The idea of ‘militarism’ has at best an uncertain status in contemporary international studies. The concept is academically marginal: strategists, and with them many historians, eschew the term because of its associations with political opposition to military power as such. Moreover, even those who recognize its histo rical relevance may doubt its contemporary applicability: thus Volker Berghahn and Hugh Bicheno (Berghahn and Bicheno, 2009) suggest that ‘[w]ith the end of the Cold War, the concept of militarism has lost most of the ideological steam that seemed to make it worth discussing. It is rarely used to describe present-day systems and policies but instead is seen as a phenomenon of the past to be examined with the tools of the historian.’ Contrary to this judgement, the idea has undergone a revival in the last decade, but (potentially reinforcing the criticism that it is a political rather than scientific concept) in the hands of critics of the global war on terror. For Michael Mann (2003) and Andrew Bacevich (2004), US militarism involves excessive reliance on military power, out of kilter with its more limited economic, political and ideological capabilities. Although polemically framed, these are serious analyses (Mann’s derives from his major study of power (1986, 1993), which involves probably the most extensive analysis of ‘militarism’ in recent scholarship) which have demonstrated anew the concept’s potency. Thus criticisms of the very idea of militarism are as misguided as some of the simpler polemical uses of the term, because the concept refers to enduring and important sociological realities. If ‘militarism’ has done little serious analytical work in International Relations (IR), this suggests a failure of the field to address some important questions. The most common usage of the term, proposed by Vagts (1973: 11), refers to the state of society that ‘ranks military institutions and ways above the prevailing attitudes of civilian life and carries the military mentality into the civilian sphere’; this idea is frequently summarized as ‘glorifying’ military power. Mann (2003: 16-17) defines it more broadly as ‘a set of attitudes and social practices which regards war and the preparation for war as a normal and desirable social activity’. It is sociologically unenlightening to restrict the meaning to ideology: the core idea is the ‘carrying’ of military forms into the civilian sphere, and this is not merely a matter of ‘mentality’ or ‘attitudes’ but (as Mann notes) of ‘social practices’. Moreover, the military forms which are carried may not necessarily

‘rank military institutions and ways above the prevailing attitudes of civilian life’, let alone ‘glorify’ war in a simple sense. While the differences between ideologies which glorify war and those which do not are significant, glorifying and non-glorifying ideologies may equally justify war and military power, and in this sense have the same core social function. Therefore we need a common term for all such ideologies, which we may distinguish by qualifying adjectives (e.g. fascist militarism, democratic militarism). Furthermore, a narrow ideological definition separates ideology from the social relations of which it is a part. The core meaning of ‘militarism’ should be specified not in terms of how military practices are regarded, but how they influence social relations in general. Militarism develops not just when ideas of war are strong, but when military relations widely affect social relations and practices. Hence I have proposed (Shaw, 1991: 9-15) that militarism denotes the penetration of social relations in general by military relations; in militarization, militarism is extended, in demilitarization, it contracts. Another advantage of this approach is that it pre-empts the political abuse to which the ideological definition easily lends itself. Invariably, ‘their’ warmaking and war-preparation is aggressive, destructive, glorifies war and is ‘militarist’, ‘ours’ is defensive, humanitarian, does not glorify war and is not ‘militarist’. The deeper sociological definition removes the simple negative connotation and in principle allows that arguments justifying militarism and militarization may be plausible, although it retains critical potential. It is possible to use it in a coherent social-scientific manner, in ways which tie it neither to a particular political critique nor to the analysis of the past.