Syllable and diphthong classification in the medieval Welsh bardic grammars

ABSTRACT

The medieval Welsh bardic grammars, known as ‘Gramadegau’r Penceirddiaid,’ include an extensive system of classification to describe syllable and diphthong types. While much of the rest of the linguistic description in the bardic grammars is heavily Latinate, this section is apparently innovative and oriented towards the demands of bardic composition. The syllables and diphthongs section is extensively revised over the course of its transmission, and either expanded or contracted depending on the aims and purposes of its editors. This article examines the two earliest revisions, found in Peniarth MS 20 (c.1330) and Bangor MS 1 (mid-fifteenth century) as evidence of the changing function of the grammars over the course of the fourteenth and fifteenth centuries. A case is made for the increasing use of the grammars as practical pedagogical documents from the mid-fifteenth century.

KEYWORDS:

• Bardic grammars
• Welsh
• gramadegau’r penceirddiaid
• grammar
• medieval
• poetry

Acknowledgments

I am grateful to Paul Russell for reading through a draft of this article, and to the two anonymous Language & History reviewers for their comments on the piece (and especially their assistance with the translation). Of course, remaining errors are my own.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ The complex internal system of rhyme and alliteration in medieval Welsh poetry; for good descriptions in English, see Lewis (2015, 30–45); Rowlands (1976, xxvii–xlix); Williams (1953, 243–247). The standard treatments can be found in Welsh in Roberts (1973); Morris-Jones (1925).

² It has been generally repeated that the grammars were probably first compiled as a response to the Edwardian Conquest of 1282, during which time the concern of the literary classes was the preservation and codification of traditional material (Matonis 2004, 155–156; 1990, 287–288) or perhaps in this case, the re-definition of panegyric poetry after the loss of its traditional patrons (Gruffydd 1995, 20). More recently, however, the suggestion has been made that at least one section – on letters – must have been compiled sometime earlier in the thirteenth century, and indeed it is possible that the entire grammatical portion circulated independently at an earlier period than has previously been allowed (Charles-Edwards 2016, 159–160). In any case, the grammars were certainly complete and in circulation by c.1330, the date of the earliest manuscript witness (Peniarth MS 20), discussed in greater detail below. All manuscript dates taken from Huws (Forthcoming). I am grateful to Daniel Huws for granting me access to this work prior to publication.

³ For the most complete list currently available of the manuscripts in which the grammars are found, see Williams and Jones (1934, xiii–xvi; xlvii–lviii) [hereafter GP] but note that some texts are missing from the list and some details are out of date.

⁴ The validity of these names as attributions is somewhat dubious, but it is a useful way to distinguish what are generally considered to be the ‘first’ and ‘second’ recensions of the bardic grammars. For a summary of the problem of authorship and attribution, see Matonis (1990, 283–86).

⁵ An edition of the text is found in Jones (1923–1925).

⁶ ‘[c]rynodeb o ran bwysig o’r ddysg draddodiadol a gyfrennid yn ysgolion y penceirddiaid’.

⁷ GP 1 (Red Book); corresponding phrase in other manuscripts occurs at GP 19 (Llanstephan MS 3), GP 39 (Peniarth MS 20); GP 89 (Pum Llyfr Kerddwriaeth). The phrase is absent in the Bangor MS 1 copy; this is discussed below.

⁸ Marotta notes that in Keil’s volumes, ‘only the work by Terentius Scaurus and Velius Longus is missing a section De Syllaba’ (2015, 55).

⁹ The Classical grammars list six; the Welsh, for the most part, list seven. The six-letter variant in Bangor MS 1 and Peniarth MS 20 is discussed briefly below.

¹⁰ For an excellent description of this section and discussion of its Latin affinities, see Matonis (1981, 132–136).

¹¹ Throughout this paper, I follow the translations for lleddf and talgrwn proposed by Lewis (2018, 226–227). I am additionally grateful to Barry Lewis for first suggesting to me the need to move away from the tendency to (mis)apply modern linguistic terminology to medieval Welsh grammatical theory. As such, the translation of talgrwn departs from more recent treatments, in which it is called ‘rounded’ (Matonis 1981, 134), ‘round at the end’, ‘end-round’, (Charles-Edwards 2016, 156), ‘rising’ (Russell 2016, 165). Ceri Lewis’ suggestion of ‘compact’ or ‘neat’ (1997, 63) is a closer approximation.

¹² In his discussion of the syllables and diphthongs section, Ceri Lewis has noted that the preservation of this distinction even in multisyllabic words is evidence of the grammars’ tendency to preserve archaic prosody. For his full discussion of syllables and diphthongs in the bardic grammars, see Lewis (1997, 61–65), for trwm ac ysgafn in multisyllabic words, see especially Lewis (1997).

¹³ Tawdd ‘liquefying’ consonants are as follows: <d, f, l, m, n, r, s>. Charles-Edwards has noted that tawdd is a calque on Latin liquidus <l, r, m, n>but the letters included in this category ‘correspond to Donatus’ semiuocales.’ (2016, 150).

¹⁴ Mud ‘mute’ consonants are as follows: <b, c, g, h, k, p, q, t, (ll), (x)>. Charles-Edwards notes that mud is a calque on Latin mutus (2016, 150). Letters in brackets do not occur in this category in the Red Book of Hergest or Peniarth MS 20 versions.

¹⁵ kanys penngamu a wna un o’r bogalyeit tu a’r llall. Text taken from the Red Book of Hergest version, GP 2.

¹⁶ This translation is informed by the discussion in Charles-Edwards (2016).

¹⁷ i.e. it describes vowel quality only: ‘A vowel whose quality changes perceptibly in one direction within a single syllable’ (Matthews 2007, s.v. ‘diphthong’),.

