(2023)

[NB: Numbered headings repeat the numbering found in Paul Herrick’s “Job Opening: Creator of the Universe—A Reply to Keith Parsons.” Struckthrough headings in the main text are those that Parsons has skipped over or consolidated together for commentary.]

Part I

1. The Issue
2. Definitions
3. Parsons’ Argument
4. What I [Herrick] Shall Argue
5. Parsons’ Target: Varghese’s “Gap” Argument
6. Stating the Gap Argument
7. Parsons’ First Critique of the Gap Argument: Deflationary Argument #1
8. A Reply to Parsons’ Deflationary Argument #1
     8.1 Why Infinite Regress Does Not Work in This Case
     8.2 Analysis of This
            8.2.1 Preview
     8.3 So What is “Odd” About the Gap Question?
9. Parsons’ Second Critique of the Gap Argument: Deflationary Argument #2
     9.1 The “Principle of Parsony”
10. Why the Principle of Parsony is Unacceptable
11. Another Try: The Physicalist Interpretation of POP
     11.1 Preview

Conclusion to Part I

Part II

12. Does the Principle of Sufficient Reason Justify asking Varghese’s Question?
     12.1 “No” Argues Parsons
13. A Reply to This Argument
     13.1 Self-Subsuming Explanation
     13.2 Stage One: The Elemental Modal Distinction: Necessary and Contingent Truth
     13.3 Stage Two: From Truth to Being
            13.3.1 Contingent Being
            13.3.2 Necessary Being?
     13.4 Stage Three: Reflection
     13.5 Stage Four: Solution

Preliminary Evaluation: The Duty of Nonexplanation

14. A Critique of This: The Problem of the Infinite Re-Asker
     14.1 How to Stop the Infinite Re-Asker
15. A Powerful Objection to All of This: The Dilemma Argument
     15.1 The Dilemma for the Theist
            15.1.1 Ockham’s Razor
16. A Solution Proposed: The Logic of Personalistic Explanation
     16.1 Freedom and (Lack of a) Sufficient Cause
     16.2 Explaining Free Choices
     16.3 Problem Solved?
17. Back to the Necessary Being Hypothesis: A Closer Look
     17.1 The Philosophical Advantage of a Necessary Being Hypothesis
18. The Argument Continues: Theism vs. Grand Unified Theory
     18.1 Analysis of This: The First Inference
     18.2 The Second Inference
     18.3 Is Theism Explanatory at All?
19. A Further Worry
     19.1 Five Models of Explanation
20. One Last Time: In Defense of Theistic Explanation
21. Concluding Thoughts

Is it True?
Personalistic Explanation
The Demand for Complete Explanation
Conclusion

Part I

In 2009 Paul Herrick published on the Secular Web a critique of my paper from 2006, “No Creator Need Apply: A Reply to Roy Abraham Varghese.” My essay was a response to cosmological arguments offered by Varghese in his introduction to the book Cosmos, Bios, and Theos (1992), which he edited with Henry Margenau. Herrick’s response, “Job Opening: Creator of the Universe—A reply to Keith Parsons” is so lengthy—four times the length of my original essay—and contains so much detail, that it is only now that I am retired that I can find the time to give this critique the attention it deserves and offer a reply. This is why this response is so belated.

I very much appreciate the depth and meticulousness of the response to my essay. While I shall argue that my original arguments emerge unscathed, indeed strengthened, from an examination of these objections, it is a rare privilege and an honor to have one’s work analyzed with such care and intelligence by so capable a philosopher as Herrick.

I shall not recapitulate the earlier points and arguments from Varghese or summarize my initial essay but shall move right to an examination of Herrick’s text. I do frequently have to repeat points made in the initial paper when I think that they were neglected or misunderstood. Herrick has conveniently subdivided his remarks into short sections, and I find something to contest, clarify, or add to in almost every section. Therefore, I shall simply list each of his sections and follow it with my rebuttals to the points I find objectionable. I shall retain his titles to the sections, so that the pronoun “I” in the heading of a section refers to Herrick, not me. In citing Herrick’s text or my original essay, I will simply refer to the page or pages of the documents as posted on the Secular Web when printed in portrait layout on standard 8.5 × 11 inch letter paper.

