘ll take three plates from the cupboard. (h3) You‘ve the name of a good employer. (h4) Amn‘t I like a scarecrow? (h5) ‘Twouldn‘t be a good thing. I. “Non-standard” usage (i1) She take three plates from the cupboard. (i2) She been taking them home ever since. (i3) I asked for today‘s special and she putting plates on the table. J. Lexical items (j1) How‘s the craic? (j2) That amadán put eggs in my bag. (j3) Don‘t be cnamhshealing! The feature-based categories were based on the Keane corpus, 7 with general reference to Filppula (1999) for grammatical features of the southern HE dialect and to Dolan (1999) for the lexical items. Non-Irish-specific features, such as (a), (c1), (g2), (h1), (i1) 8 , and a ‗dummy‘ sentence (c3), which is not a type of sentence produced in the variety, are also included for reference. 9 The individual sentences were carefully constructed, with particular attention to the numbering of the sentences in the questionnaire. Sentences could not be too idiomatic but needed to be imaginable so that morphosyntactic aspects would be highlighted by the respondents when giving their judgements. A couple of sentences were replaced or 7 The primary data was collected from John B. Keane‘s plays and letter series written in the 1960s and 1970s; this is referred to as the ‗Keane corpus‘. It consists of examples extracted from his works. The examples are sorted by the grammatical features or categories: ‘Tis(~it is)… sentences, there… sentences, fronting, complementation, cliticisation, relative clauses, small clauses, use of prepositions, HE lexicons, tense/ aspect relates and the do be form etc. The Keane corpus as a whole comprises over 18,000 words. It consists of the examples from John B. Keane‘s (1928-2002) play-scripts and letter series. He is known as a major Irish writer with many successful plays and books (Smith and Hickey 2002). The following is the list of his works cited in this paper, headed with their abbreviations: SIV Sive (1959), SRG Sharon‟s Grave (1960), HHM The Highest House on the Mountain (1961), MYM Many Young Men of Twenty (1961), FLD The Field (1966), STD Letters of a Successful TD (1967), RES The Rain at the End of Summer (1968), CHT The Chastitute (1981). The use of this corpus, containing play-scripts and everyday usage of the spoken language from a couple of decades ago, is nonetheless effective for the purpose of investigating ongoing change. 8 (i1) was included as a non-standard feature which is often observed in varieties of English. For example, the lack of the third person singular present-tense marker -s is one of the common ‗mistakes‘ of learners‘ English. (i1), which seemed to stimulate respondents‘ grammatical awareness, was listed as (1) on the questionnaire sheet. (i1) was followed by the corresponding grammatical pattern (a) so that respondents could readily start the task of judgement. 9 While sentences (h2) and (h3) are based on examples in the Keane corpus, these instances may be seen in other varieties of English. The sentence (c3) is a ‗dummy cleft‘ since there is no such example of ‗It is (adjective) that ….‘ in either HE or the standard variety. Tamami Shimada 58 altered after a pilot survey to avoid factors that might interfere with direct judgement of the morphosyntactic features under examination. For instance, the pilot version included She takes three plates from the dresser for (a) as an unmarked feature-free sentence. This sentence, however, was judged by informants as Irish, contrary to my expectation, because of the use of the word dresser. Thus, the neutral or less-culturally-marked word cupboard was employed instead of dresser or press. 10 The survey comprised eight pages in total, including the cover letter, one fact sheet for respondents‘ personal information, two pages concerning speakers‘ sociolinguistic awareness of morphosyntactic and lexical forms in HE, and four pages asking about speakers‘ attitudes and social orientation. 11 The questionnaire was time-consuming for respondents, but priority was given to the quality of information at the cost of efficient survey management. Respondents were anonymous and were asked to complete the questionnaire after prior instruction by the author and key supporters (local people who cooperated with the present author‘s work as informant/ consultants). The survey was supplemented by interviews with informants. Thus, the survey yielded both questionnaire data and, though sometimes hesitatingly, meaningful comments based on speakers‘ intuition which could not be obtained by anonymous surveys. Speaker feedback offered qualitative support for the results of this survey. 4. Linguistic forms in a sociolinguistic dimension The evidence gathered from the survey suggests a sociolinguistic significance of linguistic forms in HE. 12 The results of the survey are summarised in this section with particular attention to the selection of linguistic forms and the direction of change in HE. For the relevant discussion in section 5, the major research concerns are addressed in the light of three salient findings (4.1, 4.3 and 4.4). The discussion of the rise of extra- 10 This might be reminiscent of Labov‘s (1973) examples illustrating the difficulty that can beset attempts to tap native speaker intuition on syntactic structure, in this particular case concerning the so-called ‗positive anymore‘, Interviewer: Can people say around here We go to the movies anymore? Subject: We say show, not movies. (cited by Milroy and Gordon 2003: 175) This kind of dialogue also occurred in the elicitation concerning syntax. In passing, I came to learn in one informant session that speakers of Southwest-HE say pictures, not show or movies. Actually, a respondent gave the following sentence: She took three plates off the press. (Listowel, born in 1950s, male) in his note of ‗what we would say‘ for (a). 