The primary aim of this chapter is to investigate the use of analogies, images and models in ethical discourse, their modalities and their justification. How far can they, or should they, be replaced by arguments that do not suffer from the shortcomings of analogies? I shall concentrate on two of the richest sources of analogical reasoning in ethics, from the two ancient societies that are the usual focus of my investigation, Greece and China. How are analogies used, which types are privileged, and how were they evaluated by those who employed them or who criticised their use? For that last purpose we shall need to consider more broadly the ideals of reasoning that were proposed. Greece and China evidently represent only a small section of the possible subject-matter, but they offer opportunities to assess different approaches and ambitions, and they enable, in particular, some typical Western preoccupations to be placed in perspective, thereby contributing to my strategic aim in these studies.

Arguably, in ethics, as in most other fields of inquiry, analogies, models and images provide one of the commonest modes of argument to recommend a point of view. No one will deny that they are extremely pervasive, both as explicit comparisons and as implicit ones. The latter are usually labelled metaphors – as in Lakoff and Johnson's classic study (1980) The Metaphors We Live By – although I have expressed my reservations about the use of that term, not least as it tends to be employed to downgrade a certain kind of language use by contrast with a ‘literal’ or ‘strict’ acceptance.

But if my general point about pervasiveness is accepted, there is nothing surprising in the fact that analogies and images are so common in early Chinese and Greco-Roman ethics in particular. The interesting questions relate not to the fact that there are lots of comparisons of different types, but rather to which ones were favoured (my first-order question), and secondly, and more importantly, to how they were deployed, for example how self-consciously and how critically (my second-order one).

If we ask the first, simpler, question of which analogies, models and images are privileged, I have first to say that to review these at all thoroughly would require a monograph or two, so I must be highly selective. We may make a start by referring to some of the recurrent cosmological images found, for each of the main types carries ethical and more particularly political implications. At first glance at least, the similarities between two of the main groups of such images in China and in the Greco-Roman world are very striking. Both ancient civilisations often picture the cosmos as a state, both often also do so as a living being (the third common Greco-Roman image of the cosmos as an artefact is less often found in China). The kind of state used as a model, monarchic, oligarchic, democratic, even anarchic, differs as between different Greek and Roman writers, while the Chinese generally think in terms of the monarchic political system that was always their ideal.