Cambridge Catalogue  
  • Help
Home > Catalogue > The First Writing
The First Writing
Google Book Search

Search this book


  • 25 b/w illus. 3 maps
  • Page extent: 436 pages
  • Size: 247 x 174 mm
  • Weight: 0.89 kg

Library of Congress

  • Dewey number: 411.09
  • Dewey version: 22
  • LC Classification: P211 .F62 2004
  • LC Subject headings:
    • Writing--History

Library of Congress Record


 (ISBN-13: 9780521838610 | ISBN-10: 0521838614)


1   Overture to The First Writing

The theme of this book is the first writing systems – early cuneiform, proto-Elamite, Egyptian (all prior to 3000 BC, by some centuries), Chinese (prior to 1000 BC), and Maya and its Mesoamerican precursors (beginning ca. 500 BC). These are the scripts that first represented units of sound and meaning in systematic fashion, if in contexts that often elude full study because of deficient evidence and opaque coatings from antiquity. The first writing is involved with so many different processes – administrative development, sacred display, the establishment of consultable precedent, statements of being, even the hermeneutic probing and decipherment of such writing systems – that it deserves repeated attention, more than once in any one generation of scholars. The present book also looks more briefly at what might be called “secondary inventions,” including late ones like Scandinavian runic. Such systems allow us to understand how the processes of primary and secondary creation differ fundamentally or conform to similar patterns.

   This book came into existence for several reasons. The earliest writing records voices of antiquity with an immediacy to the past that lies beyond potsherds and lithics. Those voices come to us through momentous steps in representation that implicate, in structured fashion, sound, meaning, and sight. Many specialists have noted patterns of convergence and divergence around the world, with the understanding that universals, if present or detectable, involve similar conditions and strategies rather than pat laws of script invention and use. The moments of decisive, singular change (the “history” in the title) and general trends (“process”) require judicious acts of weighing and evaluation. The good news is that the topic of the first writing remains a productive mine of insight. There is still much to do.

   Earlier works managed to plumb the subject in part yet not always in ways that are fully satisfactory. An otherwise excellent work edited by Wayne Senner (1989) came from a lecture series that arose, it seems, from little to no communication between the authors. By scanning the book, a reader benefits from comparisons that, because of the volume’s manner of preparation, would have escaped its many contributors. By the same token – I use the term advisedly, with an apology to cuneiformists – general works on writing systems enrich our debates (Coulmas 1989, 1996; Roy Harris 1986; Sampson 1985), but several of them place too much emphasis on narrow issues, such as rectifying in acerbic fashion what are felt to be misconceptions about Chinese (DeFrancis 1984:133–220, 1989:89–121).

   More recently, Peter Daniels and William Bright (1996) have produced a massive, edited review of the world’s writing systems, including handy introductions to the study of writing systems, origins, and decipherment. The parallel format throughout the volume promotes a high degree of comparability and facilitates the detection of specific parallels, very much on display in Bruce Trigger’s contribution to the present volume; John Robertson, too, benefitted from this magisterial source. The Daniels and Bright volume highlights the need to focus on the behavior of units in script, such as syllables and morphophonemic particles. As such, it fuels a comparative approach to the study of writing systems, although there remains a strong measure of disagreement about the nature of such comparisons. Are they best left vague and general, with a sensitivity to historical and cultural setting, or should they commit to something along the lines of Joseph Greenberg’s enthusiasm for linguistic universals, with which such studies potentially have much in common (e.g., Justeson and Stephens 1994)? More than likely, both approaches need attention. At its least effective, depending on the chapter, the Daniels and Bright volume reduces writing systems to skeletal checklists of attributes and sample texts. As in all such books, the chapters on Mesoamerican systems are, to a specialist’s eye, out-of-date because of the rapid pace of work in that region. One wonders if similar problems impair other chapters.

The need for conversation

In my view, a strong edited volume should be neither a set of chapters fashioned at different times and places nor a vision imposed by any single editor, whatever his or her sagacity or level of energy. There must be a conversation between different kinds of experts, possibly conducted via email or other media of communication but done best of all in a seminar format, in the heat and negotiation of direct conversation. This is what took place at Sundance, Utah, over a three-day period in the Spring of 2000. The editor assembled a group of scholars to present papers before an audience of, at most, other contributors and selected scholars invited from Brigham Young University and a few other universities. The emphasis was less on polished presentation than discussion. Houston sent briefing papers to each participant, including sets of questions worth posing and, perhaps, answering. After the seminar, Houston distributed a set of summary perspectives and specific guidance for each chapter. This book results from that interaction, some comments from anonymous reviewers, and further thinking on the part of the editor and his gracious, ever-patient authors.

