Intuitions Regarding Kevin Scharp’s Critique of William Lane Craig’s Natural Theological Arguments from Probability Thresholds and Divine Psychology

Kevin Scharp (Scharp) has lodged two arguments against William Lane Craig’s (WLC) natural theological arguments I think are plausibly false.[1] First, Scharp argues that in giving a philosophical argument with premises which are theologically neutral but a conclusion that is theologically significant, there must exist independent justification for thinking that (on the Kalam cosmological argument) God would want to create a universe (if God would not want to create a universe, on his view it is implausible to suggest that God then is probably the Creator of the universe). Secondly, Scharp argues that claiming to ‘believe’ requires a context-dependent probability threshold with respect to how much confidence a person has in the belief in question. So, first I shall argue that Sharp’s first argument against WLC is plausibly false. Secondly, I will show how the desire to have probability thresholds with respect to confidence in beliefs is not significant. In conclusion, I shall suggest that Scharp’s two contentions are implausible and irrelevant.

Scharp first argues from ‘Divine psychology’ (God’s psychology) that the burden of proof lies on the theist to show that God would want to create the universe (considering only the Kalam argument for spatial considerations here). To be clear, I will outline the Kalam argument as Dr. Craig typically presents it, that is, deductively:

  1. Whatever begins to exist has a cause
  2. The universe began to exist
  3. Therefore, the universe has a cause

Even if all three premises are correct, Scharp assumes, the argument does not yield a convincing case for the conclusion that God exists since the theist must show in addition that God would want to create the universe. While I can understand Scharp’s motivation for this claim (often we explain things from intentions i.e., the child takes the cookie from the cookie jar because he/she has the intention/desire to do so), I think his contention here is false—or minimally plausibly false. I shall illustrate this through an analogical argument.

Suppose there exists a water bottle on a desk in an office. Further, suppose that Daniel, the detective, must find out whose water bottle it is. So, Daniel takes specimens from the bottle cap to detect finger prints as well as other background information i.e., the name ‘Emily’ written on the door. So, given the specimens from the bottle cap, the name ‘Emily’ written on the door and so on, Daniel concludes that the water bottle is Emily’s. It is natural to assume, based on the evidence, that whether or not Emily forgot the water bottle, hated and so abandoned the water bottle or even intentionally left the water hoping for someone else to drink it, the fact of the matter is that the evidence points towards—with plausibility—Emily being the person who was responsible for the water bottle in the office. Now, here is the analogy. While having direct access to Emily’s private mental life i.e., seeing her intentions/desires, it might be easy to see her direct intention—in part—responsible for her leaving the water bottle in the office, Emily’s private mental live and all it entails is independent of whether or not the evidence (or arguments) point to her being the best explanation of the water bottle being in the office. Now, if a natural theological argument shows with plausibility that God exists i.e., the Kalam argument, it does not follow that direct access to God’s psychology (what Scharp calls ‘Divine Psychology’) is necessary or sufficient to show that God exists; if the evidence points towards the existence of God, God’s desire to (or not to) create the universe is independent of the conclusion of the argument. Thus, when the conclusion is reached ‘God exists’ it is both misguided and mistaken to suggest that the theist has a burden of proof to show that God would want to create the universe (since it is not clear that that has any decisive or philosophically significant role). So much for Scharp’s first contention.

Scharp continues and argues that WLC must show that a person’s belief in God must meet a certain probability threshold to be a genuine belief i.e., 51% is not enough.[2] WLC has responds to this by suggesting that that it is arbitrary to suggest that confidence levels must meet probability thresholds to be beliefs  inasmuch as it is psychologically and qualitatively implausible to do so (introspectively), to which Scharp responded that it is simply a matter of context. I suggest that this is highly improbable. In order to give an accurate qualitative account of one’s ‘confidence’ of some belief, one must devise a truth-conducive method of assessing one’s confidence of the belief itself. This method will either be introspective or scientific.

