The topic of this article is the legal philosophical foundation of John Griffiths’s sociology of law. Griffiths has developed his foundation of sociology of law in discussion with three positions: legal realism, Hart and Dworkin. These three positions give three different answers on the question ‘what is law?’. In the first part Griffiths’s discussion of legal realism is analyzed. From the outset, a legal realistic approach to law has the benefit of its strong focus on the empirical determinants of predicting the outcomes of cases. Problematic, according to Griffiths, is a naïve instrumentalism, often related to legal realism. The second part on Hart’s theory discussed Hart’s notion of rule-following as the core of Griffiths’s sociology of law. Also the different perspectives on law are discussed. According to Griffiths, Black’s extreme external perspective is problematic, but Hart’s moderate external perspective is also not suitable for the external comparative purpose of sociology of law. In the third part, Dworkin’s theory is discussed. Griffiths, in my opinion, unsuccessfully, tried to reconcile Dworkin’s theory with legal positivism. Dworkin’s theory is an interpretive theory from the participant’s point of view, which makes it hard to use it as an adequate foundation of an empirical theory of law. For a sociologist of law, choosing an adequate conception of law is just as important as the choice for an empirical method. The contribution of Griffiths to sociology of law is in this sense unique and of great value for the sociology of law.

This contribution focuses on John Griffiths’ relentless attempt at developing a general theoretical perspective for socio-legal studies. Hence, attention to Griffiths’ important contributions to legal pluralism and the social working of law approach is paid only in passing. Similar to a much earlier assessment, the analysis of Griffiths’ proposal in this contribution is quite critical. Measured against five criteria this author deems important for any socio-legal theoretical framework, the verdict is that Griffiths’ proposal falls short of all of them. The analysis itself focuses primarily on Griffiths’ attempt to redefine the subject for socio-legal studies in terms of social control, the way he uses the concept ‘law’, and his primary focus on rules and rule following. One overall conclusion is that Griffiths remained a legal scholar to a much greater extent than he would have liked.

A central ambition that Griffiths expressed rather frequently was to realize progress in the sociology of law by formulating informative theoretical propositions and testing them empirically according to the maxim of the critical-rational metatheoretical program of Karl Popper. Our analysis of Griffiths’s contributions suggests, however, that he actually refrained from following Popper’s path: to put a Problem – formulate a Theory – testing that provisional answer by empirical Observation. Instead, Griffiths focussed mostly on the rigorously clear formulation of concepts accordingly to his strong philosophical inclination.

Similar to the more current Regulator Intermediary Target (RIT) model, Griffiths’ Social Working (SW) theory points to the relevance of intermediaries for explaining rule following behavior. In this article, the author applies both theories (RIT and SW) concerning the role of intermediaries in rule following to explain developments in Dutch prostitution policy: the non-implementation of the emancipatory, sex workers’ rights based approach, and its replacement by a more repressive policy of closing down sex facilities. The analysis shows that although both theories contain useful starting point for explaining these developments, the SW theory’s special value is its acknowledgement of how regulatory intermediaries operate in a social field with existing social rules and a specific balance of power. Such rules and power relations have put barriers to the implementation of the Dutch prostitution policy as formulated in 1999. As illustrated in the article, the SW- theory offers more tools than the RIT- model for an analysis of how legal rules work in practice.

John Griffiths’ social working approach of legislation tries to estimate the direct effects of laws which prescribe certain behavior. The basic idea of the approach is that rule-guided behavior (direct effect) is influenced by the different groups citizens belong to. Griffiths refers to these groups using the concept coined by Sally Moore (1971) ‘semi-autonomous social fields’. Although Griffiths never formulated hypotheses regarding the relation between SASFs and direct effects, the article explores two of them: If the relevant SASFs accept the new norm, direct effects will occur; and if the relevant SASFs are not ‘though’ (and don’t accept the new norm) direct effects will occur. These two hypotheses are related to the results of smoking bans in bars in the Netherlands. The acceptance of the smoking bans in bars is low. The thoughness of the SASFs in bars and their organization differ in time and so did the compliance with the smoking bans. Because this article is not based on research that depart from the hypotheses, further research based on the hypotheses is needed to draw firm conclusions. The article is rounded up with a plea to use Griffiths approach in impact assessments of legislation.