Is there a conflict between the fine-tuning argument and the design argument?

I don't think (P) follows from (I) and (L). Given (I), which I see no reason to dispute, I see

(P') If God wants to create a universe with life similar to that which we now have, he would have to create a world with fine tuning.

I'm not sure that generates the type of problem you have in mind, though. Do you think I'm missing something here?

Ted wasn't arguing that I and L entail P. He was arguing that I and L, given P, lead to a conflict. P is assumed.

I think this is a misrepresentation of both arguments. The argument goes as follows: It's unlikely that we'd have constants such as what we've got if there weren't a God. Therefore, there probably is a God. That's something like a probabilistic/inductive analog to a modus tollens. If no God, then not these constants. Not not these constants, so God.

That's not the same thing as this argument: If there's a God, then it's likely that there will be constants like what we've got. We do have these constants, so there's likely to be a God. That's the probabilistic/inductive analog to affirming the consequent. If God, then these constants. Constants, so God.

The way you framed the biological design argument goes the same way, logically speaking. If you rework the arguments so that they're not fallacious in that way, I think the conflict disappears. One says God is more likely given one fact (the existence of a universe making life even possible). The other says God is more likely given a different fact (the existence of life in the way we've got it).

I�d be relieved if the conflict disappears on a different construal of the argument. I don�t see, however, that the statement of the argument is a misrepresentation of how the argument is supposed to go. We have data (the fine-tuning data and the biological design data). We know that given all the ways things could go it�s incredibly unlikely that this data obtains. And we know that it�s much more likely that we�d have this data if God exists than if God didn�t exist. So this data provides evidence that God exists. The fine-tuning argument needs the claim that God would create a world like this (a life-permitting world) to get the desired claim that it�s much more probable that we�d have this data if God then if no God. The design argument needs the claim (I think) that it�s improbable that complex beings would exist to get the claim that it�s incredibly unlikely that we�d have the biological design data. But that�s just (L) and (I). Thus, given (P), we have the conflict.

Also, I intend the argument to include the claim that the theistic hypothesis (H) makes the data much more likely than (~H). If a hypothesis just made the data more likely and we inferred that that hypothesis was true on that basis alone I�d have similar worries about the inductive analog to affirming the consequence. But once we compare the other hypotheses and learn that the one makes the data much more likely than the others, we have strong evidence that that hypothesis is true.

To argue that the conflict disappears it needs to be shown that either there�s no conflict between (L), (I) and (P), or that the fine-tuning argument or biological design argument can be formulated without a commitment to (L) or (I), or that (P) is false. Now in general the fine-tuning argument needs the claim that the hypothesis of theism increases the likelihood of the fine-tuning data. One way to get this claim is to argue that God desires to bring about a world like this. The biological design argument needs the claim that it�s incredibly unlikely that complex organisms exist. But this seems to require the claim that it�s improbable that complex beings exist in our universe.

I think Jeremy's point relieves the apparent conflict in the two arguments. If I am reading him correctly, then it seems the way (I) and (P) are currently stated need to equivocate on the terms probable and improbable in order to have genuine conflict.

On the one hand, in premise (I) improbability is relative to complex life existing given atheism or methodological naturalism. On the other, (P) proposes probability given theism. So, the design argument is arguing that complex life is improbable if there's no God, while the fine-tuning argument claims that there are enough precise parameters for life that it points to a designer. It seems that probability and improbability are not construed exactly the same in both formulationsm, which Ted needs to show incompatibility with the premises. Hence, there is no genuine conflict.

As I said before, I think P is false and unnecessary for either design argument. The reason I think it's false is that God doesn't need there to be a logically prior probability of life's existence for God to create life. Whether through naturalistic processes overseen through providence or through special creation, God's creation might still go beyond what would be probable given how one might expect things to develop given the initial conditions and probabilistic reasoning. In fact, that's what the biological design argument is assuming.

The fine-tuning design argument is in some ways a different animal. The biological one is saying that the kind of design we see is unlikely without a designer. The fine-tuning one is saying that without the constants life would be impossible, and the constants are unlikely without a designer. That second step may be where the proposed conflict is coming in.

The fine-tuning argument says that if God is to create human beings (or something anything like us), there would have to be certain constants in place. Those sort of constants don't necessitate or even make probable by themselves the development of complex organisms. They are necessary conditions, though. So one argument says life would be highly improbably if there's no designer, and the other says a necessary condition of life would be highly improbable if there's no designer.

Something like L is therefore crucial for anyone who believes in a creator, since fine-tuned conditions are necessary for life at all. That's perfectly consistent with saying that those conditions don't make life probably in the least. So I'm not sure why there should be a conflict, and I think it's just my resistance to a principle like P.

I didn't see John's comment before writing that last one. I don't know if that expresses all my worries, but it gets to the main one in a much simpler way that I did.

John and Jeremy, thanks for the comments; hopefully this helps clarify the argument.

I claim that there�s a conflict between the following three principles (for the sake of clarity I�ve reformulated (L) and (P)):

(L) God desires to create a universe for the purpose of bringing about the existence of intelligent moral beings.

(I) It�s incredibly unlikely that complex life would arise in our universe.

(P) If God desires to create a universe for the purpose of bringing about the existence of intelligent moral beings then God will bring about a universe that makes probable the existence of complex beings.

Let me address John�s concerns first. I take it that John thinks that the probability claims in (I) and (P) conditionalize on different things. (I), though, should be understood as saying that given our universe (its initial conditions and fundamental laws and its natural development) it�s incredibly unlikely that life would arise. The consequent of (P) should be read as claiming that God will bring about a universe (initial conditions and fundamental laws and natural development) that makes probable the existence of complex beings. I don�t think that we should include God�s activity (divine intervention) in evidence we conditionalize on in (P) {see below}.

This leads into questions about (P). It�s right to be worried about (P). I think, however, (P) follows from an intuitive principle about rational action and the theistic hypothesis. Here�s the principle of rational action:

(R) If S desires to bring about a state of affairs for the sole purpose of achieving goal G then if S is fully rational and S is not limited in power then S will bring about a state of affairs that makes probable the obtaining of G without S�s further intervention.

Since God is fully rational and not limited in power, I think (P) follows. The clumsy clause at the end of (R) �without S�s further intervention� provides the reason why I don�t think we should include divine intervention in the evidence tacitly appealed in (P).

I'm not trying to be dense (seriously), but I'm still not convinced there is a genuine conflict. Taking account of divine action or not in (P) and (I) still seems important, even if it is not construed as on-going or continued intervention by the Almighty. Perhaps an illustration will help.

Suppose one stumbles across an object in the woods. Lo and behold it is a perfect copy of Shakespeare's Hamlet. It is most improbable that this orderly and designed object could have just fallen together by random luck. However, upon close examination one can deduce how it could come from a printing press. In fact, judging by the quality of the binding and print, it seems highly probable that this object is indeed the product of a professional press.

My point is that the use of probable and improbable can be used of the same object when differentiating its hypothetical source. This seems to be how the design and fine tuning arguments work. They seem to go hand-in-hand. Moreover, I'm not sure that continued intervention on behalf of an entity's originator is going to make a difference. Am I advancing this discussion or is this a step backwards?

I�m not sure I understand the claim �the use of probable and improbable can be used of the same object when differentiating its hypothetical source�. If you mean that we can judge an event probable on certain evidence and improbable on different evidence then I completely agree. For instance, it�s improbable that I draw a 3 of hearts from a standard deck of cards, but it�s probable that I draw a 3 of hearts from a pile of 3 of heart cards. It would be a mistake then to think that the claim it�s improbable that I draw a 3 of hearts from a standard deck conflicts with the claim it�s probable that I draw a 3 of hearts from a pile of 3 of hearts. However, there would be a conflict between the two claims: (a) it�s improbable that I draw a 3 of hearts from a standard deck and (b) it�s probable that I draw a 3 of hearts from a standard deck. I intend the two claims (I) and (P) to be understood in this latter sense {see my earlier post}.

Also to allay a potential source of confusion I�m not claiming that the fine-tuning data somehow conflicts with the biological design data. Also, the conflict I�m interested in does not make use of the claim that the fine-tuning data is incredibly improbable. The conflict with the fine-tuning argument and design argument I�m thinking about is that a crucial premise of each argument seems to conflict with principle (P).

Does that help at all?

So what is it with respect to in the two cases?

(I) It is improbable that complex life would exist in this universe.

That doesn't say, but that improbability is not an improbability given the existence of God. It's an improbability given certain features naturalists would be happy to concede.

(L) God would create a universe much like ours (a life-permitting universe).

As far as I can see, we're now saying something structurally equivalent to:

It is probable that complex life would exist in this universe, given the existence of God.

So one of them takes the existence of God as a given, while the other doesn't (and must not if it's to achieve the desired result). That's why I'm having trouble seeing why you think they're both taking the same claims as given. They seem quite deliberately not to be doing that.

This will have to be a short post�I�ll follow it up later if it doesn�t answer your question.

The claim (L)--God desires to create a universe for the purpose of bringing about the existence of intelligent moral beings�by itself does not imply that it�s either probable or improbable that the universe support life. Rather (L) notes God�s reason for bringing about a life-supporting universe, viz., God desires that intelligent moral beings exist. The conjunction of (L) and (P) entail that God brings about a universe that makes probable that complex life would arise. This implies that there�s a universe that makes probable that complex life would arise�given such and such initial conditions and fundamental laws, it is probable that complex life arises. Call this universe u.

So the conjunction of (L) and (P) yield:
It�s probable that complex life arise in u given initial conditions, fundamental laws, and natural development.

But (I) states: it�s improbable that complex life exists in our universe given initial conditions, fundamental laws and natural development.

I imagine principle (P) is the offensive principle. One way to understand the principle is that if God brings about a universe so that intelligent life would exist then he would favor a deck-stacking universe over a interventionist universe. The biological design argument argues that we live in an interventionist universe. Whereas given (P) and (L) theists should favor a deck-stacking universe. (Leibniz expressed similar sentiments in the Leibniz-Clarke debates.)

Ted,

You wrote:

> I claim that there�s a conflict between the following three
> principles (for the sake of clarity I�ve reformulated (L) and
> (P)):
>
> (L) God desires to create a universe for the purpose of bringing
> about the existence of intelligent moral beings.
>
> (I) It�s incredibly unlikely that complex life would arise in
> our universe.
>
> (P) If God desires to create a universe for the purpose of
> bringing about the existence of intelligent moral beings then
> God will bring about a universe that makes probable the existence
> of complex beings.

OK so far.

> Let me address John�s concerns first. I take it that John thinks
> that the probability claims in (I) and (P) conditionalize on
> different things.

I think that's a sensible reading. In the context of the design argument, I take it that (I) conditionalizes on the absence of God's existence. That is, (I) is implicitly conditional:

(I) If there is no God, then it's incredibly unlikely that complex life would arise in our universe.

I don't think defenders of the design argument need commit themselves to anything stronger than this; certainly, not to anything like:

(I') Complex life arising in our universe is improbable, *simpliciter*.

As for (P), well, it is explicitly a conditional, and so what is asserted in the consequent is conditioned on the antecedent. Roughly speaking:

(P) If there were a God (with such-and-such a purpose), then there is at least some likelihood that there be a universe that produces complex beings.

And again (as you point out below), I don't think defenders of the design argument need commit themselves to anything stronger than this; certainly, not to anything like:

(P') If there were a God (with such-and-such a purpose), then it is *probable* that there be a universe that produces complex beings.

All that is needed for the design argument's modest conclusion that the universe is evidence for God's existence, is to note that the probability of the universe's existence given theism is at least somewhat more likely than the probability of the universe apart from theism. What we're trying to do here is adjudicate among respective hypotheses, given the relevant evidence.

BTW, in my view, there's no need for the design arguer to refer to God's activity in the *consequent* of (P), which is why I slightly reworded it the way I did. I don't know if that has a bearing on your comments about divine agency.

In any event, what's important to point out, and what I think both John and Jeremy have pointed out, is that both (I) and (P) are conditional probability assessments. And since they each conditionalize on different things (theism not being the case; theism being the case), *prima facie* there's no conflict between them. In addition, they can't be combined to militate against (L). (More on this below.)

You continue:

> (I), though, should be understood as saying that given our universe
> (its initial conditions and fundamental laws and its natural
> development) it�s incredibly unlikely that life would arise.

