138 The emergence of Irish English ..     The case for contact should be considered across all linguistic levels. However, those authors who have been examining this recently, Corrigan, Kallen, Filp- pula and McCafferty, to mention the more prominent among them, have not considered phonological factors in their investigations, despite the benefits for analyses from doing this (Hickey 1990: 219). If one looks at structures which could be traced to transfer from Irish, then one finds in many cases that there is a correspondence between the prosodic structures of both languages. To be precise, structures which appear to derive from transfer show the same number of feet and the stresses fall on the same major syntactic category in each language (Hickey 1990: 222). A simple example can illustrate this (in (9) the Irish equiva- lent is given which is not of course the immediate source of this actual sentence as the speaker was an English-speaking monolingual). (9) A don’t like the new team at all at all. (WER, M55+) [] N´ı thaitn´ıonn an fhoireann nua le hA ar chor ar bith. [] [not like the team new withA onturnonanything] The repetition of at all at all creates a sentence-final negator which consists of two stressed feet with the prosodic structure WSWS (weak–strong weak–strong) as does the Irish structure ar chor ar bith. This feature is well established in Irish English and can already be found in the early nineteenth century,e.g. in the stories of John Banim (1798–1842) written in collobaration with his brother Michael. Consider now the stressed reflexives of Irish which are suspected by many authors (including Filppula 1999: 77–88) of being the source of the Irish English use of an unbound reflexive. (10) An bhfuil s´e f´ein is tigh in niu? [interrog is he self in today] IrEng: ‘Is himself in today?’ ‘Is he himself in today?’ The strong and weak syllables of each foot are indicated in the Irish sentence and its Irish English equivalent above. From this it can be seen that the Irish reflexive is monosyllabic and, together with the personal pronoun, forms a WS foot: s´e f´ein [he self]. In Irish English the equivalent to this consists of a reflexive pronoun on its own: himself, hence the term ‘unbound reflexive’ (Filppula 1997c), as no personal pronoun is present. If both the personal and reflexive pronoun were used in English, one would have a mismatch in prosodic structure: WS in Irish and SWS (he himself)inIrish English. One can thus postulate that the WS pattern of himself was interpreted by speakers during language shift as the prosodic equivalent of both the personal pronoun and reflexive pronoun of Irish s´e f´ein andthususedas an equivalent of this. Later a distinctsemanticisation of this usage arose whereby the unbound reflexive came to refer to someone who 4.2 The case for contact 139 is in charge, the head of a group or of the house, etc. (see section 4.4.2.3 for examples). Another example of prosodic match can be seen with the well-known imme- diate perfective of Irish English which corresponds, in the number of stressed syllables, to its Irish equivalent. (11) a. She’s after breaking the glass. [] T´as´ıtr´eis an ghloine a bhriseadh. [] b. He’s after his dinner. [] T´as´etr´eis a dhinn´eir. [] This consists in both languages of three ortwo feet depending on whether theverb is understood or explicitly mentioned (it is the number of stressed syllables which determines the number of feet). In both languages a stressed syllable introduces the structure and others occur for the same syntactic categories throughout the sentence. This prosodic similarity would also have helped people during language shift to recognise the new English structure – in the speech of others – as an equivalent to the Irish structure they already knew. Thus cases of individual transfer would have spread easily, ultimately becoming established as community-wide features which were then passed on to following generations. A similar prosodic correspondence can be recognised in a further structure, labelled ‘subordinating and’ (see section 4.4.6.2), in both Irish and Irish English. (12) a. He went out  and it raining. ‘He went out although it was raining.’ b. Chuaigh s´e amach  agus ´eagcurb´aist´ı. [went he out and it at putting rain-] Again there is acorrelation between stressed syllable and major syntactic category, although the total number of syllables in the Irish structure is greater (due to the number of weak syllables). The equivalence intonationally is reached by having the same number of feet, i.e. stressed syllables, irrespective of the distance between them in terms of intervening unstressed syllables. And again, it is a stressed syllable which introduces the clause. A prominent feature in Irish is the lack of a word for ‘yes’ and ‘no’. Questions are replied to in the affirmative or negative by using a form of the verb be,inthe negative if required. (13) An bhfuil t´uagdulgo dt´ıancluiche am´arach? [ is you-SG going to the match tomorrow] T´a.