Bryn Mawr Classical Review

Bryn Mawr Classical Review 2005.07.12

Stephen D. Houston, The First Writing: Script Invention as History and Process.   Cambridge:  Cambridge University Press, 2004.  Pp. xviii, 417.  ISBN 0-521-83861-4.  $90.00.  

Reviewed by Roger Wright, University of Liverpool (
Word count: 2096 words

[Authors and titles are listed at the end of the review.]

Stephen Houston's edited volume of essays derives from a meeting held at Sundance, Utah, in 2000, which aimed to collate and contrast current scholarly assessments of the invention of writing systems in different parts of the world. This apparently neutral description already begs a number of interesting questions, such as "what is writing?", "what is invention?", "what is a system?", and one of the great virtues of the collection is that the editor, despite of his own admission being tempted to do so, refrains from coming up with a universally valid account of the development (intentional or not) of the phenomena which will later develop into writing systems. Most of the specialist scholars included here are eminent in their fields, and the whole beautifully produced volume will be of interest to all those studying early civilizations.

The volume as a whole is directly focussed on the first invention of writing in any culture -- Sumerian, Chinese, Egyptian and Maya in particular -- so other civilizations are omitted, including the great European Classical and Medieval cultures with their own particular complicating phenomena (such as the way that their systems based on phonographic principles became increasingly logographic over time as a result of inevitable diachronic changes in pronunciation), which would more directly interest readers of a review in this space; but we can all study this collection with interest and profit.

This volume comes into the rare category of those that have changed my mind. In his introduction Houston declares that the invention of scripts has little to do with linguistics or even directly with language, and I was startled into disagreement. In his "Final Thoughts", three hundred pages later, he repeats this view, and by then the reader can see what he means: that the invention of writing is a special case. That is, it has become clear within the volume as a whole that the emergence of the earliest writing needs to be explained by more than merely linguistic considerations. Many societies exist without writing, and are not necessarily as a result reduced to wallowing in a state of administrative chaos. Some use pictorial modes without developing writing, a point neatly made in Elizabeth Hill Boone's account of non-linguistic graphic communicative representations such as the Aztecs' pictorial codices (with reference also to mathematical, musical, logical and chemical notation, and to the double helix model of DNA); in the words of Houston's concluding thoughts, the cultural and historical setting of the earliest writing matters (349). Historical Linguistics, which is increasingly prepared to welcome sociolinguistics as an ally, would agree now with this view more than Houston realizes; it has become a commonplace to say that there is no text without a context. The reasons and contexts of the development of writing in separate circumstances are what this volume collectively aims to explain, without demanding, expecting or discovering anything approaching a universally valid model. Comparisons are illuminating, but there is no need to assume the same answers in different parts of the world.

The search for universals has distorted some of the earlier research in these intellectual domains, so it is a relief to note that such a search is largely absent from the minds of the contributors; and, where present, it is largely argued against. The fact, which seems to be established, that the earliest Mesopotamian writing had an administrative purpose can be true without any implication that the earliest writing elsewhere must have had the same social and cultural function as it did in Ur. This general ability to remain relaxed in the face of those who instinctively demand predetermined accounts of historical evolution is also evident in the general view expressed here of Ignace Gelb's supposedly all-encompassing vision of how writing systems develop over time. In Gelb's perspective (expressed most accessibly in his "A Study of Writing", Chicago, 1963) the development of scripts follows a pre-established pattern from the pictographic to the logographic, from there to the syllabic and then to the phonographically alphabetic (which is thereby seen as the pinnacle of all possibilities). Even if this were an accurate diachronic representation of what always happens, which it is not, it would be seen now as a contingent summary of disparate events rather than a necessary path of history. As several contributors point out, or at least imply, the fact that Japanese has probably the most fiendishly complex conglomeration of orthographic details in the world has not in itself impeded their economic and social development. Any idea that reading in one direction horizontally or vertically is somehow more natural than others is put into doubt by the fact that the Aztec pictograms (studied by Boone) proceed from the bottom right, reading left and then right in an upwards boustrophedon reminiscent of the traditional "Snakes and Ladders" board game. Nor, as this volume cogently exemplifies, is there any a priori need to see writing as something which was probably invented just once (monogenesis), as many scholars, including myself, have been led to claim. Perhaps it was in the Old World, but it is not in any way unprofessional to argue that it was not. Thus the development of Mesopotamian and Egyptian writing systems may or may not have had some connection; it is conceivable that Mesopotamian systems might have something to do with the much later development of writing in China; but none of these can plausibly be said to have inspired the idea of developing some kind of writing in Central America. Gelb solved that problem by declaring that the systems in use in pre-Hispanic America do not count as writing, and the desire to disprove this claim (which he does successfully) seems to have been what motivated Houston to initiate this whole enterprise in the first place.

The particular accounts of specific contexts should perhaps be read before the general assessments. Jerrold Cooper's sensible chapter on Mesopotamia does not consider the earliest written signs related to administrative book-keeping, but focuses on the progression from those to a later stage which implies some kind of established reading public and therefore a substantial social investment in training. Robert Englund's chapter on the "probably undecipherable" Proto-Elamite of 3,300-3,000 B.C.(from Susa in what is now Western Iran) is a tour de force; he concentrates on reconstructable numerical systems, adducing many photographs and diagrams of his data. Unfortunately, his Proto-Elamite, unlike the early Mesopotamian tablets, has no later evidence available in a direct linguistic line of descent which could be used to clarify the early documentation.