Well, I can't speak for everyone, but most defenders of the design argument I have come across would construe (I) along the lines of the conditional probability spelled out above. It would be pretty self-defeating for this kind of argument to just say that our present universe is improbable, *simpliciter* (i.e., (I')). The whole point is to say that the hypothesis of theism makes the observed evidence more probable than it *otherwise* would be, and that 'otherwise' marks out the alternative conditional probability spelled out in (I).

> The consequent of (P) should be read as claiming that God will
> bring about a universe (initial conditions and fundamental laws
> and natural development) that makes probable the existence of
> complex beings.

Well, I don't really think the design arguer has to put in a reference to divine activity in the consequent of (P). But even if he did, I don't see how that point sustains the charge that (I) and (P) together militate against (L).

> I don�t think that we should include God�s activity (divine
> intervention) in evidence we conditionalize on in (P) {see below}.

But in my estimation, that is exactly what any sensible defender of the design argument *should* do. Conditionalizing on God's activity in the antecedent of (P) brings out the conditionality of the probability claim, and that is precisely why it is so useful. We want theories that predict the evidence (and are thereby confirmed by the evidence).

The reason why we want both (I) and (P) as assumptions, is because we want to say that the probability of the universe given theism is greater than the probability of the universe given naturalism:

P(U/G) > P(U/N)

In short, it's unclear to me how (I) and (P) conflict. Wouldn't any atheist agree that the probability of a red ball given red-ball-factories is at least somewhat greater than the probability of a red ball in the absence of any red-ball-factories?

More to the point, you state that the problem is that "if (P) is true then given (I) the theist has reason to believe that God did not create this world." But that would only be the case if we were to (wrongly) construe (I) as (I'). Then your conclusion might follow by *modus tollens*:

(P) If [there were a God (with such-and-such a purpose)], then [there is at least some likelihood that there be a universe that produces complex beings].

(I') It is unlikely that there is a universe with complex beings.
--
(~L) Therefore, there is no God (with such-and-such a purpose). [(P), (I'), MT]

But fortunately, the design arguer only needs (I), not (I'), and so the *modus tollens* doesn't follow. And so there's no fear of deducing anything that threatens your original (L), it seems to me.

Interestingly enough, even given (I'), it's not clear the above *modus tollens* goes through, for it doesn't follow from the fact that it is *unlikely* that there be such a universe, that there is *no likelihood* that there be a universe like ours. IOW, (I') doesn't constitute the denial of the consequent of (P).

What's confusing (at least to me) is that above you speak of "God's activity" as the "evidence we conditionalize on in (P)." But I think this gets things backwards. The antecedent of (P) isn't 'evidence' for the consequent, or evidence of any sort. The *observed universe* (or fine-tuning constants) is the evidence. What (P) states is that given the truth of a particular hypothesis (a God with a particular purpose), we have some reason to expect the evidence we do in fact have (a universe with complex beings). The 'evidence' belongs in the consequent, not the antecedent.

BTW, in a later comment you bring up the issue of a 'deck-stacking universe' vs. an 'interventionist universe'. Is this a problem additional to the alleged problem of (I) and (P) leading to the denial of (L)? Or is it an attempt to state that problem in a different way?

Sorry for the length of that reply, BTW!

The Reader's Digest version is that once we recognize that the probabilities enshrined in (P) (suitably revised) and (I) are both conditional, it seems doubtful that we can generate an argument that undermines (L). (I') *might* lead us to deny (L), but (I) would not.

I think Greg recognized what's really at issue here. Here is the problem (very roughly) as I originally posed it when Ted and I discussed it on a bike ride.

C=our universe had the initial conditions it did. G=the good aimed at in creating, e.g. rational moral agents. T=theism

On the one hand, the FTA folks looks at *their data*, C, and says that pr(T/C) is high largely b/c pr(C/T) is high, this latter is so b/c if God intends to create a universe efficiently then he�ll create a universe u such that pr(G/C) is high. [I think most FTA folks, certainly myself, think of God as wanting to create a universe likely to have the right result, not just *more* likely than chaos.]

On the other, the BDA looks at *their data*, G, and says that since pr(G/C) is low and explanation of G is called for. God�s special intervention, I, is the explanation, i.e. pr(G/I)=high, this can only be so if pr(G/~I) is accordingly low, but that is to say that pr(G/C) is low.

So FTA says pr(G/C) is high but BDA says pr(G/C) is low. There is no equivocation here, so there is a real disagreement.

Trent Dougherty wrote:

> On the one hand, the FTA folks looks at *their data*, C, and says
> that pr(T/C) is high largely b/c pr(C/T) is high, this latter is
> so b/c if God intends to create a universe efficiently then he�ll
> create a universe u such that pr(G/C) is high. [I think most FTA
> folks, certainly myself, think of God as wanting to create a
> universe likely to have the right result, not just *more* likely
> than chaos.]

Thanks for this. I finally understand what Ted was getting at in the final paragraph of his last comment above, in his contrast between 'deck-stacking universes' and 'interventionist universes'.

Re: the above, I'm not sure why the FTA requires that pr(G/C) is high. All the FTA requires is the assumption that C needs to be within a very narrow range in order to *permit* the emergence of G. That is, C is a necessary condition for G. For the FTA to be sound, one doesn't need the additional assumption that C makes G likely (even if C makes G more likely than it would otherwise be). So one needn't be saddled with the assumption that pr(G/C) is high. Why couldn't the FTA remain agnostic on that latter claim, and thus preserve compatibility with the BDA? He's got a good argument even if he only employs the weaker assumption.

You say that the reason why pr(C/T) is high is because of God's commitment to efficient universe creation, and this in turn gets cashed out in terms of pr (G/C) having to be high for any universe he creates. Isn't this rather contentious? First, I think there are some fairly good arguments against the thesis that God must be 'efficient' in his universe creation (I'm thinking here of some material Tom Morris presents in _Anselmian Explorations_). Second, the hangup can't possibly be that the deck-stacking FTA exemplifies efficiency while the interventionist-BDA does not, for the FTA implies a *great deal* of 'inefficiency': a fairly lengthy, tortuous process over eons that ultimately emerges in life (compared with, say, six days or instantaneously). So the contrast here is a matter of degree at best, not in principle. But if so, then there is little reason for the FTArguer to think that C *by itself* makes G likely, even if C makes G more likely than it would otherwise be.

In his original presentation Ted mentioned that part and parcel of the BDA is the notion of God "guiding the process of evolution along." If this is the case, I don't see why such 'guiding' at the very concrete biological level, with respect to the emergence of the human species, is incompatible with the obtaining of the much more fundamental and general laws specified in C. Why couldn't the obtaining of C provide the necessary context in which the subsequent guiding takes place?

In this respect, one who advocated both FTA and BDA would simply hold to two 'phases' of fine-tuning, broadly speaking, wherein God fixes crucial constants and forms of fundamental physical laws at the very beginning, and then brings about certain biological events at a very specific level much later. At this point, does the objection really come down to: "God wouldn't go to the trouble of fixing C within narrow bounds, if he in fact just intervenes later and gets the result he wanted"? I think that needs to be argued, at the very least. It seems to me that once we rebut the 'efficiency' contention, and go agnostic here, we're free to posit two distinct arguments from two distinct phases of 'fine-tuning' broadly construed.

> So FTA says pr(G/C) is high but BDA says pr(G/C) is low. There
> is no equivocation here, so there is a real disagreement.

I think I see why you would think that, but for the reasons stated above, I don't think the FTArguer need accept that pr(G/C) is high.

Sorry about the atrocious grammar. I changed my subject phrase and didn�t revise accordingly. It sounds pretty funny though, like �So these Irish guys walks into a bar and they says�� :)

�I'm not sure why the FTA requires that pr(G/C) is high. All the FTA requires is the assumption that C needs to be within a very narrow range in order to *permit* the emergence of G. That is, C is a necessary condition for G. For the FTA to be sound, one doesn't need the additional assumption that C makes G likely (even if C makes G more likely than it would otherwise be)�

Sure, you can use the data any old way you please. My point was that it is common for FTA advocates to use the stronger version assuming the efficiency principle. Yes, there is lots of �inefficiency� in the FTA version but not of the relevant kind. All I was adverting to was �further tinkering� like the 18th Century deists objected to. I think there are non-deistic reasons to object to further tinkering, one�s related to Augustine�s doctrine of *rationes seminales*. It is plausible that it is more fitting that God should create in �one fell swoop�, poetic even: God plants a universe seed or lights a universe bomb. IF one is committed to such a principle of efficiency then one WILL want the stronger FTA that makes it likely that C leads to G, even that C is sufficient for G. Or, one can stick with C merely being necessary and use the (sloppy?) two-stage approach you mention above. I took Ted to be assuming the efficiency principle. It is still an interesting piece of logical cartography that one committed to the relevant efficiency principle does not have the option of using the BDA in their cumulative case argument along side the SFTA. As a proponent of the relevant efficiency principle, this came as something of a shocking realization: �I can�t use BDAs anymore!� I think there are probably lots of people with my same commitments who haven�t made the realization, so I think it�s a pretty interesting and relevant result.

Greg,
Thanks for all the comments! I�m glad you see what I was trying to get at in my original post.

About the efficiency principle, see my earlier post on deriving (P) from an intuitive principle of rational action. I�ve been thinking about the efficiency principle and I don�t yet see why theists should reject it. Theists may accept the efficiency principle but deny (L)�that God�s primary reason for creating is to bring about the good of rational moral beings. I�d have to see how this went, but my concern is that denying (L) knocks out the justification for theism predicting C.

Ted, that�s a good thought. I have the same sort of worry. The only way that pr(T/C)>pr(T) is if pr(C/T) is somewhat high. But why think that if C is merely life-*permitting*. Why should learning that raise pr(T)? So I�m going to pick a possible world out of the array of possibility, at random. What are the odds that it�s a theistic one? You say n. Now I tell you the universe has initial conditions C such that C is *necessary* for rational, moral agents (G), but not *sufficient*. If it seems to you that God would want to create a world with G, I don�t see why learning C would raise pr(T). In fact, it might just *lower* it!

I want to expand on that sentiment just a bit. Let I=Got intends to bring about G, C=The initial conditions of u are such that it is unlikely that u will contain G. I think pr(T/I.C)=low! Why would God bring about a world which is unlikely to achieve his goals? You'd have to believe there was some I* like "God intends to bring about G in stages of progressive tinkering." It's quite reasonable to reject that though. I'm not sure what could motivate it.

Now of course even if pr(T/I.C)=low! it could be higher than pr(~T/I.C) and possibly count as some small evidence for T as part of a cumulative case. I have two things to say about that. First, it seems to seriously reduce the value of the FTA, to justify Ted and I calling it the "Weak Fine Tuning Arg" (WFTA). Second, it seems perfectly reasonable to me for a person to have a confidence assignment such that pr(~T/I.C) > pr(T/I.C). So for a potentially wide range of people, the WFTA won't even be some small evidence for T. I can honestly say that in the scenario above where I believe I and C, I'd bet that ~T over T, thinking "If God were going to create a world with G, he'd write it in from the beginning, so this must be a ~T world." So for my confidence assignments, C would not be evidence for T. Though it's reasonable to doubt my assumption about writing in, I doubt I'm *too* idiosyncratic that way, so I think this is a fascinating result. Perhaps someone can offer some reason against the efficiency principle.

[The following reply is probably overkill. Take it as random musings on a related path.]

Trent wrote:

> My point was that it is common for FTA advocates to use the
> stronger version assuming the efficiency principle.

Perhaps I need to broaden my vision somewhat, but when I examine, say, Swinburne's Appendix B in _The Existence of God_, his typical claim throughout the piece is the weaker one: that C (the initial boundary conditions) is *necessary to permit* the emergence of G (rational moral agents, or, at least, the emergence of "bodies which are suitable vehicles for the embodiment of conscious beings" (302)). Here is perhaps the clearest statement of that point (my asterisks):

[1] "One must go on to add that even if the necessary conditions for intelligent life are satisfied, it is still, in Barrow and Tipler's estimate, only likely to occur very rarely... The infrequency of hominoid life or of intelligent life generally is, however, *irrelevant to the argument*. The argument appeals *only* to the fact that certain conditions are *necessary* for intelligent life at all" (310-311).

This seems to state the point fairly clearly, although following is a representative sampling of quotes that hammer it home again and again (my asterisks):

[2] "... laws and boundary conditions had to lie within a certain range *if* there was to be life at all" (303).

[3] "Carbon-based life *requires for its stability* a moderate range of temperature and pressure�and, for its purposes to have much effect, a solid planet on which to live" (305).

