Linguistic history of Insular Celtic languages and Proto-Celtic origins

Linguistic History Of the Insular Celtic Family and Proto-Celtic

The Celts were ancient people in Europe who spoke the Celtic languages forming a branch of the European languages including other languages which are unknown but which have been associated with Celtic cultural traits in archaeological evidence. Celtic is used in contemporary times to describe the languages and cultures of Ireland, Scotland, Wales, Cornwall, the Isle of Man and the French region of Brittany however the Celtic language family includes the Gaelic languages of Scottish, Irish and Manx and the Brythonic languages of Welsh, Breton and Cornish.

The Celtic language family is said to be a branch of the larger Indo-European language and it has been hypothesized that the Celtic prototype language may have derived from the Pontic-Caspian steppes. By 600 BC the Celts had split into several groups of languages and had spread across Central Europe as well as Ireland and Britain and the Iberian Peninsula. The Celtic language is believed to have spread to Britain and Iberia during the first half of the first millennium and that this language developed over the centuries into the Celtiberian, Goidelic and Brythonic languages. There is a debate as to whether the Goidelic and Brythonic descend from a common Insular-Celtic language or if instead these languages are reflective of two waves of separate migration.

Figure 1

I. Two Branches of Celtic Language

The work of Fortson (2005) entitled: Indo-European Language and Culture" states that the Celtic languages "...hold a special place in the early history of Indo-European linguistics because they presented the first real challenge to the nascent science."(p.274) the surviving Celtic languages are un the form of unbroken traditional language and are stated to be confined "to a small corner of northwestern Europe -- Irish Gaelic in Ireland, Scottish Gaelic in Scotland, Welsh in Wales, and Breton in Brittany, located in northwest France with the number speaking these languages being one million or less.

Following the conquest of Gaul or ancient France by Julius Caesar and the subjugation of Britain by Claudius a century later "most of this Celtic-speaking territory was assimilated to the Roman world" and the dominant language became that of Latin which eventually resulted in the dying out of Gaulish and the other Continental Celtic languages. (Fortson, 2005, paraphrased)

The other branch of Celtic is that of Insular Celtic which includes all modern Celtic languages is stated to have "continued to flourish in the British Isles" and most particularly in Ireland since the separation of Ireland from Britain by the Irish Channel served to insulate Ireland from the affect of the Romans and from the Anglo-Saxons as well. It is stated that Ireland is "the home of the first vernacular literature written in medieval Europe" or otherwise stated literature written in a language which was not the Church's official language. (Fortson, 2005, p.275) Fortson (2005) writes that the cultural hegemony of the English expanded and "reduced the number of Celtic speakers in Ireland and elsewhere in the British Isles." (p. 275) the same is stated to be true in France since Breton is spoken by British Celt descendents falls under the French cultural dominance.

III. Two Branches of Insular Celtic

Fortson (2005) writes that Insular Celtic is comprised by two subbranches: (1) Goidelic and (2) Brittonic. (p.280) Goidelic is stated to contain "Old Irish and its descents: Irish Gaelic, Scots Gaelic, and Manx." (Fortson, 2005, p.280) Brittonic (Brythonic) is stated to contain "welsh, Breton, Cornish and perhaps the language of the Picts." (Fortson, 2005, p.280) Both Goidelic and Brittonic are referred to quite often as 'Q-Celtic' and 'P-Celtic' due to the "respective treatment of the PIE labiovelars: Goidelic turned the labiovelars into velars, while Brittonic turned them into labials." (Fortson, 2005, p.280) Continental Celtic languages are stated to appear the same on a basic level as other old Indo-European languages while Insular Celtic and specifically Irish and Welsh "look bizarre when compared to languages like Greek and Latin" stated to be due primarily to "a massive set of sound changes that occurred in rather quick succession over the course of a few centuries." (Fortson, 2005, p.280)

