DOI: 10.4324/9780415249126-P045-1
Version: v1,  Published online: 1998
Retrieved October 26, 2021, from

4. Modus ponens fallibilism

Modus ponens fallibilists accept, along with the sceptic, the deductive closure principle. But they attempt to turn that principle against the sceptic. Like relevant alternatives fallibilists, they take as a starting point the strongly intuitive claim that we do know many things about the world. They then note that, given the closure principle, it follows that we know the falsity of sceptical alternatives. For example, I now know that I am looking at my computer monitor. I also know that my looking at a computer monitor precludes my being a brain-in-a-vat. It follows by the closure principle that I know I am not a brain-in-a-vat.

Is this piece of reasoning legitimate? One might challenge those who reason in this way to explain how we know sceptical alternatives are false. How, for example, do I know I am not a brain-in-a-vat? After all, the sceptical problem arises because we seem to lack any reason for believing sceptical alternatives are false. These alternatives are constructed so as to make it impossible for our evidence to count against them. Presumably, our recognition of this explains, at least in part, our intuition that we fail to know sceptical alternatives are false.

One way for the modus ponens fallibilist to try to meet this challenge is to claim that I can know:

  • not-h: I am not a brain-in-a-vat

by inferring it from:

  • q: I am looking at my computer monitor

According to this way of proceeding, even though none of my evidence for q counts in favour of not-h, it does not follow that I have no reason to believe not-h. For that reason can be q itself. Since I know q (on the basis of my visual evidence) and I know that q entails not-h, I can infer not-h from q and thereby come to know not-h.

Is this reasoning legitimate? Let’s compare it with another case. Suppose I park my car in front of the market and go inside. Although I am not currently looking out the window I can still know:

  • p: My car is parked in front of the store

Can I then come to know:

  • r: My car has not been towed away

simply by inferring it from p? Notice that p entails r. It seems, none the less, that I would already need to have sufficient evidence to know r before I could infer p. And if my initial evidence is insufficient for me to know r, I cannot infer p and so I cannot infer r from p.

The modus ponens fallibilist reasoning concerning sceptical alternatives looks suspicious because it seems like the reasoning in the parked car case. Intuitively, I need to have reason to believe not-h before I can infer (and thereby come to know) q. Thus I cannot first infer q and then go on to infer (and thereby come to know) non-h.

Another version of fallibilism argues for the claim that we know sceptical alternatives are false by appealing to principles of inductive inference. One version of this view argues that the hypothesis that the familiar world of objects exists is the best explanation of our sensory evidence (and so a better explanation than sceptical alternatives). This licenses an inference from our sensory evidence to the familiar-world hypothesis (see Inference to the best explanation). We can thereby come to know that this familiar world exists. And since we know that the familiar-world hypothesis rules out the sceptical alternatives, it follows by the closure principle that we know sceptical alternatives are false.

The burden for this view is to say why the familiar-world hypothesis is a better explanation of our sensory evidence than any sceptical alternative. This is not easy to do since sceptical alternatives are designed to explain our sensory evidence. Proponents of the view that sceptical alternatives provide inferior explanations often appeal to pragmatic considerations like simplicity and conservatism. But there are several problems with this approach. Even if we could establish that the familiar-world hypothesis is, for example, simpler than any sceptical alternative, why should we think that this supports the claim that the hypothesis is true? Unless this crucial link can be made, it is not clear how this response to the sceptic can succeed (see Theoretical (epistemic) virtues).

Moreover, often arguments that the familiar-world hypothesis is the best explanation of our sensory data are quite sophisticated and complex. This raises the worry that only those who are philosophically sophisticated enough to follow such an argument can have knowledge of the external world.

Citing this article:
Cohen, Stewart. Modus ponens fallibilism. Scepticism, 1998, doi:10.4324/9780415249126-P045-1. Routledge Encyclopedia of Philosophy, Taylor and Francis,
Copyright © 1998-2021 Routledge.