[4] "The present consensus of evidence is that certain a priori very unlikely features of laws are *necessary for the occurrence of* carbon-based life" (305).

[5] "I suggest, nevertheless, that such readers read through these pages; they will get the main message, that laws of nature need to have very special features indeed *if carbon-based life is to occur*" (305).

[6] "Given the four fundamental forces and the basic array of fundamental particles (photons; leptons, including electrons; mesons; and baryons, including protons and neutrons), the strengths of forces and masses of particles have to have to each other ratios within certain narrow bands if the larger chemical elements, including carbon, *are to occur at all*, and the Pauli exclusion principle has to hold" (305-306).

[7] "Their small value *makes possible* the long chains of molecules, such as DNA, which make life *possible*" (306).

[8] "Our judgements as to just how narrow are the ranges within which crucial variables of boundary conditions and some of the constants of scientific laws have to lie *in order to permit* the evolution of intelligent life must be very tentative" (310).

[9] "All that is clear is that, within the kind of region of laws and boundary conditions for which we can get some feeling of proportions, the range *allowing life* is probably very small indeed" (311).

[10] "... and then consider the force of the further evidence (e) that the initial conditions and constants of laws had just those values which *allowed life to evolve*" (311).

Now what's striking here is how the opening paragraph of the Appendix seems to obfuscate the main point enshrined in [1]-[10] above. While the *argument* actually presented in the Appendix only relies on the weaker assumption about the *necessary conditions* of life (see above), that argument is actually *advertised* as making use of a much stronger claim, that the boundary conditions actually make life *probable*:

[11] "In this appendix I seek to examine the force of an argument to the existence of God from the widely discussed fact that the boundary conditions of our universe and the laws of its evolution are of a very special kind which alone could lead to the evolution of intelligent life *and in fact make it probable*" (300).

[12] "In considering the force of this evidence, I shall assume that there is a universe and that there are laws of nature, and shall consider whether the fact that those laws have a particular form, and the boundary conditions of the universe have a particular form which alone permit the evolution of intelligent life *and in fact make it probable*, provides a good C-inductive argument for the existence of God" (300).

Since (as far as I can tell) the body of the Appendix only makes use of the weaker assumption about the evidence (C is necessary for life), it seems fairly misleading to characterize the argument *at the outset* as making use of the stronger assumption (that C makes life probable). Perhaps this was an oversight on Swinburne's part. In any event, it makes clear how important it is to make clear which assumption one is appealing to, in making the FTA. Perhaps other influential formulations are equally misleading as to what is actually driving the argument; I don't know.

More on your 'inefficiency' comments, later, if time permits. (I see you've also posted a few more things while I was composing this.) And thanks for your observations; your spelling out of the variables C, G, and T in your earlier comment greatly clarified things for me. I'm happy to accept your (and Ted's) larger point that certain formulations of the FTA could tend to undermine the BDA.

Trent wrote:

> Yes, there is lots of "inefficiency" in the FTA version but not
> of the relevant kind.

Well, I'm not so sure. See below.

> All I was adverting to was "further tinkering" like the 18th
> Century deists objected to. I think there are non-deistic
> reasons to object to further tinkering, one's related to
> Augustine's doctrine of *rationes seminales*.

Perhaps there are. But I think it's important to point out that if these are objections to the BDA, then they hit a much broader target. Any orthodox Christian (Jewish, Muslim) theist is fairly strongly committed to divine intervention (in *some* significant sense) subsequent to creation, an intervention specifically designed to achieve divine purposes in a variety of ways and on a multitude of occasions. So "reasons to object to further tinkering" wouldn't simply reflect poorly upon the BDA, but upon those religious traditions in which most philosophers of religion are situated. So it's not a problem (if it is a problem) that uniquely afflicts the BDA. In short, if a Christian *et al* had to endorse the 'efficiency principle' in order to make a plausible cumulative case for vanilla theism, he'd have to have resources sooner or later to rebut the legitimacy of such a principle, at least if he wanted others to take his more specific religious commitments seriously.

> It is plausible that it is more fitting that God should create in
> "one fell swoop", poetic even: God plants a universe seed or
> lights a universe bomb. IF one is committed to such a principle
> of efficiency then one WILL want the stronger FTA that makes it
> likely that C leads to G, even that C is sufficient for G.

It seems to me that considerations of 'fittingness' are just a restatement of the efficiency concern. And on that score, it strikes me that efficiency is only a virtue in a creative agent if he is saddled with constraints on either time or resources. It is because human beings labor under both that *we* find efficiency to be a virtue, but there's little reason to extend that to God.

As for satisfying our more 'poetic' impulses, it's hard to see how any FTA -- which secures the emergence of its rational moral agents only via the march of natural selection across the globe -- could do the job, requiring as it does the chaos of intense competition, and as a consequence eons of mass suffering, species extinction, and nature in general as red in tooth and claw. The 'seed' metaphor tends to hide this brutal fact :-) But then on this telling FTA wouldn't seem to have much of an edge over BDA in the department of fittingness or aesthetic pleasingness. Surely instantaneous creation of complex moral agents would trump both scenarios on that score.

Perhaps there are goods realized in the lengthy, complicated FTA scenario that would not be realized by instantaneous creation of complex moral agents, goods that are largely indiscernible to us. In the face of such ignorance, the safer route is to make use of the weaker assumption: that if *one* of God's purposes was to secure complex moral agents, then *at the very least* he would be interested in bringing about conditions necessary to the purpose. Appealing to conditions necessary to one part of a complex goal is different from appealing to conditions sufficient for the achievement of the entire goal (or that make its achievement probable). No doubt God *is* interested in bringing about these more substantive conditions. Still, the FTA doesn't have to appeal to them in order to have a good case. (This has been questioned in recent posts; I'll try to address it if time permits.)

> Or, one can stick with C merely being necessary and use the
> (sloppy?) two-stage approach you mention above.

I'm not sure why this is 'sloppy' (i.e., initial fine-tuning plus subsequent guidance of biological evolution). It would only be sloppy if the mere emergence of complex moral agents were the *only* goal of the divine program for creation, and if God labored under time and resource constraints in achieving that goal. Since all of that is doubtful, then for all we know a whole host of goods are realized in this scenario that would not otherwise be realized, and the means chosen are the means best suited to the purpose.

> I took Ted to be assuming the efficiency principle. It is still
> an interesting piece of logical cartography that one committed
> to the relevant efficiency principle does not have the option of
> using the BDA in their cumulative case argument along side the SFTA.

Yes, it's a clever, even brilliant point; certainly not one I had ever seen before. But I do think the "relevant efficiency principle" is too strong for most theists interested in defending a package that is compatible with their larger religious commitments.

> As a proponent of the relevant efficiency principle, this came
> as something of a shocking realization: "I can't use BDAs
> anymore!" I think there are probably lots of people with my
> same commitments who haven't made the realization, so I think
> it's a pretty interesting and relevant result.

Yes, I agree. I hope I didn't give the impression of trivializing your larger concerns. I apologize if I did.

Ted wrote:

> Thanks for all the comments! I'm glad you see what I was trying
> to get at in my original post.

Thanks to you!

> I've been thinking about the efficiency principle and I don't
> yet see why theists should reject it.

Let me know if the post prior to this one doesn't clarify things a bit.

> Theists may accept the efficiency principle but deny (L)�that
> God's primary reason for creating is to bring about the good of
> rational moral beings.

I'm a bit confused here. Your spelling out of (L) above doesn't seem to be saying the same thing as your previous (and revised) specification of (L):

"(L)--God desires to create a universe for the purpose of bringing about the existence of intelligent moral beings."

Which version of (L) are you endorsing?:

(L1) God's primary reason for creating is to bring about the good of rational moral beings.

(L2) God desires to create a universe for the purpose of bringing about the existence of intelligent moral beings.

It seems to me that (L1) is more sweeping than (L2). (L2) is compatible with God having a variety of purposes in creating the universe, including but not limited to the mere existence of intelligent moral beings. It might be that other purposes are in play (perhaps centering on God himself), indeed purposes that are more 'primary'.

> I'd have to see how this went, but my concern is that denying
> (L) knocks out the justification for theism predicting C.

Well, I have no interest in denying (L2). But I'm not sure I have to, since I don't accept the efficiency principle, nor that pr(G/C) is high.

Trent wrote:

> Ted, that's a good thought. I have the same sort of worry. The
> only way that pr(T/C)>pr(T) is if pr(C/T) is somewhat high.

Perhaps I'm missing something, but why would one think this? As long as pr(C/T) is greater than pr(C/~T), then C is evidence for T, and thus pr(T/C)>pr(T). Have I got that wrong somehow? Only if one thought that C by itself made T *probable* would one need to hold that pr(C/T) must be high.

> But why think that if C is merely life-*permitting*. Why should
> learning that raise pr(T)?

These are two distinct issues, of course: whether pr(C/T) is high, and whether C raises pr(T). I'm only committed to the latter, not the former.

> So I�m going to pick a possible world out of the array of
> possibility, at random. What are the odds that it�s a theistic
> one? You say n. Now I tell you the universe has initial
> conditions C such that C is *necessary* for rational, moral
> agents (G), but not *sufficient*. If it seems to you that God
> would want to create a world with G, I don�t see why learning
> C would raise pr(T). In fact, it might just *lower* it!

Why? C is *a priori* improbable. So if we have a theory on offer that increases its probability (rather than leaving it inexplicable), then *ceterus paribus* C is evidence for the theory. If, as is certainly the case, God is interested in bringing about those conditions which are necessary and sufficient for his purposes, then *a fortiori* he is interested in bringing about the necessary conditions. That gives him interest in bringing about C. Now, if the alternative explanations of u give us *no* reason to expect C, but if T gives us *some reason* to expect C, then C raises pr(T) (even if it doesn't make T probable). Have I gone wrong here?

Trent wrote:

> I want to expand on that sentiment just a bit. Let I=God intends
> to bring about G,

OK.

> C=The initial conditions of u are such that it is unlikely that u
> will contain G.

Why would any FTA want to appeal to *that*? :-) It may no doubt be *true* that C by itself doesn't make it likely that u will contain G. But so what? What matters is that God is interested in bringing about C (among *other* things, of course) as a means of achieving G. And as long as T gives us some reason to expect C, then it raises pr(T). If it doesn't raise the probability of the alternative theories (because they don't lead us to expect C), then C constitutes evidence for T and not for these other theories.

Your original specification of C was:

"C=our universe had the initial conditions it did."

That's compatible with your more wordy version of (C) above, but there's no need for FTA to appeal to that version.

> I think pr(T/I.C)=low! Why would God bring about a world which
> is unlikely to achieve his goals?

I agree, he wouldn't. But it doesn't follow from the fact that (1) "God brings about conditions necessary for the achieving of his goals" that therefore (2) "God brings about a world which is unlikely to achieve his goals." (1) is compatible with (3) "God brings about conditions sufficient for the achieving of his goals," which in turn rules out (2). The fact that FTA employs (1) doesn't mean that it employs or even requires (2).

> You'd have to believe there was some I* like "God intends to bring
> about G in stages of progressive tinkering."

Perhaps. At a minimum you'd have to distinguish necessary from sufficient conditions, regardless of when or how they are realized.

> It's quite reasonable to reject that though. I'm not sure what
> could motivate it.

I've tried to address this a bit in a previous post.

> Now of course even if pr(T/I.C)=low! it could be higher than
> pr(~T/I.C) and possibly count as some small evidence for T as
> part of a cumulative case. I have two things to say about that.
> First, it seems to seriously reduce the value of the FTA, to
> justify Ted and I calling it the "Weak Fine Tuning Arg" (WFTA).

Perhaps the kind of conclusion the FTA should be aiming for is not that it shows that theism is probable *simpliciter*, but that theism is much more probable than any of the alternative explanations of the evidence on offer. So, if we have to make a choice, we know what choice to make, that is, what the rational choice would be. I'm not sure I'm saying anything incompatible with what you say above, however.

> Second, it seems perfectly reasonable to me for a person to have
> a confidence assignment such that pr(~T/I.C) > pr(T/I.C). So for
> a potentially wide range of people, the WFTA won't even be some
> small evidence for T.

First, I think you want to rewrite that. If, as you say above, "I=God intends to bring about G," then since I entails T, pr(~T/I.C) would be fairly high, right? :-)

Second, if the absence of theism doesn't give them any reason to expect C, whereas theism gives them at least *some* reason to expect C, doesn't that mean that C is evidence for T and not for ~T?

