His key example here is ancient China, which, by the first century BC, seems to have developed a very efficient and "modern" state structure in which highly trained bureaucrats held office on merit, and were able to impose the will of the emperor over a huge area. In spite of various episodes of decay or collapse, this "strong state" model was powerful enough in the Chinese case to prevent the formation of those "strong society" elements (such as autonomous religious organisations, or other forms of association within the state) that could have supplied the other ingredients of modern accountable government; and so China was held back from modernity by being too modern, in one way, too soon.

Fukuyama does not believe in uniform historical processes, and is not a determinist, or at least not an economic one. (There are touches here and there of biological determinism-claiming that humans are programmed for sociability, or violence, or status-seeking-but these are just invoked as background conditions, and play little direct role in the historical argument.) One of his recurrent themes is the importance of ideas, and especially of religious beliefs. India, for example, is described as having had exactly the opposite trajectory from that of China, with a strong society and a weak state, for reasons deriving entirely from the peculiar nature of Brahmanism: it was this belief-system that ossified society into a set of immobile occupational groups, and limited the military caste in such a way as to prevent rulers from raising large armies for state-building purposes.

The oddest feature of Fukuyama's argument is his invocation of the "rule of law". What we normally mean by this is a system in which laws are stated in general terms (i.e. not as ad hoc expressions of a ruler's will) and enforced impartially (i.e. not in ways that vary according to the status of the individual subject). What Fukuyama means by it is that there is some other set of norms or values that stands above the law-making of the state, validating those laws, or, in some cases, overruling them.

Sometimes he seems to have in mind the very specific (and very American) political idea of constitutionalism — in which case, when Britain had a legal doctrine of "parliamentary sovereignty" à la Dicey, Britain was like the worst sort of despotism, a state utterly without the rule of law. But at other times he accepts almost any belief in a higher set of norms as satisfying this condition, so that a primitive tribe which appoints a new chief on its shaman's say — so is, by placing the authority of the spirit world above mere political authority, asserting a commitment to the "rule of law", that essential ingredient of modern statehood. Clearly, something has gone wrong with the argument here.

With Fukuyama's third component, accountability, we come much closer to familiar territory. So much work has been done already on the development of democratic institutions in Western Europe that he can do little more than offer a summary of it. And this, despite his disavowal of Whig history, has a very Whiggish flavour to it, with its strong focus on the English "Glorious Revolution" of 1688: Macaulay himself might have blushed to write that William  of Orange was simply "brought from Holland and placed on the throne".

Fukuyama may be no more a historian than he is a philosopher or a scholar; but he has a talent for framing big questions and summarising large debates, and there is an honest seriousness about his whole approach that commands respect. The answers he gives are sometimes unconvincing, but never simply glib. So, while the book may be less than the sum of its parts, the experience of reading it is often invigorating, sometimes intriguing, and very seldom dull. And of how many other political scientists could that truthfully be said?