1. The Issue

Herrick mostly fairly states the issue as I see it. Science cannot disprove the existence of a divine creator, but it can make otiose such a metaphysical posit. If science can provide all the explanation that the universe needs, i.e., all the explanation that it is reasonable to demand, then a supernatural creator becomes an ontological dangler, and is liable simply to be trimmed off. In short, a creator must have a job to do. If he does not, if he becomes a supernumerary in our broadest accounts of what is, he will rightly be sent into retirement with the other supernatural entities once thought to populate our world but now unneeded.

Herrick correctly states my charge to the defenders of the creator hypothesis: Show that a creator is needed. Show that something essential is missing from our worldview if it does not contain such a postulated entity. Indicate how broadly accepted canons of logic or rational inquiry are violated if no such posit is made. In practice, this means that the defender of a creator hypothesis must, to quote my earlier essay, show that there is “a gap for God that cannot be closed by the progress of science, a domain for divine activity in the world that advancing science cannot seal off, marginalize, or supplant with naturalistic explanations” (p. 1).

One bone to pick: Herrick asks the question, presumably put into my mouth, “Does the success of science at explaining the world mean that it is no longer reasonable to believe in God?” The irrationality of belief in God is not implied by my earlier essay. To have implied that I would have to have indicated that the only possible rational grounds for belief in God are the kinds of arguments I criticize there. But I never say or imply this. No, on the contrary, though I think that all cosmological arguments fail, I do not deny that some people might rationally believe in God. To impute to me the aim of trying to show that belief in God is irrational is to charge me with a burden of proof that I do not accept. My aim in the essay is the much more modest one of showing that certain ways of arguing for a Creator do not work and thereby supporting the claim that the universe needs no help to exist.

Just to be absolutely clear, here are two questions:

Is it reasonable to believe in God?
Is it reasonable not to believe in God?

I have nothing to say about the first question. I answer “yes” to the second question and my reason is that once we imagine the work of science done and that the universe has all the explanation that science can give, there is no intellectually compelling reason to ask for further explanation in terms of God.

2. Definitions

Most of Herrick’s definitions are unproblematic, but a couple of them require comment:

First, by “ultimate reality,” Herrick means “something that all other things depend on for their existence, but that does not depend on anything else for its existence” (p. 3). Herrick then says that advocates of scientific materialism, the view he attributes to me, consider “prime, unoriginate [sic]” matter to be ultimate reality. I would like to distinguish between ultimate reality in the diachronic sense and in the synchronic sense. The primordial singularity (or whatever original reality is postulated by a particular cosmological theory) can be seen as the ultimate reality in the sense of a temporally original state that existed at the beginning of time and is the causal progenitor of subsequent reality. However, since that original state no longer exists, then, obviously, the current universe no longer depends upon it. On the other hand, a necessary condition for the continuing existence of present realities, the solar system, for instance, would be the ongoing operation of fundamental forces of nature, like gravity. Take away gravity and things immediately start to fall apart. My claim, contra Varghese, is that the God hypothesis is not needed either (diachronically) as an original cause at the beginning of the universe or (synchronically) as a continuing cause (creatio continuans) required to sustain the existence or functioning of physical entities.

Second, the position that Herrick attributes to me is “scientific materialism.” I do not know exactly what Herrick means by that term. I never claimed it for myself. In the original essay my aim was not to promote a positive thesis so much as to criticize a theistic one. If the upshot was materialistic, then it was mostly so by default.

I prefer to characterize my position as “metaphysical naturalism” (MN). MN is more a thesis about what isn’t rather than what is. Metaphysical naturalists, qua naturalists, cannot say what the reality ultimately is; we think that science probably has yet to find it. We do think that the ontological inventory does not contain items such as gods, ghosts, angels, genies, demons, souls, mana, juju, vital forces, Platonic Forms, or any other supernatural, transcendent, spiritual, or occult entities, powers, or phenomena. We bet that if reality has a fundamental, “rock-bottom” level, that it comprises entities that are impersonal, nonmental, and nonteleological (we might happily admit that personal, mental, and teleological things exist, but not at the fundamental level). We base this bet on the enormous and consistent success of natural science in replacing theories that explain in personal, mental, and teleological terms with theories that explain in impersonal, nonmental, and nonteleological terms.

Though its ontological claim is thus rather negative, MN does make a positive epistemological assertion. MN holds that if and when physics finishes its job, i.e., when we have the final, most comprehensive physical theory, then the task of explaining the world is reasonably regarded as finished and there is no rational requirement to push on to the invocation of supernatural entities. MN holds that the original or most basic features of the physical universe, whatever they are, constitute a completely satisfactory terminus for explanation and an appropriate resting place for our curiosity. The world is sufficient. No Creator need apply.