> I can honestly say that in the scenario above where I believe I
> and C, I'd bet that ~T over T, thinking "If God were going to
> create a world with G, he'd write it in from the beginning, so
> this must be a ~T world."

Again, let's say he *does* write 'it' in from the beginning. "It" is a lot of things, not just C. C by itself may not ensure G or even make it likely, but as long as God has some reason to bring about C (and this is compatible with saying that he has reason to bring about a lot of *other* factors to bring about G), and the alternatives to theism give us no reason to expect C, then I don't see any reason to "bet that ~T over T".

> Perhaps someone can offer some reason against the efficiency
> principle.

Perhaps someone can offer some reason *for* it :-)

I'm done for the day, BTW, to the great relief of many I'm sure ;-)

Let me point you to a particular passage which�I think�sheds some light on the issue. Given this passage, it seems that Swinburne advocates Pr(G/C) is high (and thus appeals to the SFTA). Further my reading of this passage suggests that Swinburne doesn�t think the BDA works.

(p. 303) �He [God] could do so [i.e., bring about a universe containing intelligent beings] either directly or indirectly, making the world with boundary conditions and scientific laws **such as to give rise to** intelligent organisms. All the evidence accumulated by scientists over the past 200 years shows overwhelmingly that present-day intelligent organisms evolved gradually from inanimate matter in accord with scientific laws over thousands of millions of years. So God did not produce intelligent organisms directly. But if all the evidence is that the occurrence of boundary conditions and laws *such as to permit and *make probable** the evolution of intelligent organisms are a priori very unlikely that is evidence that God brought them about�. **He made an intelligent-organism-producing universe.** With e as laws and boundary conditions **such as to make likely** the evolution of intelligent organisms, h as the hypothesis of theism, k as the existence of a universe governed by some laws of nature or other, P(h/e.k) > P(h/k), indeed [much greater]."

OK, I lied. This is the last post ;-)

Ted wrote:

> Let me point you to a particular passage which�I think�sheds some
> light on the issue.

Yes, I read that passage too. However, I think it's significant that (1) it is the *only* passage of its kind in the entire body of the Appendix, and (2) if you look closely, the claims made there ("give rise to"; "make probable"; "make likely") are not so much as *argued* from *any* of the specific fine-tuning data he presents. Perhaps I've overlooked something, but I'm pretty confident of this conclusion. Given this, I tend to place this passage in the "incautious overstatement" category. The citations [1]-[10] I provided seem to exhibit the typical characterization of the data throughout the Appendix. Indeed, he seems in [1] to go so far as to specifically *deny* that he's going beyond necessary conditions for life.

Well, the book has recently been revised again, in response to fifteen years of further criticism in the literature. Perhaps this issue has been addressed :-)

> Given this passage, it seems that Swinburne advocates Pr(G/C) is
> high (and thus appeals to the SFTA).

Well, it's interesting. The quote you provide essentially deletes my [2], which occurs at the end of this paragraph. There, he says that:

[13] "In the main body of the book I commented (p. 136) that the occurrence of life was a rare event in the vast spaces and many millennia of world history, and so supposed that its occasional occurrence would not be unexpected given the operation of some laws of nature or other, and so not very strong evidence of a Creator. I did not take seriously the fact that laws and boundary conditions had to lie within a certain range *if there was to be life at all*; and for that reason I underestimated the strength of an argument of this kind." (my asterisks)

In other words, the reason why he now thinks the argument from "the occurrence of life" is stronger than he previously thought, is that he "did not take seriously the fact that laws and boundary conditions had to lie within a certain range if there was to be life at all." If that's the evidential hinge on which his newfound appreciation of the argument turns, notice that it falls far short of asserting that the laws and conditions make life probable.

But I admit that certain passages of the Appendix appear to be in tension with other passages. My only point is that while he clearly *argues* for the 'necessary condition for life' construal of the fine-tuning data, his occasional 'makes life probable' construal seems to hang in thin-air, unsupported by any of the data he actually cites. So the argument as presented -- in distinction from the argument as misleadingly advertised ;-) -- seems closer to my reading.

> Further my reading of this passage suggests that Swinburne
> doesn't think the BDA works.

Not sure why you think that. Could you elaborate? Swinburne envisions the FTA, which concerns the emergence of "bodies which are suitable vehicles for the embodiment of conscious beings" (302), as compatible with his earlier argument in ch. 9 from consciousness. And there he is clear that there is *no* scientific explanation of consciousness. That is, human consciousness is an aspect of God's overall goal in producing complex moral agents, but that aspect is *not* explicable in terms of the initial laws and boundary conditions of the universe. You need more. I'm not saying that constitutes 'subsequent tinkering'. I'm just saying it's significant :-)

But I think it *does* constitute 'tinkering'. Note how he says, under the heading "The Scientific Inexplicability of Consciousness":

[14] "I now wish to argue that embodied agents of the kind which we find on earth, would be unlikely to come into existence in the normal course of things, that is, *but for the action of a God*. I argue that, even given the existence of an orderly inanimate universe, the prior probability of the existence of embodied agents, conscious beings who bring about effects through intentions (in the way analysed in Chapter 2) is very low, *unless a God intervenes in the natural order to bring about their existence*" (160-161, my asterisks)

There you have it, I think. The FTA, which concerns bodies alone, is (at least intended to be) compatible with the argument from consciousness, which concerns the emergence of consciousness by divine intervention into the natural order. These are different arguments which capitalize on different things, and I still don't see the incompatibility.

Near the end of the section he presses home the point:

[15] "In suggesting that the evolution of conscious beings is not explicable by normal scientific processes, I do not wish to deny that there is a story of the gradual evolution of conscious beings out of inorganic matter. My only claim is that this evolutionary story is not to be explained in scientific terms, for there are no scientific laws correlating the bodily and the mental, only correlations of too limited a scope to constitute laws. I do not wish to deny that there is a satisfactory scientific explanation given by the biological theory of evolution of the evolution of more and more complex beings which interact in increasingly complex ways. But there is not, I have argued, a scientific explanation of their increasing complexity of physiological organization and behavioural response giving rise to the particular conscious life to which it does give rise and which subsequently in part determines what happens in and through the body. So then the argument of this section has been designed to show that the existence of conscious beings would not have come about by normal physical processes" (173).

Duh.

I wrote to Trent, three comments ago:

> First, I think you want to rewrite that. If, as you say above,
> "I=God intends to bring about G," then since I entails T,
> pr(~T/I.C) would be fairly high, right? :-)

I should have said, of course, "low" rather than "high".

We must have been writing at the same time! See you tomorrow. BTW-what's the prior probability that we would post back-to-back 17-point replies? Was that fine-tuned or what?

[1]�Any orthodox Christian (Jewish, Muslim) theist is fairly strongly committed to divine intervention (in *some* significant sense) subsequent to creation,�

Right, but those things *require* intervention b/c they concern beings with free will. The creation of the conditions for beings with free will does not.

[2] �efficiency is only a virtue in a creative agent if he is saddled with constraints on either time or resources�

But there are also aesthetic aspects of efficiency, e.g. it�s simplicity which is comparable in some ways to a certain gracefulness. This could motivate God�s concern in how he creates regardless of his lacking time and resource constraints.

[3] �nature in general as red in tooth and claw�

Indeed, and I think this is some of the best evidence against theism. However, I�m not going to let relative ease of theodicy guide my inferences. Good point though, I really worry about this part. However, BDA folks have got the same problem, it�s just that God did it directly rather than indirectly, which seems to make it worse. Following van Inwagen, I�m willing to allow enough chance in the plan to account for the nastiness. Or else I can just do what most theists do with evil and say there was a good but inscrutable reason why it was on the whole better to do it this way. I think this is a *general* problem and I don�t think my views on the matter make it worse.

[4] �Surely instantaneous creation of complex moral agents would trump both scenarios on that score�

Augustine addresses this head-on in the prime passages about the doctrine of rationes seminales (a version of which I see myself as promoting). In fact, the RS doctrine is formulated precisely to harmonize what you say with the progressive account in Genesis. On the one hand, Augustine, like you deduces that�given efficiency concerns�instantaneous creation makes the most sense. Yet, on the other, Genesis clearly says creation took place progressively. His harmonization, the RS doctrine, is that God *did* create the U instantaneously, in that the moment he created the �seed� the job was done. It just had to unfold. I think the RS doctrine does the trick. I think it�s brilliant, poetic, and true.

[5] �Yes, it's a clever, even brilliant point; certainly not one I had ever seen before. But I do think the "relevant efficiency principle" is too strong for most theists interested in defending a package that is compatible with their larger religious commitments.�

Why, thank you. It really keeps coming back to the EP. I think I�ve addressed the �larger commitments� issue above. If that�s the only consideration contrary to EP, I feel good about EP.

[6] �As long as pr(C/T) is greater than pr(C/~T), then C is evidence for T�

Right, but I have worries even about that which I expressed below the above in the original and which will recur shortly.

[7] �These are two distinct issues, of course: whether pr(C/T) is high, and whether C raises pr(T). I'm only committed to the latter, not the former.�
They are indeed two separate issues, Ted and I think theists *should* be committed to the former, we hope to motivate that more. Some of what I say below�I hope�does. The main problem is that if pr(C/T) isn�t somewhat high, then the explanatory power of T drops correspondingly. The whole point of explanatory power is the directly proportional relationship between pr(h/e) and pr(e/h). If the latter is low e can indeed still count as evidence for h, but that�s not saying much. My being alive at the time is evidence�in this contrastive sense you are using�that, for any given murder that took place at the time�that I did it. However, you�d never get me indicted. And you won�t get God indicted on charges on creation unless pr(e/h) is somewhat high, not just contrastively higher than pr(e/~h). I think this is a close cousin of what Ted was trying to say.

[8] �if the alternative explanations of u give us *no* reason to expect C, but if T gives us *some reason* to expect C, then C raises pr(T) (even if it doesn't make T probable). Have I gone wrong here?�
Absolutely not, you are exactly correct, of course. However, the point I just made comes into play here. The WFTA provides *some* evidence�I�m not denying that�it�s just that if C is going to be *much* evidence, then pr(C/T) needs to be someThat high. Technically, as you say, the degree to which C confirms to T is the degree to which pr(C/T)>pr(C/~T). Since, as you say, the latter might be super-low, the former can be very-low and C still be evidence for T. But what I tried to say was that I�m not so sure about that inequality. Let�s think in frequentist terms again, just as a heuristic. Take the class of worlds W where pr(G/C) is very low. To the extent that we have confidence in an EP, to that extent we will not think the proportion of worlds in W which are created by God will be much greater than the proportion which are random. For any w such that w e W, we�re not going to have much reason to suppose that it is one which God would bring about. Now you make a very good point that every life-producing world is ipso facto a life permitting world and I take that to show pretty well that the proportion of worlds in W which God could be expected to bring about is greater than the compliment. However, all that means, again, is that C meets the technical Bayesian definition of �evidence� and that�s just not very impressive. We�d need to know that the proportion of worlds in W was *quite a bit* bigger than its compliment to infer that it was very good evidence. I think when ordinary folks talk of �evidence� they mean what a Bayesian means by �pretty good evidence.�

[9] �Why would any FTA want to appeal to [C=The initial conditions of u are such that it is unlikely that u will contain G]? It may no doubt be *true* that C by itself doesn't make it likely that u will contain G�
The FTA advocate *wouldn�t*, but the BDA advocate *must* which was our original point about there being conflict. Also, I don�t think it *is* true, but the defense of that is another story. I�m not sure about your last statement, either I�m reading it wrongly or it contains a typo.

[10] �as long as T gives us some reason to expect C, then it raises pr(T). If it doesn't raise the probability of the alternative theories (because they don't lead us to expect C), then C constitutes evidence for T and not for these other theories�
See comments above on �evidence.� I just don�t think the WFTA provides much evidence at all. I could be persuaded otherwise if I had some reason to expect God to create with saltation. I just don�t see any motivation for that at this point.

[11] �> I think pr(T/I.C)=low! Why would God bring about a world which
> is unlikely to achieve his goals?
I agree, he wouldn't. But it doesn't follow from the fact that (1) "God brings about conditions necessary for the achieving of his goals" that therefore (2) "God brings about a world which is unlikely to achieve his goals."�
But it does with the relevant EP, so it all keeps coming back to that.

[12] �> You'd have to believe there was some I* like "God intends to bring
> about G in stages of progressive tinkering."
Perhaps.�
I think this might be a bigger problem than you realize for WFTA, for in addition to providing scant evidence, it complicates T (by adding parameters) even further reducing the extent to which C is evidence for T.

