- There are two languages in competition in East Sutherland: Scottish Gaelic and English. Each is, of course, represented by a particular local variety: East Sutherland Gaelic and East Sutherland English. Although Scottish Gaelic in many other varieties will outlive East Sutherland Gaelic, I will speak here of a language death situation, since East Sutherland Gaelic is in competition not with another variety of Gaelic but with a totally different language, English.
- I coined the term semi-speaker in an article on “Grammatical Change in a Dying Dialect,” Language, XLIX (1973), 413–438; see p. 417. For a demonstration that community judgments of semi-speaker status have empirical validity in East Sutherland, see my article “The Problem of the Semi-speaker in Language Death,” International Journal of the Sociology of Language, forthcoming.
- A sixteenth speaker was actually included, an “exile speaker” who had been entirely fluent as a teenager but had spoken very little Gaelic in the forty years since she left her home village. Since she falls clearly into neither group, her results are excluded here.
- The full range of replacements in lenition is as follows for East Sutherland Gaelic (V= in the environment of a following vowel, C= in the environment of a following consonant, I=in the environment of a following front vowel, U= in the environment of a following back vowel): L/ph/</f/; L/thV/, L/sV/, L/čhI/, L/šI/>/h/; L/th;C/ø L/čhU/, L/šU/, L/khI/ < /ç/; L/khU/ < /x/ (Embo and Brora also have L/t/, L/kU/ < /x/, and Golspie has L/t/, L/kU/ < /γ/); L/f/ < ø; /m/ and /p/ in the environments L/mw/, L/pw/ < ø; /m/ and /p/ in the environments L/my/, L/mI/,L/pł/, L/py/, L/pr/, L/pV/</v/; L/kI/</y/; /k/ and /kh/ in the environments L/ky/, L/khy/<ø; L/sn/>/r∼n/.
- It should be stated at the outset that all five are commonplace and of high frequency in the normal everyday speech of any fluent speaker of ESG.
- The masculine and feminine forms of the definite article are theoretically different, the former always including a final nasal. But in ESG the final nasal is very frequently dropped in the masculine definite article, so that the article takes the form [Ɣ] for both genders. Lenition follows only in the case of a feminine, however.
- The nominative and accusative are not distinguished.
- The total number of opportunities represents the number of instances of the environment (e.g., 13 for the past tense) times the number of speakers in the category (e.g., 8 fluent speakers). Occasionally, a speaker gave more than one response, either two variants of the same item or a second lexical choice in which the lenition did or did not appear. Both responses are included in such cases. Semi-speakers not infrequently used alternative constructions or word choices, less appropriate or altogether inappropriate, as translations and thus avoided producing the item under investigation. Percentages represent failure in actual instances supplied. Thus, among semi-speakers 91 past tenses with potential lenition were called for, but only 85 were given. Of these, 9 (or 10.5%) failed to show lenition.
- A total of 5 different semi-speakers showed switching, either in the adverbial context or in the past-tense context. Switching which could not have been self-correction was displayed by 4 of them.
- For this reason only the first occurrence of this verb was counted for the purposes of tables 1 and 2 (except for the semi-speaker I. H., who switched from an unlenited to a lenited form of the verb on the second occurrence and all subsequent occurrences; the first two occurrences were counted for her). Excluding these sequential recurrences from the tables seemed preferable to the risk of exaggerating the stability of past-tense lenition among semi-speakers in the tabulated results.
- Meanings: /priš/ ‘break’, /a/ ‘he’, /ə/ ‘def. art. fem.’, ‘stick’.
- Pronouns have only one form, whether subject or object.
- It is nowadays extremely unusual for an English speaker to try to learn the local Gaelic; in earlier generations, though, when a sprinkling of fisherfolk from other parts of Scotland moved or married into the thriving local fishing population, it was not uncommon.
- Meanings: /ha/ ‘is’, /a/ ‘it’, /kle:/ ‘very’, /płɔ:/ ‘warm’; ‘It's very warm'. Note that /kle:/ (glé) requires lenition of /plɔ:/ to /vłɔ:/.
- His failure is not so extreme an aberration as this would suggest. Table 4 in Dorian, “Grammatical Change,” p. 429, shows three instances (two eventually self-corrected) of the same phenomenon. The youngest fluent speaker in the present study appears in that table as E19; he is on that occasion not one of the speakers who failed to lenite the feminine after the definite article.
- Dorian, “Grammatical Change” (see n. 2 above).
- There is some support for this possibility in a few cases from semi-speakers where gender-marked diminutives were attached to the nouns in question. Thus J. M. used a masculine-marking mutation for the noun /po:/ ‘cow’, improbable as that seems semantically; shen then confirmed that it is masculine for her by adding the masculine diminutive suffix /-an/ (similarly with one of her other lenition failures). On the other hand, she is capable of mixing signals: she used a masculine-marking mutation for /kłn'/ ‘glass’ and then added the feminine diminutive suffix /-ag/.
- See n. 17 above.
- There is already a hint of this latter in the existence of diminutive-affectionate bynames in which a few male by-names anomalously end in /-ag/ (e.g., /še:mag/, /čimag/, /čɔnag/ “Jamesy’, ‘Jimmy’, ‘Johnny’), and, much more rarely, a female name may end anomalously in /-an/ (e.g., /magan/ ‘Maggie’). See Nancy C. Dorian, ‘By-names in East Sutherland,” Folia Linguistica, III (1970), 128.
- The twelfth case was the one instance of the expected development: presence of /-u/ and failure of lenition.
- This is borne out by their failure to produce the dependent conditional as well in the tests.
- The verb throw has initial /čh/ in Embo, but either /čh/ or /th/ in Golspie and Brora. Both lenite to /h/ before a front vowel. Thus, of the 14 instances of /hilig/ ‘threw’, at least 8 (those provided by the four Embo semi-speakers) represent the lenition of /čh/; undoubtedly, some of the others do also.
- The terms are Oftedal's. See Magne Oftedal, “A Morphemic Evaluation of the Celtic Initial Mutations,” Lochlann, II (1962), 93–102; Norsk Tidsskrift for Sprogviden-skap, Supplementary Volume VI, Oslo, Universitetsforlaget.
- It is not the case that the ESG semi-speakers are merely adults who have brought their imperfect mutational system with them from an imperfect childhood bilingualism. Most of them were very much better Gaelic speakers than they are now—some even very nearly monolingual in Gaelic—up until entering school.
WORD
Volume 28, 1977 - Issue 1-2: Celtic Linguistics--1976