It follows that to address MN the theist must show that the God hypothesis is needed, that is, the theist must show (a) that either the natural world or some element or aspect of it needs a supernatural explanation, i.e., that in principle naturalistic explanations are insufficient, and (b) that the theistic God is the best candidate for that required explanation. On the contrary, I hold that theistic hypotheses are spurious because (a) the explananda for which they are invoked require no explanation, or (b) natural explanations will suffice, or (c) the theistic “explanation” given is rife with insoluble problems.

3. Parsons’ Argument

This section begins inauspiciously:

Is modern science making belief in God logically unnecessary? Is the success of science in explaining the world rendering belief in God unreasonable? Parsons answers yes to both questions… (p. 3)

Actually, I ask neither question. Neither would I give an affirmative answer to either. If by “Is modern science making belief in God logically unnecessary?” Herrick means “Is modern science making belief in God no longer intellectually compelling?” I would answer “no.” I think that belief in God has never been intellectually compelling. As for the second question, again, I make no claim at all about the unreasonableness of belief in God. I am an atheist so I think that “God exists” is in fact false, but I do not hold, and I am puzzled how I could be construed as holding, that the belief that God exists is unreasonable if arguments such as Varghese’s fail.

The conclusion of my paper was (I thought) stated clearly:

It appears, therefore, that the choice between theism and atheism boils down to a choice between ultimate brute facts—God or the most basic features, whatever they might be, of the physical cosmos. I have argued in this paper that the choice of God is not intellectually compelling, that it is entirely rational and reasonable see the universe as needing no cause and capable of existing with no metaphysical support. The universe, insofar as it is explicable, is explicable in naturalistic terms. No logical or metaphysical principle, at least none mentioned by Varghese, requires that the world have a creator (Parsons, 10-11; emphasis added).

To say that it is reasonable to believe that p in no way entails or implies (or even hints) that it is unreasonable to believe that not-p. Herrick offers no citations or quotations to justify his claims about my basic claim. He simply offers a gratuitous stipulation about my conclusion.

My chief objection here arises from another claim Herrick attributes to me:

Parsons’ opening argument can now be put more clearly. I shall suppose that by “a God with nothing to do” Parsons means “a God with no explanatory role to full in our best theory of the universe.” If science were to reach a point where everything has been explained by a completed and well-confirmed physics, in such a way that nothing is left unaccounted for, then there would indeed be no explanatory need whatsoever to suppose that God exists. (p. 4)

But no theory, not even a Theory of Everything, will literally explain everything. Even a TOE will have to leave its ultimate terms of explanation unexplained. That is just the way that explanation—all explanation—works. When you explain A’s in terms of B’s, the B’s, at least for the time being, are left unexplained. No, my point is that the ultimate terms of physical explanation, whatever they turn out to be, may rightly and reasonably be regarded as brute facts, i.e., as ultimate realities that do not have any further explanation, but just are. So, God will have no explanatory work to do with respect to any x if either (a) x has a complete and correct physical explanation, or (b) x is a brute fact requiring no further explanation.

Herrick next offers a seven-step argument that he thinks is implied by my brief comments in my first paragraph. Actually, I have no objection to anything said in the first three steps of the argument. However, consider steps four and five:

4. In contemporary scientific cosmology, eternal universe theories do away with a singularity or beginning point in time and contain the explanatory resources to eventually completely account for the existence of the material universe as a whole so as to leave “nothing for a Creator to do or to have done.”
5. These theories are either well confirmed, or are rapidly on their way to becoming well confirmed. (p. 5)

Taken jointly, these premises appear to imply that I am committed to the correctness of eternal universe cosmologies. Not so. I hold that my arguments in particular (and MN in general) do not depend upon the correctness of any particular sort of physical cosmology, but are quite compatible with, for instance, cosmologies postulating an absolute beginning, an eternal, eternally recurring, or eternally inflating cosmos, or the sort of finite but unbounded cosmos proposed in Stephen Hawking’s A Brief History of Time.