IV. Requirements for Proper Subclassification of Languages

The work of Tristram (2007) entitled: "The Celtic Languages in Contact" states that proper subclassification of languages which belong to a single branch of a language family requires knowing "not only which isoglosses they share, but also whether those isoglosses represent shared innovations, rather than archaisms, and also whether they are exclusive" or even extinct language branches. Furthermore "the exclusive shared innovations must not be the result of language contact occurring after the initial separation of the subclassified languages." (Tristram, 2007, p.93)

Tristram writes that the genetic subclassification of the Celtic languages has not fully been determined because of the fact that "Gaulish, Lepontic, and Celtiberian are not attested well enough to clarify their relations to the Insular Celtic (IC) languages as well as the fact "that it is unknown which shared innovations of IC are inherited from a putative common IC protolanguage and which are more likely to be the result of later language contacts." (p.94) Tristram questions whether there ever was a period in Insular Celtic language history in which the languages spoken in sociolinguistic conditions were "favorable to the creation of language areas." (2007, p.94)

Tristram reports that two factors facilitate the areal spread of contact-induced language changes which has been revealed in a comparative analysis and are as follows: (1) widespread bilingualism or multilingualism with regular patterns of exogamy between groups in contact; and (2) the absence of sharp sociolinguistic division between high and low varieties of the language in contact. (2007, p.95)

Tristram writes that the "sociolinguistic situations in which languages are in contact are likely to converge structurally are not necessarily the same as those in which large-scale borrowing of lexical material takes place." (2007, p. 94) it is acknowledged that there are many loan words from British into Goidelic and that this is a mutual loaning of words between the two.

Tristram questions whether any phonological and grammatical features that Goidelic and British share can be determined to have resulted from language contact following the two branches being initially separated and states that answering this question requires establishing whether it "can be demonstrated that the British and Goidelic acquired some common phonological and grammatical features after they had already developed as clearly different languages (or groups of closely related languages)..." (2007, p.95) According to Tristram "this can be done if one can reconstruct a reliable relative chronology of linguistic developments for both British and Goidelic." (2007, p.97) in fact, Tristram holds that if it can be demonstrated that intensive borrowing of structural features occurred between the branches of Insular Celtic "then all common features of British and Goidelic are just as likely to have developed through contact, as they are likely to have been inherited." (2007, p.97)

In the Goidelic language it is stated that the articulation of "stressed short mid-vowels (e and o) is raised to I and u respectively, if there was a high vowel (i or u) in the following syllable, as in Lat. cocina (with a long I pronunciation) > *kogina > *kugina > Olr. Cuicenn "kitchen." This change is stated to be "attested in some Ogam monuments." (p.98) it is stated that the I and u in Goidelic are lowered to e and o respectively, when the following syllable contained the low vowel a (long a). A similar change occurred in British in which only stressed I and us were lowered and the lowering was caused by original long a and by the final -- a in Latin loanwords. This change is not Common Insular Celtic because it postdates raising in Goidelic and raising is not Common Insular Celtic sound change." (Tristram, 2007, p.100)

Tristram writes that in Goidelic "syncope is a completely regular process" which impacts every second syllable of a polysyllabic word, counting the last syllable (following the syncope). And in all likelihood occurring at the last of the Ogam period in the middle of the 6th century. Paraphrased) Stated to be a common morphological innovation "was the creation of conjugated prepositions or preposition nouns from earlier prepositions which were followed by inflected forms of pronouns both in British and Goidelic languages and personal pronouns merged with prepositions into "conjugated prepositions." (2007, p.101) the a preposition governs pronominal dependents it is conjugated for person but the forms of conjugated prepositions are stated to be different in British and Goidelic and this is true "even if the prepositions themselves are etymologically cognate." (Tristram, 2007, p.101)

Tristram holds that conjugated prepositions "must have been created at the time when personal pronouns were still fully inflected in Goidelic and British" which subsequently cause the loss of the inflection in this word-class "in both branches" which in British extended to "all pronouns, nouns and adjectives." (2007, p.102)

