Kant's Politics in Context by Riedar Maliks (Oxford University Press, 2014) might on its own terms seem something of a philosophical oxymoron. For as Maliks reminds us constantly, Kant himself saw his political reflections as deriving from pure reason. In opposition to conservative political thinkers, such as Burke and, more remotely, Hume, Kant did not want to be tainted (as he would see it) with any hint of a derivation of political principles from convention or tradition. Experience, Kant says, is ‘the mother of illusion’. In line with his thorough-going (critical) rationalism, reliance on experience and tradition would leave him open to the charge that what he was proposing was merely a product of changing historical contexts. Indeed towards the end of his career, in the 1790s, Kant deployed a distinction between the ideal respublica noumenon – not, he insists, a ‘empty figment of the brain, but rather the eternal norm for all civil organisation in general’ – and the respublica phaenomenon, the merely historical animal under which we actually live.

Under which ‘we’ actually live? It must be admitted that there is a slight degree of unclarity here. For Kant is adamant that in general any state, with any hint of legitimacy, is better than anarchy, for human dignity requires the external freedom which is possible only under law. Even though all states have their origins in what Maliks calls ‘violent power grabs’, they may, in ‘Malik's words, ‘work themselves clean through history’. We don't, in Kant's view, have to look for any actual contract, Magna Carta, Glorious Revolution, 1776, or whatever, to see that a state is legitimate: we have only to look and see whether a rationally grounded populace would consent to whatever the situation is as guaranteeing enough external freedom to allow for the exercise of (some?) human dignity. If it is, that is enough to work on, and to obey, even while arguing for reform where this is necessary. (Kant was all for argument as opposed to violent action.) And, robustly and sympathetically, Kant is also adamant (against romantics such as Herder) that the state should not be seen as a device for producing happiness or other perfections. These are private matters, on which people disagree, and should be allowed autonomously to work out their own paths, subject to the laws and freedoms given them by the state in which they live.

Kant disappointed his more revolutionary followers by arguing strongly against revolution. No doubt 1789 and the events following may have played a part here, but this stance was not inconsistent with his basic view of the state as, once constituted, something we have a duty to obey, and his anti-perfectionism. Also, between the lines, as Malik shows, he does permit the removal of self-appointed illegitimate despots, on the grounds that certain set-ups would be so anarchic as not to amount to a state at all. In such circumstances (anarchy) there would be an obligation on the part of the population to create a genuine civil condition, as Malik puts it, one which we would be obliged to recognise.

Here, though, we reach the crux of Kant's politics. In Malik's succinct exposition of the problem, for Kant ‘the state cannot be created out of anarchy by lawful means. In order to lawfully found a state, a community must agree through the general will. A binding general will, however, presupposes a juridical condition that only exists post-founding’. This is, of course, precisely the objection of Burke and the conservatives to any republican contract theory, even one as attenuated as Kant's, and it is not clear that Kant, having deprived himself of experience, tradition and convention, can provide an answer. Still, as Malik continues, ‘this’ (absence of juridical condition) is why individuals may use coercion to bring about the founding’, presumably to avoid anarchy and to allow the exercise of human dignity under a regime of external freedoms.

Modern readers of Kant may be disappointed to find that he takes a very theoretical stand on the general will which, following Rousseau, underlies much of his political thinking. It is what (rational) people ought to choose, not what they necessarily do choose. As far as voting goes, this general view allows Kant to distinguish between active citizens and passive ones. The active ones are those who are not dependent on others, so may be expected to vote more independently. Passive ones, who will not have the vote, include women and labourers, who are apparently too dependent on others to be relied on to vote independently. The general will would also sanction the confiscation of property from the nobility and the church if it were the case that the ‘opinions’ on the basis of which these holdings had been accumulated were no longer widely held.

Of course, as Malik interestingly demonstrates, in these and other areas, Kant was very much responding to the tenor of his time, to his historical context in other words. And despite his a-historical philosophical standpoint, he is in fact a progressivist about history, in opposition to Moses Mendelssohn with whom he argued on the matter. He believed that with phenomenal respublicae approaching noumenality more and more, as people begin to appreciate the benefits of autonomy and dignity, states themselves would become less aggressive and more pacific, more (dare it be said?) concerned with trade and commerce. In such a world the more republican states would form themselves into a voluntary league, and we would be on the road to perpetual peace. Republicanism and internationalism thus went hand in hand for Kant, without, it should perhaps be underlined, renouncing their sovereignty.

This last point is helpfully taken up in Onora O'Neill's collection Constructing Authorities: Reason Politics and Interpretation in Kant's Philosophy (Cambridge University Press, 2015) Baroness O.Neill presents Kant as holding a precarious balance between realism and idealism on internationalism, and indeed on revolution. She quotes Kant as saying that ‘permissive laws of reason… allow a situation of public right afflicted with injustice to continue until everything has of itself become ripe for a complete overthrow or has been made almost ripe by peaceful means.’ Nevertheless there is (or could be, or should be) progress towards perpetual peace and international accord, if only because there are times (the Enlightenment?) when ‘the one who foretells events shapes and creates them’.

But it is not only events that we, or some of us, shape and create. The very authority of reason in ethics and politics is, in O'Neill's reading of Kant, ‘instituted or constituted – constructed – by human agents’ (or perhaps by some human agents). The situation, as envisaged by O'Neill, is one in which unco-ordinated agents, united in very little normatively or otherwise, lacking anything like a plan or reason, nevertheless construct something to frame what is moral and rational, before they are reasoners or citizens. And what they construct in the ethical area is presumably the categorical imperative in some form or other, and in the political the thought that coercion or political authority is justified when we have coercive institutions (states) which ‘could be consented to by all who live under them, whereas their rejection could not be consented to by all’.

O'Neill points out that this condition on a justified political system would in fact rule out set ups such as monarchy, dictatorship, feudalism and caste societies. But might it not also rule out wildly inegalitarian societies, which the likes of Piketty and President Hollande could not consent to? Some might be happy with that. But what about states which permit homosexuality which some of our compatriots are very far from being able to consent to? Of course, a lot hangs on could here, because, as we have seen, Kant is adamant that the type of rational agreement he is speaking of is not necessarily what people in the respublica phaenomenon would actually go along with. His politics is, after all, an exercise of construction in pure reason, and pure reasoners, who agree with Kant, (or pure reasoners who agree with Kant – commas may be important here) are not always easy to find outside the academy.