My article was a critique of what I termed associational thinking, illustrated by reference to examples concerning markets, bureaucracy and gender. I counterposed associational thinking, with its suspicion of abstraction, to counterfactual thinking, which is concerned not with what things happen to be associated together but with which social objects necessarily depend on one another and which do not. While it is clearly vitally important to know how social objects are associated with one another, and what consequences flow from this, it is also important – for practical and political, as well as academic, reasons – to know whether they have to be associated thus, or whether things could be otherwise. Thus, in response to phenomena which are widely found in association, such as capitalism and various forms of gender order, the counterfactual questions would be: is each a necessary condition of the existence of the other, or could one exist without the other? As a second theme of the paper, I argued that a modified version of the distinction between system and lifeworld was useful in understanding the particular associations under discussion.