John Baines considers the earliest manifestations of written Egyptian, which the latest evidence suggests might predate the Mesopotamian or at least have appeared independently. Baines recognizes that pictographic and even logographic methods need not be tied to any particular language: "nothing on the tags or pots unquestionably requires a reading that would work only through the Egyptian language"; and, conversely, "nothing points to the U-j script's writing any language other than Egyptian" (164; "U-j" is the name given to the tomb which supplies many of the data). Not very helpful, the reader may think, but Baines is almost certainly right for this early stage. Later on the situation is different, but "its inventors could not have foreseen that it would later develop into a nearly universal recording medium" (175), and anyway subsequent stages in the Egyptian scripts seem to have been the results of deliberate reforms rather than gradual evolutions. And, to put us firmly in our place, "the developers of writing there were surely more intelligent than the generality of their modern interpreters" (183), so the evidence should be the field of cultural historians rather than philologists. Or both. Some of the excellent photographs adduced by Baines are also reproduced on the dust cover.

The longest and the shortest chapter are both on the same data of c.1200 B.C. from China, respectively by Robert Bagley and Françoise Bottéro. The latter is the easier to understand. These data are inscribed on the inside of turtle shells and the burnt remains of animal shoulder-blades. Having two articles on the same data was a good idea, for it serves to show how little is clearly established. For example, Bagley calls the sign system used "standardized"; Bottéro says it was not. Bagley deduces that the invention must have happened quite a long time before the date of the surviving evidence, with intermediate texts being lost because they were prepared on biodegradable material, the ostensibly sudden arrival of these phenomena into the view of the modern palaeolinguist being the result of "an abrupt decision to inscribe durable surfaces" (226); Bottéro deduces that these surviving written forms were prepared very close to the time of their invention. Bottéro also suggests that the idea of writing, although none of the actual details, is likely to have come to China from further West (259).

Williams' chapter on the earliest runes, which were invented by 150 A.D., is unlike most of the others in its ability to avoid almost all theory; the earliest runes seem to have been used as nametapes, with no inherently magic overtones, and are based on knowledge of the existing Roman alphabet without fitting that exactly.

Houston's own chapter on Central America, more specifically on Zapotec, Isthmian and early Mayan evidence ("none of which supports an optimistic prognosis for full interpretation", 274), starts by making the good point that the pictographic and logographic methods are useful in multilingual contexts, since they could be read aloud in any of the available languages rather than being tied to the words of any individual language. This may explain why such systems (which Houston refers to as "open") increased as Meso-American society became more multicultural, in apparent defiance of Gelb's preferred directions of development. Thus "the singular property of Zapotec writing is that it can barely be shown to record the Zapotec language or an earlier form of it" (294; in this sentence the word "Zapotec" has a cultural and geographical sense rather than metalinguistic), and Early Mayan script is "thoroughly saturated with logography" (305).

In addition to Houston's introduction and final thoughts, and Boone's study of non-linguistic notation, there are two further general chapters. Bruce Trigger gives a concise cultural history of writing, in which "writing had multiple origins and developed in a far from unilinear fashion" (61). This account is excellent; Trigger realizes that "no writing system ... records all the linguistic structure of speech" (44), usually omitting such phenomena as relative speed, intonation and sandhi, so it is has been inappropriate for others to imply that "to be useful, recording systems had from the beginning to be able to record everything" (49). This point is well exemplified in most of the other chapters; criticizing Maya writing for not including syntax is as inappropriate as criticizing our Roman alphabet for not indicating intonation (which is often as important an element as syntax for the understanding of speech). John Robertson's chapter on speech and writing concentrates on an analysis of signs, icons and symbols (following the American linguistic philosopher Charles Peirce). Robertson surprisingly takes the Gelb line debunked elsewhere in this collection ("the most efficient means of representing any spoken sign ... is by spelling", 21). There are also some mistakes; for example Robertson refers to "a silent [e]" (32), which is a contradiction in terms, since any symbol in square brackets by definition represents a sound and is thus by definition not silent (he seems here to mean a silent letter "e"). The next two pages manifest more misconceptions concerning phonetic script which the limitations of this review medium do not allow the reviewer to exemplify. No great matter, since readers interested in the scripts of ancient civilizations will probably skip this chapter anyway.


Stephen D. Houston, "Overture to the first writing" (3-15)

John S. Robertson, "The possibility and actuality of writing" (16-38)

Bruce G. Trigger, "Writing systems: a case study in cultural evolution" (39-68)

Jerrold S. Cooper, "Babylonian beginnings: the origin of he cuneiform writing system in comparative perspective" (71-99)

Robert K. Englund, "The state of decipherment of Proto-Elamite" (100-49)

John Baines, "The earliest Egyptian writing: development, context, purpose" (150-89)

Robert W. Bagley, "Anyang writing and the origin of the Chinese writing sytem" (190-249)

Françoise Bottéro, "Writing on shell and bone in Shang China" (250-61)

Henrik Williams, "Reasons for runes" (262-73)

Stephen D. Houston, "Writing in early Mesoamerica" (274-309)

Elizabeth Hill Boone, "Beyond writing" (313-48)

Stephen D. Houston, "Final thoughts on first writing" (349-53)

[[Plus a huge list of bibliographical references (354-94) and an index (395-417).]]

Read Latest
Index for 2005
Change Greek Display
Books Available for Review

HTML generated at 13:30:11, Friday, 03 April 2009