[13] �Perhaps the kind of conclusion the FTA should be aiming for is not that it shows that theism is probable *simpliciter*, but that theism is much more probable than any of the alternative explanations of the evidence on offer.�
I�m not claiming that even the SFTA yields pr(T/C)>.5 I�m claiming that it needs to be the case that pr(C/T)=high needs to be the case even for SFTA to be a good C-inductive argument. Note that Swinburne appears to think this as well (TEG 1st ed., 147-148). In fact, he says that �It appears that P(e/h.k) equals 1� (ibid.). where e=the laws of nature.

[14] �theism is much more probable than any of the alternative explanations of the evidence on offer. So, if we have to make a choice, we know what choice to make, that is, what the rational choice would be. I'm not sure I'm saying anything incompatible with what you say above, however.�
Indeed it is not, in fact, it expresses my sentiments exactly. That�s all Swinburne was going for anyway (cf. _Faith and Reason_). I think this is the bottom line. Good point.

[15] �First, I think you want to rewrite that. If, as you say above, "I=God intends to bring about G," then since I entails T, pr(~T/I.C) would be fairly high, right? :-)�
Not so fast! You�re assuming S5 principles, I�m not. I was standing outside the array of possible worlds and so in some of those worlds there�s a God and in some there aren�t. So I could be true de re, yet T be false in the w at hand.

[16] �Again, let's say he *does* write 'it' in from the beginning. "It" is a lot of things, not just C. C by itself may not ensure G or even make it likely�
I�m assuming quite the opposite. Part and parcel of the RS doctrine is that �It� is all in C. In fact, I think there is a deduction from C to G, but I don�t need that thesis. I do think pr(G/C)>>.5 and am willing to defend that.

[17] �> Perhaps someone can offer some reason against the efficiency
> principle.Perhaps someone can offer some reason *for* it :-)�
Well, I�ve suggested that the elegance of simplicity is a consideration in it�s favor. That�s not saying *much* but it is something. I�d like to hear something against it.

Greg, thanks for all the good thoughts. The main thing that has come of considering all your points is that it all depends on the EP. As a result, I�d really like to hear your thoughts on that. I�ll see if I can�t make a defense of it. The problem is that it just seems obvious to me, the hardest things to defend.

It�s a new day and there�s philosophy to do! :)

Greg�let me say again that I really appreciate all your comments. Your comments are quite helpful and very good. Given my time-constraints, I can only reply to a few things (alas).

[On the EP�the efficiency principle]

Greg writes:
"it strikes me that efficiency is only a virtue in a creative agent if he is saddled with constraints on either time or resources. It is because human beings labor under both that *we* find efficiency to be a virtue, but there's little reason to extend that to God."

I agree that for beings with limited resources efficient action is better. I do not, however, want to reason to EP through any principle like that. As you point out it that will not work to get EP. Rather I conceive of EP as a principle of rational action. Rational action is efficient. Now, why think that? Well, I confess, when I consider it, it just has that �ring of truth� {probably, one of the more dangerous things to appeal to when trying to convince others!}.

Let me try to motivate it a bit. Rational action (or at least the species of rational action needed to get the relevant EP) is action based on reasons. When a rational agent acts on the basis of reasons the agent brings about a state of affairs for the purpose of achieving the agent�s goal. If an agent brings about a state of affairs for the purpose of achieving the agent�s goal and yet it�s not likely that that state of affairs will achieve his goals then some explanation is called for. For instance, if I want a cup of tea and take tea bags, water, and the kettle and put them in the dishwasher, then some explanation of my action is called for. Why, because, given that I want a cup of tea putting all those things in the dishwasher does not make it likely at all that I�ll get a cup of tea {Feel free to come up with another example}. I submit that we explain that by either noting that the agent is irrational in some way or that the agent lacks the relevant knowledge or power. If that�s the case then I think an efficiency principle is true.

[On Swinburne and the BDA]

Greg writes,
[in response to my claim that further my reading of this passage suggests that Swinburne doesn't think the BDA works].
"Not sure why you think that. Could you elaborate?"

Ha! I had thought about qualifying my claim to exclude the argument from consciousness but didn�t for reasons of efficiency! I�m not yet convinced that the argument from consciousness is a BDA, though the passages you cite pull me to think something like BDA reasoning is going on. Note, however, that at the beginning of chapter 9 on the argument from consciousness Swinburne seems to distinguish teleological arguments from other arguments�one of which is the argument from consciousness. Swinburne writes, �Teleological arguments� appeal to the universe�s having some very general over-all characteristic. The subsequent arguments which I shall consider, appeal to more specific features of the universe� (p. 152). This led me to think that the argument from consciousness is not a BDA style-argument.

Prior to your post I had conceived of the argument from consciousness as committed to the premise that science *cannot* explain how consciousness arose or, for that matter, consciousness itself. The BDA on the other hand allows for the possibility that consciousness arose naturalistically, but that it�s incredibly improbable. But the last passage you cite [15] makes me hesitant about this. Do you agree though that apart from the argument from consciousness Swinburne seems to indicate that the BDA does not work?�I�m thinking of a BDA using facts about non-consciousness complex organisms.

[More to come tonight� Cheers!]

An amplification of Ted here:

He writes: "an agent brings about a state of affairs for the purpose of achieving the agent�s goal and yet it�s not likely that that state of affairs will achieve his goals then some explanation is called for."

I think this points to an important fact about a plausible EP: it's defeasible. It's open to a good explanation of why there was a deviation from effeciency (though one wonders if the resulting action still wouldn't be considered perfectly effecient *given the agent's goals*). So there are two consequences from this for the advocate of the WFTA. First, we need to *hear* the explanation why God's first step in bringing about G is an action C such that Pr(G/C) is low. Second, and this point comes from Ted (it is in fact a point that Swinburne himself emphasized quite a bit in class), this further complicates T and so lowers it's prior and so eventually it's posterior probability.

In re-reading through the posts (which printed out at 26pp!) I caught an inference that seemed crucial to an argument against a concern of our's about the WFTA. I�ll keep this post to this one point.

Greg writes, �If, as is certainly the case, God is interested in bringing about those conditions which are necessary and sufficient for his purposes, then *a fortiori* he is interested in bringing about the necessary conditions.�

I may be wrong, but I don�t think this inference is valid b/c it contains an intentional operator �interested in bringing about��especially if you add *merely* necessary conditions. Intentional operators are bad candidates for preserving logical relations. Here�s a quick counterexample:
I desire to have three priests over Father Ted, Father Dougal, and Father Jack but it does not follow that I desire to have Father Jack over. For I enjoy seeing Ted and Dougal pick on Father Jack. But Father Jack by himself is rather boring. (He just sits in his chair all day and drinks). In general S may be interested in bringing about necessary and sufficient conditions for S's purposes but fail to be interested in bringing about mere necessary conditions.

OK, I intentionally took a day off so as not to burden the blog with too many of my posts. (Six posts in a day, especially from a newcomer, is a bit much for most, I imagine.)

First, a reply to the most recent comment from Ted:

>> If, as is certainly the case, God is interested in bringing
>> about those conditions which are necessary and sufficient for
>> his purposes, then *a fortiori* he is interested in bringing
>> about the necessary conditions.
>
> I may be wrong, but I don't think this inference is valid b/c it
> contains an intentional operator "interested in bringing about"
> �especially if you add *merely* necessary conditions.

But I didn't stuff *merely* necessary conditions in the consequent, just necessary conditions. It doesn't follow from the inference as I've stated it above, that if God is interested in bringing about conditions which are necessary and sufficient for his purposes, that he is *merely* interested in the necessary conditions, or that he is interested in bringing about *merely* the necessary conditions, or *only* the necessary conditions. I'm making a much more modest claim, one that I thought was fairly uncontroversial: if God is interested in bringing about X and Y, then he's interested in bringing about X. It doesn't follow from this that he's interested in bringing about a state of affairs in which *only* X is the case.

> Intentional operators are bad candidates for preserving logical
> relations. Here's a quick counterexample: I desire to have three
> priests over Father Ted, Father Dougal, and Father Jack but it
> does not follow that I desire to have Father Jack over.

Doesn't it? You desire to have Father Jack over, but *in a certain context*, namely, when he's getting ribbed by Ted and Dougal. It seems to me that the latter is compatible with the former, and indeed entails it. That's why your fuller description of your overall desire -- "I enjoy seeing Ted and Dougal pick on Father Jack" -- entails the presence of Jack.

It all comes down to the fact that I never said that God would *merely* want the necessary conditions to come to pass, just that he wants them to come to pass (presumably, because they are a necessary means of something *else* coming to pass).

The upshot is if God intends to create a life-producing universe, then *a fortiori* he intends to create a life-permitting universe (although that is not *all* he intends). And the fine-tuning data (boundary conditions and forms of laws) ensure the latter, but not necessarily the former. I am not aware of any version of the FTA that argues *from the specifics of the fine-tuning data* that said data makes likely a life-producing universe. Rather, the characteristic claim from such data is that, if life anything like ours were to be possible at all, the conditions and laws would have to be within such-and-such a range.

Think of it this way. If I'm interested in building a treehouse for my kids, I think I'm interested in giving it a floor, even if the presence of the floor doesn't ensure or even make likely that there is a treehouse. I think it would be positively irrational to want to build a treehouse but not be interested in giving it a floor. But that doesn't mean that I'm interested in *merely* the floor, or something like that.

Trent wrote:

> BTW-what's the prior probability that we would post back-to-back
> 17-point replies? Was that fine-tuned or what?

Not sure :-) But just to clarify, I was only numbering the Swinburne citations, for easy reference later (if desirable). I wasn't trying to number my own points (which I think were fewer than 17 :-)

>> Any orthodox Christian (Jewish, Muslim) theist is fairly
>> strongly committed to divine intervention (in *some*
>> significant sense) subsequent to creation,"
>
> Right, but those things *require* intervention b/c they concern
> beings with free will. The creation of the conditions for beings
> with free will does not.

Then I'm not sure I understand what the objections to "further tinkering" really amount to. Apparently, there are 'deistic' and 'non-deistic' (i.e., Augustinian) objections to further tinkering. I think the 'deistic' objections are fairly global, aimed at *any* further intervention once the universe gets going according to the laws God set up. So if as you say bringing about divine purposes with respect to free-willed-beings *requires* divine intervention, then the deistic objection is a non-starter. That means all we have left is the non-deistic, Augustinian objection (which I'll get to below). So consider my remarks above, and the paragraph in which they were embedded, to be aimed at the deistic objection.

BTW, just to clarify, above you speak of "the creation of the conditions for beings with free will," and how we don't need divine intervention for this. Are these necessary, or sufficient, conditions? I think you mean sufficient, since your point is that God's bringing about C makes it likely that the beings specified in G (complex moral agents) will emerge, correct?

>> efficiency is only a virtue in a creative agent if he is
>> saddled with constraints on either time or resources
>
> But there are also aesthetic aspects of efficiency, e.g. it's
> simplicity which is comparable in some ways to a certain
> gracefulness. This could motivate God's concern in how he
> creates regardless of his lacking time and resource constraints.

So the point is that God is concerned with looking graceful to *us*, and this means he must be efficient? I don't buy this. What would warrant such a principle?

In addition, *for all we know*, the sum total of God's actions -- BDA 'tinkering' and all -- *would* look graceful to us, if only we were aware of the full range of purposes that God is trying to accomplish by his actions. So I don't think it follows from the fact that a portion of God's actions, or a specific action taken out of context, looks 'ungraceful,' that therefore the divine program for creation is ungraceful. Or do you really intend the stronger claim that each and every divine action must appear graceful to us, from the limited perspective by which we consider it?

>> nature in general as red in tooth and claw
>
> However, BDA folks have got the same problem, it's just that God
> did it directly rather than indirectly, which seems to make it
> worse.

Well, my point was that "FTA wouldn't seem to have much of an edge over BDA in the department of fittingness or aesthetic pleasingness," not that it doesn't have an edge at all.

Although I'm unclear on how BDA really does 'make it worse'. Unless we want to go Socinian ;-), the FTA-God surely *knew* that the 'nature red in tooth and claw' scenario would unfold over eons of time among non-free-willed creatures, and yet he brought about the conditions for that anyway. How is that any better, morally speaking, than God guiding evolution 'directly'? Indeed, I don't think the BDA posits that God guided *everything* directly, just at crucial points, so that further diminishes the gap between the two scenarios, with respect to direct divine involvement in bringing about pain and suffering. And to top it off, perhaps on the BDA scenario the points of divine guidance *lessened* the competitive chaos (and ensuing pain and suffering) that would have otherwise obtained. At this point, I can't see how FTA has an edge over BDA, *at all*, such that, say, someone committed to the FTA would have a good reason to distance himself from the BDA.

