Proquest Information and Learning: Foreign text omitted PHILOLOGISTS and historical linguists are well aware that the synchronic irregularities of language may be considered completely when viewed from a historical perspective. That is, the so-called forms of the modern language are often the expected or regular outcome of some process (be it phonological, morphological, or other), and thus do not strike the specialist of historical grammar as being particularly unusual. However, it is sometimes the case that upon elucidating the regularity of some synchronically irregular form, one is confronted with a diachronic irregularity, i.e., an outcome that is unexpected when viewed within the larger context of the general evolutionary tendencies of that language. These points can be clearly demonstrated by the imperfect indicative. A synchronic analysis of the imperfect indicative reveals only three irregular verbs, shown below in (1): forms of these three paradigms are considered synchronically irregular because they deviate in their formation from all other second and third conjugation verbs, whose imperfect forms are comprised of the infinitival root plus the suffixal morphemes -ia, -ias, etc. (e.g., comer 'to eat' - comia, comias, etc., vivir 'to live' ~ vivia, vivias, etc.). Following this general synchronic pattern (or rule) of verb formation in the imperfect indicative, one would expect the verbs ir, ser, and ver to exhibit patterns such as ir ~ *ia, *ias, etc., ser ~ *sia, *sias, etc., ver ~ *via, *vias, etc. However, the synchronically irregular forms in (1) are easily accounted for when viewed in a historical light. forms of ir and ser are direct descendents of their ancestral Latin forms, which have been passed on orally, from generation to generation, for over two millennia, and with very little change, other than the usual loss of final /-m/ and /-t/ (found in all verbal paradigms),1 as shown in (2): In the case of Sp. ver, one needs only to glance at the Early Old paradigm shown in (3) to explain the synchronically anomalous veia, veias, etc.: Old infinitive was disyllabic, and the imperfect was formed regularly from it, thus, ve-er -> ve-ia, just as creia is and was derived from creer, leia from leer, poseia from poseer, and proveia from proveer. Thus the historical linguist might initially consider the forms of the imperfect indicative of Sp. ver to be perfectly regular, or at least the expected outcome. This attitude is expressed by Lathrop: The irregular imperfect veia reflects the regular outcome of the imperfect of veer in Old Spanish (176). However, a closer examination of the paradigms of ir, ser, and ver brings to the fore the following four irregularities. First, in the case of ir and ser, there has been a shift in stress in the first and second persons plural from Latin to Spanish, thus: Lat. …