Does the no miracles argument for scientific realism commit a fallacy? I've read the claim that because science is abductive, and the no miracles argument proceeds via abduction, it is circular.

Is it viciously circular?

3 Answers 3


Someone like Quine would embrace that it's circular. Naturalized epistemologists tend to be coherentists, who tend to get charged with having circular (or criss-crossing-but-never-essentially-grounding) justifiers.

  • I'd be inclined to see it as an example of the power of strange loops & tangled hierarchies, which are essentially coherentist. Specifically because of reinforcing feedback loops, which provide something more substantial and dynamic than circularity alone, and nested explanatory structures that cross-check to support consilience
    – CriglCragl
    Commented Jan 3, 2021 at 14:35

Of course it is. What is a miracle other than a complete departure from what is expected. Well, then what is an outlier? The 'no miracles' argument cannot be observed.

It can be proposed as a falsifiable principle, but then it is kind of already falsified. Various things like the unexpected relevance of mathematics to physics come across as miraculous: they are positive results we might wish for, already true, and still inexplicable when they were most relevant. Newton surely considered the power of mathematics a gift from God.

It can be theorized that we would not have reached our Western level of faith in the reliability of results if we did not already have a principle of the miraculous order of the universe.

  • 1
    not my downvote btw
    – user6917
    Commented Dec 27, 2016 at 16:54
  • 1
    Actually, I deserve the downvote, I have responded to the wrong argument.
    – user9166
    Commented Dec 27, 2016 at 20:02

There is an intriguing response by S. Psillos ("Scientific Realism: How Science Tracks Truth", 1999). The No Miracles Argument seems to assume that abduction (better, inference to the best explanation) is reliable, which is precisely what antirealists often question. Now Psillos' idea is that provided an externalist epistemology, realists need not commit to such an assumption. Instead he proposes to straightforwardly employ inference to the best explanation:

Accept realism about our best-established scientific theories as the best explanation of the empirical success of science (No Miracles Argument), and from realism infer to the factual reliability of inference to the best explanation in turn, viz. as the basic pattern of scientific reasoning which sustains the initial No Miracles Argument.

This new argument is rule-circular in that the conclusion asserts the reliability of an inference rule by which it was attained. But the whole point is that rule-circularity is not problematic within an externalist epistemology, which permits the use of a reliable inference rule irrespective of the the user's ability to vindicate it. So the circularity is not vicious.

There are problems with such a response. For example, externalism is controversial. And if antirealists did not accept the conclusion of the No Miracles Argument in the first place, why would they accept it now (and proceed arguing from it)? After all, no new evidence was provided.

You must log in to answer this question.