> I think this is a *general* problem and I don't think my views
> on the matter make it worse.

Right, but I wasn't claiming that your views on the FTA made it worse, but rather that they clearly didn't show that *BDA* was any worse than FTA, in terms of satisfying 'our more poetic impulses'.

>> Surely instantaneous creation of complex moral agents would
>> trump both scenarios on that score
>
> Augustine addresses this head-on in the prime passages about
> the doctrine of rationes seminales... God *did* create the U
> instantaneously, in that the moment he created the "seed" the
> job was done. It just had to unfold.

Keep in mind that I'm still not conceding the principles of efficiency or aesthetic pleasingness that are apparently driving Augustine's 'harmonization' of instantaneous creation with the Genesis account. My comment above was simply temporarily conceding these principles for the sake of argument, to see if they really help adjudicate between FTA and BDA.

Given your explanation of Augustine, I take it that you're distinguishing between *the divine creative action* satisfying 'our more poetic impulses' and *the result of the divine creative action* satisfying 'our more poetic impulses'. My 'nature red in tooth and claw' business was an assessment of the latter, not the former.

(Of course, on this view, it's difficult to see how Augustine has effected a 'harmonization' of instantaneous creation with the 'progressive creation' of Genesis. For what Genesis repeatedly construes as a progression of *divine creative acts* (God said, and there was, God said, and there was...), Augustine apparently construes as a *progressive result* of an instantaneous divine creative act (the growth of the 'seed' that was created).)

In any event, your point then is that on (your construal of) the FTA, the divine creative action that brings about C is instantaneous, even if it has fairly messy consequences over eons of development. Whereas on BDA, the divine creative action (at least with respect to human beings) is spread out over time via subsequent divine intervention, and is thus non-instantaneous. And therefore, if we're committed to a principle of efficiency or aesthetic pleasingness, advocacy of FTA is in tension with advocacy of BDA.

Let me know if I've stated things correctly in the preceding three paragraphs, and then I can perhaps sharpen my critique.

> I think the RS doctrine does the trick. I think it's brilliant,
> poetic, and true.

Well, I'm still not clear as to how it's to the point. Augustine's harmonization *presupposes* the relevance of principles of efficiency and aesthetic pleasingness. Thus, it can't be construed as evidence *for* them.

Consider again the context in which you first deployed Augustine:

>>> All I was adverting to was "further tinkering" like the 18th
>>> Century deists objected to. I think there are non-deistic
>>> reasons to object to further tinkering, one's related to
>>> Augustine's doctrine of *rationes seminales*.

But why should we think that Augustine's RS is a *reason* to object to further tinkering? Is it not a harmonization that already assumes we have good reason to object to further tinkering? So I'm at a loss as to what the appeal to Augustine accomplishes, in terms of bringing out a tension between FTA and BDA.

BTW, what do you think of this paragraph that you (perhaps inadvertently) skipped at this juncture of your reply? (Not that you have the responsibility to be exhaustive.):

>> I'm not sure why this is 'sloppy' (i.e., initial fine-tuning
>> plus subsequent guidance of biological evolution). It would
>> only be sloppy if the mere emergence of complex moral agents
>> were the *only* goal of the divine program for creation, and if
>> God labored under time and resource constraints in achieving
>> that goal. Since all of that is doubtful, then for all we know
>> a whole host of goods are realized in this scenario that would
>> not otherwise be realized, and the means chosen are the means
>> best suited to the purpose.

The upshot is that *even if* I concede some 'efficiency' principle, it still doesn't follow that the BDA-further-tinkering scenario somehow violates it. That would only follow if G were the only divine purpose at hand. But there's little reason to think that.

>> These are two distinct issues, of course: whether pr(C/T) is
>> high, and whether C raises pr(T). I'm only committed to the
>> latter, not the former.
>
> The main problem is that if pr(C/T) isn't somewhat high, then
> the explanatory power of T drops correspondingly... And you won't
> get God indicted on charges on creation unless pr(e/h) is
> somewhat high, not just contrastively higher than pr(e/~h).

I'm still not sure this is a fatal result, since I think of the FTA as slotted into a larger cumulative case. I'm thinking along the lines of Swinburne's project, where clearly pr(C/T) (or more generally, pr(e/h)) does *not* have to be somewhat high:

"A theory may not make the occurrence of its data e very probable, and the theory may not have very high prior probability; but if it provides the only possible explanation of e, or the only remotely plausible explanation of e, and if the occurrence of e uncaused would be a great puzzle, then e is very strong evidence for the theory... h leaves unanswered questions, but a rational detective will accept it because it is the only remotely plausible hypothesis. Things are like this, I am suggesting, with theism. The phenomena which we have been considering are puzzling and strange. Theism does not make their occurrence very probable; but nothing else makes their occurrence in the least probable, and they cry out for explanation. A priori, theism is perhaps very unlikely, but it is far more likely than any rival supposition. Hence our phenomena are substantial evidence for the truth of theism" (EG, 2nd ed., 289-290).

For compactness I've unfortunately elided out his detective/courtroom illustration, which I think helps to show the fairly modest kind of standards I'm bringing to the table wrt the project of natural theology.

Much of the rest of the discussion depends upon this "pr(C/T) is high" assumption, so I'm going to avoid it for the time being. (I might return to it, as I've written up some stuff on it, to the effect that I'm still not sure I have to *deny* pr(C/T), just your way of getting there. But let's put that aside.

> Let's think in frequentist terms again, just as a heuristic.

I must confess I simply didn't follow this paragraph, perhaps because I'm not accustomed of thinking of Bayesian evidence along the lines of frequentism. I'm sure you're making a good point, but I'm just not getting it.

> Now you make a very good point that every life-producing world
> is ipso facto a life permitting world and I take that to show
> pretty well that the proportion of worlds in W which God could be
> expected to bring about is greater than the compliment. However,
> all that means, again, is that C meets the technical Bayesian
> definition of "evidence" and that's just not very impressive.

Well, I think it's exceedingly impressive, *given* the prospects of the alternative hypotheses for the existence of a life-permitting world. But again, perhaps we're construing the natural theology project different here.

>> Why would any FTA want to appeal to [C=The initial conditions
>> of u are such that it is unlikely that u will contain G]? It may
>> no doubt be *true* that C by itself doesn't make it likely that
>> u will contain G
>
> The FTA advocate *wouldn't*, but the BDA advocate *must* which
> was our original point about there being conflict.

Well, I just reread those two paragraphs and it still looks like you were talking about versions of the FTA through and through. I guess I totally misunderstood.

> I'm not sure about your last statement, either I'm reading it
> wrongly or it contains a typo.

I'm saying that the advocate of FTA could concede that God's bringing about of C (the boundary conditions and forms of laws) doesn't make it likely that u will contain G. That premise doesn't damage his case, because he is appealing to other premises compatible with it. Or were you talking about another statement of mine?

>> as long as T gives us some reason to expect C, then it raises
>> pr(T). If it doesn't raise the probability of the alternative
>> theories (because they don't lead us to expect C), then C
>> constitutes evidence for T and not for these other theories.
>
> See comments above on "evidence." I just don't think the WFTA
> provides much evidence at all. I could be persuaded otherwise if
> I had some reason to expect God to create with saltation. I just
> don't see any motivation for that at this point.

OK. There's not much more I can say here. In the absence of a reason to accept the efficiency principle, it seems to me that everything the FTA says is compatible with everything the BDA says. And, indeed, as I've tried to point out above, even with the efficiency principle it doesn't follow that the BDA scenario is 'inefficient'. Finally, since none of the other theories explaining C are remotely plausible, I think that makes it strong evidence for the only theory which remains.

>>> I think pr(T/I.C)=low! Why would God bring about a world
>>> which is unlikely to achieve his goals?
>>
>> I agree, he wouldn't. But it doesn't follow from the fact that
>> (1) "God brings about conditions necessary for the achieving
>> of his goals" that therefore (2) "God brings about a world
>> which is unlikely to achieve his goals.""
>
> But it does with the relevant EP, so it all keeps coming back
> to that.

I don't understand. How would EP get you (2) from (1)? If God brings about necessary conditions *and* he's efficient, then he brings about a world unlikely to reach his goals??

>>> You'd have to believe there was some I* like "God intends to
>>> bring about G in stages of progressive tinkering.
>>
>>Perhaps.
>
> I think this might be a bigger problem than you realize for
> WFTA, for in addition to providing scant evidence, it
> complicates T (by adding parameters) even further reducing the
> extent to which C is evidence for T.

Perhaps this discussion is becoming too complicated, because now I'm hopelessly confused. Further above, you seemed to say that these paragraphs of yours were about the BDA, but now you're saying they're about the WFTA. I'm having a hard time keeping track of *which* assumptions you think need to be made by *which* argument. Again, this may all be due to me.

> I'm not claiming that even the SFTA yields pr(T/C)>.5 I'm
> claiming that it needs to be the case that pr(C/T)=high needs
> to be the case even for SFTA to be a good C-inductive argument.
> Note that Swinburne appears to think this as well (TEG 1st ed.,
> 147-148). In fact, he says that "It appears that P(e/h.k) equals
> 1" (ibid.). where e=the laws of nature.

I only have the 2nd edition, so I'm not so sure where this appears. Is he really talking about general standards for good C-inductive arguments? In his first chapter on inductive arguments, he's fairly clear about what qualifies:

"So (for background knowledge k) an argument from e to h will be a correct C-inductive argument if (and only if) P(h/e.k) > P(h/k)" (16).

"If P(h/e1.k) > P(h/k) then the argument is a good C-inductive argument" (17).

Nothing here about P(e/h.k) needing to be high. As long as h gives you *some* reason to expect e, then the standard for a good C-inductive argument is fulfilled. And these are the modest standards for a good C-inductive argument that are eventually applied, for example, to the cosmological argument:

"Since P(e/h.k) > P(e/k), by the relevance criterion P(h/e.k) > P(h/k), and so the argument from the existence of a complex physical universe to God is a good C-inductive argument" (131).

All it takes is for P(e/h.k) to be greater than P(e/k), not that P(e/h.k) has to be close to 1.

I'll still try to find your quote, however, even though it jars against everything I *thought* I knew! :-)

>> theism is much more probable than any of the alternative
>> explanations of the evidence on offer. So, if we have to make a
>> choice, we know what choice to make, that is, what the rational
>> choice would be. I'm not sure I'm saying anything incompatible
>> with what you say above, however.
>
> Indeed it is not, in fact, it expresses my sentiments exactly.
> That's all Swinburne was going for anyway (cf. _Faith and
> Reason_). I think this is the bottom line. Good point.

Ah, a point of fairly significant agreement. Perhaps I should have read this paragraph first before starting my lengthy reply!

So given what I say in the material you cite above, would you say situating the FTA in *that* context somehow trivializes the argument?

> Not so fast! You're assuming S5 principles, I'm not.

OK. I was also being humorous :-)

> Well, I've suggested that the elegance of simplicity is a
> consideration in it's favor. That's not saying *much* but it is
> something. I'd like to hear something against it.

Well, as I tried to bring out near the very beginning of this reply, I'm not sure why God has to *be* elegant (given that he has no time and resource constraints), or why he has to appear 'elegant' to us. And I'm not sure that his actions *wouldn't* appear elegant, given further information as to his fuller goals for creation.

Ted writes:

> I agree that for beings with limited resources efficient action
> is better. I do not, however, want to reason to EP through any
> principle like that.

OK.

> As you point out it that will not work to get EP. Rather I
> conceive of EP as a principle of rational action. Rational
> action is efficient.

Since I spent far too much time today on that other reply, let's just accept the EP for the sake of argument. How does it show in the slightest that God would not guide human biological evolution? The only reason I would think that, is if I thought that the *only* state of affairs God was interested in was the mere existence of complex moral agents. But why think that?

In addition, human evolution itself poses a problem for *any* version of the FTA, if you subscribe to this defense of the EP. Here's God, and his goal is to create complex moral agents. So does he do the *rational* thing and create them instantaneously? No, according to the FTA and Augustine, he plants a 'seed', which over the ages eventually produces human beings. How is this rational? As you put it, "Some explanation of my action is called for."

