Language, style and metre of the BDL texts

Language, style and structure in the texts as composed

It is difficult to define the original linguistic complexion of any of our medieval Gaelic ballads. Not only are our earliest texts likely to be one or, more probably, several removes from the original versions, but the ballads were not composed according to strict rules governing language and metre as was the case with bardic verse. It is thus hard to say whether certain forms which look 'modern' may or may not have been in a poem as composed, unless these forms are to be excluded on metrical grounds (hardly of much help given the free nature of ballad metre), or for reasons of dialect (e.g. vernacular Scottish Gaelic features found in the BDL texts clearly could not have been an original part of ballads composed in Ireland).

Previous studies of ballad texts make it obvious that the rhymes utilised by a composer could be influenced by vernacular pronunciation (see DF III,  pp. cxv ff.), and this is borne out by certain examples in BDL which preserve Irish dialectal forms evidently reflecting the poet's own speech (e.g. BDL VIII, 14 n; BDL XX, 78 n; BDL IX, 256 n). Given such flexibility, the language of the ballads may be said to be a form of Early Modern Irish which makes concessions to modern developments and to regional peculiarities where and when necessary, but which nevertheless retains a basic relationship to the classical language. It is, in effect, a less arcane variety of the latter.

Several factors probably governed the extent to which a ballad approximated to, or deviated from, the rules, language and style taught in the bardic schools. Chronology is likely to have played its part; as the influence of the schools declined, so the standards of classical language would become less relevant, and modern forms would increasingly find a place. Doubtless also the calibre and training of the composer would be significant, but the convention generally governing the 'authorship' of ballads does not allow us to have a clear picture of the types of poet who composed the ballads; only in one case in BDL can we be reasonably confident of having identified the poet (see BDL XXIII). In composing, the poet may have considered his potential audience, judging whether they might prefer a poem which paid relatively close attention to classical standards, or one which told (or retold) a story with minimum regard for technical accomplishment. While it is now difficult to establish the role of an audience in determining the form of a ballad, there is some evidence to suggest that the language and style of a ballad may have been determined to some extent by its theme. This is apparent from the BDL corpus.

 

Theme and style

Stylistically, the ballads in BDL may be divided into three broad groups:

(a) Ballads which have a register close to that of bardic panegyric, and are generally correct in syllabification, sometimes with a 'stricter' approach to metre. The three narrative elegies, BDL XXIII, 'The Death of Conlaoch', BDL XIII, 'The Death of Diarmaid', and BDL XXVII, 'The Death of Fraoch', come clearly within this category. Of these, the 'strictest' is BDL XXIII, which evidently started out as an apologue in a bardic poem; the rhyme-scheme here includes two internal rhymes in the second couplet of most quatrains (in deibhidhe), with a high proportion of lines showing alliteration, although not with the strictness of full dán díreach. The fact that this apologue was accepted as a ballad suggests a well defined and popular taste for verses of this sort. The composer of BDL XIII uses his metre much more freely, but sometimes provides one internal rhyme in the second couplet (strict rannaigheacht mhór demanding at least two internal rhymes in each couplet) (Knott, Irish Syllabic Poetry, pp. 13 ff.). Alliteration is often present, but not to anything like the extent required by dán díreach. The metre of BDL XXVII is freer still, resting content with aicill for the most part, and occasional alliteration. Nevertheless, BDL XXVII preserves a very close structural similarity to BDL XIII, both poems first narrating the circumstances of the hero's death, and then describing his qualities in the rhetoric of the panegyric code (MacInnes, 'Panegyric Code', TGSI, L (1976-78), pp. 436-498). The Scottish origin of BDL XIII and the Irish origin of BDL XXVII would suggest that their composers were following a widely established pattern for ballads on elegiac themes.

Not all the ballads on elegiac topics in BDL conform to these principles, however. BDL XXII, 'The Death of Oscar', is a rather diffuse and shapeless composition, lacking the concise and evocative phraseology of the poems referred to above, as well as their technical artistry. It seems likely that it employed six-line quatrains of deibhidhe from the beginning. BDL XXVI, on the death of Fionn, resembles a catalogue of the type discussed in the next section, although it employs the rhetoric of panegyric.

(b) Ballads which are essentially catalogues of warrior attributes, names, or weapons. When the catalogues are of warrior attributes (as in BDL XII, XXI and XXVI), they tend to have lines of six-syllables or less, giving an impression of swift movement, and it may be significant that BDL XXI appears to be related to a 'run' in the prose tale Cath Finntrágha. In BDL XII, where the metre is trian rannaigheachta móire (41414141), hypermetric lines are apparent, and modern, non-classical forms are common. Catalogues of warrior names (and place-names) tend to have seven-syllable lines (as in BDL I, III, IX, XIV, XVI etc.), and they are highly formulaic in their style, giving little oportunity for involved syntax or complex sub-ordination.

Of a rather similar complexion to these catagories are the two Oisean/Patrick debates, BDL X and XI. Formulaic structure is very obvious in BDL X. Both poems have a very high proportion of hypermetric lines which cannot be corrected readily, and both BDL texts attest quatrains in metres other than the predominant metre of the poem. In BDL XI, two of these quatrains (qq. 2 and 112) are integral to the argument, and are likely to have been part of the poem as originally composed. Such metrical confusion may well be a sign of lateness. The register of both poems is noticeably lower and much simpler than that of poems in group (a) above.

