It is generally taken for granted that the present-day pronunciation of Arm. p' t' k' represents the original status: Meillet (Esquisse 23) simply states this view without discussion: ‘Il y avait trois séries d'occlusives ... chacune d'elles existant sous forme de sourde non aspirée, de sourde aspirée (c'est-à-dire où l'explosion était suivie d'un souffle) et de sonore'. In spite of the general agreement upon this matter, it seems worth while to take it up again and test the common view of p' t' k', since this view appears to be based only on the present-day situation.

One of the fundamental changes from Proto-Indo-European to the Germanic dialects is that from a verbal system based on aspect to a system based on tense. By the time of our Germanic records the change was accomplished; aspect to be sure was still conveyed in the Germanic languages, but through morphological processes that were not consistently applied to all verbs, such as the ga- prefix. Tense, however, was a category marked in every verb. Moreover, unlike the PIE situation, where the imperfective (present) was sometimes marked to distinguish it from other aspects—a situation maintained in Gk. tithēmi ‘I place’ beside the unmodified stem form of the aorist éthēka T placed'—in Germanic only the preterit was marked, with rare exceptions like Go. standan beside stōþ, bidjan beside baþ. Accordingly the Gmc. verbal system differed from the IE system in the grammatical meaning which it characteristically expressed, and in its method of marking that meaning.

It has been generally held that Gothic possessed a formal aspectual system which was very much like the system in OCS. In each language a simple imperfective is thought to have become perfective by the addition of a preverb—in Gothic especially by the addition of ga-, which in most instances did not change the meaning but merely the aspect of the basic verb (PIAG 103). And in each language every perfective verb, simple or derived, had future meaning in the present tense; though in Gothic, because there was no equivalent to the OCS iterative, a perfective verb also had present meaning. This tense ambivalence of the Gothic perfective present has been referred to as a flaw in an otherwise clear-cut aspectual system.

The present article is intended to display a procedure of morphemic analysis that makes maximum use of morph-segmentation, and classifies morphs into morphemes in terms of recurrent partials within a fixed frame of reference. The frame of reference consists of a text in which substitutions are made at various points and the resulting other changes are observed. Thus, in English we find one boy, one girl, one man, but two boys, two girls, two men. From this one separates out the partials one, boy, girl, man, two, -s, and replacement of a by e (this example is deliberately put in orthographic terms); -s and the replacement of a by e are then grouped together in a class of morphs, a morpheme. We can now give this morpheme a label, if we wish, such as plural, but we could just as well call it something else; the referential meaning (meta-linguistic or ‘real’ meaning) does not concern us in the analysis, but there is no need to ignore it in choosing our labels.

Though considerations of meaning in linguistics can be replaced, up to a point, by rigorous structural procedures, i.e. procedures involving solely the kinds and order of the elements of the language under investigation, they cannot be replaced by distributional procedures, despite the claim recently made by Harris. Distributional procedures may be sufficient to establish the rules by which all longer expressions (especially sentences) can be constructed out of the elements, but they are inadequate for the establishment of certain other rules that would mirror the so-called logical properties and relations of sentences and other expressions.