11 The second survey (Sheet B), which is not analysed in this paper, is designed to examine speakers‘ orientations and identity, and their relationship with linguistic evaluations. 12 See Shimada (2010b) for a detailed sociolinguistic description of the respective forms in the twenty-six examples. Morphosyntactic Features in Flux 59 linguistic meanings is included in 4.2, since this underlines the importance of speakers‘ sociolinguistic awareness. Extra-linguistic meanings are formed by the awareness raised in linguistic interactions in the community, and this can affect the use/ disuse of linguistic forms marked in the awareness dimension. 4.1. Marked and unmarked forms The first finding is given as a general observation of the results of the subjective judgments. The survey confirms that different features were judged differently. Finding 1. Morphosyntactic forms are unevenly marked in speakers‘ subjective judgements of ‗use‘/ ‗non-use‘, ‗Irishness‘ and ‗bad grammar‘. Different judgements on different forms are noted as significant as confirmed in Table 1. Table 1. The number of respondents who marked a given example in each question. 13 14 Lexical items, in general, gained more marks than morphosyntactic features in terms of use, non-use, bad grammar and Irishness. Of the morphosyntactic features, the do be form of (e1) and (e2) (named the E group) was highly conspicuous. The less marked feature was the so-called perfect sentence (the G group). In the judgement of bad grammar, the do be form was exceedingly marked. A have perfect sentence and a type of there… sentence, besides lexical items, were least judged as bad grammar. In respect of Irishness, lexical items were considerably more marked than 13 The result of question (iii) about intelligibility (i.e. ‗Don‘t understand the meaning of the sentence‘) is not included in Table 1, since the current focus is sociolinguistic awareness. Regarding incomprehensibility, the lexical example (j3) cnamhshealing was outstanding (n=23), although this word tended to be understood by the majority of the Listowel respondents who were born before 1960. The survey confirmed that (h4) amn‟t I …? failed to be understood by younger Cork respondents. 14 These figures do not show non-positive yes. That is, if respondents said ‗all‘, ‗same as Q1‘, ‗all except…‘ and the like, instead of specifying the number, they were not counted since such responses did not exhibit ‗markedness‘ in terms of speakers‘ awareness. Tamami Shimada 60 the other groups. Of the morphosyntactic features, (c2) ‟Tis lovely she is was marked in the Irishness category (n=28), and (d3) There was a great housekeeper lost in you (n=18) and (h5) having ‟twouldn‟t in the sentence initial position followed (n=18). There is an obvious division between use and non-use judgements depending on the featured groups. The majority of respondents reported their use in the D group (There… sentences), the F group (be after Ving/ NP sentence), the G group (perfect sentence), the H group (cliticisation), and to some extent in the J group (lexical items). By contrast, the B group (non-canonical constituent order), the C group (cleft-like sentence), the E group (the do be V-ing/ AdjP form) and the I group (‗non-standard‘ usage) were regarded as non-use. It is noted that two examples in the E group were both judged as non-use, far more often than the other groups. A significant difference among examples in the same group was observed especially in the B and C groups. 15 4.2. Extra-linguistic meanings of linguistic forms Table 1 shows different items mapped differently onto a sociolinguistic dimension, where Irishness and Standard are two salient categories to describe speakers‘ awareness. In communication settings, meta-linguistic interactions rise by means of speakers‘ perception of Irishness and bad grammar. 16 The use of linguistic forms, being differently located in awareness, entails meta-linguistic interactions. Linguistic forms in HE, both lexical and morphosyntactic forms, can obtain extra-linguistic meanings in addition to the linguistic meanings by virtue of speakers‘ awareness of Irishness and Standard. The do be form, the most marked grammatical form in speakers‘ awareness, is a prominent example by which the construction of extra-linguistic meanings can be lucidly explicated. In SwHE, do be V-ing/ NP functions as a habitual marker, where do and be are joined together and have no prosodic prominence in this do and be combination. Speakers confirm the habitual meaning in contemporary varieties of SwHE. (1) We do be praying for you in our prayers, whenever we get the notion to kneel. ‗We usually/ always pray for you in our prayers, whenever…‘ 15 The inconsistency in the C group is interesting, as this could lend sociolinguistic support for the claim that ‟tis lovely she is is a separable construction from clefts of the type in standard varieties of English. See Shimada (2010a: 73ff) for a ‗noncleft‘ analysis. For the B group, see Finding 2 below in 4.3. 