Here we have an awkward chronological clash regarding the subject of cursivity between the literary references and preserved texts, involving the period c.350 to 300 bc. There are though some even earlier, if sporadic, stirrings. One learns a lot from writers who have made, or think they have made ‘errors’ — and this is not the place to enter on the thorny question of defining orthography; there is room for a doctoral thesis on the topic if only any classical archaeologist-cum-epigrapher would take on such a banal task. First, I repeat a point I have made regarding a much earlier text (Johnston and Jones 1978: 104–105), on an Athenian or Attic amphora of 625–600 bc on display in the British Museum (Figure 7) — where the cutter of an owner’s graffito started incising, correctly, an omicron, but finished it off as a sigma, which should in fact have been the following letter; he confused himself, but only because he could mistake a rounded arc as the top of a sigma. One could have wished for some such happenstance in the earliest long text from the Mediterranean Iron Age, the much discussed ceramic ‘Nestor’s cup’ from Pithekoussai of c.720 bc (Jeffery 1990: pl. 47, 1; Figure 8); but it does not quite happen — an omicron in the first line is corrected to an epsilon in a somewhat ugly manner, and in the second line an omitted nu is rather more deftly inserted just below its proper place and an epsilon was half cut before the inscriber realised the letter should be alpha; but at least it all demonstrates a concern for what was perceived as accuracy.
Here we are getting back to near the origins of Greek alphabetic writing; for whatever reason and at whatever precise period around 900–800 bc, some Greek-speakers geometricised the Semitic alphabet, fitting the somewhat casual angles of the Semitic signs to the visual human-made representations on artefacts prevailing in the contemporary Greek world — patterns involving straight lines, regularly at 90o or 45o, and circles — the Geometric style. Where that particular style was weakest, on Crete, where far more luxuriant and inventive pictorial designs were common enough, we find the weakest such adaptations of the Semitic letter forms. For example, Crete is one of the areas that retains some of the complexity of the Semitic yod in its iota — unlike the majority of the rest who boldly adopt a simple vertical stroke. Such a curly iota is found in a 7th-century graffito on a pot from Knossos (Johnston 1996), for example. In contrast, on the island of Thera, which is clearly dependent on Crete for its alphabet, the iota appears not only in painted texts — most strikingly a ‘doll’s-house’ (Jeffery 1990: 470, A and pl. 79) and the unpublished example of the deceased’s name painted on the foot of a similar Athenian oil amphora used as a grave marker, both c.600 bc. It is also important to note examples occuring in rock-cut graffiti. Occasionally on these we find the same letters made of either curving or straight strokes, or indeed single letters employing both, in the same text (e.g. Inglese 2008: 469, 473; Figure 9).
This form of mixed usage continues. Curving lines are cut on stone, and on pots, straight lines are used in painted texts, with little particular pattern of usage that I can observe. The overall framework, however, always remains the geometricised set of signs of the period of origin, as noted at the beginning of the previous paragraph.