Over the last twenty years, Political Scientists have increasingly turned to the concept of ‘authoritarianism’ to help explain the unexpected tenacity of non-democratic regimes in the Middle East. Rather than examining the causal factors which putatively resulted in the absence of Arab democracy, this recent literature has instead sought to identify the mechanisms by which existing regimes have ensured their survival. In the sub-field of Comparative Politics, scholars have studied the electoral manipulations, judicial biases and coercive tactics employed by the state for its own ends (Brownlee 2007; Bellin 2004; Posusney and Angrist 2005). In Historical Sociology, attention has focused on the social pacts, class alliances and economic formations which underpin authoritarian modes of governance (Hinnebusch 2006; Heydeman 1999; Pratt 2007). Despite the advances of recent work, this literature’s reluctance to scrutinize the normative understandings and ontological assumptions embedded in the term ‘authoritarianism’ has limited its ability to interrogate fundamental questions about authoritarian politics. This paper argues that the shortcomings of existing conceptualizations of authoritarianism are implicitly derived from the underlying models of social science upon which recent work in these two sub-fields is based. The paper proposes that a third model – that of interpretive social science – can generate substantive new insights into the workings of authoritarian politics in the Middle East. The paper identifies three widespread assumptions in the existing literature: (1) an essentially liberal definition of authoritarianism; (2) an exclusively macro-level, elitist and institutionalist perspective on politics; (3) a predominantly voluntaristic account of the structure/agency debate. After outlining the basis of interpretive (as opposed to empiricist or critical realist) social science (Yanow and Schwartz-Shea 2006), the paper explains how interpretive approaches reformulate these three assumptions, namely by: (1) revealing the normative foundations and power relations embedded in the term ‘authoritarianism; (2) refocusing attention on informal micro-politics without losing the link to macro-level structures; and (3) reformulating the structure/agency problematic in terms of narrative explanation. The paper concludes that not only do interpretive methods have the potential to gather richer empirical data about authoritarian politics than empiricist or realist approaches, but that the interpretive model of social science can draw attention to key theoretical and substantive questions about Middle Eastern authoritarianism that are underexplored in the existing literature.