The Glasgow Review Issue 4
COMET and The House Among the Stars: Scottish Texts via the Internet
This paper aims to introduce the small but growing list of Scottish texts — at the moment, primarily dramatic texts — which are freely accessible for research purposes via the World Wide Web on the Internet. Since late 1994, Glasgow University’s COMET project has started to gather texts — principally, but not exclusively, Scottish texts — and make them available world-wide electronically, with the permission of the authors or their estate 1. The present paper focuses on one available text, the, otherwise unpublished, translation by Bill Findlay and Martin Bowman of French Canadian playwright Michel Tremblay’s play The House Among the Stars (La Maison Suspendue) 2. I hope to illustrate some of the possibilities of exploiting the COMET electronic ‘list’ by considering the representation of different varieties of Scots in this translation.
OWERSETTINS INTIL SCOTS
There is, of course, a long tradition of translations and adaptations into Scots of literature in a variety of languages. The tradition goes at least as far back as the sixteenth century, and Gavin Douglas’s powerful version of The Aeneid, and it might be argued that the motivation for translating prestige literature into Scots has remained constant for half a millenium. Douglas complains of the ‘penuryte’ of Scots in comparison to Latin, the ‘maste perfite langage fyne’, but by translating Vergil’s masterpiece into Scots, he demonstrates that the base vernacular is indeed capable of rising to the demands of great literature. This point has had to be repeated again and again, given the subsequent loss of status of the lowland tongue as Scotland became first a member of the United Kingdom, and then a ‘stateless nation’ (McCrone, 1992). In the twentieth century, eminent translations include Hugh MacDiarmid’s embeddings of the lyrics of Blok and others in A Drunk Man Looks at the Thistle, William Lorimer’s translation of The New Testament in Scots, William Neill’s Tales frae the Odyssey o Homer, Liz Lochhead’s Tartuffe, Edwin Morgan’s Cyrano de Bergerac, and even Matthew Fitt’s free rendition of Led Zeppelin’s Stairway Tae Heaven. The varieties of Scots used in these translations vary -- from the literary medium, Lallans, in MacDiarmid and Neill, to the urban demotic of Lochead, Morgan and Fitt -- but each work demonstrates that the particular variety of Scots used is capable of being the vehicle for serious literature.
Bill Findlay and Martin Bowman’s several translations of Michel Tremblay’s plays make an honourable addition to this long tradition. Their source texts differ from Gavin Douglas’s insofar as Tremblay’s plays are written mainly in a variety of Québécois French, Joual (literally, ‘horse language’), which is itself far from being considered a ‘perfite langage fyne’. Indeed, one of the attractions of Tremblay to his translators was the perceived fit between the low-status Joual as a dramatic medium, and the lowly status of modern Scots:
...in the final analysis the key attraction was our confidence that Scots would prove an especially effective translation medium, allowing us to get closer in letter and spirit to Tremblay’s Montreal Québécois than would prove possible using English.
(Boaman and Findlay, 1994: p.66)
The parallels that they note between Scots and Joual seem compelling:
Similar to the position in Scotland with Scottish-accented English, North American-accented standard French is spoken in Quebec; but, and again similar to Scotland, varieties of urban and rural vernacular speech co-exist with standard speech and make a significant contribution to Quebec’s literature and culture.
As we shall see, the presence in Scotland of a literary Scots, ‘Lallans’ complicates the picture more than this quotation would allow. Nevertheless, The House Among the Stars is a fascinating text to consider from a linguistic point of view because it attempts to capture the variety of Québécois speech by making use of a range of Scots speech styles. The play tells the story of three generations of a French Canadian family, and it is set (simultaneously) in 1910, 1950 and 1990. Scenes from these three time periods are interwoven, and each develops and resonates with the others as the play progresses. The setting is constant: all the events take place in the rural home of the oldest generation, which is then used as a summer house by the succeeding generations, who are urban working class, and then middle-class academics, respectively. The three generations are distinguished by their speech variety, which means that the language of the characters is effectively a history of their family as it moves from country to city, and from peasantry, to working-class, to middle-class. Indeed, a major attraction of this play to the translators was the opportunity it afforded to demonstrate the versatility of Scots:
The play was a gift for us as it allowed for three distinct idioms: a rural Scots for 1910, an urban variety as in The Guid Sisters for 1950; and a Scottish English for the 1990 characters. With family roots in Banffshire and Angus, the translators welcomed the opportunity to try and suggest the sound of the Northeast in the rural characters’ voices. Curiously, The House Among the Stars has had more success in Scotland than Quebec, having had two professional productions within a year (the Traverse in October 1992; Perth in August 1993) whereas in Quebec it has not been produced professionally since its créationin 1990.
