All preserved examples of Aegean writing are on clay tablets or other administrative documents that have been recovered from destruction layers of buildings destroyed by fire, or on small-scale objects made of durable materials whose survival was dependent on archaeological chance. It is clear that various kinds of perishable materials, such as leather, wood or papyrus, could have been, and almost certainly were, used to write on. The nature of the scripts themselves provides one indication that this was the case. The signs of Linear A and B are quite complex and consist of curved as well as straight lines. They are therefore more suited to writing or painting with ink on papyrus or pottery than to being inscribed in semi-dried clay (Chadwick 1976: 27; Palaima 2003b: 171). This suggests that some form of ‘paper’ support was used from an early period. It is possible that the development of Linear A from the Archanes script, which, as far as we know, could only be written by pressing seals into clay, should be seen in relation to the availability of imported papyrus or the acquisition of a technology for turning animal skins into a suitable writing surface. Since contacts between Crete and Egypt go back to the Early Bronze Age, it is not unlikely that papyrus was imported and used as a writing material also on Crete. A type of sealing known as the single-hole hanging nodule consists of a triangular lump of clay which had been formed around a knot at one end of a string. Hallager has argued that nodules of this type were attached to the string used to fasten papyrus documents (Hallager 1996: 198–199). However, the only things that are known for certain about the function of the single-hole hanging nodules is that they were attached to string and that they must have been used for some specific purpose in the Minoan palatial administration. The majority of them were inscribed in Linear A, usually only with a single sign, which may have indicated the category to which whatever they were attached to belonged.
A type of sealing known as flat-based nodules has more certainly been regarded as evidence for lost types of texts. These are lumps of clay with one or more seal impressions; in some examples impressions on the base of the nodules show that they had been attached to pieces of very thin leather which had been tightly folded and tied with thin string (Figure 6). This has led to the suggestion that leather or even parchment was in common use for writing (Hallager 1996: 135–158; Weingarten 1983: 38–42; see also Chadwick 1976: 27–28; Krzyszkowska 2005: 156; Schoep 2006: 56, n. 2; Shelmerdine 2008: 12; Younger and Rehak 2008: 175). It is in fact hard to imagine what other than written documents the folded and sealed pieces of leather could have been.
Like papyrus, leather and parchment, wood has few chances of survival in the Greek soil. Evidence from later periods of Greek antiquity and from contemporary Egypt and the Near East shows that wood could be used quite extensively for different type of supports for writing. Wooden boards covered with stucco and textile, which could be written on in ink, were used in Egypt from the Old Kingdom onwards (Cribiore 1996: 65). For all we know, similar wooden boards could have been in common use on Crete. Numerous inscriptions on stone survive from Greece in the historical period, testifying to the use of writing in official contexts and public display. However, textual evidence indicates that large whitened wooden boards were also used for public notices in Athens (Thomas 1992: 83). None of these survives, but as pointed out by Rosalind Thomas, wooden boards rather than stone might in fact have been the main medium for official inscriptions. There is no reason why wooden boards could not have been used in the Bronze Age as well for a similar purpose. The possibility that monumental public inscriptions existed in the Bronze Age Aegean cannot be completely ruled out, despite the total absence of evidence. It can, however, be considered certain that stone was not used, as in that case one would have expected some fragments at least, if not entire texts, to have survived.