[t  ɑ ] ∨ N´ıl.[n j il j ] ∨ [is] [not-is] 140 The emergence of Irish English The single-word verb forms are frequently spoken with a fall–rise intonation (indicated by ∨ ) and this was evident in the speech of the informants recorded for A Collection of Contact English (see section 4.4). (14) Are you getting support from the EU for sheep farming? (RH) Iam ∨ (CCE-S, M60+) Afall pattern (without the rise in t´a and n´ıl)isfound with a stressed short vowel which occurs when negating something in the past. (15) a. An raibh t´u riamh i Meirice´a? N´ı raibh. \ (CCE-S, M60+) [ were you ever in America] [not was] b. Did your brother work on the farm as well? (RH) He did not. \ (CCE-W, M75+) Ye t another case, where prosodic equivalence can be assumed to have motivated anon-standard feature, concerns comparative clauses. These are normally intro- duced in Irish by two equally stressed words n´a mar‘thanlike’,as in the following example. (16) T´as´eibhfad n´ıos fearr anois n´a mar a bh´ı. [is it further more better now not like  was] ‘It’s now much better than it was.’ Several speakers from Irish-speaking regions, or those which were so in the recent past, show the use of than what to introduce comparative clauses. (17) a. It’s far better than what it used to be. (TRS-D, C42–1, F) b. To go to a dance that time was far better than what it is now. (TRS-D, C42–1, F) c. Life is much easier than what it was. (TRS-D, C42–1, F) d. They could tell you more about this country than what we could. (TRS-D, M7, M) It is true that Irish mar does not mean ‘what’, but what can introduce clauses in other instances and so it was probably regarded as suitable to combine with that in cases like those above. From the standpoint of prosody than what provided a combination of two equally stressed words which match the similar pair in equivalent Irish clauses. The use of than what for comparatives was already established in the nine- teenth century and is attested in many emigrants’ letters such as those written from Australia back to Ireland, e.g. the following appeared in a letter from a Clare person written in 1854: I have more of my old Neighbours here along with me than what I thought (Fitzpatrick 1994: 69). It is also significant that the prosodi- cally similar structure like what is attested in the east of Ireland where Irish was replaced by English earliest, e.g. There were no hand machines like what you have today (SADIF, M85, Lusk, Co. Dublin). 4.2 The case for contact 141 ..   Despite the typological differences between Irish and English there are nonethe- less a number of unexpected parallels which should not be misinterpreted as the result of contact. Some cases are easy, such as the homophony between Irish s´ı /ʃi / ‘she’ and English she (the result of the vowel shift of /e/to/i/inEarly Modern English); see remarks in section 5.6.Asimilar homophony exists for Irish b´ı ‘be’ and English be, though again the pronunciation of the latter with /i /isdue to the Great Vowel Shift. Other instances involve parallel categories, e.g. the continuous forms of verbs in both languages: T´am´eagcaint l´ei [is me at talk-  with-her] ‘I am talking to her.’ Indeed the parallels among verbal distinctions may have been a trigger historically for the development of non-standard distinctions in Irish English, i.e. speakers during the language shift who found equivalents to most of the verbal categories from Irish expected to find equivalents to all of these. An example of this is habitual aspect, which is realised in Irish by the choice of a different verb form (habitual b´ıonn versus non-habitual t´a). (18) B´ıonn s´eagcaint l´ei. [is- he at talking with-her] IrEng: ‘He does be talking to her.’ ‘He talks to her repeatedly.’ The possible pathway by which an equivalent to this arose in Irish English is discussed in the section on the habitual in 4.4.1.4.3. Another coincidental parallel between the two languages involves word order, despite the differences in clause alignment which both languages show. In both Irish and English prepositions may occur at the end of a clause. A prepositional pronoun is the most likely form in Irish because it incorporates a pronoun which is missing in English. (19) An buachaill a raibh m´eagcaint leis. [the boy that was I at talk-  with-him] ‘The boy I was talking to.’ Further parallels may be due to contact which predates the coming of English to Ireland. For example, the use of possessive pronouns in instances of inalienable possession is common to both English and Irish. (20) Ghortaigh s´eaghl´uin. [injured he his knee] ‘He injured his knee.’ This may well be a feature of Insular Celtic which was adopted into English (Vennemann 2000, 2001), especially given that other Germanic languages do not necessarily use possessive pronouns in such contexts, cf. German Er hat sich am Knie verletzt, lit. ‘He has himself at-the knee injured.’ 142 The emergence of Irish English ..      