The problem with introspection is that while generalizations can be made ‘I believe some belief B to be more probable than not’, it is difficult to see how—with any precision or measurable precision—one can give a percentage of the confidence from introspection i.e., ‘I believe B with a confidence of 54%.’ A further problem arises, though. What is the threshold of decimal points necessary to be given? Should it be four decimal points? Ten decimal points? Decimal points ad infinitum? Indeed, there is at work here the problem of designating arbitrary numbers for confidence in a belief (while a generalization is significantly more reliable). Thus, there must be a scientific approach to retrieving a qualitative analysis of confidence in a belief. Here, though, the problem is worse than that of introspection. For inasmuch as ‘confidence in a belief’ is a person’s orientation to that belief (or propositional attitude), it is unclear why it is possible to think that a person’s orientation or propositional attitude are qualitative notions. The burden of proof, then, lies on Sharp for providing a method.

It is also unclear whether or not the natural theological arguments should be taken disjunctively or conjunctively. For instance, even supposing Scharp’s probability threshold requirement for confidence to have a ‘belief’, it is unclear how this is interpreted in light of the plethora of natural theological arguments.[3] If one was, hypothetically, 51% sure of the truth of the premises in the Kalam argument and so ‘not believing’ in Scharp’s view, what happens when, say, two other arguments are included? But, further, suppose that one held to all three natural theological arguments with a confidence of 51%. Would that person, then, believe in God? It seems that once 51% is met, God’s existence is more probably than not (setting aside considerations of reformed epistemology i.e., Alvin Plantinga). Interestingly enough, WLC, after the debate, even notes the point I am making here with respect to taking natural theological arguments conjunctively:

“I didn’t think of this in the dialogue, so I didn’t say it. But afterwards, this is exactly right, and it occurred to me. Timothy McGrew, who is a professor of philosophy at the University of Western Michigan, emphasizes that even deductive arguments that, say, make God’s existence 20% probable (that’s all, just 20%), if you accumulate these arguments – 20%, 15%, 30%, 35% – pretty soon, as you say, the cumulative probability of these independent arguments is way over 50%. This is the way a cumulative case is built in a court of law, isn’t it? No single piece of evidence might be enough to convict beyond reasonable doubt, but when you put all of the cumulative evidence together then it can be beyond reasonable doubt that the accused is guilty. So the very fact that I got around ten arguments, each of which increases the probability of God’s existence, would, I think, make it very plausible to think that this does give you great confidence that God exists.” [4]

Once a person has good reason i.e., thinks it more probable than not, that God exists, it seems that they believe that God exists. To suggest that this is not so shifts the burden of proof to the objector to come up with an adequate account of the threshold of probability for confidence of belief for the natural theological arguments taken disjunctively and conjunctively.

I think that Divine psychology and considerations of devising probability thresholds for beliefs based on confidence levels are irrelevant to natural theological arguments. [5]

[1] These comments (and the full debate) can be found at Retrieved at April, 2 2016. Also, two side notes should be made. First, while I disagree with Scharp on a host of issues (mainly assertions he made), I am restricting myself to two and not others–the laws of physics/chemistry/biology as inconsistent with miracles, for instance, I am leaving out. Secondly, many of Scharp’s arguments are, in my view, dismissive and ad hoc and so I will not deal with them here. For instance, WLC’s so-called ‘weakness problem’ is just a manifestly false understanding of WLC’s arguments. For instance, as WLC notes in the debate, the threshold of probability with respect to each premise is lowered for the sake of appealing to a wider audience; in WLC’s actual view, he thinks the premises far more probable than ‘more probable than not.’ Scharp’s response to this is to insist the weakness problem and dismiss what is, on my view, WLC’s intellectual humility. To take another example, Scharp commits a classic Appeal to Authority Fallacy in the debate so as to undermine WLC’s claims regarding confidence levels by appealing to ‘Contemporary Epistemology’ (without any reference).