16 See Jaworski and Coupland (2004) for the term ‗metalanguage‘ and the use of ‗meta-‘. Morphosyntactic Features in Flux 61 Importantly, speakers not only possess the knowledge of the intralinguistic meaning of a linguistic form but also of the extra-linguistic association between the form and its social connotations. The do be form is generally considered to be a deviation from the Standard, which is constructed in speakers‘ minds, and bears further negative social connotations such as ‗not-well educated‘ or ‗for poor people‘. It follows that the grammatical judgment of ‗correct‘ or ‗wrong‘ is further associated with social indicators such as education and income. This was borne out by elicited comments such as the following: ―The impression I would have is, yes, the person is… The age of the person is important. If ‘tis an old person, I would smile and ‘tis condescending smile. […] I feel superior. […] but if ‘tis the things when my pupils in the school said to me, I would correct them, you know, and would say ‗no, that is not correct‘‖ (2004, Cork, age: 50s, male); ―people who say it mostly got very little chance to go to school through poverty in the past. Now in 2004 Ireland is a rich country and you will not hear it at all‖ (2004, Listowel, age: 70s, female). These comments suggest that speakers tend to associate the use of do be with absence of a knowledge of the normative grammar learned in school. The use of do be has thus served as a criterion of education and socio-economic status. It is therefore important to note that, as well as any other social medium that yields symbolic value, such social connotations are evoked by linguistic features, being shared by members of a speech community. The connotations can exert an inevitable force on the unconscious selection and avoidance of particular linguistic forms. 4.3. Difference in the same syntactic feature category Another finding related to differential judgement of a syntactic feature category may further inform discussion on the direction of language change in the Irish context. In the do be form and the be after form, for example, the types of complement do not significantly affect speakers‘ judgements. To be precise, do be V-ing displays a similar judgemental tendency as do be NP; and be after V-ing is evaluated similarly to be after NP. However, in the examples of non-canonical constituent order or socalled fronting (Group B), there is a significant difference in judgement according to the phrasal type of the fronted constituent. In SwHE the marked constituent order is a significant syntactic device for signalling informational saliency, as is illustrated in (2). (2) Mike: (Entering). What were you doing, then, around the house? Looking here and there and walking on your toes! Pats: Thinking to steal a few eggs I was, but I changed my mind and said to myself that I would ask first before I went stealing. [SIV35] Tamami Shimada 62 In the highlighted clause, a higher value of information is placed on the first minimal constituent thinking to steal a few eggs, which means that this constituent is salient in the information structure. Saliency is syntactically expressed by marked constituent order, often along with phonological prominence. Non-finite VP, Pred-NP, Obj-NP, PP can be in the sentenceinitial position in HE. (3) Non-f. VP Gone to buy the wedding cloths they are. [SIV34] (4) Pred-NP Bloody good firing it was, too! [STD10] (5) Obj-NP Fifty pounds Dota gave to buy the clothes and the drink for the wedding. [SIV34] (6) PP Into jail ye should be put, a brace of dirty beggars. [SIV24] In the questionnaire, two types of the sentence were included. (b1) PP From the cupboard she takes three plates. (b2) Non-f. VP Taking three plates she is. This B group (b1, 2) was judged as generally non-used. It might be that the respondents found it difficult to picture a non-contextualised scene from a sentence because this syntactic pattern is closely related to the expression of the informational salience in context. Noteworthy is that speakers‘ judgements of (b1) and (b2) diverged. Despite the corpus-based diversity of the phrasal category for the fronted constituent, the judgements were inconsistent. The VP-fronted sentence Taking three plates she is is regarded as bad grammar with Irishness less marked, while the PP-fronted sentence From the cupboard she takes three plates is comparatively free from the linkage of Irishness and bad grammar. This is summarised under Finding 2. Finding 2. Speakers‘ judgements within the same grammatical feature sometimes differ, for example, non-finite VP-fronted vs. PP-fronted sentences. Finding 2 seems to offer an important clue for examining the relationship between linguistic forms, on one hand, and speakers‘ awareness of Irishness and Standard, on the other. Consider first what leads to the diffusion of the judgement between the VP-fronted and PP-fronted sentences. Table 2 clarifies the discrepancies in the judgements of bad grammar and