   It is important to frame the mentality that created the seminar, the needs and expectations that urged it into existence. In the first place, it was clear to me that most scholars of early scripts are fascinated by writing in general. The series on writing produced by the British Museum has only enhanced that interest by presenting specialist knowledge in brief, readable treatments (e.g., Moore 2000). With that interest, however, has come a growing understanding that comparison is best done, not by a single person, often misconstruing data and debates in areas far from their speciality, but within an interactive setting that convenes authorities from diverse fields. From this will emerge, it is hoped, a shared vocabulary and a sharpened sense of problems and prospects relating to each system. Dialogue cannot exist unless scholars tune to the same wavelength.

   At times it seemed probable during our gathering that more meetings would be needed before such a wavelength could be found and clear signals discerned. Disciplinary boundaries and preoccupations die hard, and anthropological regard for cross-cultural comparison can seem, to some, over-generalizing and distorting. As an editor, I struggled with a desire to get others to frame questions as I do. Eventually I understood that this might not be a good thing: other voices and approaches need to be heard. By scholarly predisposition, anthropologists tend also to be suspicious of “presentism,” the idea that past mentalities are easily accessed by acts of empathy or that present-day terms and categories correspond to those in the past. For example, several participants felt that “utilitarian” was a non-problematic concept, while others considered “magic” or apotropaic functions to be as “utilitarian” as any accounting of emmer harvests (cf. Postgate, Wang, and Wilkinson 1995:475). By similar predisposition, humanists and philologists are inclined to question an enterprise that casually poaches evidence from classical disciplines going into their second or third century of existence. Yet, as this book shows, the encounter is worthwhile. The conversation must proceed.

The “ideological model” of writing

There were other potent motivations for this book. As became clear at Sundance, there is increasing discomfort with, and outright rejection of, the gradualistic and unievolutionary models of script development championed by Ignace Gelb and others (1963; see also Damerow 1999b:4–5; Schmandt-Besserat 1992; cf. Michalowski 1993b). Instead, much evidence indicates that script developed step-wise, in rapid bursts – at our scale of analysis, we can see this in terms of a single human lifetime, albeit with subsequent modifications that extend over centuries. Depending on region, some of these developments were independent of other scripts, with each traveling along distinctive trajectories. This process loosely resembles Niles Eldredge and Stephen Jay Gould’s model of “punctuated equilibrium,” in which rapid speciation interrupts long periods of stasis. Gould himself has developed the notion of “punctuational change” to describe the sudden effects of historically unique moments – contingency or, in his words, “wondrous and unrepeatable particulars” – that radically alter the world (Gould 1999:ⅹⅻ).

   As a result of these changes, there is now a deepening influence of what has been called the “ideological model of literacy.” It sees writing as both a concrete object and the product of social practices within particular cultural settings and power structures (Street 1993:7). From this vantage-point, writing can be understood principally as a system of communication with subtle, seldom-understood relations to orality, as in the performance or “recitation literacy” attested in Mesoamerica (Houston 1994a:29–31). Naturally, this would also have an impact on the pace of script creation, for the logical reason that systems of communication require rapidly coalescing systems of graphs to encode such messages: what good is a lone syllabic sign when many are needed to record a language, or one logograph when the totality of lexemes in ancient languages compels, even in the most limited register, a far broader range of symbols? In no known case does a logograph exist without other, contrastive ones. What comes to the fore in considering the ideological model are indigenous concepts of what writing was and what writing does, who was entitled to read and write (implying gender relations as well as those of social status), and the predicaments of cross-generational transmission. Ensuring that a script endures must involve the strategies of pedagogy and apprenticeship.

   The backdrop that this perspective replaces, the “autonomous model of literacy,” sees writing as a technology with consequences that erupt independently of social setting (Street 1993:5). Invariable cognitive effects are thought to occur when writing, especially alphabetic writing, appears (e.g., Goody and Watt 1963), a conclusion that has met with massive disapproval from some quarters (cf. Halverson 1992; Houston 1994a) and continued defense from others (Goody 2000:5–9). Elsewhere, David Noble has underscored the deep-seated amour in occidental thinking for the perfectability and uniform results of technology (Noble 1999; see also R. McC. Adams 1996). Writing would not appear to lie outside those notions. A subset of mechanistic views of writing is the theory that the properties of certain languages, especially those emphasizing single-syllable morphemes, lent themselves to script invention (Daniels 1996b:585). William Boltz, a valued participant in our seminar, has extended this theory by suggesting that principally monosyllabic languages, such as Mayan, Chinese, and Sumerian, were more likely to possess homophones and were thus further inclined to explore the graphic possibilities of rebus (Boltz 2000:4, 15). From this process came writing. Daniels’ and Boltz’s proposition stands in acute contrast to the ideological or cultural model by situating the origins of graphic communication in features of language, leaving to the side other motivations that would seem more applicable. Moreover, Mayanists would find two propositions in Daniels’ and Boltz’s theory difficult to accept: that Maya writing was the first in its general region (it is not, deferring to Zapotec, later Olmec developments, and, probably, Isthmian); and that Mayan languages are structured like Chinese (they are not, requiring a large number of prefixes and suffixes; see Houston, ch. 10, this volume).

