Rationalism was a method of philosophy in which the central methodological principle was that there was some way in which the mind could directly grasp the truth, independently of experience. The idea itself probably has its roots in Plato, who famously asserted that all knowledge was already present in the mind and in need only of remembering (in the Meno), and that the mind had direct access to the world of forms (in the Republic). Later the idea was taken up by Descartes, Leibniz, Spinoza, and, in his own unique way, Kant. Motivating these thinkers was the idea that knowledge was something that couldn’t be false, and since nothing that certain could come from the senses, knowledge, if it existed, had to come from somewhere else. Ultimately rationalism turned out to be a less than ideal method, for reasons that I will outline below. But the point here is not to exorcise the ghost of rationalism, rather the point is to illuminate that the flawed idea behind rationalism, that the mind has direct access to the truth, is behind some arguments employed by more modern authors. Thus our goal is to reveal these arguments as fundamentally unsound by pointing out how they rely on the central tenant of rationalism, and thus have the same kind of flaws.
The fundamental reason that rationalism has fallen out of favor is that it just doesn’t work. First of all there isn’t (wasn’t) a consensus even within the rationalists themselves. But if pure reason was supposed to lead to the truth how can this be possible? Either the greatest minds reason imperfectly on a large scale, in which case rationalism is doomed since no one is smart enough to actually get anywhere with it, or reason alone is insufficient to reveal the truth. And, more importantly, rationalism has never produced a testable doctrine that has been empirically confirmed. The physics of Descartes and Spinoza and the monadology of Leibniz have all been shown to be bad theories about the world. Of course the more philosophical and less testable positions have held up better, but even a large number of them seem questionable or in error to modern readers. Thus the rationalist projects that have been attempted seem to have failed, and thus it is reasonable to conclude that there is some flaw with rationalism itself.
Furthermore the existence of a priori knowledge is questionable, and without such knowledge rationalism obviously is flawed. The problem with a priori claims is that fully consistent alternatives to these supposed a priori truths can be constructed, and it is thus impossible to say, without reference to experience, which of these alternatives is true. Famously such an alternative undercut Kant. Kant insisted that we could know a priori that Euclidean geometry was the geometry of the world through a kind of transcendental argument. However, it was later shown that the world does not actually have an Euclidean geometry, despite its supposed a priori status. And even the a priori status of logic has been thrown into question, since it has been shown that there are perfectly consistent alternatives to classical logic. Although we still think that classical logic is probably the best description of truth, at least macroscopically, it is clear that because alternatives exist the truths of logic aren’t a priori, since logic itself then isn’t a priori.
And, most tellingly, rationalism would seem to make truth relative. This is because different people may claim direct the intellectual perception of contradictory facts; that these facts must be a priori, and so on. (That such conflicts will occur will become clearer when we examine the various rationalist-style arguments below.) This implies that either truth is relative, since different people have intellectual access to contradictory truths, or that people may incorrectly think that they have direct intellectual access to truth. But if people can be mistaken about their intellectual access to truth then we have no way to ensure that we are not actually making this mistake. And so intellectual perception would then require some kind of empirical confirmation. Either way rationalism is undermined.
So the idea that we have direct access to the truth in some way doesn’t work. But there are five kinds of arguments that are used in modern philosophy that depend on this very assumption. Thus they are in fact invalid arguments, and should be discarded. Let me deal with each of them in turn by showing how this assumption is essential to their justification.
Argument from intuitions
An argument from intuitions involves accepting some premise because it is intuitively true, or rejecting some conclusions (and thus arguing against one of its premises) because its conclusion is unintuitive. Of course not all arguments from intuitions are flawed. If the intuitions in question are intuitions ultimately based on empirical knowledge (like my intuition that the sun will rise tomorrow) then they may be a shorthand for a perfectly reasonable argument. What I mean to attack here are arguments form intuitions that rely on intuitions without a basis in experience. For example, it would, in my opinion, be illegitimate to reject a claim about the nature of personal identity because it claims something unintuitive about identity over time. Identity over time is not something that we have any direct access to, so it is not that our intuition about how it works can be formed by exposure to it, they are formed by exposure to what other people have to say about it. If our intuitions about it were accurate they would have to come from some direct intellectual perception of it; this is the rationalist underpinnings of the argument, and why we reject it.
Argument via conceptual analysis
An argument through conceptual analysis is an attempt to establish a claim based on careful explication of some of our concepts. Usually this means considering what the concept means, when it seems to apply and when it doesn’t, and so on. This would be like, for example, making a claim about heat by thinking hard about what heat means. Again, like with the argument from intuitions, this assumes that somehow our concepts are guaranteed to be accurate reflections of the world, which assumes some direct intellectual access to the truth. But of course we recognize that our concepts may be imperfect reflections of the world, contra-rationalism, and hence that we can’t draw conclusions about the world from our concepts alone.
Argument via identity in counterfactual situations
Arguments that invoke identity in counterfactual situations are those that ask us to consider a world in which some person or thing is missing a particular property, and then draw conclusions from the fact that it is still the same person or thing despite missing that property. (Such arguments occur in connection with essential properties and the philosophy of language.) This erroneously assumes that we have some direct intellectual access to facts about identity in across various possible situations. Again, there is no justification for believing that we have access to such facts, unless identity is seen as completely a matter of convention, in which case only equally conventional truths could be arrived at by such arguments.
Argument via identification of conceivability with possibility
Arguments that identify conceivability with possibility argue either that because we can conceive of some situation it is possible, and hence a counterexample to some theory, or that we can’t conceive of some situation, and hence it isn’t possible and not a counterexample to some theory. (This occurs most often in connection with the philosophy of mind.) This assumes that what we can conceive tracks what is possible (that we have direct intellectual access to possibility), which doesn’t seem to be the case. I can conceive of heat moving through a vacuum, but heat, being molecular motion, cannot actually move through a vacuum; it is conceivable but impossible. We could possibly save the identification by arguing that we don’t know what is in fact conceivable, but this too would make the argument erroneous.
The open question argument
Finally we have the open question argument, which denies the identification of one thing with another by pointing out that it seems open to debate whether those two things really are identical. (First used by Moore in the context of ethics.) This assumes that we have direct intellectual access to the truth of identity claims; again, this seems false. It might seem like an open question whether heat is molecular motion (certainly we can consistently doubt it), but heat and molecular motion are nevertheless identical.