We shall shortly consider in more detail how the three varieties of Scots are ‘suggested’ in the Scottish version. Before doing so, something should be said about the text available via COMET.
THE HOUSE AMONG THE STARS ON COMET
One problem for scholars of Scottish texts, particularly scholars outside Britain, is a lack of availability of texts, particularly contemporary texts. In Tremblay, for example, we have a playwright whom the Herald newspaper has described as ‘the most regularly staged playwright in Scottish theatre’, with five plays staged between 1989 and 1994, sometimes more than once -- and whom the Guardian newspaper has described as ‘the best playwright Scotland never had’ (ibid., p.62). Yet only one of the Scots translations has been widely published -- an edition of The Guid Sisters (London: Nick Hern Books, 1991). There have been other editions of The Guid Sisters (Exile Editions 1988) and Forever Yours, Marie-Lou (LadderMan Playscripts, 1994) — but neither of these is easy to obtain. One of the prime motivations of the COMET project is to make such texts accessible by as wide a readership as possible. The possibility of disseminating them electronically via the World Wide Web was obviously attractive to us, and fortunately the translators and Michel Tremblay’s British agent welcomed the idea with enthusiasm, despite the fact that the direct financial benefit in electronic dissemination is nil.
After seeing and being extremely impressed by the Perth production of The House Among the Stars, I approached Bill Findlay to ask whether he and his collaborator would ‘donate’ the text to COMET. He sent us a copy of the typescript with pencilled emendations, and this was scanned into machine-readable text, and lightly edited. It is not tagged: the COMET team feels that it is important to get the corpus up and running, so that different researchers can use it in different ways. Accessibility was our first priority. The play has two alternative endings, reflecting an ‘original’ and ‘revised’ Québécois version. The Traverse and Perth Theatre opted for different endings, and both are given in the COMET text.
VARIETIES OF SCOTS
As McClure (1979) observes, the varieties of Scots available for writers to use result in complex stylistic effects. Colloquial varieties such as rural Scots, urban Scots and Scottish English can be used to ‘place’ characters with reference to their geographical location (country versus city) and social class (working-class versus middle-class). These varieties are described in some detail in Aitken and McArthur (1979). To these colloquial varieties must be added the written, literary medium, sometimes known as ‘Lallans’ or ‘synthetic Scots’. Lallans contains elements of Scots from different geographical regions, with the addition of various archaisms and neologisms. As noted above The House Among the Stars embodies a dramatic attempt to use different varieties of Scots to differentiate among three generations of a family, separated by time, regional identity and social class.
The play is of course a play, and we are dealing here with ‘suggestions’ or ‘representations’ of three varieties of Scots, written for dramatic productions in the central belt of Scotland. Kirk (1992/3, p.90) warns of the dangers of too great a focus on literary texts when making claims about Scots speech:
If researchers are to study Scottish English seriously, it seems essential that they should get away from the notion that only literary uses count. No matter how reflective or realistic of speech literary texts may be (cf. Kirk 1988), they are still no substitute for genuine spoken data.
As we shall see, a close examination of The House Among the Stars bears out Kirk’s caution; but even so, literary texts are interesting to many of us in their own right, particularly in a case like this where the use of different varieties of Scots has such a central dramatic function.
