In Wittgenstein's “Language Games” learning a language means learning how to exist inside differing language games, in which words are used in differing ways. Wittgenstein holds that the observer must simply clarify the odd bit of language usage and hand over for further investigations to the sociologist. Morrison notes that this paradox is simple: if the activities of individuals only gain sense through participation in social practices then we have to look at the latter as the sources of our explanations and therefore we have to continually contextualise both what we are investigating and our own position. Stating that rules such as law can only be understood by embedding them in society - as Hart, following Wittgenstein, claims - does not offer an answer about their origins. This finding leaves Hart with terrible flaw in his theory. He tries to amend with this possibility of infinite regress by recourse to an ultimate social practice of recognising law which he calls the rule of recognition. He cannot, however, tell us how to find the rule of recognition; instead it is left as a presumed social practice. Hart also wishes to keep to a legal positive definition of law to save our capacity for moral analysis. Therefore it is open to Hart to say that his rule of recognition is simply what he will use as the basis of his description of the legal system; but he must then accept that this leaves open and unanswered what constitutes the actual nature of the rule of recognition. Thus different societies may have different rules of recognition; the problem then is that such an expression as the rule of recognition needs to have some limitations imposed upon it, otherwise it cannot serve the role it plays in Hart’s theory.
Hart is giving an account of how we publicly justify the recognition of the legal rules; it does not necessarily follow that this is how privately we (or even the judges) identify them. We may, indeed, pose the question of...