Comparative linguistics in seventh-century Ireland: De origine scoticae linguae

ABSTRACT

De origine scoticae linguae (DOSL, also known as ‘O’Mulconry’s Glossary’) is an etymological glossary dating from around the late-seventh or early-eighth century. It discusses the origins of about 884 Irish words, very often deriving them from Latin, Greek or Hebrew. As such it represents the earliest etymological study of any European vernacular language. Despite this, however, the text has to date been almost completely ignored for its significance in the history of linguistics. This article analyses the authors’ methods, particularly with regard to the semantic and formal components of etymologies, and argues that the text shows considerable coherence, both internally and in relation to its sources and models in the Graeco-Roman linguistic tradition. It argues that DOSL is a serious work of scholarship that represents a milestone in the historical development of comparative linguistics.

KEYWORDS:

• Etymology
• comparative linguistics
• Irish
• Latin
• Greek
• Hebrew
• glossaries
• Auraicept

Acknowledgments

In memory of Anders Ahlqvist, whose support and friendly advice were always appreciated. Thanks to the anonymous readers for many helpful suggestions.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ For example Robins (1997, 85); Law (2003, 192); and several encyclopaedia articles.

² The earliest part is edited in Ahlqvist (1983) (see review by Roe 1988); the core and commentary together in Calder (1917). For discussion, see Acken (2008; Acken 2013); Ahlqvist (1989, 2016); Burnyeat (2007); Hayden (2011, 2012, 2013, 2014); Hofman (2013); Poppe (1995–1997, 1996, 1999, 2002); Thurneysen (1928).

³ This representation of ogam is moot, given that Old Irish (including the Auraicept itself) was in fact written using Latin script (see Poppe 2018; Moran 2018).

⁴ Over 100 entries are three words long: e.g. OM 92 Áu ab aure (Áu ‘ear’ from [Lat.] auris ‘ear’). Entries with more than 90 words are OM 213, 288, 403, 537. In addition, OM 830–830k is a longer, complex disputation on the word neimed ‘privileged, sacred’ and its many applications. I retain the siglum ‘OM’ (for ‘O’Mulconry[’s Glossary]’) in reference numbers since this is well-established in the scholarship.

⁵ The core of the text was afterwards supplemented with a second stratum of entries (about a quarter of the whole), probably sometime in late-ninth or early-tenth century.

⁶ The texts that represent milestones in the history of linguistic enquiry into other vernaculars tend to focus primarily on grammatical description. Since these were invariably themselves based on Latin models, some elements of comparavitism can be detected, thought this is more often than not implicit. See e.g. Aelfric’s Latin–Old English glossary (c. ad 1000; ed. Zupitza 1880); the Icelandic ‘First Grammatical Treatise’ (mid-twelfth century; ed. Benediktsson 1972; Haugen 1972); French and Occitan grammars appearing from the thirteenth century, Welsh in the fourteenth. For surveys of this material see Law (2003, 190–209); Hayden (2017).

⁷ Ambrose, Cassian, Augustine, and Cicero have a much thinner presence. Virgil probably refers to the commentary by the grammarian Servius. For discussion on all of these sources, see Moran (2019, 43–50).

⁸ See Amsler (1989, 136–147). Since DOSL is an anonymous text, I use the term ‘authors’ here on the assumption that it was a collaborative effort.

⁹ Online edition in Bauer, Hofman, and Moran (2017).

¹⁰ I adopt here the terminology used in Buchi (2015). In Amsler’s categorisation (1989, 23), he uses the term compositio for internal creations (following Klinck [1970, 45–70]), but does not have a distinct category for borrowings.

¹¹ Text and translations from Moran (2019), where further commentary can also be found on each entry. Non-Irish words in these citations are marked in italics in order to signal where code-switching occurs. For convenient cross-reference, I cite Irish headwords in the translations according to their form in the Electronic Dictionary of the Irish Language (eDIL, www.dil.ie).

¹² For another interpretation of this headword, see Moran (2019, 344).

