Thursday, November 12, 2009

Lange on Laws and Lawmakers

Here in St. Louis, we are having a reading group on Marc Lange’s interesting Laws and Lawmakers: Science, Metaphysics, and the Laws of Nature (OUP, 2009).

The book is motivated by the observation that there seems to be a tight connection between laws and counterfactuals. As everyone knows, laws support counterfactuals in a way that accidental truths do not. Lange offers a novel and interesting way to account for this link: he argues that laws are a “stable set of truths” in a way that sets of accidental truths are not. What this means is that unlike sets of accidental truths, all laws are still true under any counterfactual supposition that is consistent with their all being true.

Lange begins with this connection between laws and counterfactual and, in the course of the book, turns it into an ontological dependence between laws and counterfactual truths. He posits the existence of ontologically primitive “subjunctive facts” and argues that those are the truthmakers for law statements. I have only read the first two (out of four) chapters, so don’t ask me how to make sense of the notion of a “subjunctive fact” yet. But I do have a comment on the early part of the book.

Lange discusses some apparent counterexamples to his thesis of the “stability” of laws. Such counterexamples include David Lewis’s view that at least when dealing with deterministic situations, any ordinary counterfactual supposition requires a violation of the laws (a “miracle”).

(Lange discusses these counterfactuals only in a 9 pp. long footnote (Chap 1, fn. 29). That seems odd. It seems that given the relevance of these counterexamples to Lange’s view, he should have discussed them in the main text.)

I’m not entirely convinced by Lange’s response to the counterexamples (and also by some of the arguments in the main text that are similar in structure). Obviously I cannot do justice to the issue here, but very briefly, Lange seems to argue along the following lines: true, there are counterfactual suppositions that are consistent with the laws but under which the laws are violated, at least in some “worlds”; however, intuitively, under the same counterfactual suppositions, the “worlds” in which the laws are violated are more dissimilar from our world than the “worlds” in which the laws are not violated, and this still supports the stability of laws. Even granting the relevant intuitions, I suspect that the intuitions about similarity between “worlds” that ground the relevant judgments are generated, in turn, by knowledge of which truths are and are not laws. In other words, the counterfactual judgments that are supposed to support the claim that laws form a stable set (in a way that accidents do not) seem to depend on knowing which truths are lawful and which are not. So I am not convinced that Lange has found a noncircular way to connect laws and counterfactuals, contrary to what he suggests in his book. (Circularity is not a problem, unless you want to ontologically ground the laws in the counterfactuals, as Lange does.) Am I missing something? If anyone has insights on this matter, I’d be interested to hear them.

(Another odd choice: John Roberts is in the same department as Lange at UNC Chapel Hill and has published extensively on laws [including his own book on laws: The Law-Governed Universe, OUP 2009]. Even though Roberts is acknowledged at least four times in the book for having contributed this or that point, none of Roberts’ publications are cited by Lange. It’s hard not to wonder how that came about.)

All this being said, Lange’s book is thought-provoking and worth reading.


  1. Dear Gualtiero,

    I don't know Lange's book, so I cannot give a qualified commend. But nevertheless, one thing which came to my mind ist the following:

    Why do you think that the epistemological point has any bearing on the ontological one? It might very well be the case that we only find out the counterfactual truths by virtue of knowledge of the laws, but nevertheless the laws migt be ontologically grounded, so to say, in these truths.
    (Analogously to: We only find out which properties there are by encountering instances of these properties, but nevertheless these instances are ontologically grounded in the abstract properties - at least on a universialist account.)

    Obviously on such an account the epistemological question remains: How we know which universal statements are laws and which aren't is still a problem. But this leaves the ontological relations untouched.

    Best wishes.

  2. Of course I meant "comment" not "commend". Sorry for typo(s).

  3. Why do I think that the epistemological point has any bearing on the ontological one?

    Thanks for the excellent question! (Sorry for the delayed response.)

    It's not so much that I think the epistemological point has any bearing on the ontological one. It's that if we don't have access to the truth of the counterfactuals independently of our knowledge of which are the laws, I lose track of why on earth we should draw the ontological conclusion in the first place.

    Lange's main argument for his ontological conclusion, if I understand correctly, is that we have strong intuitions about laws being necessary in a certain way (consistent with their contingency in another way). The goal is to save these intuition, and the appeal to counterfactuals is supposed to be a way to find truthmakers for laws that explain their necessity. But if our judgments about counterfactuals are in turn explained by our beliefs in laws, I no longer see why we should believe that the counterfactuals explain the necessity of the laws.

    Does this make sense?

  4. As I understood it from your first post I thought that Lange's thesis is that there are such strange things called "subjunctive facts" which make law statements true (call this LLT - Lange's lawmaker thesis).
    Now I thought that your point is that we only have knowledge of the subjunctive facts in virtue of our knowledge of the laws. And therefore you conclude we cannot ground the laws in the subjunctive facts.
    If that was your point (of course in a nutshell) then it seems to me that you think that epistmeological questions (How do we know which subjunctive facts obtain? Answer: Via our knowledge of laws) have bearing on ontological ones (Which entities are ontologically fundamental? Laws or subjunctive facts? Answer: It cannot be the subjunctive facts, because we only know them via knowing the laws). And this would not strike me as a convincing argument against LLT.

    But, of course, I agree that this alone does not make LLT plausible. There is a need for arguments for his suggestion or, in your word, "why on earth we should draw the ontological conclusion in the first place". (Maybe he simply takes facts to be truth-makers of ordinary statements and maybe he also believes that there are mathematical facts which make mathematical statements true. Therefore we should also in the case of counterfactuals take it to be the case that there are certain facts, i.e. subjunctive facts, which make them true until there are no reasons to the contrary. But I am guessing here, since I do not know Lange's work. I hope at least there are some arguments somewhere in his book.)

    So, if Lange only wants to defend LLT I see no problem resulting from your remarks, but if his goal is indeed the explanation of our intuition of necessity regarding laws he should indeed say something more, especially if he admits that such judgements are based on similarity considerations regarding possible worlds. Here I think you are perfectly right in saying that these judgements are based on our knowledge of laws. (But then he should not say that he argues for an ontological thesis, since "intuition of necessity" is an epistemological category and therefore with regard to this explantory hypothesis your arguments do present a though charge; at least as far as I can see)

    I also have another question: Does Lange really say that laws are ontologically grounded in counterfactuals? This would be strange, since counterfactuals are linguistic entities (sentences or propositions) and laws are non-linguistic general facts (or are they genreral propositions and therefore also of linguistic nature? - I must admit I'm always confused by the subtle differences). As I understand it laws could be grounded in non-linguistic subjunctive facts (but those are different entities than the corresponding counterfactuals), but not in counterfactuals. (But of course it would be possible that knowledge of truth-values of law-statements depends on knowledge of truth-values of counterfactual statements.)