Terrace systems are one of the most characteristic features of the modern landscape of southern Greece. The question of whether this seductive lattice-work of dry-stone walls can be projected back into antiquity has intrigued scholars for some time, and in recent years discussion has become particularly lively.1 Here I shall argue that the terrace systems now visible are modern (i.e. probably post-medieval) for the most part. The ancient systems of cultiva­ tion in Greece (and Italy) for which we have a reasonable amount of information largely used other means of keeping soil on hill­ sides. For farming, as for most other aspects of Greek and Roman life, our sources emerge from the wealthiest sector of society. It is virtually impossible to ascertain securely how peasants attempted to solve the problems of soil movement and erosion, even if we can sometimes make plausible guesses. This is not to say that terracing was not practised in antiquity; on the contrary, I am sure it was. What I am suggesting is that we cannot now ‘see’ ancient agricultural terraces, for two reasons: (1) terracing is virtually invisible in the documentary sources; more importantly,

(2) the landscape itself has been constantly worked and reworked since antiquity, so that any ancient terracing systems have been either obliterated or incorporated into modern terracing in such a way as to render them undatable. I remain unconvinced by any claimed finds of agricultural terraces dating back to classical anti­ quity or beyond.