V. Common Special Imperfect Tense Shared by British and Goidelic

Tristram (2007) writes that there is a shared special imperfect tense in both British and Goidelic and that no traces of this are found in the Continental Celtic languages to date. There are however, some divergences stated in the "use of the imperfect in the two branches" leaving little doubt "that the parallels in the formation and use of the imperfect in British and Goidelic are accidental.' (p. 102)

However, Tristram states that the majority of the endings of the imperfect are non-related in British and Goidelic etymologically "so the Proto-Insular Celtic imperfect cannot be reconstructed" which is clear when comparison is conducted between the two paradigms of the conditional in Olr. And MW of the PCelt. Verb *kar- "to love" as shown in the following labeled Figure 2.

Figure 2

Olr.

MW

1. sg.

no-carainn carwn

2. sg.

no-cartha carut

3. sg.

no-carad carei

1. pl.

no-caramais carem

2. pl.

no-carthae carewch

3. pl.

no-cartais cerynt

Source: Tristram (2007)

VI. Developments in Insular Celtic Languages

The Insular Celtic languages are stead to have developed "a rather rigid VSO order just at the time when Vulgar Latin tended towards a fixed SVO word order. It is conceivable that the VSO order in Medieval IC is just a compromise between the conflicting tendencies in the development of fixed word order in VL and Early IC." (Tristram, 2007, p.103) Tristram states that the second position of enclitics in sentences is presumed to have been inherited from PIE via Proto-Celtic then the IC sentences containing enclitics (E) could have one of the structures as follows:

V-E (S O)

V-E (O S)

P-E SVO

P-E VSO

P- EOVS

Furthermore, free word order in sentences without any enclitics was still possible and it is possible that verb-initial structures "could have been generalized at this stage, presumably by extension of the V-E SO patterns and the previously existing structures in which the object preceded the subject (P-E SOV) could have been eliminated because they are impossible in VL, which tended to become a rather rigid SVO language at the same time." (Tristram, 2007, p.104)

Fortson (2005) writes that the look of Insular Celtic insofar as its similarity to the old Indo-European was "dramatically altered by a series of changes to the vowels, in particular the loss of most final syllables, the loss of internal vowels and umlaut." (p.281) the majority of these changes occurred in the common Insular Celtic period however some changes were "later parallel and independent developments." (Fortson, 2007, p.281) the inherited IE verbal morphology is stated by Fortson to be "preserved intact in Insular Celtic" however this is true more so in Goidelic than in Brittonic. (2007, p.281)

VII. Proto-Gaelic

The work of Fortson (2007) states that Proto-Goidelic is the prehistoric ancestor of Irish and that this was spoken in Ireland by the Christian era's beginning and perhaps even earlier. The Ogam (Ogham) contained the earliest preservation of the Irish language contained on stone inscriptions written with "strokes and notches chiseled along and across a central line, usually the edge of a stone." (Fortson, 2007, p.282)

Ireland's conversion to Christianity in the fifth century was accompanied by an introduction to the Roman alphabet and the alphabet was taught to Irish clerics by monks in western Britain. During the fifth and six centuries witnessed were "significant alterations to the cultural landscape of the Emerald Isle" as well as changes that were radical in the Irish language and it was "during this time that the Irish changed from looking roughly like Gualish or Latin to looking like Irish." (Fortson, 2007, p.283)

Summary and Conclusion

The examination of the Insular Celtic and Proto-Celtic linguistics has revealed that there are two subgroups of the Insular Celtic language which are those of Goidelic and Brittonic. The Brittonic Insular Celtic includes Welsh, Breton and Cornish and it is believed that it may even contain the language of the Picts. Goidelic and Brittonic are referred to a 'Q-Celtic' and 'P-Celtic' upon the basis of the respective treatment of the PIE labiovelars. This study has also examined the Proto-Goidelic language which is the Irish language's ancestor from prehistoric times and which was spoken in Ireland prior to the advent of the Christian era.