‘JUST AS IN THE case of the paintings of olden days, unless they were inscribed, one did not know what each thing was.’ Whatever the ‘old paintings’ that Aristotle had in mind, the reference is certainly to large-scale work; and we know from Pausanias’ detailed descriptions that the practice of inscribing still prevailed when Polygnotos and Mikon were executing their famous murals, something over a hundred years before Aristotle's time of writing. In fourth-century parlance, their work could perhaps already be counted as ‘ancient’. The analogy that Aristotle is making is with definitions which are insufficiently precise and exclusive to do their job effectively. We note that this purports to be a statement of fact rather than an inference: literally, ‘it used not to become known’ what the paintings showed. Whether or not we believe that Aristotle was right about this, it is at least clear that he regarded the practice as obsolete and no longer necessary in his own times. What is more, when we turn our attention to vase-painting, we shall find evidence to support this temporal distinction in general terms.

The interplay of image and word had long been ubiquitous in the culture of ancient Greece. But there are very few places where the two come so close together as in the painted inscriptions on Greek vases: indeed, inasmuch as the inscriptions at times seem to be located with a view of filling gaps in the figure-scenes, the word can actually become a part of the image. This was a phenomenon that had a fairly rapid growth, then a pronounced peak, then a steady decline. Even at its peak, in the high archaic period, it was a minority practice among vase-painters; yet it was widespread enough, within and beyond Athens, to pass with little extended comment from scholars nowadays. By the full archaic period, it had become rather rare and by the fourth century, as Aristotle's parallel suggests, even more so.

In 1990 appeared Henry Immerwahr's long-awaited Attic Script: A Survey. The title of the book hardly conveys the fact that vase-inscriptions heavily preponderate in its content, though it fairly represents the treatment that follows, which is epigraphical first and last.