Further, in the final step of the argument, the conclusion offered is, once again, stronger than anything I claimed. Herrick takes my conclusion to be:

[T]hanks to the progress of science, belief in the existence of God (or a Creator of the universe) is either epistemically unjustified right now, or is fast becoming so. (p. 5)

No. My claim in the original essay is that nonbelief in a creator is epistemically justified, not that belief in a creator is epistemically unjustified. Once more: The former does not entail the latter. Even if the sorts of cosmological arguments I criticize in my essay comprehensively and catastrophically fail (which they do), it does not follow that there are no rational grounds for belief in God. Alvin Plantinga (2000) and John Hick (2004) each offer very different ways of justifying belief in God that make no appeal to explanatory hypotheses. Now, I have qualms about both Plantinga’s and Hick’s projects, but those arguments do not enter at all into the essay Herrick criticizes. Herrick says that my argument rests on the following unstated assumption:

A: It is not reasonable to suppose that God exists unless there is an adequate explanatory reason to do so. (p. 5)

But I never make that assumption, and absolutely nothing that I argue in the paper turns on that assumption.

In fact, my opening paragraph is not even a philosophical argument, but a historical judgment. Gods have histories like anything else; they are born, they flourish, and they die. Gods seemingly die not because of philosophical debunking but because they become irrelevant. Gods, like humans, need something useful to do, or people start to forget about them. Has anyone ever disproved the existence of Zeus? I don’t think so, and, if not, why don’t people still worship Zeus? The reason is that at some point in history Zeus stopped doing anything useful for people. Nothing that mattered for human life seemed to require Zeus. Invocations of Zeus began to seem quaint and old-fashioned, a holdover from a less sophisticated and more ignorant era. The same thing has happened and is happening to the Judeo-Christian God.

A charming story told by John Searle (1999, p. 35) illustrates the ways that we—all of us—have changed. In Venice there is a medieval church named Madonna del Orto, the Madonna of the Orchard. The story goes that when the church was being built, it was originally to be named San Christoforo. However, a statuette of the Madonna was found in an orchard next to the cathedral site. To the people of the day, it was obvious that God had placed the Madonna statuette in the orchard to signify his will concerning the name of the new church, and so the church was named. Today, even a devout person finding a statuette of the Madonna next to the site where a church was being built would very likely dismiss it as a coincidence. Nietzsche’s madman was right; God is just not the vital, immediate, constant, crucial presence in our lives as he was for our ancestors a few centuries back.

Natural science has played the major role in this historical de-centering of God. People naturally and plausibly used to see the hand of God in innumerable phenomena that now even popes and TV preachers see as natural events. Such movements as “Scientific Creationism” and “Intelligent Design Theory,” as their leading advocates candidly admit, are efforts to reverse the modern marginalization of God by naturalistic science and clearly put his hand back into the governance of the world. So far, all that such efforts have produced is egregious pseudoscience (e.g., attempts to account for the fossil record in terms of the Noachian flood), or, at best, God-of-the-gaps arguments that seek refuges for God in the lacunae of scientific accounts (which is, e.g., what appeals to “irreducible complexity” really amount to). The problem is that science progressively closes these gaps, and gap-apologists must then beat a hasty and undignified retreat.

My challenge to theistic philosophers and apologists like Varghese is therefore straightforward: Produce a gap for God that science cannot close. Show that there is explanatory work that needs to be done but that science, in principle, cannot do. Show that MN is wrong to hold that the natural universe is the ultimate reality. My argument in the original essay is simply that arguments such as Varghese’s totally fail to meet that challenge.

4. What I [Herrick] Shall Argue

Herrick quotes my aim in my original paper (p. 4). I will quote my own statement:

I shall examine Varghese’s argument in detail and show why it fails. In particular, I argue that there is no intellectual difficulty in postulating an initial state of the universe as an ultimate brute fact, and that Varghese’s arguments to the contrary are flawed. (By ultimate brute fact” I man a primordial or original state of affairs, one which, though its existence is logically contingent, is not caused by dependent upon, conditioned by, reducible to, or supervenient upon any other prior or more basic entity or state of affairs.) (p. 2)

My aim in the paper is therefore (a) to debunk a set of arguments given by Varghese, and (b) to maintain that there is no reason not to postulate the initial state of the universe as an ultimate brute fact. In judging Herrick’s critique of my paper, I will therefore ask: (a) Does he show that my criticisms of Varghese are unsound? And (b) does he show that it is in any way unreasonable, irrational, or epistemically improper to postulate the initial state of the universe as an ultimate brute fact? If I find that Herrick fails to show these things, I will conclude that his critique fails.

