Studies of writing as a social practice often emphasise its social context rather than content, especially in relation to its origins. For Goody, Ong, Havelock and others, writing (especially alphabetic writing) transformed society (e.g. Goody and Watt 1968; Havelock 1986; Ong 1982). In contrast to early accounts of the origins of writing that emphasised form (pictures to signs), contextual studies into the origins of writing emphasise the function that early writing fulfilled within a given society. Such studies are often ‘teleological’ in their conclusions — writing arose as an ‘imperfect’ form, ‘incomplete’ in relation to its later manifestations. Perhaps most familiar here is the Egyptian hieroglyphic system, whose origins are often sought in late Predynastic funerary contexts, notably that of Tomb U-j at Abydos (e.g. Dreyer 1998; see also Baines 2004). Rather than seeing the tiny labels or pots with large painted signs as the first intimations of greater ‘things to come’ in the fully-fledged hieroglyphic system, Piquette (2007) and others prefer to see these as part of a late 4th-millennium BC context of material practice (see also Piquette, this volume; 2008). A similar argument can be made for the earliest clay ‘documents’ in Sumer and, perhaps, for the earliest Aegean script use as part of a set of elite practices of display, rather than the first, imperfect steps towards a means of administrative control (cf. Bennet 2008: 5–6; Flouda, this volume; Schoep 2006: 44–48). In most cases, so the argument goes, the basic need was to deal with the amount and complexity of data to be recorded. Postgate, Wang and Wilkinson (1995), for example, argued that writing always occurs because of a need to record economic data and that the different forms it took are products of taphonomic processes that differentially preserve certain materials (cf. also Pye, this volume). Houston perceptively points out, however, that the “materiality of script differed by cultural setting” (Houston 2004: 350, his emphasis). It is difficult to imagine, for example, that we are missing extensive collections of clay documents from late Pre-Dynastic Egypt or masses of perishable papyrus texts from later 4th-millennium BC Mesopotamia.
Denise Schmandt-Besserat (1996) famously derived writing from accounting practices already millennia old by the time the first numerical tablets were produced in Southern Iraq and Iran, the system only subsequently being enriched by the development of signs with phonetic values (cf. also Cooper 2004). Her argument emphasises that function and content are not necessarily co-extensive. A system (like the khipu, or her early tokens, for example) responded to a need to record and organise information, one of a number of material practices, while a phonetic element was introduced to make clear elements that could more effectively be realised through language, such as the names of institutions, divinities or individuals. Postgate (1994: 51–70) points out that cuneiform writing took centuries to acquire the range of uses that we now regard as de rigueur for any self-respecting writing system. Here a distinction between mechanics and content is important: the late 4th-millennium BC recording system elucidated by Schmandt-Besserat and others did not contain within it the germs of the Epic of Gilgamesh. More recently, the printing press, derived from the technologies of wine production (the screw press) and sealing/stamping in the 16th century, defines the way we view text on screen using radically different technologies.
Emphasis on origins is important in another sense, in that writing — in the narrow sense — was not ‘invented’ each time it appeared; there were a limited number of original ‘inventions’: Egypt and Mesopotamia, although many see them as linked (e.g. Postgate 1994: 56), China and Central America. From these origins, it then spread, in the case of the Old World both to east and west, although not always as fully-formed systems (given the inertia of convention), but sometimes as the ‘idea’ of writing. The latter point implies a knowledge of the principles of a system and its social role. Is the invention of writing a one-way process, like the adoption of agriculture or urbanism? Systems can be lost (cf. Baines et al. 2008), but more often through replacement (most spectacularly evident in the spread of systems based ultimately on the Phoenician alphabet through much of the Old World). A particularly striking example is the replacement in the early 1920s of the Ottoman script by its western cousin, the Roman alphabet, as part of Kemal Atatürk’s westernisation programme for the newly-formed Republic of Turkey (e.g. Lewis 1999: 27–39). A counter example to replacement is, of course, the loss of the syllabic Linear B script in the Aegean, unlike in Cyprus, where a syllabic script lived on alongside the novel alphabet. This example is a salutary reminder that social forces can outweigh material practices; in the Aegean it is most likely that oral practices lived on, while written died out.