¹⁸ GP 3.

¹⁹ Matonis points out that these all end with rounding (1981, 135).

²⁰ A type of rhyme in Welsh where the final consonant (if there is one) is retained, but the vowel is switched out (e.g. rhan/llon); talgrwn diphthongs can all rhyme with one another (e.g. llaw/glew; ewch/dowch) and lleddf diphthongs can all rhyme with one another (e.g. llaeth/doeth), but the gwib diphthong cannot form proest with any other diphthong. See Morris-Jones (1925, 254–62), from which the examples in this footnote are taken.

²¹ Matonis labels these as ‘ending non-back’ (1981, 135); Ceri Lewis points out that these all ended in the semi-vowel i in Old Welsh, which further suggests the antiquity of the phonetic rules preserved by the grammars (1997, 63–64).

²² Note that this definition does not occur with the rest, in the main section on syllables and diphthongs, but rather in the earlier section on letters. Discussed below.

²³ In the Red Book of Hergest, b, c, g, h, k, p, q, t, x (GP 1); in Llanstephan MS 3, b, c, g, h, k, p, q, t, ll (GP 19).

²⁴ GP 1, Red Book of Hergest. Corresponding section in Llanstephan MS 3 is at GP 19.

²⁵ It is possible that the inclusion of <ll> amongst the mut letters reflects a development or correction in Llanstephan MS 3 and Balliol College MS 353 from the Red Book of Hergest text, comparable to the impulse towards correction in the Peniarth MS 20 text (discussed below).

²⁶ Unless they are C (tawdd) + C (tawdd), or C (mud) + C (tawdd), in which case they are cadarnleddf, as given in Table 1. Note that this is a subcategory of the talgrwni‘abrupt’ syllable, which above is claimed to encompass all syllables with one vowel and any number of consonants.

²⁷ a thywyll datkanyat arnei (GP 41).

²⁸ Balliol College MS 353 contains the same triad. The grammar in the Balliol manuscript has not been edited, but can be accessed online at <http://image.ox.ac.uk/show?collection=balliol&manuscript=ms353>.

²⁹ Text taken here from Peniarth MS 20; corresponding triad can be found in the Red Book of Hergest (GP 16) and Llanstephan MS 3 (GP 36).

³⁰ In (1913, 26), Morris-Jones gives examples of uncontracted forms in the work of Dafydd ap Gwilym, Iolo Goch, and Lewis Glyn Cothi.

³¹ GP 90: Pann vo ssilldaf a’i dechrav yn wib, a’i diwedd yn gadarnleddf, val y mae ffravtr, honno a elwir dipton wibleddf ‘When a syllable begins with straying, and ends with strong tapering, as in ffravtr [friary], that is called straying tapering.’

³² This portion of the text in square brackets in [1] is taken (by Jones 1923–1925) from the Roger Morris copy in Peniarth MS 169. Jones takes the following text [1–13] from Bangor MS 1.

³³ The text in square brackets [13–15] is taken by Jones (1923–1925) from the fragmentary copy in Peniarth MS 191.

³⁴ ‘crynodeb digon amherffaith’.

³⁵ It is perhaps worth noting that the section on lleddf a thalgrwn in the modern poetic handbook Cerdd Dafod does not break lleddf down into three different subcategories: rather, the definition of lleddf and talgrwn syllables is basically simplified into the different diphthong classes, with special attention paid to the sequence <wy> (reminiscent of (5) above). From a practical, modern pedagogical standpoint at least, these subcategories are not useful or necessary. (Morris-Jones 1925, 235).

³⁶ ‘syniad cywiraf am y gramadeg fel yr oedd pan ddaeth allan o law Einion ei hun’.

³⁷ For example, both Peniarth MS 20 and Bangor MS 1 [3] give the maximum number of letters permitted in a syllable as six, ac arwyd uchenait (GP 40) ‘with a sign of aspiration’ (i.e. <h>); all other copies give seven as the limit. Likewise, the definition of hir/byr occurs at the beginning of these Peniarth MS 20 and Bangor MS 1, rather than at the end.

³⁸ I have benefitted from digital access to the Peniarth MS 20 manuscript at the National Library of Wales website <https://www.llgc.org.uk/en/discover/digital-gallery/manuscripts/the-middle-ages/chronicle-of-the-princes>, which has perhaps allowed for a closer glimpse of this piece of marginalia than was available during the preparation of the GP edition in 1934.

³⁹ I am grateful to one of the anonymous Language and History peer reviewers for drawing my attention to this possibility.

⁴⁰ For example, most notably, Lewys Glyn Cothi’s copies of his own poems, collected in Peniarth MSS 70, 109 and found in parts of Peniarth MS 40 and the Red Book of Hergest (Johnston 1995, xxvii).

⁴¹ Dafydd Nanmor and Gwilym Tew; mentioned in Smith (1998).

⁴² Gwilym Tew again, who copied the version of the Dwned found in Peniarth MS 51: see Jones (1980) for more. Dafydd Johnston notes likewise that Ieuan ap Rhydderch of Ceredigion (1390–1470) owned a copy of the bardic grammars (2005, 236).

⁴³ ‘Ni fu llewyrch mawr ar farddoniaeth Gymraeg yn ystod deugain mlynedd cyntaf y bymthegfed ganrif, a gwrthryfel Glyndŵr oedd yn bennaf cyfrifol am hynny.’

⁴⁴ I am indebted to Peredur Lynch, both for raising the possibility of this connection to me, and more generally for many long conversations about the subject of this article.

Additional information

Notes on contributors

Michaela Jacques

Michaela Jacques is a PhD student in the Department of Celtic Languages and Literatures at Harvard University.