Herrick announces two aims of his rebuttal:

In this paper I shall defend two claims. First, Parsons’ crucial fourth premise is false. Second, philosophical theism, far from being vulnerable to the continued progress of science, rests on a rationally satisfying and philosophically attractive logical basis that cannot, in principle, be overturned by the continued progress of natural science. Theism cannot even be overturned by the “Holy Grail” of contemporary physics a completed, well-confirmed grand unified “theory of everything… (p. 6)

Fair enough; I merely ask the reader once again to note my above qualifications about my conclusions in the original paper and that my arguments about eternal universe theories are hypothetical, not categorical: If an eternal universe cosmology is correct (a possibility raised by Varghese), then there is no reason not to regard the basic components and laws of the physical universe as ultimate brute facts needing no further explanation.

5. Parsons’ Target: Varghese’s “Gap” Argument

6. Stating the Gap Argument

7. Parsons’ First Critique of the Gap Argument: Deflationary Argument #1

I could correct and qualify several things Herrick says in these sections, but I have already done quite a bit of that, and I would like to cut to the chase. Hence, I will just skip over these sections without further comment and get to the issues where Herrick and I directly clash.

8. A Reply to Parsons’ Deflationary Argument #1

Suppose that the universe is infinitely old. Suppose further that each state of the universe is explicable in scientific terms. The nature of scientific explanation is a very large and much-debated topic (see, e.g., Salmon, 1990). Here I shall suppose that the terms of scientific explanation are broadly causal, that is, that successful scientific explanations will accurately represent some aspect or element of the objective causal structure of the world, where causation can be taken either deterministically or probabilistically. I shall further suppose that each state of the universe is explicable in terms of causes operating at earlier stages of the universe, and that those previous states in turn are likewise scientifically explicable, back ad infinitum. Varghese appears willing to make these suppositions, so I merely ask: “What, then, still requires an explanation? What needed answers have we failed to give?” If, ex hypothesi, every state of an infinitely old universe is scientifically explicable, what mystery remains such that we must posit something beyond? What rationally compelling question have we failed to address? What epistemic sin or dereliction of epistemic duty have we committed if we take the universe as a brute fact?

The reply, of course, is that the series of states as a whole has no account. Herrick, reflecting on an example proposed by philosopher William Wainwright, argues that explaining the existence of each member of an infinitely long series of humans would not explain the existence of humanity taken as a whole (p. 10, 11). Suppose that humans had existed through infinite time such that each human’s existence was explicable in terms of a pair of parents, and each parent in terms of another pair, and so on back ad infinitum. Would the existence of humanity, as a whole, be thereby explained? Herrick says not:

[T]here are two good reasons to maintain that the existence of such an infinite series as a whole would lack an explanation even if each individual member of that series in principle has one…. To reason from the claim that each part of the series has an explanation to the conclusion that the whole series is thereby explained commits a composition fallacy. (In general, a composition fallacy occurs when someone reasons that since each part of a whole has property p, it follows that the whole must have property p as well…. Second, and more importantly, it seems crystal clear, upon sustained reflection, that even if no one individual member of the series lacks an explanation, the following very general question about the series as a whole remains totally unanswered: “Why do human beings exist?” In other words, “Why does this past-eternal series of humans as a whole exist? Why is it not the case that no human beings ever existed? Or, to put it still another way: “Why does the inventory of all that actually exits include a (past-eternal) series of human beings. (p. 11)

So, does my critique of Varghese commit a composition fallacy? Not at all. My argument is not that each state of the universe has an explanation and therefore the universe as a whole has an explanation. On the contrary, I insist that the universe “as a whole” has no explanation and needs none. I take it as a brute fact. That is my whole point. My argument is that if each state of the ex hypothesi infinitely old universe is scientifically explicable then there is no need—no rationally compelling demand—for further explanation. We have satisfied the legitimate demands of reason without raising any specious questions about “the whole.” That there is an eternal universe with the fundamental features—the basic physical constituents and laws—that it has is just a brute fact that needs no further account.

