Investigations in item stability: In pursuit of the optimal meaning list for use in the initial stages of the comparative method
Item statusRestricted Access
The lack of a widely-accepted, objectively-defined standard list of ‘basic’ meanings for use in the initial stages of the comparative method is identified as a priority in the resolution of areas of unnecessary subjectivity in historical and comparative linguistics. A methodology is presented, capable of ranking meanings by a score fully representative of the four features identified as necessary for a meaning to be considered optimal for use in the initial stages of language comparison: maximal item stability, maximal resistance to replacement of form by borrowing, maximal conceptual simplicity, and maximal universality. The stability of 67 meanings is quantified using a procedure described, but not adequately implemented, by Dolgopolsky (1986) and Lohr (1999); the results are integrated with Tadmor et al.’s (2010) borrowed, analyzability, and representation scores, to form a composite score by which the meanings are ranked. The resultant ranking, while not representative of the definitive list of meanings optimal for use in the initial stages of the comparative method, owing to the limitation on the number of input meanings, demonstrates the viability of the methodology presented here. Statistical results are presented to support the hypothesis that there is a strongly significant relationship between item stability and variation in stability; however, contrary to expectations and conflations evident in the literature, no evidence is found to support the hypotheses that there are strongly significant relationships between item stability and item borrowability, analyzability of form, or item universality. Finally, the results are used to test the validity of the glottochronological hypothesis of a constant rate of replacement; no support whatsoever is found to support this hypothesis.