Writing Systems: A Linguistic Approach (review)

This is the best available textbook for a course in writing systems, but it is uneven. The core comprises eleven descriptive chapters, each on one or a few scripts; each chapter concludes with a bibliographic paragraph, a list of terms introduced, and well-conceived exercises. Ch. 3, ‘Chinese’ (20–49), is extensive and detailed. Ch. 4 treats ‘Japanese, Korean, Vietnamese’ in fewer pages (50–78), more than half of them devoted to Japanese, so that the entire Vietnamese alphabet is not even shown; moreover, a lot of space is devoted to a rather forced analogizing of Jpn. on (borrowed) and kun (native) readings of kanji (borrowed Chinese characters) to phenomena of English spelling—which cannot even be skipped over, as it recurs throughout the book. The only chapter where R seems not to have consulted a specialist about details, Ch. 5 ‘Cuneiform’ (79–96), is mercifully brief and cannot be relied on. Chs. 6, ‘Egyptian’ (97–114), and 7, ‘Semitic’ (115–44, mostly on Hebrew with a bit on Arabic and even less on Ethiopic), are necessarily condensed. Chs. 8, ‘The Greek alphabet’ (145–69, including both prealphabetic Greek scripts and the Greek-derived alphabets of the Christian East), 9, ‘The Roman alphabet’ (170–84, including Finnish and Scots Gaelic), and 10, ‘English’ (185–98), are clear and accurate presentations of the history of the Western alphabets (though the typesetter has made quite a hash of the German Fraktur examples (182)). Ch. 11, ‘The Indian abugida and other Asian phonographic writing’ (199–232), is uniquely and commendably detailed (but of modern scripts it treats only Devanagari, Burmese, and Tibetan, with a bit on Mongolian plus Bengali introduced in an exercise). ‘Maya’ (Ch. 12, 233–46), like so many accounts, devotes as much space to the intricate but well-understood calendar as it does to the difficult logosyllabic writing system. Ch. 13, ‘Other writing systems’ (247–68), includes Cherokee, Cree, runes, ogham, Pahawh Hmong, and, unaccountably, Blissymbolics, an ideographic notation system that is not writing.

Blissymbolics is not writing by R’s own definition: ‘the use of graphic marks to represent specific linguistic utterances’ (2). This appears in Ch. 1, ‘Introduction’ (1–8), a lightning survey of basic notions, including the distinction between spoken and written language. Ch. 2, ‘Theoretical preliminaries’ (9–19), introduces a number of technical terms, including the indefensible grapheme ‘a contrastive unit in a writing system’ (10). (Why are the characters of Chinese considered graphemes (28), rather than the recurring phonetic and semantic components of characters? or even the seven basic brushstrokes with which they are written?) Ch. 14, ‘Classification of writing systems’ (269–79), presents some traditional classifications and those offered by John DeFrancis and Richard Sproat. In his own classification, R commendably uses the terms abjad (consonantary) and abugida (Indic-style, where the basic letter denotes Ca and other vowels are denoted by added marks) introduced by this reviewer (Fundamentals of grammatology, Journal of the American Oriental Society 110.727–31, 1990)—but claims abugida as his own contribution (274) and fails to describe the clarifications of the history of writing that the distinctions embodied by the two terms made possible.

Two recurrent annoyances mar the book. One is the mere idiosyncrasy of naming the calendar eras old and new (xvii) instead of bce and ce (or bc and ad). This is especially confusing the first time it appears (21), regarding the periodization of Chinese, adjacent to ‘Old Chinese’! The other is quite serious. On the basis of a now-mythic talk at the 1992 LSA by William Poser, never published and never even to be written down (p.c.), R claims that all scripts (except Yi) traditionally called syllabaries, including Japanese kana, Greek Linear B, and Mesopotamian cuneiform, are in fact moraic scripts. A moraic analysis of Japanese phonology is legitimate, but no phonological analysis of any Semitic language has justified the claim that, for example, Akkadian is written with a cuneiform moraography...