But how can this be? Doesn’t this leave very important questions unanswered? Paraphrasing Herrick, why does the inventory of all that actually exists include a (past-eternal) series of this sort? Why this kind of stuff? Why these physical laws? If, for instance, protons and gravity have always existed, why have we always had protons and gravity instead of some other physical (or nonphysical) reality? We can imagine there having all along been a very different reality. Why then do we have this universe rather than one of the indefinitely many others we seem to be able to conceive? Aren’t these legitimate questions?

As I note in the original essay, these are certainly very odd questions, odd not just because they are distinctly philosophical questions, but because, despite their distinguished philosophical pedigree, it is hard to see just what motivates them. Asking why we have this kind of universe seems to imply that at some point there was a real (and not just imagined) possibility that we might have had some other kind. When and where was it ever possible that we would have had a different universe?” On the hypothesis of an infinitely old universe, there was no time or place when things might actually have been fundamentally different. The universe and its laws, including its conservation laws, have always existed, so something fundamentally different has just never been a physical possibility. Therefore, on the hypothesis of an eternal universe, what exists at time t is the consequence of initial conditions, i.e., those relevant physical conditions that exist at times preceding t, and the causal laws (whether deterministic or probabilistic) that applied in that circumstance—and this has been the case for every state at every past time. For anything fundamentally different to have come about at any given time, the prior conditions and laws would have to have been fundamentally different. But they weren’t.

No doubt the response to the above paragraph would be that I am still missing the point. The point is that there might “all along” have been a different kind of universe. But what is the significance of “might” here if it is not a physical possibility? In what sense might it have been the case that we had a different kind of universe “all along?” All it seems to mean is that we can imagine there having been a different kind of world “all along.” So what? There are all sorts of imaginable but unreal circumstances that do not demand an account of their nonexistence. However, for Varghese, Wainwright, and Herrick, our ability to imagine different sorts of universes creates a deep mystery and elicits an urgent “Why?” with respect to the existence of the actual universe.

The metaphysical naturalist shrugs and responds to that “Why?” with “Why not?” Is nonexistence the natural, basic, expected state so that existence per se must be explained? As Adolf Grünbaum asked (private communication) why should we suppose the “spontaneity of nothingness?” Why isn’t something just as natural, basic, or expected as nothing? The same could be said for the question “Why this?” Why not?

Is it a violation of some epistemic duty, norm, or principle, to dismiss with such insouciance a question that others find so compelling? What would that duty, norm, or principle be? Is the existence of the actual universe, as opposed to an imagined alternative, in any sense unexpected, improbable, or puzzling? Possibility is not probability, and to say that some other reality was possible in no way implies that it had any probability of being actualized, or that it was expectable in any sense at all. For the metaphysical naturalist, then, the existence of the actual universe and the absence of other imaginable universes are not mysteries that cry out for explanation, and there is no compelling reason to address them.

Is the epistemic demand simply that every question that can intelligibly be asked ought to have an answer? But this can’t be right. Some questions are foolish, pointless, idle, or loaded in such a way that an acceptable answer cannot be given. In those cases, the right thing to do is not to attempt an answer, but to challenge the question. This is what I am doing with respect to the question “Why does this universe exist instead of some other or none?” I regard these questions as pointless, what I call “pseudomysteries,” since I can see no reason why the existence of this universe was in any sense unexpected, improbable, puzzling, or mysterious.

Though I certainly do not think that I have, perhaps it will be complained that I have trivialized the issue. It is not merely that we can imagine other universes but that our universe, even if eternal, will be logically contingent, that is, our universe is not logically necessary. There are infinitely many possible worlds other than the real world. So, why do we have our world rather than some other logically possible world? However, to say that our universe does not exist as a matter of logical necessity does not make its existence in any way unexpected, improbable, puzzling, or mysterious. It just doesn’t. Repeating what I said above: Possibility is not probability, and to say that some other reality was logically possible in no way implies that it had any probability of being actualized, or that it was expectable in any sense at all. Logical contingency is not factual contingency. That something can, without contradiction, be conceived not to exist in no way implies that its existence is in any way dependent, conditional, or provisional. It might be a brute fact.

On what kind of theory does it become urgent to ask why our world exists rather than some other possible one? It would seem to be a kind of theory that proposes a sort of scenario, something like the one envisioned by Leibniz where he imagines the Creator prior to the creation surveying all possible worlds and selecting which one, if any, to actualize. Unless we presuppose a scenario of someone or something choosing or selecting one world out of all the possibilities, then we have no basis for asking why this one was expectable rather than some other. Apart from some such selection scenario, how would we assign probabilities for the actualization of possible worlds? How would we partition our infinite sample space of probable worlds, and what weights of probability should we assign to the subintervals we create? I do not know how to answer these questions, and Herrick has not indicated any answers either.

