There are, however, some cases where the nature of the surface surely did lead to particular usages, over and above the matter of rounded pen / brush strokes already tackled above. A number of letters in various Greek and Etruscan scripts have rounded elements; in some they are in a sense secondary, beta and rho (В, Ρ) can have rounded loops, and in some scripts gamma and delta (Γ and Δ) also can have substantial rounded sections. But in all these we find angled alternatives in regular use, and to date I have detected no patterns emerging, either locally or in various media, but would wish to conduct a fuller review. Under the same heading, however, we can note a vogue to the use of squared-off circular parts of letters where more than half a circle is involved — theta, omicron, phi, omega — on bronze and stone, particularly in some areas of the Peloponnese. Polykleitos uses such forms in his signatures on statue bases (Figure 15; but this is probably not the reason why Varro (in Pliny, Natural History, xxxiv 56) describes his statues as “four-square”!). On such hard surfaces the inscribing of circular letters took three main forms, beyond that of doing one’s best with driving a chisel: turning the curve into straight lines, using individual points to make up the curve, or using some form of punch or compass, the latter often leaving a central point (Figure 16).3 Here we can at least make the observation that such usages are at best extremely rare in the painted letters of Greek vase inscriptions; I know of none, but stand to be corrected.
Let us turn back to sigma — or the three- or four-stroke sign which does duty for the main sibilant in many Greek poleis or iota in a minority. We have already noted that it can have a curving profile, whatever the medium. I just mention a few more cases, where in fact it is the sibilant that is so shown and is of three strokes. It is quite regularly used in texts inscribed into stone and metal of Rhegion in South Italy (Figure 17), but rarer elsewhere — here a most peculiar case (British Museum GR 1888.6-1.421; Figure 18), an exception that proves the rule if there ever was one — among the 2500 graffiti from Naukratis. What I discussed earlier with respect to later developments in the 4th century bc is of a different character: the so-called lunate sigma, and its cousin the lunate epsilon, quite clearly results from the rapid brush writing of the four-stroke sigma that became the major simple sibilant sign in the Greek alphabet by the late 5th century. We can see some hints of this change in a few painted inscriptions on Attic white-ground lekythoi of the middle of that century, with a minimising of the ‘heart’ of the letter (Immerwahr 1990: 158, S14; Figure 19). Here it is of relevance to note that there is no clear, immediate successor to that set, and just one possible predecessor, a truly unusual text: among the cache of baked clay tablets found in the ruins of Persepolis in ancient Media termed the Fortification tablets, there is one written in Greek (Figure 20) and the date must be somewhere around 500 bc — we have no need to debate the closer issue here (Lewis 1977: 12–13, n. 55). It is a short note of the disbursement of an amount of wine, and the sigma appears in both forms, four-bar and, on a fairly sharply curving surface, in a clear lunate form. It is indeed an oddity that is not readily explained, as Lewis notes; can it be a single surviving example of what may have been a form in widespread use on perishable media, in Media?
By way of concluding remarks, we can discern in general a very slow adoption of ‘cursive’ shapes in informal writing in the classical period, with occasional but very scattered examples of such lettering in its broadest sense in earlier periods. It seems reasonable to argue that the formal scripts of the Greek poleis in the 5th and 4th centuries bc may have acted as a brake on such change, the stoichedon system being perhaps the most effective of such devices. At the same time it is proper to look at other aspects of contemporary material culture to see whether similar forms of what we may call conservatism are apparent there. It is such a large topic that one must be selective and so cautious in drawing conclusions that are too sweeping. Modern students of the ancient world perhaps look too closely for change and innovation, overlooking such conservatism. It is mentioned most frequently, I would suggest, in the context of dating; ‘such a type is long-lasting and we should not press the evidence’ is not an uncommon sentence, though perhaps not so much used of the period we have under scrutiny, when certainly artistic change continued apace.