Irishness. Morphosyntactic Features in Flux 63 PP fronted VP fronted Bad grammar 8 21 Irishness 2 10 Table 2. Judgement of non-canonical constituent order: Obtained figures. The VP-fronted sentence (b2) is marked in terms of speakers‘ awareness; more respondents reported (b2) for their answers to non-use, bad grammar, and Irishness than they did for (b1). The reason for the disparity is supplied in some of the comments that revealed the speakers‘ awareness of Irish-Gaelic syntax. One comment on (b2) is given in the interview: ―You wouldn‘t say it in English. That‘s the way you speak in Gaelic. […] It‘s correct in Irish, bad in English‖ (2006, Listowel, male, born 1930s). The comment is well-directed, weighing Irish on the one hand and standard varieties of English on the other. The Irish sentences, corresponding to the two HE sentences judged in the survey, are given in (7) and (8). (7) From the cupboard she takes three plates. Ok. Ón gcófra a thógann sí trí phláta. [Irish] from (the) cupboard PRT takes she three plates Acceptable. From the cupboard she takes three plates. [Standard English] (8) Taking three plates she is. Ok. Ag tógant trí phláta át sí. [Irish] (at) taking three plates is.REL she Unacceptable *Taking three plates she is. [Standard English] The VP-fronted type in (8) ranked fourth in the bad grammar judgement in the twenty-six examples included in the questionnaire, while (7) ranked among the last five in the judgement of ‗Irishness‘. It can be fairly concluded that the judgemental difference between (b1) and (b2) arises on account of acceptability. In other words, certain criteria may have been formed with reference to ‗Standard‘. Knowledge of the Irish language is likely to affect judgement. This is further examined in 5.2. 4.4. Regional difference In general, the two speech communities under investigation, Cork and Listowel, can be included in the larger category of the community of SwHE on the grounds that they share the same code, while at the same time there are differences in urbanization and the type of major industry between these two places. That is, the association between a linguistic form and the meaning, manners of expression and sociolinguistic norms are shared in these places; this underpins the common sociolinguistic ascription of the two varieties. In other words, these two sub-varieties derive from fairly a coherent speech community, although occasional Tamami Shimada 64 differences are observed in vocabulary and phonology. The survey confirmed this general observation; the data from Cork and Listowel display similar tendencies across many linguistic features. However, there was a morphosyntactic feature which revealed a regional difference, as summarised in Finding 3. Finding 3. A regional divergence between Cork and Listowel is observed in the amn‟t I~? example, while many of the forms do not exhibit such divergence. Amn‟t I ~ is exemplified in (9)-(11) from the Keane corpus; (10) among these was included in the survey. (9) Amn't I the same as any other man? [HHM66] (10) God help us, amn‟t I like a scarecrow always,… [SIV 4] (11) Amn‟t I supposed to have a fortune or something? [HHM 7] The difference in speakers‘ judgements between Listowel and Cork was most prominent in the amn‟t I ~? example. The use and non-use judgements in the two places were as follows: use (n=2) and non-use (n=12) in Cork, and use (n=8) and non-use (n=8) in Listowel. In addition, four respondents in Cork, only one in Listowel, reported unintelligibility of the amn‟t I ~? example. Cork respondents, in general, were unlikely to use amn‟t I, regarding it as comparatively bad grammar, while a certain Irishness was also recognised. Amn‟t I ~? was a marked feature in Cork, since a larger number of the Cork respondents reported ―non-use‖ of this form than they did for the missing of third-person singular -s ending. Listowel speakers, in contrast, tended to report their use of amn‟t I~? and associated this form more with Irishness than with bad grammar. Interestingly, in Listowel, amn‟t I~? was not so marked as the Two of us‟ll ~ example (h2) and the cliticisation of the non-auxiliary verb have (h3). The ‟twouldn‟t example (h5), in turn, was the most marked item in Listowel. The data shows that Listowel speakers, both younger and older, are likely to use ‟twouldn‟t and associate it with Irishness. To Cork speakers, on the other hand, ‟twouldn‟t is nothing but a type of unmarked cliticisation. 5. Linguistic forms in flux On the basis of the observations in section 4, one may now discuss what causes the changing pattern of features in HE, and what affects speakers‘ sociolinguistic awareness. It was confirmed, in 4.1 and 4.2, that linguistic forms in HE are mapped onto a sociolinguistic dimension with the two salient societal categories of Irishness and Standard and that particular linguistic forms can have extra-linguistic meanings by virtue of speakers‘ Morphosyntactic Features in Flux 65 awareness. The judgement by speakers may be derived from speakers‘ knowledge of these meanings. Given this, let us now examine what linguistic properties induce the formation of the extra-linguistic meanings. Findings 1-3 in section 4 are here used to consider the direction of change in relation to speakers‘ awareness. In this section, what determines subjective judgements is discussed with particular reference to tense-aspect forms. 