The Cambridge History of English and American Literature in 18 Volumes (1907–21).
Volume II. The End of the Middle Ages.
§ 6. Alleged Celtic Contribution
It has been argued that an additional cause of the differences between Early and Middle Scots is to be found in Celtic. Interaction has been assumed because the Lowlander and Highlander were brought into a closer, though forced, association in a unified Scotland, or because the anti-English policy of the former, threw him back, no matter with what feelings, upon his northern and western neighbours. There are, however, serious objections to the general assumption and to the identification of many of the alleged borrowings from Celtic. In regard to the first, it must be kept in mind (a) that the only possible interaction, literary or otherwise, was with the Gaels of the west and south-west; (b) that the inhabitants of Strathclyde and Galloway were, to a certain extent, Romanised Celts; and (c) that race-antipathies, as shown in The Flyting of Dunbar and Kennedie, were a strong barrier to linguistic give-and-take, especially in grammatical structure and orthography. On the marches there would be borrowing of words, perhaps even breaking down of inflections and phonetic change. There is evidence of such effects in the initial [char]f for quh (hw) of the pronoun, at the Aberdeenshire end of the “Highland line”; but changes of this kind do not affect the literary standard, or every dialect of the spoken language.
The alleged contributions from Celtic are (a) verbal and (b) orthographic, perhaps phonological. The first are admittedly of the slightest, and are being gradually reduced. In the second a contingency is assumed which, as in the case of central French interference, was the least likely to happen. The closest intimacy is necessary before one language, especially that which is dominant, permits modifications of its grammatical and orthographic habit. Our chief authority on Lowland dialects has described some of the salient variations of Middle Scots, “in the form of words, and consequently in their written form,” as “due mostly to Celtic influence.” While it may be admitted that Middle Scots was not “founded upon precisely the same dialectic type as the written language of the early period,” it is by no means clear that buik, moir, glaid, etc. for older northern forms, the loss of t as in direck, or its addition as in witht, the inserted mute l in chalmer (or chaumer, as pronounced), rolkis (rocks) and waltir (water), the t in the past part. as defamet, or in the adverb, as in frawart—that any of these things are the result of the Low-lander’s unconscious affectation of “Ersch” speech. The onus probandi lies with the supporters of this view. At present no evidence has been produced: it will be surprising if it can be produced.