By Geoffrey Sampson

Rethinking Writing

By Roy Harris

The Athlone Press, 254 pp., £45.00

ISBN 0 485 11547 6

Published July 2000

 

 

I have never understood why such a fuss is made about the linguist Ferdinand de Saussure.  He made remarkable advances in reconstructing Proto-Indo-European, but only a few specialists care about that today.  What Saussure is famous for are ideas about the conceptual foundations of linguistics, which he included in a university lecture course, and which impressed some of his students sufficiently for them to turn their notes into a book after Saussure’s death.  Many of the ideas did need stating at the time.  Saussure pointed out that studying a language as a system at a given period is different from studying its evolution over history, and he noted that the elements of a sign system take their value from their contrasts with other elements.  These things are true, we all see that they are true, and most of us have moved on.

 

But the French-speaking world has a habit of elevating intellectuals to hero status, so that their lightest words are endlessly reinterpreted – with approval, or critically.  Saussure is included in this pantheon, and Roy Harris is a member of his cult.  This book argues that Saussure, and indeed the Western intellectual tradition more generally, have misunderstood the nature of writing.  Writing has been seen as a mechanism for recording speech, when it is much more.

 

It is true that 20th-century linguists downplayed writing in favour of spoken language.  They wanted to focus on the language rules which people obey spontaneously, rather than the artificial rules imposed by stylists; some linguists believed that spoken language is innate in mankind, whereas writing obviously is not.  By now, though, linguistics has largely given up this unbalanced treatment, and I cannot see that Saussure had special responsibility for it.

 

Harris quotes odd-sounding remarks by Saussure about French spelling pronunciations (for instance sept with the T pronounced before a consonant) being “teratological” rather than natural developments.  Long passages of the book satirize Saussure’s description of Chinese script:  how, Harris asks, might one verify Saussure’s statement that Chinese characters represent spoken words directly and ideas only indirectly?  “Is it a matter of determining for how many Chinese the characters represent in the first instance spoken words …?  And would this be an investigation of Chinese people’s opinions, or of their neurolinguistic processes?”

 

Even if we accepted Harris’s objections, what would we be saying here?  A Swiss academic made short-sighted comments about phonetic spelling and Chinese script in a lecture to students ninety years ago.  Not exactly “hold the front page” stuff, I’d say. If Harris had not accorded Saussure guru status, there would be nothing to discuss.

 

Furthermore, in the Chinese case the confusion is Harris’s.  Saussure’s statement is straightforwardly correct, as demonstrated for instance by the writing of Chinese synonyms.  If Chinese characters represented ideas, the characters for synonyms should be the same or similar; but the words are different, and correspondingly the characters for Chinese synonyms are often entirely unrelated.

 

Even the remark about teratological pronunciations may have made more sense than Harris appreciates.  A leading idea for 19th-century linguists was that languages develop according to biological laws of growth. That view is not tenable now, but this does not mean that it was silly in its day.  If historical changes in words were normally governed by internal laws like those which control the development of an organism, then “teratology” would not be a bad metaphor for the special changes induced by external factors such as spelling.

 

Nothing in Harris’s discussion suggests that he has empathized with Saussure’s train of thought.  Harris treats Saussure as a culture-hero, yet he does not seem to understand why Saussure said what he did.  A paradox indeed.

 

 

 

Geoffrey Sampson is Professor of Natural Language Computing at the University of Sussex.  His most recent book is Empirical Linguistics (January 2001).