5.1. Visualised judgemental tendencies To clarify the difference of subjective judgements in linguistic forms, consider tense-aspect forms, with the examples of do be V-ing, be after Ving, be V-ing time-Adv and a lexical item, craic, for reference. (e1) I do be taking three plates from the cupboard. [do be habitual] (f1) I am after taking three plates from the cupboard. [be after perfect] (g1) They are visiting here many years. [be V-ing time-Adv] (j1) How‘s the craic? [lexical item] The meaning of (f1) is close to that of the sentence ‗I have (just) taken three plates from the cupboard‘ in standard varieties, although the HE example conveys the emotion-explicit nuance of ‗hot news‘. The meaning of (g1) is most likely to be paraphrased in a standard variety as ‗I have {visited / been visiting} here for ten years‘. The lexical item craic, meaning ‗fun‘, is etymologically of Old English origin, while this word is believed to come from Irish, as illustrated in 2.3. This belief seems to stimulate the use of the phrase for those who have a bilingual knowledge of Irish and English. 17 Figure 1 visualises the judgemental tendencies of the four examples. The example (j1) gained the highest points in the answer to the use question; its high degree of marking as possessing Irishness is almost equal to other two lexical items of Irish origin. The lexical item craic displays the opposite tendency against the do be form. The do be habitual in (e1) and the be after perfect in (f1) are both marked in awareness, but (f1) is considered comparatively favourably while (e1) is unfavourably judged in light of the informants‘ subjective reports of non-use. Remarkably, the be perfect continuous example They are visiting here many years in (g1), though its use being comparatively confirmed, is unmarked in terms of the speakers‘ awareness of Irishness and Standard. (g1) is the most unmarked in speakers‘ awareness despite its distinctiveness from a crossdialectal perspective. 17 ‗How‘s the craic? ‘, underpinned by many ‗use‘ notings, was one of the most popular expressions which speakers wrote down for the requested example of ‗Irish English‘ in the author‘s questionnaire survey in 1999. See also 2.3. Tamami Shimada 66 Figure 1. Judgement tendencies of the four tense-aspect forms. 5.2. Morphosyntactic conformity Given the different judgemental tendencies regarding different linguistic forms, one of the most intriguing concerns is what makes the difference in speakers‘ awareness. This can be discussed with the four examples in 5.1. For example, one could explain why do be is marked as bad grammar with the dominant report of non-use. Likewise, be after is, on the other hand, in comparison reported as use, while Irishness is equivalently recognised in these two forms. 18 In a similar vein, the question arises why be perfect continuous is not marked in speakers‘ awareness; markedness being as little as with the have perfect. The be after perfect, however, attracts speakers‘ attention compared to other perfect expressions. How are these tendencies to be explained? A persistent clue to such questions may be the excessive marking of non-use and bad grammar in the do and be combination. Interviewees frequently revealed a strong reaction to the do be form for its bad grammar, as testified in the written comment from a Listowel respondent: ―We never say ‗do be‘ or ‗does be‘ [; it is] considered very bad grammar‖ (Lis- 18 Interesting data concerning the ‗acceptability‘ of these two types of sentence in HE is provided by Hickey (2007), who undertook a questionnaire on the acceptance of sample sentences to younger respondents. Hickey (2007: 207, 234) reports for the test sentence She‟s after spilling the milk a 96% acceptability rate in County Kerry (n= 24) and 93% in County Cork (n=84) and for She does be worrying about the children 13% in Kerry and 6% in Cork. Morphosyntactic Features in Flux 67 towel, born in 1980s, female). Do be and does be are stigmatised in this way. Some respondents acknowledged the habitual category in Irish grammar and the Irishness judgement of the do be form was as high as the be after form. Speakers‘ judgements, however, do not take advantage of this reference to Irish. The reason is explained by the intuitional imprinting of normative grammar and conventionalisation. A more essential reason can be deduced from interviews and comments with the respondents: The sequence of do and be causes unacceptability of this form in their speech. That is, the sequence of two auxiliaries, do and be, does not meet the criterion of Standard by a morphosyntactic constraint in the linguistic knowledge of speakers. In other words, the form of do be does not meet morphosyntactic conformity in the light of their awareness of the grammatical rules of the Standard. The be after V-ing form is a contrasting example, with its adherence to morphosyntactic conformity. The be after form, having two types of the complement V-ing and NP, does not violate English morphosyntax, despite the distinctiveness from Standard English. The noun categories including gerund follow a preposition, which conforms precisely to English morphosyntax. It is important to note