Epicurus thought of justice as an agreement to prevent people harming and being harmed.

Morality has in the past made progress when we broadened the category of things we weren't permitted to harm (animals, 'infidels'); saw through some delusions and rationalisations about what harms are good for people themselves (prison punishment, hysterectomies for unhappy 1950s wives); and readjusted our for-the-good of others criteria so as to demand only reasonable sacrifices (ceasing to use children as handy chimney sweeps).

I had the idea that there were secret laws of the universe that could explain the baffling human reality around me, and that philosophers maybe had the key to them.

We are now returning to the 18th century empirical approach with the new interest in the evolutionary basis of ethics, with 'experimental' moral philosophy and moral psychology. As a result, we understand better why moral formulas are experienced as ineluctable commands, even if there is no commander and even if the notion of an inescapable obligation is just superstition. So moral philosophy has made huge progress.

For the chemists, who wanted to manufacture new medicines and elixirs and transform base substances into noble ones, the notion that there was no metaphysical barrier to doing so - it was just a matter of getting the particles into new arrangements - was encouraging. That was the Baconian programme.

You can reasonably make the intellectual journey from thinking it's permissible to eat shrimp to thinking it's not permissible, or vice versa, whereas our slavery journey was uni-directional. We are as certain we are not going back to that old kind of slavery as we are that we aren't going back to the geocentric universe.

Epicureanism did inspire libertine culture in isolated sects, but Epicurus himself rejected an ethics of sensory indulgence, and he would have disowned latter-day 'Epicureanism' as a fussy, expensive, unphilosophical approach to eating and drinking.