ANALYSING THE HOUSE AMONG THE STARS
The analysis given here is fairly rudimentary, and is meant mainly to illustrate the kind of basic investigation that can be performed on other COMET texts. The House Among the Stars was down-loaded from the COMET files onto a floppy disk, and then it was analysed using the freely-available TACT text analysis program. This program allows us to look quickly and easily at orthographic, phonological and lexical variation in the play. First of all, I wanted an alphabetically-ordered list of all the lexical items used in the play. This list highlights obvious lexical correspondences such as aboot/about; aw/all; gress/grass; etc These spellings obviously indicate phonological differences. Moreover, the use of specifically Scots lexical items such as shoogle; thrang; numpty; breenge; etc can easily be identified and collated. Scots grammar is less easily investigated using a word-list, although the presence of grammatical markers (such as the topic-marker see; and various question-markers, such as whitwey and howfurno) can be quickly displayed. In short, the alphabetical display can be used to categorise items according to what we know about the character of different Scots speech varieties.
Finally, we can then use a KWIC (ie Key Word In Context) concordance to look at the categorised items, and, in this play, to see which character is using the item. This information tells us how the various items are distributed amongst the characters, and so amongst the three generations
RESULTS AND INTERPRETATION
Data on pronunciations and lexical choice is given in the Distribution Tables 1-17 (Appendix). Any interpretation based on them obviously derives from the written text of the play: a ‘suggestion’ of different regional and social identities can also be given through the adoption of particular accents by the cast of any particular production. However, the text of the play does give clues to how this differentiation might be accomplished, both by suggesting Scots pronunciations of different items, and by the distribution of different types of Scots vocabulary.
A cursory glance at the tables shows that there is a greater gulf between the representation of Scottish English and the Urban/Rural Scots than between the Urban and Rural Scots varieties. An exception to this general rule is seen in Table 1, where the Scots phoneme /x/ is confined to the 1910 characters. This consonant can be classed as a ‘stereotype’, its presence alone marking the most conservative Scots speakers. The Urban Scots speakers lose the /x/, but retain the Scots vowel /o/ in thoat and broat which correspond to the /ç/ of Scottish English thought and brought. Here the three varieties are most distinct.
Otherwise, most of the pronunciation features found in the 1910 columns of Tables 1-15 are also found in the 1950 column but not in the 1990 column: in Table 8, for example, the Scots /i/ phoneme which derives from OE /E/ is indicated by the <ei> and <ee> spellings in breed, deid, deefnin, meesured in the 1910 column, and by deid, heid, pleesure in the 1950 column. Where the pronunciation ‘bread’ is used by a 1950 character, the item is in quotation marks, indicating that an ‘affected’ accent is used deliberately (a similar occurrence in the opposite direction is evident in Table 6, which shows a 1990 character adopting the Scots pronunciation ‘intae’ for effect).
However, there is evidence of greater variability in the representation of the 1950 characters’ pronunciations. This is particularly evident in Table 11, which indicates a consistency of pronunciation of the back vowel /ç/ before /r/ and /n/ in the 1910 characters. This consistency is not found in the 1950 characters: here pronunciations switch between caur/car, hauns/hand, staun/stand, and understaund/understand. Moreover, where the 1910 characters have the pronunciations daurk and whaur, the 1950 characters have the pronunciations dark and where. This variation in the 1950 characters’ pronunciations might have different interpretations: it might be taken to represent social variation or pretention within the working-class Scots characters; or it might be taken to represent the process of transition from Urban Scots to Scottish English (cf Macafee, 1994). Whatever the interpretation, the data does suggest that one feature which distinguishes the Urban Scots speakers from both the Rural Scots and the Scottish English speakers is the variability of the pronunciations. This variability seems an accurate portrayal of the kind of ‘code mixing’ also found by Macaulay in the speech of working-class Ayrshire speakers (Macaulay, 1991: 262):
What seems to be clear from the interviews is that all the lower-class speakers have a stylistic choice available to them in a number of variables that is not available in the same way to the middle-class speakers.
Macaulay’s point here is that middle-class speakers’ adoption of conservative Scots forms is much more pronounced stylistically than working-class speakers’ adoption of Scottish English forms. Scottish English forms might be adopted by working-class speakers for reasons of mild emphasis. Certainly, the occasional Scots adoptions by the 1990 characters in the play is much more striking than the Scottish English usages of the 1950 characters (see the discussion of vocabulary distribution below).