that this morphosyntactic conformity is distinguished by semantic congruity, which refers to the meaning of a given form. If this matches that of Standard English, then the form agrees in semantic congruity. Speakers of SwHE are highly conscious of morphosyntactic conformity, but not of semantic congruity. A form that satisfies morphosyntactic conformity is generally accepted insofar as the form has semantic distribution in the grammatical system of SwHE. Finding 2 shown in 4.3 helps further explain morphosyntactic conformity in relation to speakers‘ subjective judgements. The VP-fronted example Taking three plates she is was judged as relatively bad grammar, in proportion half as often as do be V-ing; the PP-fronted type From the cupboard she takes three plates was judged as less bad than the VP-fronted example. This supports the view that English morphosyntactic constraints, involving speakers‘ awareness of Standard, affect speakers‘ judgement. Despite the fact that Irish allows for this syntactic pattern and many speakers of HE know about this (see (8)), the acceptability in Irish, while fostering awareness of Irishness, does not overturn the judgement of bad grammar. On the contrary, this acceptance in Irish morphosyntax may sometimes highlight disconformities in English and veer to the violation of the morphosyntactic conformity. The fronting of the non-finite VP, thus, does not sufficiently satisfy morphosyntactic conformity in SwHE. 5.3. Forms in flux: Some changes in progress Two forms, namely amn‟t I~? and the do be form, are noted here for their potential implication for change in progress. Local differences in Finding 3 in the previous section are noteworthy for the consideration of lan- Tamami Shimada 68 guage change in the speech community as seen in the data from two places. Milroy (1987: 86) notes that ―it seems reasonable to assume that insight can be gained into the process of their formation if a set of data from a city is compared with a set from a surrounding area‖, then citing the remark of Labov (1972: 300), since ―linguistic change and rapid dialect mixing appear to be a general characteristic of urban dialects‖. In the present work, the data of Cork City is assumed to represent an urban variety and Listowel Town a semi-rural variety. The example Amn‟t I like a scarecrow? was comparatively judged as ‗bad grammar‘ in the Cork survey but was rather unmarked in Listowel. This may indicate a potential shift of the amn‟t I form towards disuse owing to normative awareness, with the alternation of aren‟t I (e.g. Aren‟t I like a scarecrow? ), although this form actually remains in use. 19 Amn‟t I, which highlights variations in flux in rural and urban varieties, may serve as an example for investigating the dynamism of HE. This explains how knowledge about a standard variety affects use/ disuse of linguistic forms. The do be form is another interesting example for discussing change in progress. A linguistic form once labelled ‗bad grammar‘, often along with the social meaning of ‗not well-educated‘ and ‗for poor people‘, is a warning light. The attention that the form then attracts may have hindered speakers from using this ‗stigmatised‘ form. 20 The do be form, generally speaking, seems to be moving into disuse, if we focus on the contemporary situation. This is confirmed by the dominant tendency in speakers‘ reports of non-use. It is true, however, that the majority of the SwHE speakers, both urban and rural, even the younger generation, have recognition or knowledge of the linguistic meaning of the do be form. The linguistic knowledge of speakers, moreover, is formed by reference to the Irish language as follows: ―The phrase ‗I do be…‘ is a direct translation from Irish. In the Irish language there are 2 present tenses - ‗I am‘ and ‗I do be…‘, but in English there‘s only one. But this direct translation isn‘t used in all parts of the country‖ (2006, Listowel, born in 1980s, female). 19 There are examples including recent usage from a national newspaper. For example, And all the whole juggling our sense of, ―What the hell am I doing here? Why amn‟t I living in Vegas with a stripper? ‖ [Jul 28, 2009, The Irish Times, written by a columnist who is originally from Dublin and lives in Cork]. and […] I never thought I‘d get it because in some ways I thought the view would be ‗sure, amn‟t I getting enough? ‘…. [Feb 01, 2008, The Irish Times, part of quotation of the speech by a Clare musician]. 20 It is, however, not intended to imply that the do be form is or will be entirely lost in this dialect. There are SwHE speakers who internalise this construction in their grammar. A characteristic linguistic feature of which speakers are aware of its unfavourable social connotations may come to be assigned covert prestige (cf. Labov 2001 ‗The nonconformity hypothesis‘). Morphosyntactic Features in Flux 69 This respondent listed only the do be example in her response to the question of Irishness, but she did not identify this form as ‗bad grammar‘ unlike the majority of older respondents. Younger speakers today are more generous towards this form, though the database was too small to demonstrate a generational tendency. Thus, it can be assumed that the do be habitual was in active use until the 1950-1960s, but today, in