Wasow 1975 argues that three problems about rule ordering arise in the proposal of Harman 1972. Two of these are in fact illusory, and the other can be met by means of the convention that variable movement rules can also move there. The interaction between Tough Movement and Dative Movement provides additional positive evidence for the proposal in question.

[IE pure N-grades of the and types yield Alb. ul/lu and ur/ru, except where Ә appears to have been inserted (perhaps because the form in the N-grade was difficult to pronounce). The extended grades arising in this way fall together with original R-grades, which on becoming stressed or restressed yield Albanian combinations with the vowel a. IE N-grades of the and types are indistinguishable from extended grades and R-grades; they yield âm and ân in the Geg dialect, ëm and ën in Tosk. The article presents the evidence for these developments in a large number of etymologies, many of them new.]

Although the Greek language shows a considerable number of forms with present reduplication in ι, yet in no sense do these constitute a living category. They are almost a closed circle of old and inherited verbs, subject to all kinds of analogical changes which give evidence of failure to understand their structure, while even new forms show by their very appearance that they were not consciously felt as reduplicating forms. Thus πταμαι (Moschus, etc.) for πτoμαι was clearly patterned after σταμαι, for if felt as a reduplicating form it would have appeared as ∗πταμαι. At the same time this analogical formation shows that itself was no longer analyzed as reduplicated. We may in fact say that the whole category of reduplicating presents, however clear is its existence from the historical point of view, was psychologically a nonentity. The individual reduplicated verbs were either structurally obscure to the speakers of the language, or else they were analyzed wrongly, i.e. not in accordance with their origin. Thus no doubt μí-μvω was usually conceived as μíμμ-vω because the latter conception is supported by a number of verbs like τμ-vω, and, moreover, by the probable division of syllables in pronunciation (μíμ-μ-vω rather than μí-μvω while there was nothing at all to assist in the apperception of the original weak form of the root μv-. Even those reduplicated presents which are clearest to the philologist, e.g. δí-δω-μι and τí-θŋ-μι must have been wrongly analyzed or seemed structurally obscure to many speakers. That e.g. was sometimes conceived as , with thematic vowel, is shown by the analogical of the New Testament. On the other hand, e.g. were certainly often taken as thematic formations and analyzed τíθ--τ and τíθ--ται in the as is made clear by of D at Luke 8:16, and still more by the following thematic forms in papyri: and cf. Robertson, Gram. Gr. N. T. 318.