Search

1.1. In our historical work, we must be ever aware that reconstructing chunks of a protolanguage-be they phonemes, lexemes, or whatever-and tracing the grammatical development of attested daughter forms-be there one or many-are two necessary aspects of the same method of proof. On the one hand, we can justify a reconstruction only to the extent that it is a necessary starting point for historical development into some attested formation. On the other hand, several lines of historical reasoning from attested formations must predict and determine a specific reconstructed form. Kurylowicz, we are told, summed up this fact about doing etymologies by the formula, "Every com