relative decline, it is taking on unfavourable social connotations associated with its conspicuous non-standardness. Furthermore, such negative connotations may be disappearing due to disuse; this is slightly but nonetheless definitively indicated by the generational analysis of results. In the survey, Listowel respondents born before 1935 and between 1960 and 1980 and Cork respondents born after 1976 did not associate the do be form with Irishness; the younger generations meanwhile displayed a relative absence of the ‗bad grammar‘ judgement. In sum, awareness of Standard, which is constructed from normative school grammar and models of Standard English, sometimes assigns negative social connotations to the features that deviate from the speaker‘s criteria of Standard. However, those connotations may eventually disappear as a result of increasing disuse caused by the excessive markedness of the given linguistic form. The examples of amn‟t I~? and do be, revealing a certain status of being in flux, seem to attest the significance of normative permeation. It can be assumed that particular social settings activate awareness of Standard towards non-Standard forms which violate morphosyntactic conformity. Awareness of Irishness, largely due to actual knowledge of the Irish language, may determine the social connotations of a particular feature, contributing to change in HE via the actual use/ disuse of forms. It should be noted that extra-linguistic meaning and connotation is socially dependent by nature, reflecting the values cultivated in a given speech community. In the survey, the evaluation of linguistic items could vary according to the fluctuation of the attitudes towards Irishness and Standard in the sociocultural context. Whether a positive or a negative connotation is associated with a particular linguistic form depends on the direction in which social attitudes are shifting. 6. Conclusion This paper has described sociolinguistic aspects of morphosyntactic forms in SwHE. It has illustrated speakers‘ awareness of the forms based on a survey, drawing on some significant findings, discussed the direction of change. Morphosyntactic forms in SwHE are mapped differently onto a sociolinguistic dimension, where Irishness and Standard are two salient categories in speakers‘ awareness. The findings such as divergent judgements within the same syntactic feature category (non-finite-verb-phrase vs. preposition-phrase fronting), the regional difference of amn‟t I~? , and Tamami Shimada 70 the contrast of two tense-aspect forms, do be and be after, lead to the suggestion of morphosyntactic conformity to explain the factor that is relevant to the selection of linguistic forms. Examination of the tendencies in the subjective judgement of grammatical forms suggests that whether a given sentence meets morphosyntactic conformity or not depends on the speakers‘ awareness of Standard, and the disconformities cause the judgement of bad grammar. The awareness of Irishness, maintained largely by speakers‘ knowledge of Irish-Gaelic, sustains the significance of the forms in which Irishness is perceived. It is likely that the forms that speakers recognise as Irish were more marked than others, usually leading to either use or non-use, determined by morphosyntactic conformity. With this conformity being established, Irishness obtains positive connotations. The balance between Irishness and Standard is the key to the direction of morphosyntactic change in HE. Acknowledgements This study is supported by the respondents to my interviews and the 2006 questionnaire. I would like to express my gratitude to the respondents and to all those who have supported my fieldwork. Particularly, I would like to thank my friends, who are key consultants: Elsie Harris, Emmet Stones, and the Keane, Clifford and O‘Connor families. An earlier version of this paper was presented at the Klagenfurt Conference on Corpus-based Applied Linguistics (CALK 14) in September 2014. I am grateful to Professor Allan James at University of Klagenfurt, the conference organizer, for encouraging me to submit this paper to the AAA journal, and for his many insightful comments since our fortunate encounter. This article is an outcome of my current research projects: ‗Diversity and continuity of Englishes: Hiberno-English in focus‘ (Research project number: 25770141) granted by Grant-in-Aid for Scientific Research (KA- KENHI) and ‗Sentence forms and meanings: Development from the comprehension of various theories‘ granted by Meikai University Miyata Research Grant 2014. References Aikhenvald, Alexandra (2001). ―Language awareness and correct speech among the Tariana of northwest Amazonia‖. Anthropological Linguistics 43. 411-430. Anchimbe, Eric A. (2007). Linguistic Identity in Postcolonial Multilingual Spaces. Newcastle: Cambridge Scholars Publishing. Bex, Tony and Richard J. Watts (1999). Standard English: The Widening Debate. London: Routledge. Morphosyntactic Features in Flux 71 Bliss, Alan J. (1972). ―Languages in Contact: Some Problems of Hiberno-English‖. Proceedings of the Royal Irish Academy 72. 