Finch and Mason (1990) have recently referred to various articles, by myself and others, concerning the hierarchy of obligations among network members, in relation to decision-making about the provision of practical help to elderly people. In their fascinating and informative case study, Finch and Mason argue that this hierarchy is best viewed as a normative belief which influences negotiation within families, rather than as a way of predicting people's behaviour in practice. I would argue that the question of the relationship between ‘statistical’ norms of behaviour (i.e. what most people do), and normative beliefs (i.e. what people think is the right thing to do), is a question open to empirical investigation. Certainly, it would be a misuse of case study methodology to suggest that failure to adhere to the hierarchy in one particular instance could be advanced as proof that it did not reflect a norm of behaviour. Statistical norms are never deterministic in individual cases, but it may still be possible to predict what the majority of people will do.