¹³ Or, cin a úr could be interpreted to mean ‘without its soil’ (eDIL 1 úir).

¹⁴ See Moran (2019, 40–43), for further details on the patterns discussed here. The prefixes é- and es(s)- both derive from the same word (*eks-), but are treated here synchronically.

¹⁵ The text in brackets seems to be dittography.

¹⁶ The meaning of this headword is uncertain (see eDIL s.v. dígnais, and cf. 2 dognas).

¹⁷ The first element in this etymological compound could also be interpreted as etar ‘between’ .

¹⁸ Or perhaps: Etsad ‘treasury’, i.e. cattle-sufficiency [< éit ‘cattle’ + soad ‘sufficiency’]. (See eDIL s.v. 1 sód, soad.)

¹⁹ Greek words in the glossary are cited in Latin script in forms that reflect late Greek pronunciation: see Moran (2011a) for details.

²⁰ Hebrew does not feature in etymologies of this type. This presumably reflects the absence of Hebrew from the comparisons of Latin and Greek found in Late Antique grammatical discourse.

²¹ For other early attestations of these technical terms, see eDIL, s.vv. comṡuidigthe, comṡuidigud, con-suidigedar.

²² Further information on other early Irish glossaries can be found at Russell, Arbuthnot, and Moran (2009).

²³ This table contains some minor corrections to that printed in Moran (2019, 38).

²⁴ The text predates the development of the grapheme v as a marker for consonantal u in Latin.

²⁵ Other evidence in the text confirms that this was pronounced as /v/, rather than ancient /w/ (Moran [2019, 81]).

²⁶ Diachronically in Irish, initial /f/ is also the outcome of earlier /w/ in inherited words and early Latin borrowings (e.g. OIr. fín < Lat. uīnum): see detailed discussion in Russell (2012), who discounts the possibility that the authors of glossary etymologies understood that this was a genuine historical sound change.

²⁷ It is notable that words beginning m- and n- are nowhere derived from words in b- and d-. If associations were based on sounds alone, the former could well have been interpreted as nasalised versions of the latter. In this case, however, there is no orthographical ambiguity to exploit: nasalised b- and d- are written mb- and nd-.

²⁸ See discussion e.g. in Baumgarten (1983, 225, 1986–1987, 2, 4); Russell (2008, 2–3, 6).

²⁹ www.oed.com/view/Entry/153742.

³⁰ The author here adopts the perspective of a modern philologist, perhaps with some irony, since the central argument of his book is that Plato’s etymologies should be taken seriously.

³¹ Baumgarten (1983, 1986–1987, 1990, 2004); see discussion in Russell (2008).

³² Baumgarten (Baumgarten 1986–1987, 2, 1990, 115).

³³ Baumgarten (1990, 115, repeated at 2004, 58, 38).

³⁴ Drawing on the etymological analyses in Matasović (2009); and Vendryes, Bachellery, and Lambert (1959–1996). None of the lists of modern etymologies presented here are exhaustive. It should also be borne in mind that etymologies for many headwords in the glossary are still unknown or uncertain.

³⁵ Keil et al. (1855–1880), ii 27.9–13: U quoque multis Italiae populis in usu non erat, sed e contrario o, unde Romanorum quoque vetustissimi in multis dictionibus loco eius o posuisse inveniuntur … ‘polchrum’ pro ‘pulchrum’, ‘colpam’ pro ‘culpam’ dicentes et ‘Hercolem’ pro ‘Herculem’.

³⁶ Perhaps coluinn here might be interpreted in its basic sense ‘body, flesh’, if the author had some corporeal sin in mind (see eDIL s.v. colainn).

Additional information

Notes on contributors

Pádraic Moran

Pádraic Moran is a Lecturer in Classics at the National University of Ireland, Galway. His research interests include ancient and early medieval grammatical and rhetorical traditions; glosses, glossaries, commentaries and scholia; and knowledge of Greek and Hebrew in the early medieval West.