If the expectation of categories in the target language which are present in the outset language is a guiding principle in language shift, then it is not surprising to find that grammatical distinctions which are only found in the target language tend to be neglected by speakers undergoing the shift. The reason for this neglect is that speakers tend not to be aware of grammatical distinctions which are not present in their native language; at least this is true in situations of unguided adult learning of a second language. What is termed here ‘neglect of distinctions’ is closely related to the phenomenon of underdif- ferentiation which is known from second-language teaching (Major 2001). This is the situation in which second-language learners do not engage in categorial distinctions which are present in the target language, for instance when German speakers use the verb ‘swim’ to cover the meanings of both ‘swim’ and ‘float’ in English (schwimmen is the sole verb in German) or when they do not distin- guish between when and if clauses (both take wenn in German). This neglect can be illustrated by the use of and as a clause co-ordinator with a qualifying or concessive meaning in Irish English. (21) Chuaigh s´e amach agus ´eagcurb´aist´ı. [went he out and it at putting rain-] IrEng: ‘He went out and it raining.’ ‘He went out although it was raining.’ To account for the neglect of distinctions in more detail, one must introduce a distinction between features which carry semantic value and those which are of a more formal character. Word order is an example of the latter type: Irish is aconsistently post-specifying language with VSO as the canonical word order along with noun + adjective, noun + genitive for nominal modifiers. There is no trace of post-specification in Irish English, either historically or in present-day contact varieties of English in Ireland. The use of the specifically Irish word order would, per se, have had no informational value for Irish speakers of English in the language shift situation. Another example, from a different level of language, would be the distinction between palatal and non-palatal consonants in Irish phonology. This difference in the articulation of consonants lies at the core of the sound structure of Irish. It has no equivalent in English and the grammatical categories in the nominal and verbal areas which it is used to indicate are realised quite differently in English (by word order, use of prepositions, suffixal inflections, etc.). An awareness of the semantic versus formal distinction helps to account for other cases of non-transfer from Irish. For instance, phonemes which do not exist in English, such as /x/ and / γ/, have not been transferred to English, although there are words in Irish English, such as taoiseach ‘prime minister’, pronounced [ti ʃək], with a final [-k] and not [-x], which could have provided an instance of such transfer. Although the /k/ versus /x/ distinction is semantically relevant 4.2 The case for contact 143 in Irish, it would not be so in English and hence transfer would not have helped realise any semantic distinctions in the target language. A further conclusion from these considerations is that the source of a sound like /x/ in Ireland can only be retention from earlier varieties of English. This explains its occurrence in Ulster Scots and in some forms of mid Ulster English, but also its absence elsewhere, although it is present in all dialects of Irish. The literature on Irish English contains remarks on the relative infrequency of the present perfect in Irish English (as early as Hume 1878; see Kallen 1990). This is a category which has no equivalent in Irish and so it is not surprising that it is underrepresented in Irish English, as seen in the following examples. (22) a. I’m seven years home now. (TRS-D, C42–2, F) b. She’s there for six years. (TRS-D, C42–2, F) Several features from Irish syntax are conspicuously absent from historical doc- uments in Irish English. That this is not an accident of the textual record was confirmed by the material in A Collection of Contact English,adata collection consisting of the English of good present-day speakers of Irish (see section 4.4). Table 4.6 lists the salient features of Irish grammar which were never transferred into English in this collection, even in discourse situations with considerable code-switching. The argument from parameter setting Irish is a post-specifying language (VSO, N+Gen, N+Adj) and the fact that English is pre-specifying (SVO, Gen+N, Adj+N) is recognised quickly by lan- guage learners, and would have been in the historical language shift as well. This recognition then blocks (and blocked in the past) the transfer of any post- specifying strategies from Irish to English. The view that the direction of spec- ification is a parameter of language, which needs to be recognised by only one setting and which is then fixed for all others, is supported by the data in A Collection of Contact English and by the history of Irish English. The question