Magic and the PSR

Take the principle of sufficient reason (PSR) to entail the claim that reality is intelligible, and ‘magic’ to describe any cause, event or state of affairs that is intrinsically unintelligible¹. Now consider the following argument:

1. If the PSR is false, then magic is possible.
2. But magic is not possible.
3. So, the PSR is true.

It seems clear that, if the PSR is true, then there is nothing in reality that is intrinsically unintelligible. The act a magician performs when pulling a rabbit out of a hat, though seemingly ‘magical’ in our technical sense, is perfectly intelligible in virtue of some trick kept secret by the magician. Indeed, it would be impossible for such an act to be unintelligible, such that the rabbit could emerge out of the hat inexplicably. But suppose we hold the PSR to be false, and in so doing we affirm that some features of reality are intrinsically inexplicable. Given such a scenario, a rabbit popping out of a hat inexplicably could not be said to be in principle impossible. Less still could we say that such a state of affairs is merely unlikely or improbable, for two reasons. First, the PSR is the precondition for the very possibility of making inferences to the best explanation (IBE)², whether in daily life or in the domains of science and philosophy. This is why, when scientists encounter some unexplained phenomenon, they take the proper stance of searching for an explanation, that is, proceeding under the assumption that there is an explanation to be had. If the PSR were false, however, no such presumption could be justified, since the in principle possibility of inexplicable events entails that, for any unexplained thing, event, or state of affairs X, the best explanation might just be that X is inexplicable.

Second, the objective likelihood of specific events depends, largely, on our ability to assign objective probabilities to them, and these probability assignments are possible in light of such events being law-governed. But if the PSR is false, then the probability of some event E’s ocurring must now take into account both the probability of E ocurring in a law-governed way — such that the laws are obeyed — as well as the probability that E occurs for no reason at all — such that the laws are not obeyed. But such a probability calculus would undermine our ability to talk of specific events as having a certain likelihood altogether³. All of this seems unreasonable, on par with the fact that, as Pruss and Rassmussen have noted⁴, we do not ever observe things coming into and out of existence from nothing, without any cause whatever. These considerations therefore give us reason to affirm the PSR.

Hillary Putnam offers what I take to be a parallel argument in his 1975 paper, ‘What is mathematical truth?’, where he outlines what has come to be known as the “no miracles argument.” According to Putnam, scientific realism — the view that the universe as described by science is real, independently of interpretations — “is the only philosophy that doesn’t make the success of science a miracle.” In other words, the fact that our best scientific theories are so successful testifies to their truth. Similarly, the fact that the things of our experience — as well as those within our range of observation — are explicable in terms of various causes, events, states of affairs etc., testifies to the truth of the PSR.

To rephrase Putnam, PSR is the only philosophy that doesn’t make the intelligibility of the world a miracle.

[1]: Something is intrinsically intelligible, or intelligible in itself, if there is an explanation to be had for why it exists, instantiates the properties that it does, and so on. Conversely, something is intrinsically unintelligible, or unintelligible in itself, if it itself is fundamentally inexplicable — a brute fact.

[2]: For more on the relevance of IBE to the PSR, see Alexander Pruss, “The Principle of Sufficient Reason: A Reassessment.” Cambridge University Press. 2006. pp. 280–284.

[3]: For a more detailed exposition of this argument, see Pruss 2006, pp. 273. Thanks to Chad McIntosh for bringing it to my attention.

[4]: See Alexander Pruss and Joshua L. Rasmussen, “Necessary Existence.” Oxford University Press, 2018. pp. 38–39.

Formally, the argument runs:

1. If it were possible for something to come into existence uncaused, then we should expect to observe things spontaneously coming into existence all​ the time.​
2. But we don’t observe this.
3. So it is not possible for something to come into existence uncaused.

Note, however, that a non-modal version of the argument is more modest, and provides a similar conclusion:

1. If things come into existence uncaused, then we should expect to observe things spontaneously coming into existence all​ the time.​
2. But we don’t observe this.
3. So things don’t come into existence uncaused.