Two areas where conservatism can be seen however are religion and trade. While it is clear that the major polis cults did enjoy new developments, especially in monumentality of the architectural environment, at a more humble level it has been argued that the increase in quantities of small mould-made terracottas in many sanctuaries in the classical period reflects a concern for stability, not change, in cult and its attendant rituals (von Hesberg 2007). On what would appear to be a completely different level we may note the relationship, or lack of relationship, between vase shapes and their suitability for stacking on board ship, i.e. ‘progressive’ cooperation between potter and trader. It is a topic that needs greater attention than can be given here, but I just note two aspects that one might consider counter-intuitive. The shapes of storage or transport amphoras respond little to the requirements for easy packing until, in crude terms, the 3rd century bc (Johnston 1984). One type, from Mende (Figure 21), oddly enough one that is mentioned by Demosthenes in a court speech involving shipments, takes on even more ‘aesthetic’ shapes in the 4th century. Here it could be said to reflect the trend in shape development in Attic red-figured ware, to what one might fairly call Victorian values, with ever slimmer stems and flowering, fragile rims; yet the late red-figured krater is exported very widely, from Spain to the Black Sea. One might argue exceptions, for example the solid Castulo cup of the 5th century bc, but also remember the elegant stemmed kylix, in near mass production from c.550 until the 4th century bc (Shefton 1996; Figure 22). Not only transport, but kiln furniture is affected (see the highly intricate kiln supports required, and made, for the better firing of South Italian equivalents of later Attic red-figured pots at Metaponto [Cracolici 2003]).
In writing there are similar tensions between aesthetic concerns and practicality, and they receive similarly mixed solutions. An informal scribble on a sherd, of c.475–450 bc from the Agora at Athens demonstrates this wonderfully (Lang 1976: C21; Figure 23), with its scratched sketch of an official stone text, complete with grid for the lettering and the heading invoking the gods, albeit the rest of the text is far more in keeping with contemporary painted names on pots. Or one can cite (Figure 24) the use of truly monumental lettering, of a lapidary kind, in some ceramic texts, where smaller size and normally hastier script are otherwise found, all I think dedications to deities (Johnston 1997: 109–111; we can add to my list published there splendid unpublished examples on Panathenaic amphora(s) dedicated in the Athena sanctuary at Kamiros on Rhodes).
I finish with a stone text to demonstrate a final aspect that has consequences for our topic. It is a ‘speaking object’, though in fact a counter-example but an interesting one. A tombstone of c.540 bc, with a text inscribed such that it rises up from the ground, as if the dead were speaking (Jeffery 1990: pl. 73, 1); very often indeed the epitaphs do treat the stele in the first person, and we find the usage very widely spread in the Greek and Etruscan world from the 7th century. Here however, rather perversely it is the unknown reader of the text who is in the first person “I am in pain looking on the tombstone of young Autokleides” (Figure 25). The material effect of the words rising from the ground is inexplicably lost.
A broad range of related texts that offer names, but of a more casual nature are the identificatory labels that painters and sometimes gem-cutters and the like put on their figured scenes. From the beginning, around 700 bc on present evidence, painters tried whenever possible to begin the label as near to the head of the relevant figure as possible, as if he or she were speaking their name. I noted earlier that something approaching the cursive sigma can be found on painted inscriptions on mid 5th-century pots, but they are rare and isolated occurrences within a group which is noteworthy however for another aspect most of them share, in that they reflect the new aesthetic of the squared off stoichedon style, with labels very often related more to the surface than the figures, and there are hints that such usage may have been derived from large-scale wall or panel painting (Figure 19).
In sum we may see in the development of script in the classical period a similar neglect of ‘pure’ technology as in other areas of everyday life, as against the remarkable intellectual input into epistemology and the arts, as we would call them, not least in the words written in the seemingly still conservative scripts. The fact that many of those words were publicly uttered in somewhat basic theatrical surroundings, before the rise of the monumental theatre in the 4th century, is a further example of such social attitudes.
3 I note here an example of surface being of importance to modern scholarship: those who advocate an early date for the transfer of the alphabet from the Levant to the Greek world (see Jeffery 1990: 426–427 for an overview) have argued that the Greek omicron with a central dot is a clear echo of the Semitic ayin, ‘eye’, which had lost that dotted iris by the 1st millennium bc. They are of course no such thing; all early omicrons, to my knowledge, do without it also.