63-82. Census (2006). Volume 1: Population Classified by Area. Dublin: Central Statistics Office Ireland. Dolan, Terence Patrick (1999). A Dictionary of Hiberno-English: The Irish Use of English. Dublin: Gill & Macmillan. Filppula, Markku (1999). The Grammar of Irish English: Language in Hibernian Style. London: Routledge. Greenbaum, Sidney (1996). The Oxford English Grammar. Oxford: Oxford University Press. Harris, John (1993). ―The grammar of Irish English‖. In: Milroy and Milroy (eds.). Real English: The Grammar of English Dialects in British Isles. 139-186. Hickey, Raymond (2003). Motives for Language Change. Cambridge: Cambridge University Press. Hickey, Raymond (2007). Irish English: History and Present-day Forms. Cambridge: Cambridge University Press. Jaworski, Adam, Nikolas Coupland and Dariusz Galasiński (2004). Metalanguage: Social and Ideological Perspectives. Berlin, New York: Mouton de Gruyter. Kallen, Jeffrey L. (1997). Focus on Ireland. Amsterdam: John Benjamins. Labov, William (1972). Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Labov, William (1973). ―Where do grammars stop? ‖ In: Shuy (ed.). Sociolinguistics, Current Trends and Perspectives. 43-88. Labov, William (2001). Principles of Linguistic Change: Social Factors. Malden, Oxford: Blackwell. Milroy, James (2003). ―On the role of the speaker in language change‖. In: Hickey (ed.). Motives for Language Change. Cambridge: Cambridge University Press. 143-157. Milroy, James and Lesley Milroy (1993). Real English: The Grammar of English Dialects in British Isles. London: Longman. Milroy, James and Lesley Milroy (1998). Authority in Language: Investigating Standard English. Third Edition. London: Routledge. Milroy, Lesley (1987). Observing and Analysing Natural Language: A Critical Account of Sociolinguistic Method. London: Blackwell. Milroy, Lesley (1999). ―Standard English and language ideology in Britain and the United States‖. In Bex and Watts (eds.). Standard English: The Widening Debate. 173-206. Milroy, Lesley and Matthew Gordon (2003). Sociolinguistics: Method and Interpretation. Malden, Oxford: Blackwell. Shimada, Tamami (2007). ―‗Irishness‘ in Hiberno-English: Linguistic hybridism and ethnolinguistic identity‖. In: Anchimbe (ed.). Linguistic Identity in Postcolonial Multilingual Spaces. 285-309 ． Shimada, Tamami (2010a). English in Ireland: Beyond Similarities. Hiroshima: Keisuisha. Shimada, Tamami (2010b). ―What grammatical features are more marked in Hiberno-English? : A survey of speakers‘ awareness and its primary details‖. Bulletin of Graduate School of Social and Cultural Systems at Yamagata University 7. 1-25. Shuy, Roger W, ed. (1973). Sociolinguistics, Current Trends and Perspectives. Washington DC: Georgetown University Press. Smith, Gus and Des Hickey (2002). John B. Cork: The Mercier Press. Tamami Shimada 72 Trudgill, Peter (1999). ―Standard English: What it isn‘t‖. In: Bex and Watts (eds.). Standard English: The Widening Dabate. 117-128. van Hamel, A. G. (1912). ―On Anglo-Irish Syntax‖. Englische Studien 45. 272-292. Morphosyntactic Features in Flux 73 Appendix: Questionnaire 2006 (Part I: Awareness of morphosyntactic forms) [The left page] From the sentences (1) to (26) on the opposite page, please choose the ones that apply to the following five statements. You may choose as many as you like. If you think there is no relevant number, please fill the bracket with ―Nothing‖. Any of your comments are welcome. Question 1: Which sentences do you think you would use yourself? Your answer: [ ] Your comment: Question 2: Which sentences do you think you would not use on any occasion (including when you are talking with your family and friends)? Your answer: [ ] Your comment: Question 3: Which sentences can you not understand the meaning of? Your answer: [ ] Your comment: Question 4: Which sentences do you think have “bad grammar”? Your answer: [ ] Your comment: Question 5: Which sentences do you think show “Irishness”? Your answer: [ ] Your comment: [The right page] (1) She take three plates from the cupboard. (=a) (2) She takes three plates from the cupboard. (=i1) (3) I am after taking three plates from the cupboard. (=f1) (4) From the cupboard she takes three plates. (=b1) (5) The two of us‘ll take three plates from the cupboard. (=h2) (6) I do be taking three plates from the cupboard. (=e1) (7) It is from the cupboard that I take three plates. (=c1) (8) Taking three plates she is. (=b2) (9) I asked for today‘s special and she putting plates on the table. (=i3) (10) How‘s the craic? (=j1) (11) You‘ve the name of a good employer. (=h3) (12) They are visiting here many years. (=g1) (13) There‘s no one can deny it. (=d1) (14) She does be lovely with her long hair. (=e2) (15) ‘Tis lovely she is. (=c2) (16) It is lovely that she is. (=c3) (17) Amn‘t I like a scarecrow? (=h4) (18) We‘ll visit here tomorrow. (=h1) (19) ‘Twouldn‘t be a good thing. (=h5) Tamami Shimada 74 (20) There was a great housekeeper lost in you. (=d3) (21) She been taking them home ever since. (=i2) (22) Tom is after his supper. (=f2) (23) That amadán put eggs in my bag. (=j2) (24) Don‘t be cnamhshealing! (=j3) (25) My sons have visited there for many years. (=g2) (26) I knew there was good news in you. (=d2) Note: The corresponding indexes are given for current convenience. They were, of course, not in the questionnaire for the respondents. Tamami Shimada Department of English Faculty of Languages and Cultures Meikai University, Japan