There is, of course, another possible explanation of the variation in usage: the fallibility of the translators, who might have been attempting to represent a particular variety of Scots consistently. Such inconsistency might be exacerbated by the fact that the COMET text is not based on a manuscript which has been prepared for publication. There are indeed some indications that some of the variability in pronunciation might be due to oversight on the part of the translators. There are a few examples where the 1910 pronunciations are more variable than the 1950 ones: in Table 9, for instance, the 1910 characters have the pronunciations wey/way, where the 1950 characters only have wey and the 1990 characters have way. This variability might have no dramatic function, and it might not represent the actual speech styles of the Rural Scots characters, but as an isolated instance, it might simply be a slip on the part of the authors. Other examples of possible ‘slips’ can be found in Table 7 where the 1910 characters have the pronunciations sound and roond/round to the 1950 characters’ soond/sounded and roond. However, as the tables show, the level of consistency in the pronunciations of the 1910 characters -- and also the 1990 characters -- is much greater than that of the 1950 characters.
In short, then, Tables 1-15 show a marked distinction between the pronunciations of the Scottish English speakers (1990) and those of the Rural Scots and Urban Scots speakers (1910 and 1950). The Rural Scots and Urban Scots speakers are much closer, at least as far as their textual representation goes, although the Urban Scots speakers are shown to have greater variation, which is manifested as a greater tendency to adopt Scottish English pronunciations alongside their Urban Scots pronunciations -- note, for example, in Table 2, the 1950 items work, world and water, alongside more conservative wance, waant and watter. The representation of Urban Scots is also manifested in phonetic spellings; for example, the 1910 instance of besom corresponds to 1950 bizzum; and 1990 nothing and another corresponds to 1950 nuthin and anuther.
The grammatical indicators considered in this investigation are few (Table 17). Macafee (1983: 48) notes the use of see as a topic-marker in Central Urban Scots; and the use of causal question markers such as how or whit wey. The use of see is evident in the speech of the 1910 and 1990 characters (but not, ironically, the 1950 characters), and causal question-markers, whitfur and whitwey are evident in the 1910 characters’ speech. Only the elaborate howfurno is present in the speech of the 1950 characters. On the face of it, these distributions might suggest a commonality of grammatical forms among the generations; but in fact, the situation is more complex, as an examination of the lexical items shows more clearly.
An examination of the lexical distributions of Scots items among the three generations (Table 16) shows a greater degree of distinctiveness, and calls into question the accuracy of the representation of Rural Scots in this play, at least as a ‘suggestion’ of the dialect of Angus and Banffshire. First of all, the density and distribution of lexical items varies among the generations. The 1990 characters have by far the fewest number of Scots items: breenge, dook, feart, numpty, and the inevitable wee. Three of these items (dook, feart and wee) also occur in the speech of the other two generations. As we have seen, the 1990 characters share with the 1910 characters the use of see as a grammatical topic-marker. The use of these Scots items by the 1990 characters is comparable to the use of Scots items by middle-class speakers, either overtly or covertly (cf Aitken, 1979). Some items, such as wee, are covert -- the speaker might not be aware that it is a particulary Scottish word. Others, such as dook, breenge, and feart have more the character of overt Scotticisms. In Aitken’s words (1979: 108):
...some traditional Scots ‘dialect’ words feature as occasional embellishments of middle-class Scottish Standard English speech on appropriate informal and formal occasions (for example, in speechmaking to provide a more or less jocular reminder that the speaker is a good Scot). Knowledge of some of these items is reinforced by reading in the vernacular Scots classics (such as the novels of Scott). Accordingly, some, at any rate, appear to be employed more often by ‘educated’ Scots speakers than by their less erudite working-class fellow-countrymen, and appear to constitute a kind of middle-class folklore of what identifies the true Scot in speech.
