O’Connor, Michael Patrick. “Writing Systems, native speaker analyses, and the earliest stages of Northwest Semitic orthographies”, in The Word of the Lord Shall go Forth, Ed Carol L Myers and M O’Connor. (Winona Lake: Eisenbrauns, 1983), 439-465.
In this article, O’Connor begins by arguing for the separation of speech from writing. Writing does not belong properly to the discipline of linguistics but semiotics. There are obviously structural parallels between the two (graph:sound, space:time), but rather than being language, writing is a “delinguistic sign system.” However, writing is often regarded as continuous with speech and therefore plays a role of language maintenance and development within a speech community. Thus, writing will reflect some features of speech as understood by native speakers.
Applied to early Northwest Semitic orthography, O’Connor notes features he considers important reflections of native speaker analysis. First, the consonantal orthography represents “the higher informativeness of consonants” by dividing the class of consonants from the smaller class of vowels, although he notes that the liquids (characterized by full sonority) are still notated along with the glides. Next, final vowel notation reflects a phonological analysis of the languages’ vowels. The use of yod for ī and waw for ū is intuitive, but the use of he for ā, ē, and ō is not as obvious. However, as part of a vowel system represented by the vowel triangle, its use is evident. The vowels have been separated into high and non-high groups. The yod and waw notate front-high and back-high vowels respectively while he notates the three non-high vowels. The choice of glides to represent vowel sounds is natural since they represent the least complex sounds. There are six glides: yod and waw, aleph and he, and lastly ayin and ḥet. The last pair are unlikely candidates due to their pharyngeal character. Phonologically, aleph is lax (as are yod and waw) while he is tense making it the best candidate for contrasting vowel sounds. The later developments are more complex. Medial vowels appeared as an extension of the final vowels, but they also were derived from the contraction of diphthongs. Neither of these developments are linguistically significant. The next two developments reflect a reanalysis of the vowel system: the use of waw and yod from derived medial ē and ō was extended systematically to all medial ē and ō, and waw and yod were extended to mark the mid-vowels in final position as well.