Abstract

Observational terms should be distinguished from theoretical ones. Sounds of spoken language are observational terms. We can hear them or record them as real entities. What are the connections between the various terms? All phones (sounds of speech) that relate to a certain phoneme share some phonetic features, and these features define a theoretical term which is the certain phoneme. The phoneme itself cannot be heard, since the realization of a phoneme is a phone, an observational entity which always has more external features.

We show that /b/, /k/, and /p/ are each one phoneme, which is realized in certain phones. The claim that each of them split into two phonemes is dealt with, and our conclusion is that there is no basis to suggest it. We show that other languages also have some phonemes that are in a similar situation, for example, English delete—deletion or produce—product. The split to two phonemes that Weiman and Rosén suggested emerged not from the facts but from false principles in structural linguistics. Rosén had later to follow it up by the idea that there are “complex roots” in Hebrew. Thus the problem is shifted from philology to morphology and to the lexicon. Some further deficiencies in structural lore are noted.

We also show that there is no need to apply to other related languages (Aramaic or Arabic) to discover theoretical forms in present Hebrew. They are not historical remnants but living entities existing in the linguistic knowledge of every Hebrew speaker.

pdf

Share