Comparatists sometimes encounter unavoidable uncertainties in their research of correspondences between the phonemes historically attested and those which have been reconstructed by comparative operations. This is just the case of thephonemes /c/ of Skr.
ric-and
ruc-, (f. ex. pt. I), and of the Hawaiian /k/(<MP*t or *nt) and /?/(<MP*k or *ηk), and some others as well (pt. II).
Linguistic reconstructs (forms or phonemes) are all, as a matter of fact, merelyfictitious, if not mere fictions at all. They are established by comparative operationsmethodical and systematic, reflecting, however, the respective ways of thinking of the reconstructors and their varying ability of exhausting the actualevidences in concern.
Hence the uncertainties and disagreements however slight in reconstructionsmay be inevitable in the last analysis; the discrepancies of the subjective withthe objective are much common and very human indeed, too. Effort to seekat any price after the narrowest possible exactness of correspondences maysometimes be erroneous, even irrational. Funther, ‘there exist in fact no trulyquantitative, scientific criteria for measuring relationship and for furnishing a clueto comparability, indeed a permission to compare, which will allow us to disregardintuitive, impressionistic judgements and non-linguistic criteria’, as Prof. E. Pulgram says in one of his beautiful articles he presented me with (Lingua, 10, 1961, p.25); also, see V. Georgiev Issledovanija po sravniterno-istori Ceskomujazykoznaniju. Moskva, 1958. pt. II (p.23-).
抄録全体を表示