of structural match Initial mutation in Irish (see last item in table 4.6)isacentral device for indicating essential grammatical categories such as tense, number, gender, case, etc. And yetitisastructural principle which is never transferred to English. The reason probably lies in its unique phonological character. There is no way of matching it to any grammatical process in English and then transferring it, something which is possible with many syntactic structures which can be mapped reasonably well onto English syntax. Other factors in neglect of features The neglect of a form in the target language may in some instances be motivated not so much by its absence in the outset language, but by some other factor. Take, 4.3 Structural features of Irish 145 for example, the lack of do support with negated have in Irish English (Trudgill, Schreier, Long and Williams 2004). Here not is cliticised onto have and not onto do which is absent in negated sentences of this type. (23) a. You haven’t much trouble at all with it. (WER, M55+) (cf. You don’t have ) b. You haven’t to dry it or anything. (WER, F55+) (cf. You don’t have ) One explanation for this is that the use of do in habitual structures (as of the nineteenth century in Irish English) may well have triggered its avoidance in sentences with negated have. Another instance of this avoidance would be the past of use to which does not occur with do in (southern) Irish English, e.g. He usen’t to drive to work, not He didn’t use to drive to work. Overrepresentation The mirror image of the neglect of distinctions is the overrepresentation of fea- tures, that is the scope of a feature in the outset is applied to the target language where this scope is usually smaller. The Irish English use of the conditional illus- trates this phenomenon. It is non-standard inasmuch as it represents an overuse compared with other forms of English, e.g. as an equivalent to the imperative or in interrogatives as with the following examples. (24) a. Would you hurry up with your tea! (WER, M55+) b. Would the both of youse get off out of here! (DER, M35+) c. Would you be able to cook if you had to? (WER, M50+) This overrepresentation also applied to the definite article (see section 4.4.4 for more details). Curiously, the indefinite article, which does not exist in Irish, is not dropped in English. This might be expected because it is known from other languages, such as Russian, that the lack of an article (here the definite article) leads to its neglect in a target language, such as English, which has one. 4.3 Structural features of Irish In terms of structure, Irish 12 and the other Celtic languages show certain features which link them typologically (Borsley and Roberts 1996). They all have the 12 Irish is amember of theCeltic languages and, along withScottish Gaelic and Manx (now extinct), it forms the Q-branch. The complementary P-branch consists of Welsh, Cornish (partially revived) and Breton. This division is derived from the treatment of inherited Indo-European k/kw which appears as /k/ in the Q-branch and /p/ in the P-branch; compare Irish ceann ‘head’, ceathair ‘four’ with Welsh pen, pedwar. The proto-Celtic language appeared on the European mainland in the last centuries BC. Between roughly 500 and 300 BC Celtic speakers moved across to the British Isles. The Celts who came to Ireland were speakers of Q-Celtic and possibly came through different routes and at separate times from the P-Celtic speakers who settled in Britain. 146 The emergence of Irish English system of initial mutation whereby such essential grammatical categories as tense, gender, case and number are indicated by altering the initial sound of a word. In grammar, all Celtic languages share the principle of post-specification (see remarks in previous section) which can be seen in the following Irish examples. (25) a. Dh´un an cail´ın an doras. [closed the girl the door] ‘The girl closed the door.’ b. T´ahataShe´ain ar an mbord. [is hat John- on the table] ‘John’s hat is on the table.’ c. Is oile´an ´alainn Inis Me´ain. [is island beautiful Inishmaan] ‘Inishmaan is a beautiful island.’ The typological structure of Modern Irish was already established by the time of the first attestations in the Old Irish period (AD 600–900). What happened in the following centuries is that the nominal inflections and complex verb forms were simplified considerably, yielding a morphologically simpler system by the end of the Middle Irish period (AD 900–1200). This was the time when the first settlers from England arrived so that it is Irish of the early modern period (AD 1200–1600) which they came in contact with. Some structures, which had an apparent influence on emerging forms of English in Ireland, were not completely established by then, notably the use of iar, tar ´eis (tr´eis)ori ndiaidh, ‘after/behind’, to express an immediate perfective ( ´ O Corr ´ ain 2006), though the initial use in Irish probably does go back to the early modern period ( ´ OS ´ e 1992 pace Greene 1979). The following sections (4.3.1–4.3.4) offer a brief description of the grammar of Irish 13 which provides the basis for the ensuing discussion of contact and language shift. Only those structures of Irish are discussed here which could possibly have influenced English in Ireland. For instance, verb forms in the past or conditional are not listed as these have never had any influence on a variety of Irish English. 