In some central and important cases, … the existence of specific power relations in the society will produce an appearance of a particular kind. Cert… - Raymond Geuss

If one thinks that ideological conceptions are an important feature of modern societies, and that the analysis of ideologies will therefore have to be an integral component of any contemporary political philosophy, Rawls’s view is seriously deficient, because it does not thematise power. The idea that seems to be presupposed by the doctrine of the veil of ignorance—namely, that one can in some way get a better grasp or understanding of the power relations in society and how they work by covering them up, ignoring them, or simply wishing them away—seems very naïve. To the extent, then, to which Rawls draws attention away from the phenomenon of power and the way in which it influences our lives and the way we see the world, his theory is itself ideological. To think that an appropriate point of departure for understanding the political world is our intuitions of what is “just,” without reflecting on where those intuitions come from, how they are maintained, and what interests they might serve, seems to exclude from the beginning the very possibility that these intuitions might themselves be “ideological.”

In many of the cases of conceptual innovation, … creating the conceptual tools is a precondition to coming to a clear understanding of what the problem was in the first place. It is very difficult to describe the transition after it has taken place because it is difficult for us to put ourselves back into the situation of confusion, indeterminacy, and perplexity that existed before the new “tool” brought clarity and this means it is difficult for us to retain a vivid sense of what a difference having the concept made. We can just barely imagine ourselves in a world in which there are no states, as opposed to local barons, warlords, clans, primitive communal forms of village organisation, etc.

Some forms of Kantianism put great weight on “what we can imagine” holding that this can be a source of insight into necessary connections. Thus various of Kant’s arguments about space and time depend on the purported fact that it is impossible for us to imagine certain things: we can know, Kant claims, a priori that space has only three dimensions because we cannot imagine it as having more than three dimensions. History in the form of non-Euclidean geometry and modern physics has put paid to that particular line of argument, but in general we should beware of depending too much on “What we can imagine?,” especially in politics. As Nietzsche puts it somewhere, sometimes the fact that you can’t imagine a situation in which things are very different from the way they are now is not an especially good argument for the claim that they must be as they now are, but, rather, represents a failure of your powers about which you should feel mildly apologetic.