The word aller has called forth a great variety of etymological explanations, none of which is completely satisfactory to the linguistic scholar. However, I make bold to offer a fresh contribution to the problem, in the hope that it will fulfill all three requirements of a good etymon, phonological, historical, and semantic.

The well-known hypothesis, that aller comes from the rapid giving of military commands, ambulemus, *amlemus, *allemus, offers serious phonological difficulties. Not only does ambulare give embler regularly in Old French, but, in addition, both ambulare and alare appear in the Reichenau Glosses. Furthermore, it is not necessary to seek a common etymon for the verb 'to go' in Romance, since the very confusion of forms existing in the conjugation of this verb in all the Romance tongues is enough warrant for positing a separate origin for the Northern French aller.

Despite all that has been written on the subject, it is still an open question whether the prevailing types of strong preterites of class VII in NGmc. and WGmc., as Ole. hét, hlióp, helt, fekk, lét, blét, etc., have a common origin with Gothic reduplicated preterites of the types haíháit, áiáuk, haíhald, faífāh, lailōt, haíhōp, etc., or descend from IE unreduplicated formations, in particular lengthened-grade aorists. In recent decades this problem has come to be involved with another one, namely the origin of the pt. pl. stem in classes I-V of the strong verb, above all in classes IV and V for Germanic as a whole, and probably most scholars who accept unreduplicated formations as the source of the forms in the category first mentioned do so for the forms in the latter categories also. It is not proposed in this paper to restate the arguments that have been put forward in support of this view,1 or in refutation thereof,2 but rather, tentatively assuming its general correctness, to reconstruct in chronological order the various developments which the theory seems to require if it is to be brought into detailed correspondence with the known NGmc. and WGmc. forms which it purports generally to explain.

In historical English grammar it is regularly assumed that Gmc. initial [f s þ] became voiced before voiced sounds in southern England at some time in the Old or early Middle English period. For the present purpose, Wright's description of the change is sufficiently representative: ‘The initial voiceless spirants f, s, þ became the voiced spirants v, z, ð in late OE. or early ME. in Kentish and the southern, especially the south-western dialects, as vader, vat, vlesch, vrend; zaule zǫule, zinne zenne zünne “sin”, ðat ðet, ðing. The modern dialects show that this voicing of the initial voiceless spirants must have taken place at an early period, because it is almost exclusively confined to native words, hence the change must have taken place before the great influx of Anglo-Norman words into these dialects. The use of the initial voiced for the voiceless spirants is now obsolete in Ken., Sur., Sus., and obsolescent in s. Pem., Hamp., and the I.W., but it is still in general use in east Hrf., parts of Glo., west Brks., Wil., Som., and Dev.‘

We acknowledge here the receipt of such works as appear to bear on the scientific study of language. The publicity thus given is regarded as a full return for the presentation of the work; no book can be returned to the publisher. Reviews will be printed as circumstances permit, and copies will be sent to the publishers of the works reviewed.

There has been a great deal of interest lately in the application of structural techniques to problems of dialect comparison. Studies in this field are rendered difficult, however, by the fact that many of the existing surveys were planned at a time when the phonemic principle was still insufficiently understood and questionnaires lacked examples illustrating points which were later found to be of great importance. As a result, many of the field records available at present are inadequate for complete phonemic analysis. If a geographical study is to yield enough structurally relevant information, it must be based on advance knowledge of the major phonemic and phonetic features characteristic of the region. A structural linguistic survey in an area not previously studied might therefore consist of two stages: (a) a preliminary series of phonological studies at selected points in the area, using informant methods of the type employed in studies of a single dialect, with a classification of dialect differences found; and (b) a detailed geographical survey with the aid of questionnaires to determine the spread of the features discovered in the preliminary study.