Skip to content
Coordinates of ThoughtHome of Philosophy
Join
Debate

Thesis The contingency argument fails because a collection of contingent things needs no explanation of its own

Derek Hollis Fellow
· 6 replies
God & the argumentsAnalyticIslamicModern EuropeanFalsafaBertrand RussellDavid HumeIbn Sina#contingency-argument#fallacy-of-composition#necessary-existent#principle-of-sufficient-reason
Follow

Its coordinates: 1.8 Natural Theology

Where it stands

The argument from contingency is one of the most elegant in philosophy, and because it comes apart cleanly into steps, I think I can show exactly where it breaks.

Roughly: everything is either necessary or contingent; contingent things depend on something else; the whole collection of contingent things is itself contingent; so it depends on something outside it, which must be necessary.

My objection is the one Hume puts in Cleanthes' mouth in Part IX of the Dialogues, and Russell pressed in his 1948 radio debate with Copleston. If I explain each of twenty particles of matter, I have explained the twenty. "The collection" is not a further thing waiting for its own explanation; it is an act of the mind grouping them. Russell's version: every human has a mother, but the human race doesn't.

If that is right, step three is the fallacy of composition, and the argument doesn't get off the ground. Each contingent thing is explained by earlier or other contingent things, possibly forever, and there's nothing left over.

I'm open to being shown that the argument doesn't need step three in this form, but I'll be surprised.

LikeReply
Follow6 replies

Join the conversation: members reply here, and joining is free. Reading is always free.

Join free to reply Sign in

6 replies

For the thesis 1

  1. Pieter van Dijkhuizen

    For Fellow

    For Derek. Even granting dependence, the infinite chain still explains each member by a prior member. Ask "why does the chain exist?" and Hume's answer is: because each link does, and each link is explained. Demanding more assumes a strong principle of sufficient reason, that there must be an explanation for the existence of everything contingent taken together. That principle is the real premise, and it's exactly what a sceptic denies.

    Helpful · 3
    1. Ikechukwu Anyanwu

      Against Fellow

      Pieter has put his finger on the real issue, and I'd say it cuts the other way. If you deny that the totality of contingent things needs an explanation, you deny the principle in a case where it seems most compelling. Every scientist assumes that things have explanations. The sceptic is happy with that principle everywhere except at the one point where it leads somewhere he does not want to go. That's not obviously more rational.

Against the thesis 2

  1. Layla Sfeir

    Against Fellow

    Derek, Ibn Sina is better defended than Russell's version of him. In the Ishārāt the point is not "each part is contingent, so the whole is" by analogy. It is that the whole depends on its parts: a totality of contingent things exists only if they do. So it cannot be necessary in itself. Then he asks what could account for it. Not the totality itself, since nothing accounts for itself; not any member, since that member would then have to account for itself too. So something outside the totality. That is an argument about dependence, not a slide from parts to wholes.

    Helpful · 1

Other replies

  1. Vikram Sethi

    Fellow

    Small logical point. Composition is not always a fallacy; it depends on the property. "Every brick is red, so the wall is red" works; "every brick is light, so the wall is light" doesn't. So the debate is whether being contingent (or being in need of explanation) is like redness or like lightness. Layla's version argues that contingency transfers because of dependence on parts, which looks like the redness kind to me. Hume's reply needs to say explanation is the lightness kind.

    Helpful · 2
  2. Omar Ragab

    Fellow

    Worth noting that the kalām version avoids this objection altogether, by a different route: it argues the universe began, so an infinite chain is not available. Of course it then has to defend the beginning, which has its own problems. But the composition objection is specifically aimed at the contingency form.

    Helpful · 1
  3. Derek Hollis

    Fellow

    I'll revise. My thesis was too quick: Ibn Sina's step is not a naked composition fallacy, it's an argument from dependence, and Vikram is right that whether contingency transfers to wholes is the real question.

    New thesis: the argument is valid but rests on a strong principle of sufficient reason, and the composition objection is best read as a refusal of that principle. Ikechukwu says the refusal is arbitrary. I still think it's principled, since we've never observed an explanation of a totality. But I accept that's where the argument lives.

    Helpful · 2