Universals Archive

Meinhof himself mentions vowel changes in Bantu perfects (in archaic Swahili the perfect stem of -PATA ‘to get’ is -PETE; similarly in Konde) but interprets them as assimilations and contractions, thus not affecting his claim.

1. Stem modification by assimilation and contraction is excluded. In German, e.g., ablauts such as TRINK(en) - TRANK - (ge)TRUNK(en) count as positive evidence (Meinhof 1936: 15), whereas umlauts à la KALB - KÄLB(er), triggered by the vowel of the suffix and hence assimilations, are not in the scope of this universal (Meinhof 1936: 46). Apparently, Meinhof does not care whether the umlaut rule is still productive or merely historical. <WS>2. "Flektierende Sprachen" according to Meinhof have four characteristics:(i) inflection, i.e. declension and conjugation;(ii) "Bildungselemente, die keine selbständige Bedeutung mehr haben" (i.e., genuine affixes, rather than function words also retaining content word uses);(iii) "Veränderlichkeit des Wortstammes", especially "innerer Vokalwechsel" (see #1676);(iv) "Mannigfaltigkeit der Pluralbildung" (i.e., plural allomorphy) (see #1673).It is possible, though not mandatory, to read him as claiming that there is an implicational chain among them:(iv) … (iii) … (ii) … (i) That is, his "flektierende Sprachen" are not necessarily "flexive" in the sense of having cumulation, otherwise considered a key property of this morphological type. In actual fact, his prime examples of "flektierende Sprachen" – non-analytic Indo-European, Semitic, "Hamitensprachen" – are.3. See also ##1959, 1960 for further correlates of "flektierend" suggested by Meinhof.