[2] Scharp argues that ‘belief’ and ‘more probable than not’ are not synonymous. It is unclear what Scharp means by ‘belief.’ If it is a matter of ‘confidence’ in some proposition or states of affairs then my arguments still show that he hasn’t avoided arbitrariness. However, Scharp has offered a thought which suggests his view to be correct, namely, his ‘Hilary example’ where he says that thinking ‘Hilary will win’ with a confidence of 51% isn’t belief that she will win. I think this is implausible for two reasons. First, while a low probability of some confidence in a belief is in one sense variable and not very confident (as he himself notes), this doesn’t show that it is not a belief at all. It only shows that it is a weak belief or a non-confident belief (and if one rejects this this only exemplifies how the term ‘belief’ being vague is detrimental to Scharp’s argument). To my mind, the sentence ‘I believe, but with not much confidence, Hilary will win’ is a meaningful sentence. Secondly, there is no standard to which ‘confidence’ will be held. What does confidence consist in? Dying for the belief? Spreading the belief around in hope that people will become confident in it? It seems that confidence in a belief is not quantitative as Scharp imagines it is; this is no surprise though, it is implausible to come up with necessary/sufficient conditions and minimum thresholds for ‘belief’ if in every case ‘belief’ will be different. Another issue here that confidence is also often subjective; 51% for one person might be sufficient to believe (perhaps they have sufficient motivation to believe and thus 51% is good enough?), while others might be inclined to think it philosophically ‘not enough’ for belief i.e., Scharp. To argue that 51% isn’t enough because it is not as solid as, say, 90%, it seems implausible to suggest that confidence in a belief is necessarily connected to a certain threshold. (An interesting counter-example are people who have confidence in their beliefs despite all reasons to the contrary). Now, one might reject this and argue, for instance, that when a doctor tells the patient ‘I am confident you have cancer’, it is not a matter of the confidence that the doctor believes—it is evidence that convinces the doctor. But, I suggest that this is not always true. For instance, it is possible that a doctor might believe in miracles and so, independent of the 95% chance of the patient having cancer, the doctor believes that the patient will get better. So, it is not true that a probability threshold can (or should) be given for certain ‘contexts.’ (To avoid this and say ‘well confidence levels must, then, meet certain thresholds for each person (for belief)’, the objector here is being ad hoc and question begging in favor of thinking there is, really, a method of devising probability thresholds for each particular confidence level).

[3] The literature here is vast. It is worth citing a few pieces of literature on this topic. The natural argument from the existence of the universe comes in at least two ways: the Kalam cosmological argument and the Leibnizian cosmological argument (often called the ‘argument from contingency’). These are two cosmological arguments are defended by contemporary philosophers of religion. For a full defense of the Kalam cosmological argument, see Craig, William Lane. The Kalam Cosmological Argument. (Eugene, Oregon: Wipf and Stock Publishers, 2000). Print. For the Leibnizian cosmological argument, see Craig, William. Reasonable Faith: Christian Truth and Apologetics. (Wheaton, Illinois: Crossway, 2008), p. 106-111. The argument from fine-tuning has received much attention in light of scientific (particularly astrophysical and cosmological) evidence. For a defense of such an argument (alongside a host of other arguments i.e., beauty, consciousness, morality (axiological), providence, religious experience, et cetera), see Swinburne, Richard. The Existence of God. 2nd ed. (New York: Oxford University Press, 1979), Chapters 7-14. For a defense of the ontological argument see, Plantinga, Alvin. God, Freedom and Evil. (Grand Rapids, Michigan and Cambridge, UK: Eerdmans, 2002), Part 2 Section c, 1-10. Plantinga also defends a thesis in positive Christian apologetics which is not natural theological per se but nonetheless still somewhat of a reformed epistemological argument for the existence of God (what he calls the ‘AC Model’), see Plantinga, Alvin. Knowledge and Christian Belief. (Grand Rapids, Michigan and Cambridge, UK: Eerdmans, 2015), Chapters 3 and 4. For a treatment on the argument from consciousness see, Moreland, J.P. Consciousness and the Existence of God. (New York and London: Routledge, 2008), Chapter 2. See also Adams, Robert. The Virtue of Faith and Other Essays in Philosophical Theology. (New York: Oxford University Press, 1987), Part IV, 16. Edith Stein has also applied her phenomenological and Thomistic (scholastic) philosophy to develop an argument from consciousness. See Stein, Edith. Finite and Eternal Being. Trans. Kurt F. Reinhardt. (Washington, D.C: ICS Publications, 2002). Print. It would be immodest here to fail to repeat that this list barely scratches the surface of the plethora of literature on this topic. (For I haven’t also mentioned the initial formulators of many of the arguments, i.e., St. Thomas Aquinas, St. Anselm, Aristotle, Plato, Leibniz, Al-Ghazali, et cetera).

[4] Retrieved April 3, 2016.

[5] These are just intuitions regarding two of Kevin Scharp’s claims. I do not intend for this post to be taken as exhaustive of WLC or Scharp’s full views; I am responding to Scharp from a different (but sometimes similar and overlapping) perspective. (Also, I am omitting some of WLC’s responses to Scharp for (i) spatial considerations and (ii) clarity).