13 Unfortunately, there is no comprehensive reference grammar of Irish written in English. A useful introductory book to consult is the short school grammar produced by the Christian Brothers, New Irish Grammar (1977). Other grammars, but written in Irish, are available, notably a fuller form of the Christian Brothers’ grammar: Graim´ear Gaeilge na mBr´aithre Cr´ıosta´ı ‘Grammar of the Christian Brothers’ (latest edition: 1999). There is also a comprehensive academic overview of the history of the Irish language from the beginning to the present day, which again is in Irish, Stair na Gaeilge ‘The history of Irish’ (McCone et al. 1994). An envisaged English version has not been published to date (mid 2006). Some other works could be mentioned in this context. ´ O Siadhail (1989)isanoverview by a good speaker of Irish but which is linguistically unreliable. Good r ´ esum ´ es are to be found by MacEoin in Ball and Fife (1993) and ´ O Dochartaigh in Macaulay (1992). 4.3 Structural features of Irish 147 ..    The verb system of Irish has been greatly simplified since the earliest period. What is left is a system with three tense distinctions – present, past and future – and a formal distinction between indicative and subjunctive. These distinctions are made by inflectional endings and, in the past and subjunctive, by an initial mutation as well. Because of the large number of former verb forms, many sup- pletive forms survive, rendering the paradigms of common verbs very irregular. Non-finite verb forms There is no infinitive in Irish. What corresponds to that of English is a non-finite verb form, traditionally known as the ‘verbal noun’ (Christian Brothers 1977: 126–30), abbreviated as ‘VN’ in the following, which can be the equivalent of an infinitive complement or the progressive in English. (26) a. Ba mhaith leis dul amach. [would like with-him go-VN out] ‘He wants to go out.’ b. T´a Brian ag foghlaim (VN) na Fraincise. [is Brian at learn-VN French-] ‘Brian is learning French.’ When this non-finite verb form occurs with sentences expressing purpose, the preposition chun ‘in order to’ is found. (27) Chuaigh N´ora amach chun m´oinafh´ail. [went Nora out in-order-to turf get-VN] ‘Nora went out to get turf.’ The infinitival phrase has the word order Obj + Verb-- with the particle a before the verb form much like English to. This applies to any such structure, irrespective of whether the sentence expresses purpose or not. For an infinitival phrase in the negative, gan ‘without’, i.e. ‘not to’, is used instead of a. (28) a. Mhol s´ed´uinn teach a cheannach. [advised he to-us house to buy-VN] ‘He advised us to buy a house.’ b. D´uirt s´ı linn gan a bheith d´ana. [told she with-us without be misbehaved] ‘She told us not to misbehave.’ Tense The tripartite division of tense of Irish – present, past, future – corresponds on first sight to that of English. However, the future is formed synthetically in Irish and the present has a greater range as it covers that of the present perfect of English. 148 The emergence of Irish English (29) a. Rachaidh m´echuige n´ıos d´eana´ı. [will-go I to-him more late] ‘I will go to him later on.’ b. T´am´e anseo le tr´ı huaire anois. [is I her with three hours now] ‘I have been for three hours now.’ It is nonetheless possible in Irish to express the continued relevance of an action to the present (30a) and contrast this with a structure ( ´ OS ´ e 1992) where this is not the case (30b). (30) a. T´aocht bpota d´eanta aige. [is eight pots made at-him] ‘He has made eight pots.’ (and will possibly make more) b. Rinne s´eocht bpota. [made he eight pots] ‘He made eight pots.’ (and is not making any more now) Analytic tense structures indicating the future are available in Irish by com- bining the verb t´a ‘be’ with the non-finite form of a further verb, much as in English. (31) T´am´eagdul go Luimneach am´arach. [is me at go-VN to Limerick tomorrow] ‘I am going to Limerick tomorrow.’ Aspect Irish has several aspectual categories which are expressed with particular struc- tures. Progressive aspect can be expressed using t´a and a non-finite verb form (the verbal noun). This can contrast with a simple present which suggests an iterative action, much as in English. (32) a. T´as´ıagscr´ıobh leabhar nua. [is she at write-VN book new] ‘She is writing a new book.’ b. Scr´ıobhann s´ı leabhar nua gach bliain. [writes she book new every year] ‘She writes a new book every year.’ To express habituality ( ´ OS ´ e 2001: 123), Irish makes use of a particular verb form b´ı (this verb conjugates in the present) which, like t´a, combines with a non-finite verb form (the verbal noun). (33) B´ıonn s´ıagscr´ıobh go luath gach maidin. [is- she at write-VN early every morning] ‘She is always writing early in the morning.’ [...]... tags in Irish is a clear case of a structural parallel to English As might be expected, tags are used in English without any difficulty, both by monolingual English speakers and by those who are still bilingual with Irish as a first language For example, no non-standard usage of tags was found in A Collection of Contact English But see section 4. 4.6.6 for comments on is it? in Irish English 4. 3 Structural... anticipation of an answer (with reverse polarity) The principle is the same in Irish and English, as can be seen from table 4. 7 Embedded questions The word order of a question which is embedded in a sentence is the same as that of the direct question This fact is of relevance to Irish English; see section 4. 4.6.7 4. 4 The grammar of Irish English 161 e a (67) a An raibh s´ s´ sta? [       was he satisfied]... English 1 2 3 4 5 6 7 8 9 10 A Collection of Contact English A Survey of Irish English Usage (included in Hickey 2004a) Dublin English Recordings (used for Hickey 2005) Waterford English Recordings (used for Hickey 2001b) A Corpus of Irish English (included in Hickey 2003a) Tape-Recorded Survey of Hiberno -English Speech – Digital (included in Hickey 2004a) Irish Emigrant Letters (available in the National... language shift to English This issue will be returned to in section 4. 4.1 .4. 2 when considering the expression of the resultative perfective in Irish English The passive in Irish In Irish, there is a semantic equivalent to the passive of English (Depraetere and Reed 2006), but it is not realised by reversing subject and object Instead, the type of structure just discussed, with t´ ‘is’ and a past participle,... times For features not recorded in previous studies 172 The emergence of Irish English of Irish English, such as were to as past habitual (see section 4. 4.1 .4. 3), twice that number was required If a feature occurred less than the required amount, this is stated explicitly below When dealing with literary records of Irish English, particular care is necessary Text type is an important consideration Obviously,... found in literature where vernacular Irish English is portrayed (Taniguchi 1956: 45 –8) Comparatives Phrases used for comparison which correspond to English ‘more X than Y’ have two elements in Irish which both begin with /n-/ as seen in the following example The initial nasal is relevant to a transfer structure found in Irish English; see the discussion in section 4. 4.5.2 (65) T´ s´ n´os airde n´ a dhearth´... speakers who were not native speakers of Irish but who lived in an area which had been Irish- speaking up to 1 64 The emergence of Irish English the early twentieth century Quotations from this set of recordings are indicated by RL (= Ross Lake, the area investigated) 2 A Survey of Irish English Usage This survey covers the areas of morphology and syntax within Irish English and was conducted over a period... sought from the informants 166 The emergence of Irish English In the following, the abbreviation DER refers to data from the Dublin English Recordings The speakers are referred to by gender and age bracket, e.g F40+ is a female informant between 40 and 45 years of age All the informants in these recordings were speakers of vernacular Dublin English 4 Waterford English Recordings This body of data is similar... period which contain parodies of Irish English These begin at the end of the sixteenth century and continue well into the nineteenth century It is, of course, necessary to stress that parodies cannot be taken to reflect accurately the then contemporary speech of the Irish But given that these texts are the only attestations for earlier 4. 4 The grammar of Irish English 167 Irish English, they have been used... characteristics of Irish English is one which has been applied by the present author (Hickey 1999b) and by Markku Filppula (2004b) in an effort to account for the similarities in vernacular grammar between the various regions of Ireland, specifically between the north and the south of the island 162 The emergence of Irish English Table 4. 8 Data sources for the grammatical analysis of Irish English 1 2 3 4 5 6 7 . the early modern period of Irish English and hence available to the Irish speakers involved in language shift to English. This issue will be returned to in section 4. 4.1 .4. 2 when considering the expression. bilingual with Irish as afirst language. For example, no non-standard usage of tags was found in A Collection of Contact English. But see section 4. 4.6.6 for comments on is it? in Irish English. 4. 3 Structural. idiomatic uses in Irish which parallel usages in Irish English (see section 4. 4.3). On.Apart from the literal meaning of ‘location on something’, this preposition is used in Irish to express the