A while back I read a book that started by telling its readers what it wasn't. Dunn's book needs the same disclaimer. This is really a book about understanding politics- it discusses, analyzes and judges different ways of understanding, not different kinds of politics. He begins by describing what he takes politics to be - essentially, conflicts within and between states - but most of the book is about how we might try to understand and change this subject matter. So the book is structured more often than not around a binary or ternary: should we be fatalists or voluntarists when it comes to understanding politics? Should be we understand politics from a Platonic or a sociological perspective? Is politics best understood with the tools of disinterested epistemology, or from the perspective of a particular political interest? Is politics best understood as brutal power plays, or as the realm of human cooperation?

If you can accept that, it's bracing stuff, until he gets around to suggesting, as he almost always does, that you take a bit from each member of his binary or ternary or quarternary, if that's a word. Which sometimes feels like a let-down.

There are other problems, too. Someone suggested that Dunn writes like Mozart. And that is true. But it doesn't mean it's clear and easy. It means that, like Mozart, it's excruciatingly difficult to follow and awesomely abstract, but sounds pleasant. Dunn has written about the early modern English political theorists a lot, and it shows. This is political theory in the *literary* tradition of Hobbes, not of utilitarian prose. It's like late Henry James. If you can't cope with that, you probably want to avoid this book. I actually quite like it.

The most serious problem is his vacillation over the importance of the State for political understanding. Against anarchists or libertarians or neo-liberals he wants to insist on the importance of the state for our political understanding; but as a clear-eyed social scientist he essentially admits that the moving force of political change over the last fifty years hasn't been the state at all: the state has been an effective tool, not the major force, for the spread of capitalist policies. I'm not sure that this is an irresolvable contradiction, but it's very hard to read his conclusion - that the creation of the modern democratic republic is humanity's most advantageous deed, and that all political understanding must start from it - without flicking back through to check whether you missed something in the previous 200 pages which were about how, basically, it's the economy, stupid.

Also, he's got the usual 'I used to be a socialist but then the USSR collapsed and now I know that socialism was always a fairy tale' pessimism that you find in older political thinkers. Newsflash! Our options are not capitalism or Stalinism. It just isn't true to say that "there is a clear surplus of conflict over co-operation in human interactions and... there will always continue to be so," 361. This might be true at certain levels of abstraction and at certain times; but in general, everyone gets on pretty well: I drive on the correct side of the road, I stop at stop-signs, I use my indicator. The exceptions and conflicts are more noticeable, but only because the cooperation is so all-pervasive. And that cooperation is reason for optimism that we can, in fact, improve our lot. ( )

Politics is inevitably disappointing. Why is this so? Politics is important and obscure and difficult? Must it be so? How can anyone even begin to understand politics? In fact, why bother to try to understand it at all? This book about politics, endeavours to answer all these questions. Politics shirks nothing, no aspect of political thought or theory. It explains first in the abstract (what is politics? etc.) and then makes this concrete, tying the ideas into a fascinating re-interpretation of Thatcher's Britain. Dunn shows how this lasted and then fell apart, in all its complexity. The focus then becomes more general, spanning ideas of state, judgment, corruption, democracy and its failings, economics, markets etc. etc. The final part is one of consolidation: what is political science; what are the implications of our and the world's current political situation and how can we use this knowledge to choose better?

The initial overview builds on abstract visions of rule and political understanding, and the collision between human purposes, drawing on Aristotle, Locke, Marx, Adam Smith, Max Weber and, most appreciatively, Thomas Hobbes. The middle eight, is a consideration of the significant political and economic shifts during the Thatcher years. He proposes that the British populous was more repulsed by Labour than attracted by the Tories; a case of omission rather than commission. Thatcher, a political "dominatrix", personally interpreted her electorate, and sought to communicate what seems with hindsight more a response than a considered ideology. The subsequent Just War to systematically refashion the economy to be internationally competitive saw economism far outstrip political advances, a disparity through selective radicalism also addressed in Larry Siedentop's Democracy in Europe. The later chapters drift around more general issues, centred on the capitalist legacy of recent history, "a low dishonest quarter of a century" according to Dunn. If Harold Wilson's week was a long time in politics, this era of "globalisation" has been an eternity. But politics can still surprise. Demonstrations of public protest can wrest back power from those who may have lost sight of their elective mission.