This were a wikked way, but whoso hadde a gyde
That wolde folwen vs eche a fote.
There is no such complete guide, for the first part of Morsbach's Mittelenglische Grammatik, Halle 1896, remains a splendid fragment, and Luick's Historische Grammatik der englischen Sprache, Leipzig 1914-, which promises a full account of the early periods, is still far from completion. Happily two distinguished scholars—Dr. Henry Bradley in The Making of English and his chapter in The Cambridge History of English Literature, vol. i, Dr. O. Jespersen in Growth and Structure of the English Language—have given brief surveys of the whole early period which are at once elementary and authoritative. But for the details the student must rely on a mass of dissertations and articles of very unequal quality, supplemented by introductions to single texts, and, above all, by his own first-hand observations made on the texts themselves.
Some preliminary considerations will be helpful, though perhaps not altogether reassuring:
(i) A great part of the evidence necessary to a thorough knowledge of spoken Middle English has not come down to us, a considerable part remains unprinted, and the printed materials are so extensive and scattered that it is easy to overlook points of detail. For instance, it might be assumed from rimes in Gawayne, Pearl, and the Shropshire poet Myrc, that the falling together of OE. -ang-, -ung-, which is witnessed in NE. among (OE. gemang), -monger (OE. mangere), was specifically West Midland, if the occurrence of examples in Yorkshire (XVII 397-400) escaped notice. It follows that, unless a word or form is so common as to make the risk of error negligible, positive evidence—the certainty that it occurs in a given period or district—is immeasurably more important than negative evidence—the belief that it never did occur, or even the certainty that it is not recorded, in a period or district. For the same reason, the statement that a word or form is found 'in the early fourteenth century' or 'in Kent' should always be understood positively, and should not be taken to imply that it is unknown 'in the thirteenth century' or 'in Essex', as to which evidence may or may not exist.
(ii) It is necessary to clear the mind of the impression, derived from stereotyped written languages, that homogeneity and stability are natural states. Middle English texts represent a spoken language of many local varieties, all developing rapidly. So every linguistic fact should be thought of in terms of time, place, and circumstance, not because absolute precision in these points is attainable, but because the attempt to attain it helps to distinguish accurate knowledge from conclusions which are not free from doubt.
If the word or form under investigation can be proved to belong to the author's original composition, exactness is often possible. In the present book, we know nearly enough the date of composition of extracts I, III, VIII, X, XI a, XII, XIII, XIV; the place of composition of I, III, X, XI a, XII, XIII, XVI, XVII (see map).
But if, as commonly happens, a form cannot be proved to have stood in the original, endless difficulties arise. It will be necessary first to determine the date of the MS. copy. This is exactly known for The Bruce, and there are few Middle English MSS. which the palaeographer cannot date absolutely within a half-century, and probably within a generation. The place where the MS. copy was written is known nearly enough for IV b, c, XII, XIV e, XV b, c (possibly Leominster), XVI, XVII; and ME. studies have still much to gain from a thorough inquiry into the provenance of MSS. Yet, when the extant copy is placed and dated, it remains to ask to what extent this MS. reproduces some lost intermediary of different date and provenance; how many such intermediaries there were between the author's original and our MS.; what each has contributed to the form of the surviving copy—questions usually unanswerable, the consideration of which will show the exceptional linguistic value of the Ayenbyte, where we have the author's own transcript exactly dated and localized, so that every word and form is good evidence.
Failing such ideal conditions, it becomes necessary to limit doubt by segregating for special investigation the elements that belong to the original composition. Hence the importance of rimes, alliteration, and rhythm, which a copyist or reciter is least likely to alter without leaving a trace of his activities.
§ 2. DIALECTS. At present any marked variation from the practice of educated English speakers might, if it were common to a considerable number of persons, be described as dialectal. But as there was no such recognized standard in the fourteenth century, it is most convenient to consider as dialectal any linguistic feature which had a currency in some English-speaking districts but not in all. For example, þat as a relative is found everywhere in the fourteenth century and is not dialectal; þire 'these' is recorded only in Northern districts, and so is dialectal. Again, ǭ represents OE. ā in the South and Midlands, while the North retains ā (§ 7 b i): since neither ǭ nor ā is general, both may be called dialectal.