Skip to content

Could God Have Set Up Darwinian Accidents?

I have a paper forthcoming in the Theology and Philosophy journal Zygon, that I thought some of the readers of this blog might find interesting. Here’s the PhilPapers entry:

John S. Wilkins (forthcoming). Could God Have Set Up Darwinian Accidents? Zygon.

Charles Darwin, in his discussions with Asa Gray and in his published works, doubted whether God could so arrange it that exactly the desired contingent events would occur to cause particular outcomes by natural selection. In this paper I argue that even a limited or neo-Leibnizian deity could have chosen a world that satisfied some arbitrary set of goals or functions in its outcomes, and thus answer Darwin’s conundrum. In more general terms, this supports the consistency of natural selection with providentialism, and makes “theistic evolutionism” a coherent position to hold.

I must return to this and ask whether it is in fact true that a fully Leibnizian Deity could realise a block universe in such a way that it met all utility functions. Apparently there is a theorem in economics and game theory that makes this problematic. Anyone Who Knows is asked to note these matters in the comments. There might be a shared paper out of it…

21 Comments

  1. Physicalist Physicalist

    I expect that if maximizing utility makes any sense at all, then there is a possible world in which all beings have their utility maximized. (The god might just create one or two very happy people, for example.)

    Any economic or gametheoretic considerations are going to deal with questions of what reliably leads to maximizing utility, but a god presumably wouldn’t be bound by these rules.

    I take it the relevant question is whether a Leibnizian world would look anything at all like ours (e.g., would it include natural selection?). Or perhaps the question is just how complicated could a world be and still have a well-defined utility function . . .

  2. Derek R Derek R

    Somewhat related (voting systems):
    http://en.wikipedia.org/wiki/Arrow's_impossibility_theorem

    Arrow’s theorem has been used in economics:
    http://plato.stanford.edu/entries/economic-justice/#ArrThe

    Also you have the Ugly Duckling theorem, which may or may not be related:
    http://plato.stanford.edu/entries/economic-justice/#ArrThe

    And don’t forget the whole Godelian tar-pit:
    http://en.wikipedia.org/wiki/Gödel's_incompleteness_theorems
    which might apply if an organism is evolving while subjected to forms of mathematical constrants (i.e. laws of physics).

  3. Derek R Derek R

    Trying again, I think my post was rejected due to too many links:

    Arrow’s theorem says that a perfect voting system is impossible. It has also been applied to economics, and probably applicable to organism survival with a little imagination.

    The Ugly Duckling theorem says that an unbiased classification is impossible. (Which may or may not apply to an organism).

    Then don’t forget Godel’s incompleteness theorems. If an organism’s evolution is based on mathematical constraints (i.e. laws of physics) then it’s conceivable that a Godelian situation might occur to which there is no known answer.

  4. I don’t have any specific suggestion to offer about this problem, but I wanted to duck in for a moment to note that this issue is just the kind of theological problem that retains its interest if you don’t take theology itself seriously because it addresses what does and does not make sense when you’re talking about optimizing a system under constraints. Of course actual system designers don’t have the full capabilities of a Leibnizian god–that’s why we’ve got rule utilitarianism, if that’s what you professional philosophers are still calling it–but designing a market is a little like setting the parameters for natural selection and lots of folks are interested in that.

  5. Jimbo Jimbo

    Is god the victim of the Saint Petersburg Paradox?

    On one hand it seems like an omnibenevolent deity has to realize all increases of utility, no matter how marginal.

    One the other, it seems like Leibniz’s God is efficient, favoring those worlds in which he gets the most bang for his buck, because one of L’s criteria of universe choice is simplicity.

    It doesn’t seem like Arrow could be a problem for an L-God, because Leibniz would allow for dictatorial preferences owing to some minutiae of his theory of action and freedom.

  6. All of these apart from the St Petersburg Paradox involve more than one agent and set of preferences, though, if I understand them. Since the L-God is fully rational, fully apprised and has a fully ordered set of utility preferences, surely he is not so constrained?

    • Jimbo Jimbo

      Whatever desiderata God uses to explain his choice of universe could possibly be modeled as players in a game. Beauty versus moral goodness; searching for some pareto optimal distribution of beauty and goodness might be a way to figure out which parameters are compossible.

      • That is what a deity might do if there were cost-benefit ratios for their utilities; but I am presuming that either a neo-L deity and an L deity both have fully ordered preference rankings. The question is whether they can realise these. Thus far, I cannot see why not.

  7. Hey John, great article. I am going into a different direction then your question, if you do not mind. 🙂 For example, the synthetic theory of evolution is based on nondeterministic probabilistic events such as positive natural selection, so how could a block universe be consistent with nondeterministic events?

  8. Also, what is the substrate of this neo-Leibnizian deity? 🙂

  9. James Goetz:
    Hey John, great article. I am going into a different direction then your question, if you do not mind. For example, the synthetic theory of evolution is based on nondeterministic probabilistic events such as positive natural selection, so how could a block universe be consistent with nondeterministic events?

    I know that you drew some sketches, but I do not feel that you clearly addressed the issue of a deterministic block universe with apparent probabilistic contingencies. Or perhaps your language was too subtle for me to see your explanation?