If you say, "Well, obviously, God had some goals *in addition to* the mere existence of complex moral agents, and those goals are realized by the lengthy process God chose," then I think you've conceded my point. The 'efficiency' objection only works against the BDA by narrowing our conception of the divine goals.

Given this, to begrudge BDA on the grounds that it posits inefficiency is just a quibble, it seems to me. If you've already swallowed the camel of several billion years of painstakingly slow evolution as somehow compatible with a perfectly 'rational' God choosing perfectly 'efficient' means, why strain out the gnat of a few evolutionary pushes here and there?

> Do you agree though that apart from the argument from
> consciousness Swinburne seems to indicate that the BDA does not
> work?�I'm thinking of a BDA using facts about non-consciousness
> complex organisms.

I'm still fairly confused here. What material in Swinburne do you think "indicate that the BDA does not work"? Can you be specific here? Or have you already and I just didn't see it?

I only brought up the passage because I thought it drew an important analogy between the BDA and the argument from consciousness: both arguments assume that in order for God to accomplish his goal of bringing about complex moral agents, he engaged in interventions *subsequent to* (in some important sense of this phrase) the Big Bang. So any objections to BDA's on interventionist/inefficiency grounds are objections to the argument from consciousness. So I'm looking for a relevant difference here, as to why a proponent of FTA should discount the one but accept the other.

Note that the intervention involved in the argument from consciousness is *not* as a means of accomplishing goals with respect to creatures who already have free will (i.e. interacting with them), but as a means of producing those creatures in the first place.

Trent wrote:

> It's open to a good explanation of why there was a deviation
> from effeciency (though one wonders if the resulting action
> still wouldn't be considered perfectly effecient *given the
> agent's goals*).

Yes, this is a secondary point I've been trying to make. Efficiency is relative to goals. Both the FTA and the BDA focus in on *specific* divine goals. The fact that they might look 'sloppy' together (from a certain point of view) says more about our limited access to the full range of divine goals for creation, than it does about God.

That's why I've been stating all of this in terms of necessary and sufficient conditions. The FTA and BDA zero in on certain phenomena either fulfilling certain goals God has. But it's no part of either argument to say that these highlighted goals exhaust the divine purpose.

BTW, I tried to show this with Appendix B in Swinburne, but there wasn't much of a reply. I have yet to find a published version of the FTA that actually argues that the mere occurrence of the fine-tuning conditions *makes it likely* that human beings would exist (rather than, *permitting* the existence of human beings). I mean, how would one actually go about *showing* this from the strength of the four fundamental forces, the charge and mass of elementary particles, the velocity of the expansion of the universe, isotropy, etc.? I can't imagine a FTA even getting off the ground with that kind of inflated claim about making human life likely. By imposing such a requirement upon the FTA (with or without accepting a BDA), I think you've virtually destroyed it.

> So there are two consequences from this for the advocate of the
> WFTA. First, we need to *hear* the explanation why God's first
> step in bringing about G is an action C such that Pr(G/C) is low.

Why? Do we need to hear from an advocate of the *SFTA* why God decided to create via billions of years of evolution, rather than instantaneously? "God wants to make complex moral agents. Therefore, the rational plan of action is to take billions of years to accomplish the goal." Huh? But if you have to confess your own ignorance at this point (of God's fuller purposes, beyond the mere existence of agents), then why can't the advocate of the WFTA make the same move?

> this further complicates T and so lowers it's prior and so
> eventually it's posterior probability.

God's complex goals for creation undermine the simplicity of theism that is needed for the purposes of this program of natural theology? Do you think an advocate of the SFTA would be willing accept the consequences of this principle?

As for Swinburne, it is in the midst of his chapter *defending* the simplicity of theism, that he says:

"Some theists have held that God created the world at a first moment of its history and imposed upon it then the laws of its future operation and thereafter left it to itself. This is the view of the deist. By contrast, in developing the theist's position, I shall assume the more orthodox view that God is at each moment of the world's history responsible for its operation at that moment of its history" (91).

Just some thoughts!

[1] �BTW, just to clarify, above you speak of "the creation of the conditions for beings with free will," and how we don't need divine intervention for this. Are these necessary, or sufficient, conditions? I think you mean sufficient, since your point is that God's bringing about C makes it likely that the beings specified in G (complex moral agents) will emerge, correct?�

Well, sort of. I don�t really want to say sufficient b/c I�m only saying that Pr(G/C)=high, not Pr(G/C)=1. But clearly you are reading me aright.

[2]So the point is that God is concerned with looking graceful to *us*, and this means he must be efficient? I don't buy this. What would warrant such a principle?"

I don�t deny that God has reason to look graceful to *us*, not that you mention it, but that�s not what I was thinking of. I was thinking of it as an a priori principle of rational action and I think that there is an inherent aesthetic of simplicity.

[3] In addition, *for all we know*, the sum total of God's actions -- BDA 'tinkering' and all -- *would* look graceful to us, if only we were aware of the full range of purposes that God is trying to accomplish by his actions. So I don't think it follows from the fact that a portion of God's actions, or a specific action taken out of context, looks 'ungraceful,' that therefore the divine program for creation is ungraceful. Or do you really intend the stronger claim that each and every divine action must appear graceful to us, from the limited perspective by which we consider it?

There�s an interesting parallel here with theistic approaches to handling evil. I think your response here is akin to a response to evil that is too skeptical for my blood. Some theists�quite a few these days, the �skeptical thesists� they�ve been called�who say, �Look, you can�t expect all of God�s actions to seem fair to us, from our limited perspective. If only we could see the grand design, it would all make sense.� I think that�s a bit of a punt on third and short. To think we have so poor a grasp on good and evil undermines much of common sense morality that I�m not willing to give up. Anyway, you know what the opponents of skeptical theism say, so I�m just saying that mutatis mutandis for the aesthetic of God�s action. If we see a universe that legitimately seems sloppy to us and we have opposite expectations from theism, that�s evidence against theism. I don�t deny that evil is evidence against theism and I don�t deny that unnecessary inefficiency is as well.

[4] �FTA-God surely *knew* that the 'nature red in tooth and claw' scenario would unfold over eons of time among non-free-willed creatures, and yet he brought about the conditions for that anyway. How is that any better, morally speaking, than God guiding evolution 'directly'?�

I think there�s a clear answer here actually. If you take the definition of what�s in God�s plan given by van Inwagen in his essay on �Chance in a World Sustained by God,� then you�d get the result that the SFTA God did not have evil in his plan, whereas the BDA God did.

[5] And to top it off, perhaps on the BDA scenario the points of divine guidance *lessened* the competitive chaos (and ensuing pain and suffering) that would have otherwise obtained. At this point, I can't see how FTA has an edge over BDA, *at all*, such that, say, someone committed to the FTA would have a good reason to distance himself from the BDA.

I think the van Inwagen point applies here. Interestingly this sort of theodicy is part of what inspired Darwin (see _Darwin�s God_).

[6] �In any event, your point then is that on (your construal of) the FTA, the divine creative action that brings about C is instantaneous, even if it has fairly messy consequences over eons of development. Whereas on BDA, the divine creative action (at least with respect to human beings) is spread out over time via subsequent divine intervention, and is thus non-instantaneous. And therefore, if we're committed to a principle of efficiency or aesthetic pleasingness, advocacy of FTA is in tension with advocacy of BDA.
Let me know if I've stated things correctly in the preceding three paragraphs, and then I can perhaps sharpen my critique.�

I think you�ve got me right here.

[7] �Augustine's harmonization *presupposes* the relevance of principles of efficiency and aesthetic pleasingness. Thus, it can't be construed as evidence *for* them.�

Sure, weren�t meant to be. I was replying to your point about the �third� option.

[8] �But why should we think that Augustine's RS is a *reason* to object to further tinkering? Is it not a harmonization that already assumes we have good reason to object to further tinkering? So I'm at a loss as to what the appeal to Augustine accomplishes, in terms of bringing out a tension between FTA and BDA.�

I�m pretty sure I wasn�t trying to use it for that purpose, but I could be mistaken.

[9] �>> I'm not sure why this is 'sloppy' (i.e., initial fine-tuning
>> plus subsequent guidance of biological evolution). It would
>> only be sloppy if the mere emergence of complex moral agents
>> were the *only* goal of the divine program for creation, and if
>> God labored under time and resource constraints in achieving
>> that goal. Since all of that is doubtful, then ***for all we know
>> a whole host of goods are realized in this scenario that would
>> not otherwise be realized, and the means chosen are the means
>> best suited to the purpose.�*** [emphasis added]

For all we know, God created the universe five minutes ago and gave us all fake memories. For all we know, God let the devil put fake dinosaur bones in the ground to test us. :)> My point is just that in my approach, those sorts of appeals are no good. Responsible judgements take into account what we *do* know and posit the *simplest* possible assumptions in positions of ignorance. This is why I keep returning to the point that adverting to what God�s purposes might be �for all we know� add parameters to T thus lowering its prior, which has adverse effects to all arguments for T. I�m not looking for the God of Abraham, Isaac, and Jacob here, not at this stage anyway. I think all that the scientific evidence can support is something much more like deism. Now I think this makes it reasonable to expect a revelation and to examine claims to revelation, then historical evidence can flesh out our concept of T. At that point the content added to T is paid for by the new types of evidence. Lot�s of people apply this two-stage apologetic, but I think Swinburne is about the only one who realizes how little you get out of the fist stage. Most people want to jump straight to the end. As a sidenote, this is a version of an old complaint against Aquinas: �There�s a prime mover, therefore God exists� and he seems to assume it�s the God of A I and J. I think this set a very bad trend.

[10] �The upshot is that *even if* I concede some 'efficiency' principle, it still doesn't follow that the BDA-further-tinkering scenario somehow violates it. That would only follow if G were the only divine purpose at hand. But there's little reason to think that.�

Aaaak! That�s an argument from ignorance! I�m trying to apply a method here that is honest to the facts and Bayes�s Theorem. I said earlier that I wasn�t going to let ease of theodicy guide my theorizing and I�m not going to let what I want the end result to be�this multifaceted God with multifaceted purposes�guide my theorizing either. I will only infer the minimum the evidence supports. You�re working with a much richer version of theism than I am here. You may be conditionalizing on historical evidence or religious experience, but I�m not. I�m a good old fashion two-stager. At the level of cosmological evidence, there�s not much warrant for further currently unknown and unspecified purposes. When we infer peoples motives from their behaviour, we should apply the principle of parsimony very strictly. For any given action, there are infinitely many belief/desire combinations which would explain it. The only reasonable thing to do is to go with the *simplest* one. This is simply a curve-fitting problem. I think you�re getting ahead of the curve. [PS, none of this was �said� in my inner voice in a less than genial tone. :)>]

[11] �I'm still not sure this is a fatal result, since I think of the FTA as slotted into a larger cumulative case. I'm thinking along the lines of Swinburne's project, where clearly pr(C/T) (or more generally, pr(e/h)) does *not* have to be somewhat high: (289-290).�

Well, that is true but it�s taken out of the appendix and I honestly don�t think there�s much force to the point in this particular context. As I said earlier, I think that for Pr(T/C)>Pr(T) by much at all we need Pr(C/T) to be high, not just >Pr(C/~T). If it could be shown that Pr(T) was a *lot* higher than Pr(C), then you�d really have something. It�s just that I�m agnostic about that, so I don�t think the comparative strategy is a good one here. Unfortunately, my frequentest analogy was meant to illustrate this, to which you replied, �I must confess I simply didn't follow this paragraph, perhaps because I'm not accustomed of thinking of Bayesian evidence along the lines of frequentism. I'm sure you're making a good point, but I'm just not getting it.� Perhaps I�ll try to restate it more clearly in the future.