4 thoughts on “Intuitions Regarding Kevin Scharp’s Critique of William Lane Craig’s Natural Theological Arguments from Probability Thresholds and Divine Psychology

  1. In your refutation of the Divine Psychology objection, you say “if the evidence points towards the existence of God, God’s desire to (or not to) create the universe is independent of the conclusion of the argument”.

    In the Fine-Tuning Argument (FTA), God’s desire to create the univers is not independent from the *premise* of the argument, which is essentially this: “it is more probable than chance or necessity that God decided to create the universe”. And Dr. Craig provides no justification for this claim except to show that chance is very improbable and that God could have reasons to design the universe. But this doesn’t show that design is more probable than chance, for perhaps God has NO reason to create the universe, in which case it’s less probable than chance. The person making the claim that design is more probable than chance does indeed have the burden of proof to justify this claim.
    And you cannot use the conclusion of the FTA in support of the idea that it’s not improbable that God would decide to design the universe, for that would be begging the question.

    In short, it’s not enough just to show that chance is very improbable. The proponent of the FTA needs to show that chance is LESS probable than design. And as Dr. Scharp points out, you cannot do this without making probability judgements (such as “it’s not improbable”) about God’s decision making on whether or not to design the universe, i.e. appealing to divine psychology.


    • Jock,

      Thank you for your response. You raise important questions and so I will do my best to respond to them in full.

      You argue that the central premise of the FTA is that “it is more probable than chance or necessity that God decided to create the universe”. To clear preliminary misunderstandings, I should note that it is not a central premise of the FTA (otherwise WLC would be arguing in a circle since he would be using his conclusion to put his argument through—which is not a very compelling method of argumentation).

      First, here is WLC’s formulation of the FTA found in the Scharp and WLC debate:

      1. The fine-tuning of the universe is due to either physical necessity, chance, or design.
      2. It is not due to physical necessity or chance.
      3. Therefore, it is due to design. [1]

      You rightly notice that WLC’s conclusion is that design is more probable than chance or physical necessity. However, your comment that “Dr. Craig provides no justification for this claim except to show that chance is very improbable and that God could have reasons to design the universe” and that “this doesn’t show that design is more probable than chance, for perhaps God has NO reason to create the universe, in which case it’s less probable than chance” I find troubling. Dr. Scharp, which you are aware, argued the exact same thesis you are arguing: potential motive is not enough to put a probability assignment through. But why think (i) God only has merely potential (and not actual) motives we can know and access and (ii) a motive is necessary? While you are right to notice that out of two hypotheses H1 and H2, even if the probability of H1 is very low, if H2 has a probability of 0, H1 is still more probable; however, it is problematic that you regard a (plausible) motive as necessary. For spatial considerations, I will reference you comments made by Tyler Journeaux, an MA student of philosophical theology at Oxford University, who has argued that it is not the case that “stipulating a (plausible) motive is necessary to get a successful inductive argument for God’s existence” and further that “if the fine-tuning argument works then it is an argument for thinking that God has some motive for creating a world (whatever that motive may be).” [2]. But, for the sake of argument I will presuppose that a motive is actually necessary. Tyler has dealt with this as well via. personal email-correspondence:

      “However, it gets worse for Scharp than this preliminary analysis suggests. The delicious irony here is that the Christian (at least) can give an account of God’s motive. In the doctrine of the Trinity one finds the idea of God whose love is shared between the persons of the Godhead; each person loves the others with a sort of outpouring of love, a self-giving love. The kind of love which is the precondition of, and makes sense of, Kenosis (κένωσις). Why would such a God want to create a world? Well, He has an internal source of motivation: to love (with all that entails). Why did he create a world of rational free agents who can choose to enter into relationships with him? So that he could love, could pour Himself out (Kenosis) in love, et cetera…His nature, exemplified by the outpouring and self-giving love of the three persons, gives God a plausible motive for creating other persons with whom He might enter into a love-relationship. He creates out of, and as an expression of, that love which characterizes his Nature as a community.”

      So, there is justification for thinking God does, in fact, have a motive for bringing the universe into being and fine-tuning it. Tyler concludes that “so long as we do not have comparably good arguments for thinking that God does not have a motive as we have for thinking that the fine-tuning cannot be explained by either chance or necessity, the argument should be considered a success.” Dr. Scharp is wrong to think that there is no motive, and you are mistaken to think WLC does not deal with this objection (visit (i), (ii) Philosophical Foundations for a Christian Worldview and (iii) The Blackwell Companion to Natural Theology for more information on the FTA (if you want to critically engage with his work further)).