(c) Narrative poems which simply relate an incident. These can have a fairly wide stylistic range, but they generally avoid ornate rhetoric of the type found in the elegies in group (a). In terms of plot construction, metrical competence and register, the finest of this group in BDL is BDL VII. More typical of the genre are BDL VI and XX, which have a fight with an invading warrior or host as their theme; BDL V, on the hunt of Slievenamon, relieves the flat inevitability of the narrative with its description of warrior arms. The general impression given by these poems is that correct line length is usually observed, although submetric and hypermetric lines may occur, and rhymes indicative of vernacular pronunciation are in evidence. The least satisfactory narrative poem in BDL, both in plot construction and metre, is BDL XVII, which appears to be incomplete. Here the end-rhymes of its Deibhidhe metre are frequently very poor, and result in an atrophied metrical form which sometimes resembles blank verse.

 

Dating the original compositions

Given the apparent connection between the themes and styles of the BDL ballads, it will be evident that undue weight must not be placed on the style of the poems when attemping to date their original form. It would be tempting to regard the stricter poems in group (a) as representing an earlier phase of verse, with the poorer specimens in groups (b) and (c) coming at the lower end of the chronological graph. The danger of such an assumption is underlined by the known date of BDL XXIII, the strictest poem in the corpus, which is 1490, and it must therefore have been composed within the lifetime of the BDL scribes. The only evidence which could place a poem in group (a) significantly earlier than 1490 is the ascription of BDL XXVII, which, if the identification were with either of the suggested poets, would place the ballad in the second half of the fourteenth century. This evidence, however, is much too shaky to build any theory on it.

One BDL poem in group (b), BDL XXI, has evidently been influenced by the prose tale Cath Finntrágha, while another in group (c), BDL VI, appears to derive directly from the tale. A date of c. 1460-1470 has been suggested for the final form of Cath Finntrágha, and it seems likely that both poems should be dated to this period. Links exist between Cath Finntrágha, BDL VI, XXI and in addition BDL XXII, and it seems possible that here we have a group of poems which were composed as part of a mid-fifteenth century refashioning of the theme of the Last Great Battle, with Cath Finntrágha providing a major stimulus.

There are few external clues for the dating of the remainder of the BDL ballads, with the exception of BDL IX. The language of the majority is no more complex or archaic than that of the Cath Finntrágha group, and it seems unreasonable to press them back beyond 1400. Gerard Murphy was inclined to place BDL VII in the early fifteenth century, noting that in the DF text 'only four or five words occur in the whole nineteen stanzas that might present difficulty to one acquainted with the modern spoken language' (DF, III, p. 153). He also placed BDL IV no earlier than the fifteenth century, and possibly as late as the opening years of the sixteenth (ibid., p. 121); he similarly placed BDL V in the opening years of the sixteenth century (ibid., p. 134). Such conclusions rest on a fair degree of subjectivity, mainly because of the lack of any obvious stratification within the language of the ballads, but there are no adequate grounds for dispute. Taken with the arguments advanced tentatively in the two preceding paragraphs, there may be reason to believe that the BDL scribes were, for the most part, tapping a fairly late phase of ballad composition.

There is, however, at least one text in BDL whose origins go back further than the fifteenth century. This is BDL IX, which has a complicated textual history, showing parts of it belong to the late twelfth century, although other parts of it may be more recent. The language of the narrative (as opposed to the catalogue) section of this poem as preserved in BDL does seem to contain earlier forms than one finds in most BDL ballads; there is, for instance, a markedly greater frequency of synthetic verb forms in this poem as it stands in BDL. Nevertheless, the BDL text does not preserve the linguistic archaisms found in the DF text (e.g. infixed pronouns), which are consistent with the late twelfth century origins of certain parts of the poem (DF, III, pp. 17-18). The BDL text therefore seems to attest a more extensive process of linguistic updating than that in DF, although the latter version has been severely truncated in transmission.

Another possible exception to the general dating of the BDL corpus as probably fifteenth century in origin is BDL XVI, with its Banner Quatrains. One senses that this poem may be a fusion of several thematic and linguistic layers which may be considerably older than the fifteenth century. This is suggested by the opaque nature of the poem as a whole, an opaqueness perhaps indicative of a long period of reshaping and reinterpretation.

Whatever lies behind BDL XVI, BDL IX is an important reminder of the way in which ballad texts draw on early themes, with continual linguistic weathering and structural modification of the ur-version. While most of the BDL texts as we now have them may have originated after 1400, the themes of several others besides BDL IX can be traced as far back as the twelfth century or before, often in prose forms; these texts include BDL XIII, XX, XXII, XXIII, and XXVII.

 

The BDL scribes' handling of metre

The principles used in the edition are such that the condition of the metres of the BDL corpus can be readily appreciated. In general, these metres are well preserved. Where irregularities are apparent, they may be accounted for in three ways: (1) the loose form of metres employed by the ballad tradition; (2) problems in the transmission of the texts earlier than BDL; (3) irregularities introduced by the BDL scribes.

Difficulties in group (1) and (2) are fairly easily distinguishable from (3), since they are significantly harder to correct or, indeed, defy correction (e.g. BDL X, 62 n, and BDL XX, 99 n). In group (3) there are two main types of metrical irregularity: (a) scribal error, involving the loss of words, phrases, or lines (e.g. BDL I, 9 n; XVII, 68 n; XX, 81 n); (b) the loss of unstressed syllables, most commonly in final position, which is characteristic of the scribes' Perthshire dialect. Group (3) (b) predominates in the MS, and examples can be picked out almost at random.

The scribes' overall understanding of metrical units is indicated at times by their punctuation or editorial marks. While they usually present one couplet of verse per MS line, they sometimes show division into lines by means of a slash (/) (e.g. BDL V, XIII); and on occasions when verses consist of six-line units, rather than the normal four-line quatrains, they call attention to these larger units by means of brackets in the margin (e.g. BDL XXII).