Sign o' the times

Writing has long fascinated Roy Harris. Ten years ago he published a remarkably polemic book on its origins, in which he questioned whether we knew what it was. Now he goes a stage further and roundly asserts that we cannot set limits to a subject until its investigation is complete. This is not merely moving the goalposts, but concealing them until you call "Goal!". By stretching his net wide enough he is able to catch some queer fish: music, knitting patterns and manuals of dance-steps all find a place in this book.

Nearly 50 years ago a distinguished American linguist, I. J. Gelb, published a famous book called A Study of Writing with the subtitle The Foundations of Grammatology. He defined writing as "a device for expressing linguistic elements by means of visible marks". Harris disagrees. He reserves the term "glottic writing" for Gelb's definition, and uses "integrational analysis" to bring nonlinguistic means of communication within his scope. This has advantages, in that it enables the Braille alphabet to be accommodated, although it is intended for "processing" (Harris's word) by touch, not sight. But, for most of us, there is a clear distinction between a sign reading STOP and a red traffic light.

In fact, Harris is primarily concerned with messages conveyed visually, and it might be better to call this book "A Theory of Semiology", and, with that reservation, it deserves to be taken very seriously. In a civilised society we are constantly assailed by writing. Signs, notices, and advertisements clamour for our attention, and use every trick of typography, colour and symbolism to attract it. All these need to be described and analysed by the investigator.

Glottic writing was, no doubt, in origin a laborious process of finding ways to use signs to indicate the sounds of speech. But once invented, it began to acquire a life of its own, independent of its spoken origin. Reading was at first always aloud, but we now regard silent reading as normal; and this allows us to recognise words as graphic units without analysing them into separate letters. It is this which enables us to cope with the vagaries of spelling systems as remote from the perceived sounds as those of English or French. It is possible to grasp the meaning of a written word without knowing how to pronounce it. Or to quote the reverse experience, when I was learning modern Greek, being already familiar with the graphic patterns of ancient Greek, I often found I had to create mentally a written form for what I heard before I could understand it. Quite clearly, writing has a life of its own, and should not be treated as a poor relation of speech.

Another example of this, which Harris does not quote, is the use of abbreviations. It is obvious that "etc" is a convenient shorthand for "et cetera", but what of the abbreviations which are themselves treated as vocabulary items? We regularly speak of "gaining a PhD" pronouncing this as "pee-aitch dee", not "doctorate of philosophy". Some abbreviations have become words in their own right, especially acronyms, such as Nato (which becomes Otan in French).

Harris is more concerned with the nonlinguistic features of writing. For instance, he has an interesting chapter on the direction of writing. Instead of our familiar left-to-right order, some scripts use right-to-left. Early Greek inscriptions sometimes combine these, writing alternate lines in opposite directions and reversing the letters, so that what was "bud" in one line would be written "dub" in the next. Early alphabets were, therefore, designed so that no letter could be confused with its mirror image, a feature which survives into our present upper case alphabet. Children learning to write still do this, but are discouraged. Chinese and other oriental scripts are often written in columns running down the page starting at the upper right-hand corner, but no script starts at the bottom of the page. In fact, Chinese can also be written in our fashion, and I am reminded of our former professor of Sanskrit, who once showed us a Mongolian manuscript. When we asked him in which direction it was written, he replied that they wrote it in columns like Chinese, but he preferred to turn it on its left side and read it across the page. There is also a story of Henry Bradley, who as a small child, taught himself to read by watching the book his father held standing in front of him at family prayers; naturally he read upside down starting at the bottom right.

Writing is, of course, a human activity with social (or "macro-social" in Harris's terminology) implications, and illiteracy is a stigma. For instance, the modern debate on orality and literacy appears in a new light on Harris's principles. He is very interesting on signatures, though he hardly mentions the problems connected with handwriting, except in an appendix. The signature as a means of authenticating a document is not a form of communication, for it is often illegible. It is simply a personal sign, which, although it can be forged, serves for most purposes to distinguish us from others. In the Far East, where idiosyncratic handwriting is not encouraged, personal seal impressions are used instead.

A short discussion of the difference between tokens and emblems is enlightening. For Harris "tokens are signs based on one-one correlations between single items. Emblems are signs based on one-many correlations, in virtue of which the 'many' are regarded as forming a single class". This is easier to follow if we quote examples. Emblems "include signatures, proper names, masons' marks, logos, trademarks and national flags". Tokens "include notches on a stick, beads on an abacus, ticks or crosses on a list". We might perhaps say that an emblem means the same again, while a token means another one of the class. This is an important distinction, but one not often observed in usage.

This book contains brief discussions of many more fascinating topics, well illustrated by sometimes surprising photographs of inscriptions of many kinds. It is a pity it is couched in opaque language. The ordinary reader may need a larger than ordinary dictionary to cope with words like "telementational", "logocentrism", "pyromantic" or "basmalah". The last is, presumably, what is better known as "bismillah". But for those willing to try, the reward is an enhanced understanding of something we often take for granted.

John Chadwick, emeritus reader in Greek language, University of Cambridge, is known for his work on the linear B script.