      Just as you suggest “[t]he person making the claim that design is more probable than chance does indeed have the burden of proof to justify this claim” so the objector has the burden of proof to show how the immensurable improbability of chance or physical necessity outweighs, probabilistically, the design hypothesis. (As a side note, if you seriously think that the design hypothesis is less probable than chance (or physical necessity), be sure to consult WLC’s objection against the chance and physical necessity hypothesis [3] relative to the probability of the design hypothesis).

      You conclude that “[t]he proponent of the FTA needs to show that chance is LESS probable than design” and “you cannot do this without making probability judgements (such as “it’s not improbable”) about God’s decision making on whether or not to design the universe, i.e. appealing to divine psychology.” If the FTA works, then the conclusion of the disjunctive syllogism is that “Therefore, it is due to design” and thus design is the most probable explanation (that is just the nature of a disjunctive syllogism). [4]

      [1] “Is There Evidence for God?” William Lane Craig vs Kevin Scharp Retrieved May 6, 2016.
      [2] Tyler has personally granted me permission to reprint partial email material here (and so all quotations arw directly from email correspondence). His comments, though, are (i) subject to his own revision and (ii) will be (in part) reprinted on his own website: . Here are two reasons to doubt stipulating a motive is necessary. (Tyler answers the question and relevance of ‘motive’ in the context of the KCA and not the FTA). (1) “Let’s suppose, for instance, that we were in a court of law where a criminal case was being brought before us. You represent the prosecution, and I am with the defense. Suppose you provide sufficient evidence to establish beyond reasonable doubt that my client is guilty, but I retort that you haven’t been able to show motive, and that without motive you can’t establish your case. Since all you need to do is convince a (reasonable) jury of my client’s guilt beyond reasonable doubt, you needn’t necessarily have any good theory with respect to motive. If my client’s fingerprints are on the murder weapon, and he has no alibi, and there’s video-footage of him at the crime scene around the time of the murder, and he was arrested with the victim’s blood all over his shirt, and so on and so forth, eventually stipulating a motive becomes unnecessary.” (2) “Suppose, somewhat irreverently, that God actually did the very thing some cocky new-atheist who thought he was being ‘cute’ dared him to do by rearranging the stars to write “I created you all. Signed: God. P.S. please stop reading Richard Dawkins’ books;” would Scharp rush to the scene and exclaim, index finger wildly pointed sky-high, “but wait, we need to prove motive before we go around saying things like that God did that!” No, surely not. Must we prove that God had a plausible motive to create a universe in order to establish beyond reasonable doubt that God did create the universe? No. The weight of evidence might be stacked in such a way that we can do without stipulating a plausible motive.”

      [3] To (i) make available for those wondering and (ii) for your philosophical consideration, I will re-print WLC’s argument against chance and physical necessity. WLC argues that physical necessity is not “a plausible explanation, because the finely-tuned constants and quantities are independent of the laws of nature. Therefore, they are not physically necessary.” With respect to chance, he argues as follows: “The problem with this explanation is that the odds of all the constants and quantities’ randomly falling into the incomprehensibly narrow life-permitting range are just so infinitesimal that they cannot be reasonably accepted. Therefore the proponents of the chance explanation have been forced to postulate the existence of a ‘World Ensemble’ of other universes, preferably infinite in number and randomly ordered, so that life-permitting universes like ours would appear by chance somewhere in the Ensemble. Not only is this hypothesis, to borrow Richard Dawkins’ phrase, “an unparsimonious extravagance,” it faces an insuperable objection. By far, the most probable observable universes in a World Ensemble would be worlds in which a single brain fluctuated into existence out of the vacuum and observed its otherwise empty world. So, if our world were just a random member of the World Ensemble, by all probability we ought to be having observations like that. Since we don’t, that strongly disconfirms the World Ensemble hypothesis. So chance is also not a good explanation.” In weighing the probability of chance, physical necessity and design, one must critically engage with the literature, not presuppose the truth of the conclusion one is most apt to accept.
      “Does God Exist?” William Lane Craig: Retrieved May 6, 2016.

      [4] You might profit from consulting Robin Collins on the FTA:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s