The first part of this quotation is particularly pertinent to the dramatic effect of the use of Scots items by the 1990 characters in this play: the adoption of Scots forms is an act of identification with the preceding generations by a class of society which has moved away from its social and geographical roots, and is now engaged in ‘finding itself’. Stylistically, as noted in the earlier discussion of Macaulay, the occasional use of Scots by middle-class speakers is more striking than the use of Scottish English forms by the working-class speakers of the 1950 generation. The second part of the quotation is not quite so pertinent to the speech of the 1990 characters in the play: their Scots forms are colloquial rather than erudite, familiar rather than pompous. However, the point about ‘a kind of middle-class folklore of what identifies the true Scot in speech’ does apply to the speech of the oldest generation.
So far the speech of the 1910 generation has been dubbed ‘Rural Scots’ unproblematically, taking a cue from Findlay and Bowman’s hint that it recreates the Scots of Banffshire and Angus in North-East Scotland. However, the speech of the 1910 characters in fact suggests little of the peculiarities of NE Scots speech. For example, since the 15th Century, Northern Scots has had the phoneme /f/ for <wh-> in such items as whaur, white, and so on. There is no indication of this widespread feature in the speech of the 1910 characters. Neither is there any indication of the unrounding of the Scots vowel /O/ to give /i/ in Northern Scots, resulting in meen, skeel, and eese rather than mune, schuil and yaise (cf Table 6). The similarities between the 1910 and 1950 Scots realisations, already noted, suggest that if the 1910 characters do speak in a Rural Scots, it is a Rural Scots of the Central belt, rather than a Northern variety. Despite Findlay and Bowman’s family roots in the North-East, and given the fact that both productions of the play were in Central Scotland, this representation of conservative Central Scots pronunciations might well have been pragmatically driven — audiences from the Central belt might have struggled more with Northern forms. Indeed, Bowman and Findlay note that audience concerns led to the modification of the dialect used in The Guid Sisters from Findlay’s West Fife to West Central Scots(1994: 67).
However, in some respects at least, the speech of the oldest generation in The House Among the Stars is less rural and more literary in character. We must be cautious about such characterisations, because, as McClure (1979) observes, what seems archaic and literary to the present generation might well have been current in the Scots speech of earlier generations. Moreover, the Concise Scots Dictionary suggests that, of the Scots items used by the 1910 characters, only chiel (a young lad) might be termed ‘literary’ in character. Even so, the speech of the 1910 characters in The House Among the Stars shares with other examples of literary Scots the collocation of Scots items from different regions of Scotland which are not now in general use throughout Scotland: for example, swee and sharn from the North-East and Shetland, fegs from Ulster Scots, glisk from Fife. The use of the exclamation fegs! is more likely to be familiar to ‘educated Scots’, in Aitken’s terms, from its presence in well-known Lallans poems, such as A Drunk Man Looks at the Thistle, than from everyday usage. Although much of the Scots used by the 1910 characters is general, current Scots, the synthesis of items from different regions is reminiscent of a literary variety rather than an ‘actual’ local variety. This observation is not meant as a criticism of this aspect of the language of the play: Bowman and Findlay are dramatists, not dialectologists. Their purposes are better served by appealing to ‘a kind of folklore’ about the Rural Scots speech of a previous generation, rather than by attempting a pedantic recreation of a particular variety -- their own expression, ‘a period ‘wash’’ (1994: 68) indicates that the strategy is deliberate.
This paper has attempted to show the value of having a text such as The House Among the Stars available for study in machine-readable form via the Internet. By using a simple text-analysis program such as TACT, it is relatively easy to investigate the ways in which varieties of Scots are used in a modern play to represent the social and regional identities of different characters, how these might develop over time, and the various tensions and correspondences between them. Moreover, as the discussion above has shown, this representation is not a matter of merely transcribing or recreating the ‘authentic’ dialects of different regions and times; rather it appeals to present-day ‘folklore’ of what constitutes older Scots speech, and attempts to convey an impression of the past. Again we must remember Kirk’s caution that the use of Scots in literature should not be taken to correspond to everyday Scots speech (Kirk, 1992/3: 90), although this caution does not detract from the achievement of the play.