   A key suggestion here is that the use of certain metaphors constrains and channels interpretation. If regarded much like McCormick’s reaper or Edison’s light bulb, writing becomes a transportable technology like any other, a series of mechanical parts detached from their fundamental communicative function. To put this a different way, the medium has come to dominate the message, along with the dialogue that prompted that message. Most volumes on the origins of writing also use the metaphorical conceit of genealogy and descent lines, as though systems of writing consisted of self-contained packages of DNA that bear multiple offspring (e.g., DeFrancis 1989:figs. 10–13). This biological conceit is an unwelcome visitor from linguistics, prompting the reflection: why are there references to “daughter languages” when no historian or anthropologist would allude to “daughter polities?” However, the ideological model is not without flaws. Its emphasis on power relations subordinates cultural values and communicative strategies to blunt mechanisms of inequality. This makes some discussions of writing seem rather like moral parables of cunning that triumphs over virtue (e.g., Larsen 1988:177). It populates the world of early writing with self-conscious actors among elites and slack-jawed dupes among non-elites. Or, in another twist, Marxists of various stripes replace the slack-jawed caricature with another, that of the passive–aggressive peasant in sullen resistance and intermittent rebellion (e.g., J. C. Scott 1976:188–189). These views do have the benefit of recognizing that literacy comprises two processes, writing and reading, and two sets of people, the communicator and the receiver, roughly paralleled by the principal features of the hermeneutic literature on intention and reception (Machor and Goldstein 2000). In addition, writing itself can be seen as a fetishizing and vitalizing act that imparted properties to an inscribed object beyond its physical existence as a textual medium, a subject that has been explored by David Stuart in unpublished work: at stake in particular are the associations between names and personal identities on the one hand and possessions or gifts on the other.

   Nonetheless, a heightened accent on cultural setting runs the risk of overlooking transcendent human solutions to common challenges: that is, how are sound and meaning noted graphically, and how do such marks relate to an external world? Hieroglyphic systems preserve those existential links, although it now appears that most primary scripts followed or began in tandem with representational systems that adhered to fairly rigid canonical conventions (see chapter by Robertson). The relation of script to image needs further discussion by scholars working in places where ancient peoples favored hieroglyphs. What, for example, is the connection of emblematic aids and their capacity for showing action to any changes that might be detected in writing? Here, as everywhere, scholars must weigh a search for general patterns against the need to respect the presence of real human diversity.

Writing and the “state”

The First Writing also responds to a long-standing argument about the inception of early scripts. In the New World, Joyce Marcus is the purest advocate of the “statist” or “political” view of writing, which treats script as a correlate of the “competition for prestige and leadership” and, even more controversially, equates complexity of writing with complexity of “state” control, an assertion belied by Mixtec data from Postclassic Mexico (Marcus 1992:435). In Mesoamerica, for example, the probable existence of “recitation literacy,” which interweaves reading with acts of verbal performance, complicates the public reach of texts to a great degree, since even a small text could extend to a wide audience by means of recitation. Conversely, some ostentatious texts, as in the Maya area, may well have been covered by now-perished textiles, making their concealment and selective but intentional failure to transmit part of the overall message (D. Stuart 1996). Here, writing is firmly distinct from reading, or, as I have stated elsewhere, “textual production” – the painting, incision, and sculpting of written signs–must be separated analytically from responses to them (Houston 1994a:28–29). Some texts may not even have been intended for human or living “readers” but simply to curse or to indicate that certain acts had been performed (Gager 1992; Bodel 2001:19–24). The important point is that writing-in-use and writing-in-context – what might be called the “pragmatics” of writing (Brown and Yule 1983:27, 35) – are complex matters that necessitate a comparable depth of interpretation.