[12] �since none of the other theories explaining C are remotely plausible, I think that makes it strong evidence for the only theory which remains.�

I�m not sure I grant your assumption here. I�ll try to restate my point here. We are interested in Pr(C/T). Now remember that we are assuming that Pr(G/C)=low. And T is�according to method--the simplest form of T: that God intends to bring about G. Now if that�s *all* we�re assuming I don�t see why Pr(C/T)>Pr(C). You make a good case with one of your necessary condition arguments, very similar to one I almost gave. I�m so lazy that instead of looking down for yours, I�m going to give mine. A necessary condition for building a house is to assemble a large stock of lumber. Now, there aren�t very many reasons to assemble a pile of lumber so even though a pile of lumber doesn�t confer much probability on there later being a house, it�s about the only good reason for doing so, so we might think that a pile of lumber is good evidence for a future house. I.e. Pr(there�s going to be a house/a pile of lumber)=low, *yet* Pr(a pile of lumber/there�s going to be a house)>>Pr(a pile of lumber/~there�s going to be a house) so Pr(there�s going to be a house/a pile of lumber)>Pr(there�s going to be a house). All fine and dandy, but Pr(H/L) can�t be *to* low. If it�s too low, then our data isn�t a pile of lumber, but some guys felling trees. Note that if it�s this low, then Pr(men are felling trees/there�s going to be a house may not be much more at all than Pr(men are felling trees/~there�s going to be a house). So on that option, the FTA is no good. But if we ratchet it back up to the level of a pile of lumber, then the BDA will never work because it�s not so unlikey that a pile of lumber will be used to make a house. I hope this is helpful, it�s sure been fun.

[13] �>>> I think pr(T/I.C)=low! Why would God bring about a world
>>> which is unlikely to achieve his goals?
>>
>> I agree, he wouldn't. But it doesn't follow from the fact that
>> (1) "God brings about conditions necessary for the achieving
>> of his goals" that therefore (2) "God brings about a world
>> which is unlikely to achieve his goals.""
>
> But it does with the relevant EP, so it all keeps coming back
> to that.
I don't understand. How would EP get you (2) from (1)? If God brings about necessary conditions *and* he's efficient, then he brings about a world unlikely to reach his goals??�

I love it! I characterize C as �that given which it is unlikely that God�s goal G will obtain� and you characterize it as �the necessary means to G�! OK, we *both* agree that the former is accurate, but, as I�ve tried to emphasize in this post, you�re sort of cheating by looking ahead to even tell that it�s a means to anything. You�re peeking at the future, that God will take these felled logs and make a house out of it. You just wouldn�t expect a house merely by seeing men fell logs, nor, say I, does the house building theory have much of an edge on myriad other reasons there might be men felling trees. Well, that�s the best I can do for now.

[14] �I only have the 2nd edition, so I'm not so sure where this appears. Is he really talking about general standards for good C-inductive arguments? In his first chapter on inductive arguments, he's fairly clear about what qualifies:�

Sorry, my bad, I meant for it to have much of comparative advantage on ~T. In other words�to use the term correctly this time�there can be a good C-inductive argument from C to T and yet T not have much advantage over T. This is b/c this is compatible with there being a good C-inductive argument from C to ~T. What�s crucial is whether c(C,T)>c(C,~T). If my lumbar-jack example above is any good at all, then it is far from obvious that that holds. Before I personally could use the WFTA I�d need some assurance that this was so. My problem is that I�m not so sure what this assurance would even look like. I�m being totally honest that I would *like* to have it and would thus be very open to receiving it.

[15] �So given what I say in the material you cite above, would you say situating the FTA in *that* context somehow trivializes the argument?�

No, I wouldn�t say that. It�s just that I couldn�t in good conscious personally use the argument for the reasons stated above. I can totally see, though, how someone could see it the other way.

[16] �Not so fast! You're assuming S5
principles, I'm not.
OK. I was also being humorous :-)�

What is this �humor� of which you speak? ;)> That�s my goatee on the end btw, for which I�m noted.

[17] �Well, as I tried to bring out near the very beginning of this reply, I'm not sure why God has to *be* elegant (given that he has no time and resource constraints), or why he has to appear 'elegant' to us. And I'm not sure that his actions *wouldn't* appear elegant, given further information as to his fuller goals for creation.

A. *because* it�s elegant, B. we don�t *have* further information.

[18] �Efficiency is relative to goals. Both the FTA and the BDA focus in on *specific* divine goals.�

Right, I�m postulating the simplest goal, until I have reason to postulate a more complicated one. I would feel inauthentic if tailored my thesis to extraneous data, it would feel ad hoc.

[19] BTW, I tried to show this with Appendix B in Swinburne, but there wasn't much of a reply. I have yet to find a published version of the FTA that actually argues that the mere occurrence of the fine-tuning conditions *makes it likely* that human beings would exist (rather than, *permitting* the existence of human beings).�

Unless you count the passages you cited from Swinburne! :)> I�m not nearly as concerned with how people *do* use the argument as with how people *should* use the argument. That�s what this is about for me. I�m concerned people have been using the argument wrong.

[20] �I mean, how would one actually go about *showing* this from the strength��

Ah! I�ve got an argument. Only a teaser for now. Given that C *in fact* led to G, in a fully deterministic universe, Pr(G/C) would be 1! Now maybe this isn�t (I�m far from convinced it isn�t). Now suppose you believe in quantum indeterminacy, not just epistemic, but ontic (like so many sheep today :P). This would make Pr(G/C) diverge from 1, however, I don�t think it would lower it to .5. I go with Einstein over Bohr. Quantum ind needn�t significantly effect macro-level predictions. As I read chaos theory, some of it supports this line of reasoning and some doesn�t. The worst thing for my line of reasoning would be if weather systems and ecosystems were really chaotic. I don�t think they are but some really smart people do. So I haven�t *shown* it, but I�ve shown *how* to show it. That�s *something*. :)>

[21] re: my high standards destroying the FTA

Methinks thou doth protest to much! After all, one of the biggest (most common on this university campus anyway and often in print) objections to the BDA are self-organization scenarios! If these objections grow in frequency and strength, you WFTA folks will come running to the SFTA! :)>

[22] �Why? Do we need to hear from an advocate of the *SFTA* why God decided to create via billions of years of evolution, rather than instantaneously?�

According to RS doctrine He *did*, but in a way which manifest a form of beauty not otherwise possible, temporal beauty. You might be able to do a dance on the head of a pin, but you can�t do it in an instant! I�ve shown you mine, now you show me yours. :)>

[23] �if you have to confess your own ignorance at this point (of God's fuller purposes, beyond the mere existence of agents), then why can't the advocate of the WFTA make the same move?�

A. I�d like to *see* the move first. B. Not all ignorance is created equal. My ignorance is of a fact we both admit: it took aeons for life to evolve. Where we differ is WFTA posits further tinkering and miracles.

[24] �this further complicates T and so lowers it's prior and so
> eventually it's posterior probability.
God's complex goals for creation undermine the simplicity of theism that is needed for the purposes of this program of natural theology? Do you think an advocate of the SFTA would be willing accept the consequences of this principle?�

*I�m* an advocate of the SFTA, and *I�m* willing to. The goal I posit is simpler than the goal you posit. So if we conditionalize on the same evidence it is guaranteed to be the case that c(E, T-me) > c(E, T-you). Parsimony is the guiding light of inference. Hypotheses that cover the data are a dime a dozen. The way to pick is parsimony.

[25] �As for Swinburne, it is in the midst of his chapter *defending* the simplicity of theism, that he says: (p. 91)
"Some theists have held that God created the world at a first moment of its history and imposed upon it then the laws of its future operation and thereafter left it to itself. This is the view of the deist. By contrast, in developing the theist's position, I shall assume the more orthodox view that God is at each moment of the world's history responsible for its operation at that moment of its history" (91).

Right, but you can�t pin deism on me or interventionism on Swin in light of this passage. He�s not dissociating himself from what I�m advocating. If he were to explain himself further, I have no doubt that he would say what van Inwagen says in the first part of �the place of chance in a world sustained by God,� i.e. that God holds in existence at each moment what particles there are and causes them at each moment to have the causal powers they have. That�s the point, not that God needs to tinker. He needs to distance himself from deism precisely because because his view has a tendency to look that way (note that it hasn�t prevented him from being accused of deism).

Once again, I appreciate all your thoughts and even though I attempt to rebut them, they go in the box and don�t just sit there, they mix around. Who knows what the result will be?

[1] (On the crucial inference)

Let�s distinguish between:
(a) merely interested in bringing about necessary conditions.
From
(b) interested in bringing about merely necessary conditions.

Greg�s claim is: If God is interested in bringing about those conditions which are N&S for his purposes then *a fortiori* he is interested in bringing about the N conditions.

My counterexample concerns the claim read as (b): If God is interested in bringing about those conditions which are N&S for his purposes then *a fortiori* he is interested in bringing about merely N conditions.

Greg writes: �I�m making a much more modest claim�: if God is interested in bringing about X and Y, then he�s interested in bringing about X.�

Well, I still think my counterexample holds. :) Intentional operators, as a rule, don�t preserve logical relations.

Re: the Father Ted case. Greg writes: �You desire to have Father Jack over, but *in a certain context*, namely, when he�s getting ribbed by Ted and Dougal�.

Yes! I desire to have Father Jack over in so far as he�s an essential member of the crew. But I don�t desire to have Father Jack over by himself, even though he is a necessary condition for that crew. This applies to the FTA in the following way: suppose God is interested in bringing about the N&S conditions for life it does not follow that God is interested in bringing about mere N conditions for life. Why? Because the N condition by itself is not what he is interested in bringing about. Compare: I�m interested in bringing about my dream house. Now a necessary condition of building my dream house is that I have a box of nails. But I�m not interested in having a box of nails (by itself). I grant that it gets me a very small way towards the goal of having my dream house, but I am not interested in a mere necessary condition�I�m interested in N&S conditions. (Maybe I�m idiosyncratic on this score.) Re: your tree house example. You write �If I�m interested in building a tree house for my kids, I think I�m interested in giving it [the tree house] a floor.� I agree, but note that you say that you are interested in giving it�the tree house�a floor. You aren�t interested in the floor by itself. The floor is good for you only insofar as it is part of what is required for the tree house. (Note BTW that there�s another sense of the phrase �what is required for� which means N&S conditions instead of just N conditions�I think this is relevant regarding some of the passages you cite from Swinburne.)

Greg writes: �It all comes down to the fact that I never said that God would *merely* want the necessary conditions to come to pass, just that he wants them to pass (presumably, because they are a necessary means of something *else* coming to pass).�

Taking in to consideration the distinction between (a) and (b)�it seems that we�re close on this one. I agree if �he wants them to pass� means he wants them to pass [insofar as they are a part of N&S conditions]. But I don�t agree if �he wants them to pass� means he wants them to pass [only insofar as they are N conditions].

[2] (On granting the EP but appealing to other purposes)

Greg writes, �The upshot is that *even if* I concede some EP, it still doesn�t follow that the BDA-further-tinkering scenario somehow violates it [EP]. That would only follow if G were the only divine purpose at hand. But there�s little reason to think that.�

You�re right that the BDA-etc- scenario does not violate the EP, given other purposes. But those other purposes must be derived from T {the theist hypothesis} itself. If it is not so derived then it complicates T and so lowers it prior probability and so weakens the WFTA by lowering the Pr(T/C). Now as you have pointed out�and as Trent and I have agreed�that the WFTA *can* still count as evidence for theism. But two quick points on this: (1) challenges exist to the Bayesian notion of evidence that make us hesitate about endorsing this as evidence for theism. Trent illustrated this nicely in one of his earlier posts. The probability that I�m guilty of a crime committed in New York is greater given that I�m in New York then if I�m not in New York. But from an intuitive standpoint this is not evidence that I committed the crime. (2) We can generate an infinite number of hypotheses by conjoining T and some purpose. It is, thus, *crucial* to the overall success of the FTA (et. al. natural theology arguments) to keep T as simple as possible. The SFTA does this: it derives that God has overriding reasons to bring about rational moral beings from T alone. The WFTA�assuming the EP�must appeal to other purposes to allow for the BDA, but that�s a problem for the reasons noted above.

[3]

Greg writes, �[L]et�s just accept the EP for the sake of argument. How does it show in the slightest that God would not guide human biological evolution? The only reason I would think that, is if I thought that the *only* state of affairs God was interested in was the mere existence of complex moral agents. But why think that?�

Hopefully what I said above helps on that score.

Greg continues: [Here I�ll paraphrase] �If that�s the only goal, why not instantaneousness creation?�

God has overriding reasons that rational moral beings exist. Rational beings require growth in knowledge, knowledge that includes learning about our environment. Suppose God created the world 5 minutes ago and all our memories, etc. We could still learn about our environment but it would not be genuine knowledge. (On this point I�m *tempted* to argue that instantaneousness creation and such would imply that God deceives us and this does not facilitate the goal of genuine growth in knowledge which is a good by SFTA lights that God aims to create.)

[4] Greg writes, �I intentionally took a day off.� Ted writes, �I�m intentionally taking tomorrow off!� :)

I'll delete them to save space on Matthew's server, but I'll leave this here so the last comment doesn't sit there making no sense.