Austin’s theory of theory of law is simple: the law follows the pattern of power; the sovereign gives commands and obeys none; the subject obeys commands; the law consists in only those commands that directly or indirectly emanate from the sovereign. Nevertheless, Austin’s theory of sovereignty is not simple at all. When we look at the relevant chapters closely, it becomes evident that Austin has two rival theories of sovereignty, one for a single person and one for a ‘determinate body’. It is only the latter that allows him to say that sovereignty lies, ultimately, with the electors, the strange conclusion of his book. But Austin’s second theory of sovereignty is not consistent with his own theory of law. Austin’s faces a dilemma. Is law - as most people take it to be - a public order of standards of conduct aiming to guide behaviour? If so, sovereignty ought to be public and intelligible. If not, sovereignty can remain a mystery to those living under it (accessible only after the event by the expert legal philosopher). For the latter reading, law and sovereignty are ‘normatively inert,’ as some of Austin’s followers claim today. But Austin does not agree with his modern followers. Austin’s second theory of sovereignty is aimed at satisfying a practical requirement of law and jurisprudence, i.e. to be in the position of publicly guiding conduct.

Many contemporary practitioners of analytic jurisprudence take their understanding of legal positivism largely from Hart, and the debates about legal positivism exist largely in a post-Hartian world. But if we examine carefully the writings and motivations of Bentham and even Austin, we will discover that there are good historical grounds for treating both a normative version of positivism and a version more focused on legal decision-making as entitled to at least co-equal claims on the positivist tradition. And even if we conceive of the inquiry in philosophical and not historical terms, there are reasons to doubt the view that a theory of the nature of law is the exclusive understanding of the core commitment of legal positivism. Positivism as a descriptive theory of the nature of law is important, but so too is positivism as a normative theory about the preferable attitude of society or theorists, and so too is positivism as a normative or descriptive theory of adjudication and other forms of legal decision-making. Those who understand positivism and the positivist tradition as being more normative or more adjudication-focused than the contemporary understanding allows are thus committing neither historical or philosophical mistakes, and little would be lost were we to recognize the multiple important contemporary manifestations of the legal positivist tradition.

Hans Kelsen’s critique of John Austin has so far attracted little attention among legal theorists. This article argues that Kelsen’s attack on Austin anticipated the key elements of Hart’s rejection of the Austinian conception of law as sanction-backed sovereign command. At the same time, the way in which Kelsen presents his critique of Austin’s conception of sovereignty reveals important differences in purpose and intention between Kelsen’s Pure Theory of Law and Hart’s legal theory. The Pure Theory of Law is animated by an ideal of legality that is alien to purely descriptive jurisprudential approaches in the Hartian tradition.

The article concludes that this difference between Kelsen and Hart merits further exploration and that it might help to show that the Pure Theory of Law is still relevant to contemporary legal theory.