   Still, a cybernetic focus on information storage needs to be acknowledged. The density of encoding surely increases with writing as does the importance of writing as a mechanism for the cultural homogenization of elites, who, the more they compete, the more they look the same (Wheatley 1971:377). Norman Yoffee puts this well, suggesting, after James Scott, that complex polities aim, through script and, I would say, iconography, to regularize themselves by ensuring “legibility,” the joint understanding, perception, and, ultimately, control of social arrangements (Yoffee 2001:768). To use a variety of metaphors, complexity unravels unless everyone follows the same “play book,” the same system of measurement, the same naming patterns, the same legal codes and modes of taxation, and, of course, the same rationalizations of difference and social hierarchy. The shared play book does not complicate life – it simplifies it, leading to a regularity of code that has now, in the twenty-first century, reached its ultimate expression in digital communication. By itemizing the products of human effort, scribes allowed such goods and services to be identified and managed. There is, however, a more subtle point as well: with these systematic expectations and joint views, people transcended purely local knowledge or any direct familiarity with interlocutors. The collective understandings transformed those who were unknown into, at least, those who were understood. In turn, the breach of those expectations – obviously idealized in part – became more evident to the supervisory polity, which could then move to censure and rectify infractions of the play book.

   Janes Scott makes it clear that the penetrative oversight of such polities often fails (1998), and it may be that the controlling traditions of premodern states worked more effectively. These traditions did not, as Yoffee indicates, come from the top or from outside but grew out of preexisting “commonalities” (Yoffee 2001:769). If so, one wonders why the invention of script would dramatically change that which seemed already to exist. How did it shape those commonalities in such a way as to make them more cogent or more penetrative? A perplexing counter-example comes from ancient Mesoamerica. How did the vast constructions of the first millennium BC in the Mirador Basin of Guatemala grow out of societies that were apparently preliterate yet displayed an astonishing homogeneity of material culture (Forsyth 1993)? Arguably, the buildings themselves, as bulky as hills, provided unavoidable daily reminders of social covenants, of how people ought to fit into larger communities, in much the same way that medieval cathedrals embodied community and covenants beyond any one person or any one generation. On present evidence, Maya script developed at about the time the Basin societies began to disintegrate (Houston, ch. 10, this volume). In no way can writing be seen to shape the societies of this area. Rather, strangely enough, its first appearance marks their time of decay. It may not be a coincidence that the development of linear script in Mesoamerica took place at the end of the Olmec civilization (Houston, ch. 10, this volume).

Episodes of script development

Archaeologists have long observed that journalists – and those scholars supplying juicy copy – forever stress the earliest and the biggest. These attributes make for clear stories, without any need for tiresome academic hedging and exposition. Two recent stories in a prominent newspaper zero in on earlier dates for supposed Chinese script in central Asia (c. 2300 BC), long before its appearance in China itself, and another set of alphabetic texts in Egypt that occur well before their emergence in the Sinai peninsula and the Levant (Wilford 1999, 2001). The underlying point of the reports is that monogeneticism applies in both cases, Egyptian writing engendering the alphabet, an unspecified central Asian script leading to the later florescence of Chinese. Although stated less explicitly in these articles, it appears that some scripts came into existence in places of cultural contact, less as blatant copying than as counter-markings pertaining to opposing ethnic or linguistic groups. Most “biscripts” of pivotal importance to decipherment, including the Rosetta Stone, Diego de Landa’s Maya abecedario, the Pyrgi tablets, and late Ogam and Latin texts in Wales, come from collisions between, and accommodations among, dominant and subordinate groups. At such membranes of contact the newly devised script often displayed radically different principles of organization from its parent or stimulus, as in the early alphabet, Sequoyah’s syllabary, Ogam or Naxi. That is, such scripts differ for reasons other than the low intelligence and phonetic obtuseness of copyists – in this, Peter Daniels’ distinction between “sophisticated” and “unsophisticated” origins of script is unhelpful (Daniels 1996b:579). “Ideological” or, in less loaded terms, “cultural” models of script invention find such matters of far greater interest than the very notions of “monogenesis” or “multiple-genesis.” In key cases, stimulus diffusion – the practice of copying or imitation at a very general level – can be impossible to prove or disprove, resting as it does on arguments about dates and relative physical proximity. More to the point, if writing is socially and culturally rooted, why are such broad, pan-millennial, and pan-continental formulations of any interest? Is their sole purpose to show that humans are imaginatively impoverished and that the proverbial mouse-trap could only be invented once? The monogenetic view can be exhilarating in its broad scope. However, in my opinion, it devalues the details of human experience by its sheer grandiosity, a scriptural equivalent to Arnold Toynbee’s lofty and somewhat discredited utterances about civilization (McNeill 1989:177).<

printer iconPrinter friendly version AddThis