PROBLEMS IN STUDY OF ENGLISH LANGUAGE 227 



our study is not at an end. We know very little about any dialect 

 when all we know is its vocalic and consonantal behavior, and, indeed, 

 when we add to that an acquaintance with some of its inflectional 

 habits. With reference to the great Middle English dialects, each 

 of which has an abundant literature and may lay claim to have been 

 at some time a literary language of some pretensions, we need to 

 know its characteristic vocabulary, the special idiomatic phrases 

 which distinguish it, its metrical system, and its syntax. If we are 

 asked how far our present codified knowledge of the dialects in 

 question meets these requirements, we shall have to hang our heads. 

 This is a matter of some concern to the literary historian as well as to 

 the linguist. It has been much the fashion to talk about "schools" 

 or " groups " of Middle English poetry. The terms may be misleading, 

 but we will not pause to quarrel with them. In the present state of 

 our ignorance about some of the things just mentioned, there is 

 constant danger of our confusing what belongs to' a dialect with what 

 belongs to a school. Nor is that all. Different works are not infre- 

 quently ascribed to the same writer on the ground of resemblances in 

 style and language which, if they prove anything, prove only that 

 the documents are written in the same dialect and employ a common 

 stock of catchwords and catchphrases. The abuse of the argument 

 from so-called parallel passages is largely due to our ignorance or 

 neglect of all dialectic phenomena except those of sounds and forms. 

 It is safe to say that arguments for identity of authorship in the 

 Middle English field are every day based on collections of parallel 

 passages of a kind that would call down Homeric laughter on the 

 heads of their accumulators if they were dealing with documents 

 and writers of our own day. Now much of this abuse comes from 

 pure neglect of logic; but by far the larger part of it must be charged 

 to ignorance, excusable and even unavoidable ignorance, perhaps, 

 but still ignorance, pure and simple. The investigator simply does 

 not know that the phrase or sentence or verse that he copies down 

 on his card is not the property of A or B or C, but of everybody who 

 spoke or wrote that dialect, and that, accordingly, it was the in- 

 evitable form of words when that idea (itself a commonplace of 

 experience or reflection) had to be expressed. 



The "vocabulary test" is pretty nearly discredited by this time, 

 so fantastic are the pranks which it has been forced to play in the 

 face of an astonished world. But the "parallel passage test" is still 

 in high favor. Yet we all know, it is to be presumed, that, for some 

 purposes, the unit of expression is not the word, but the set group of 

 words, the phrase or sentence; and that consequently the test 

 from parallel passages is often in no way distinguishable from that 

 from community of vocabulary. 



All this suggests one of the most serious desiderata of our science. 