    • If an event is probabilistic, that simply means it is something that happens as a class of events over some distribution function. Each event itself (say, a particular case of decay or a particular quantum foam event) has a real existence of unity probability once it has occurred. So it exists in a block universe as much as a determinate event. What these things are from outside the block universe is a pattern of event classes. Arguably, all secondary causes, whether determinate or not, are merely patterns in the block universe for an eternalist observer, not dynamic processes in a metaphysical sense. Mark Colyvan, the philosopher of physics, has indicated to me that this implies there is no causation at the metaphysical/physical level, only at the level of analysis in which things may be considered to be contingent (i.e., the special sciences).

      As to the substrate for a neo-L deity, I suppose it is the same or similar to the substrate for a fully-L deity; some eternal state of being. I personally am happy to rest content with the eternalist substrate for a block universe…

      • I want to make sure that I understand you. Does Colyvan’s theory imply that an eternal observer can view a block universe without exhaustive causation of the probabalistic outcomes in the universe? I believed that model for many years, which goes along the line of closed Arminianism. Currently, I doubt that model and lean toward open futurism (open theism).

        Anyway, if I correctly understand Colyvan’s theory, then it drastically diverges from the b-theory of time which implies that everything from the initial to everlasting expansion of the universe happens simultaneously while all appearance of contingencies is nothing but the appearance of contingencies in a world of causal determinism. In the case of classical b-theory, it is incompatible with Darwinian accidents but the mere appearance of Darwinian accidents.

        • Sorry for the delay in responding, Jim.

          What would a stochastic universe look like to an extratemporal observer (ETO)? Each probabilistic event would form a pattern of classes of causes (Cs) and effects (Es). In a deterministic universe, the ETO would see 100% of Cs correlating downstream to 100% of Es; in a stochastic universe the ETO would see some proportion of Cs correlating with Es. But they would see all causal relations as patterns of Cs and Es correlating to some degree.

          Arguably to the ETO, causation would simply be these patterns, and the reason why the patterns exist would have to be answered in terms of primary cause (why the universe exists as it does) rather than secondary causes (why those Cs correlate with those Es). In other words, why is the universe that is realised like that? But stochastic correlations are not a problem for the ETO – they are just partial patterns.

        • Hey John, I suppose an omniscient ETO could see all contingencies such as all possible spacetime initial conditions and all possible outcomes for each initial condition. I doubt that a finite ETO could see anything or do anything. An omniscient ETO would need wisdom and power unknown to humans to generate a spacetime universe from his originally extratemporal existence. I agree with Kant that an infinite sequence of time could have not occurred and while I conjecture that the observe universe originated from a dimensionless physics such as an omniscient, maximally potent ETO. The generation of spacetime resulted in the deity figuratively straddling a dimensionless and dimension existence. The deity always knows all possibilities and observes the settling of present time while occasionally intervening but never suspending the laws of physics. Well, that is my most recent view. 🙂

  10. eddie eddie

    I’m trying to think of the differences between darwinian accidents and divine judgements that look like darwinian accidents. Every cat that catches a mouse would have to be considered a ‘good’ cat, the mouse a ‘sinner’ and it to be god’s judgement that one should eat the other. Whereas in darwin’s world, the mouse would be a ‘good’ mouse if it had made baby mice before it got eaten, while the cat would be a ‘bad’ cat if, despite eating many mice, it failed to make any kittens.
    The evolutionary outcomes would be different, but would they be noticable?

  11. Barton Barton

    For anyone who went to John’s talk last night in Newtown; Why were all models of Gods assumed to be outside the time dimension? Can a god, according to a Theist, not be restricted to time?

    For example in the case of the Christian God – Just briefly reading the first page of the old testament- it seems to imply that god acted within time, thus being within the time dimension as you and I.

    And surely issues regarding time, and consequently chance, are critical to whether a Theist can be a Darwinian?

    I’d love to hear anyones opinion on this

    • Well I was trying to deal with the problems for a traditional theist, and since Augustine’s Confessions, God has been seen to be outside time in traditional theism (i.e., small “c” catholic theology). However, if you presume God is inside time, the view you are presenting is more like panentheism; in this case God is limited, and the issue of intervention versus secondary causes (natural law) doesn’t arise, since God is immanent in every event (that is a fancy way of saying he’s there but not the same as the event).

      A Darwinian who said that God was inside time would have to deal with the possibility that God simply could not foresee the future effectively due to quantum randomness, contingency and chaotic interactions. I can’t help everybody…

  12. About Arrow’s theorem:

    It applies to god, even if he is a single dictator.
    Think about god’s criteria as the players. God must have multiple criteria on what constitutes a good universe. These criteria don’t have to be explicitly contradictory, nor do they need to emanate from different agents. All we need for Arrow’s theorem to hold is that you have more than 2 criteria, more than 2 options, and that the criteria themselves are not well ordered (that is, that the is no single criterion which hold above all others.

  13. John S. Wilkins:
    Well I was trying to deal with the problems for a traditional theist, and since Augustine’s Confessions, God has been seen to be outside time in traditional theism (i.e., small “c” catholic theology). However, if you presume God is inside time, the view you are presenting is more like panentheism; in this case God is limited, and the issue of intervention versus secondary causes (natural law) doesn’t arise, since God is immanent in every event (that is a fancy way of saying he’s there but not the same as the event).

    A Darwinian who said that God was inside time would have to deal with the possibility that God simply could not foresee the future effectively due to quantum randomness, contingency and chaotic interactions. I can’t help everybody…

    For the most part, Christians believe that God is both transcendent to time and inside time. In the case of Augustine, he apparently proposes some type of exhaustive divine determinism that is somehow compatibilist with free will, but I suppose that nobody has yet to explain the compatibilism.

Comments are closed.