The House Among the Stars is, of course, ideal for this kind of analysis. However, the use of Scots and English in th give the fledgling COMET project a distinctive identity. As time goes on, we hope that the usefulness of COMET will become more and more apparent to an increasing number of users.
Aitken, A.J. (1979) ‘Scottish Speech: a historical view, with special reference to the Standard English of Scotland’ in Aitken and McArthur, eds. pp. 85-118
Aitken, A.J. and T. McArthur, eds. (1979) Languages of Scotland Edinburgh: Chambers
Bowman, M. and Bill Findlay. (1994) ‘Québécois into Scots: Translating Michel Tremblay’ in Scottish Language No. 13, pp. 61-81
Kirk, J.M. (1992/3) ‘Computing and Research into Scots’ in Scottish Language No. 12, pp. 75-131
Macafee, C. (1983) Glasgow Amsterdam/Philadelphia: John Benjamins
----- (1994) Traditional Dialect in the Modern World: A Glasgow Case Study Peter Lang: Frankfurt am Main
Macaulay, R.K.S. (1991) Locating Dialect in Discourse Oxford: OUP
McArthur, T (1992) The Oxford Companion to the English Language Oxford: OUP
McClure, J.D. (1979) ‘Scots: Its Range of Uses’ in Aitken and McArthur, eds. pp. 26-48
McCrone, D. (1992)Understanding Scotland: The Sociology of a Stateless Nation Edinburgh: EUP
Robinson, M., ed. (1985) The Concise Scots Dictionary Aberdeen: AUP
The tables below illustrate the distribution of certain features of Scots/English pronunciations and lexical items across the three generations of the family whose history is told in The House Among the Stars. The tables are not exhaustive; neither do they indicate the frequency with which a particular item appears in the speech of any group of characters. They simply illustrate the main differences in pronunciation and vocabulary among the generations: 1910 (Rural Scots); 1950 (Urban Scots); 1990 (Scottish English).
Table 1: /x/ is present in Scots pronunciations
Table 2: In Scots, OSc /a/ remains after /w/
Table 3: In Scots, /T/ becomes /h/
also In Scots, /h/ precedes /I/ in <it> and <is(n’t)>
Table 4: In Scots, OSc /o/ or /ç/ become /a/ in labial environments
Elsewhere, OSc /o/ or /ç/ beocome Modern Scots /o/
Table 5: Early /a/ becomes /e/ or /E/ in Modern Scots, especially before /r/
Table 6: Old English /o/ becomes Scots /O/
Scots /O/ then becomes unrounded /e/ in long environments
Scots /O/ then becomes /I/ in short environments
Table 7: OE /u:/ remains in Scots, with loss of length
Table 8: OE /e/ becomes Scots /i/
Table 9: OSc /ai/ becomes Scots /Ãi/
|stay||stey||steyed (stay *)||stay|
* - One example of ‘stay’ is spoken by a child, who is simultaneously representing two characters in two ‘time zones’, 1910 and 1950.
Back to Table
Table 10: OE /a:/ becomes /e/ in Scots
Table 11: In Scots, the back vowel /ç/ is found before /r/ and /n/
Table 12: In Scots, /v/ is vocalised, medially and finally
Table 13: OE /aNg/ becomes Scots /aN/
Table 14: Final /T/ is deleted finally, and at morpheme boundaries, in a small number of words
Table 15: In Scots, the vowels /I/ and /Ã/ fluctuate
Table 16: Distribution of Scots lexical items
Table 17: Selected Scots Grammatical Indicators
|see (topic marker)||X||---||X|
1 - The COMET project was started in 1994 with New Initiatives funding from the University of Glasgow. The Project Director is Dr Cathy Emmott. I am grateful to Ann Gow, the Research Assistant, and to Jean Anderson, the Project Manager, for their patience and technical assistance during the writing of this paper.
Back to Text
2 - I am grateful to Bill Findlay, Martin Bowman, and Michel Tremblay's UK agent for permission to make The House Among the Stars available on COMET. Information about performance rights can be found on the title page of the COMET text.
Back to Text