It appears that it is the theistic scenario itself, or something like it, that creates a demand for explanation that otherwise would not exist. “God” is not the answer to our ultimate “why.” The ultimate “why” is posed by those who have already given “God” as their answer. This is why I call the questions “Why this? Why anything?” pseudomysteries. The theistic scenario creates the “mystery” it answers. Naturalists reject the scenario.

8.1 Why Infinite Regress Does Not Work in This Case

8.2 Analysis of This

Herrick argues that a temporally infinite chain of logically contingent states, where each state is scientifically explained in terms of a preceding state (and recall that we are assuming that each explanation is broadly causal), cannot be explanatorily adequate. Herrick quotes with approval an argument by Laurence BonJour that there must be epistemologically basic beliefs because if I justify belief B by citing reason R, the justification is only provisional, and not conclusive, until R itself is justified. But if the string of purported justifications extends ad infinitum and never terminates in a properly basic belief then each stage in the string of justifications remains eternally provisional and there is no conclusive authorization for the purported justification at any stage. Herrick comments:

This seems right to me, and so does a qualified application of this general form of reasoning to an infinite explanatory regression of contingent beings, such as the one posited by an eternal universe hypothesis. When the contingent existence of A is offered as the explanatory ground for the contingently exiting object B, then the existence of B is accounted for by saying that A caused B but the contingently exiting object A itself has not been given an explanation. In that case, the existence of B is not conclusively accounted for, and further explanation is called for—the existence of A needs an account. (p. 12)

The example he adduces is when your car won’t start, and you discover that the battery is dead (p. 12). But, he says, this would be an inadequate and only partial understanding until I understand why the battery is dead. And understanding why the battery is dead will only be partial until I understand the further causes of my dead battery, and so on. He concludes:

[I]f each explanation in an explanatory series is similarly inconclusive, then at each stage in the series we lack a rationally satisfying, completed explanation. Therefore, if a series of contingent things goes back like this forever … then there simply is never going to be a completed, rationally satisfying explanation no matter how far back we go. (p. 13)

How, though, do we conclude from the alleged fact that each stage of explanation is incomplete to the conclusion that the whole infinite series of explanations is incomplete? How is this not a fallacy of composition? On the contrary, why not conclude that an infinite chain of contingent causes, qua infinite, tells the whole causal story? Considered sub specie aeternitatis, every physical state of the eternal universe will have a cause; there will be no gaps anywhere in the universal chain of causality. What is left that cries out for an account? Again, this series “as a whole” will not be logically or metaphysically necessary, but, once more, it is hard to see how this poses a genuine puzzle.

Herrick appears to be conflating two very different cognitive enterprises here: (1) Justifying a belief by tracing a chain of inferences back to a foundational belief. (2) Giving ever-deeper explanations of an event by tracing back a chain of sufficient reasons. The former kind of activity tells us when we are justified in holding a belief; the latter tells us why an event occurred by situating it at the terminus of a causal chain stretching back indefinitely. The former kind of activity must terminate in a finite distance; no such requirement is discernable for the latter. The former provides no justification until the end is reached; the latter increases our understanding with each step.

I deny the assumption behind the comparison, namely that an explanation in terms of contingent physical causes is rationally unsatisfying, incomplete, or inconclusive in the manner that Herrick claims. Consider the dead battery case. If that is in fact the correct explanation of why your car will not start, then this is the actual reason; there is nothing partial, inconclusive, or provisional about it. The explanation states an objective physical reality. Of course, we can inquire further into why the battery is dead if need or curiosity prompts us. Suppose, though, that we never find out why the battery died. Does this failure in any way invalidate, vitiate, undermine, or cast any doubt at all on the explanation of why the car won’t start? Why should it? The car won’t start because the goddamned battery is dead. Full stop.

Consider another example: The victim was killed by blunt force trauma to the head, the forensic investigators conclude. Of course, the law is very interested in finding out who the murderer was, and prosecutors will want to piece together the incidents that led up to the act of murder. Such details will be of great interest to the jury. But however these investigations turn out, or even if the case is never solved, one thing that is undoubtedly and clearly the case was that blunt force trauma was the cause of death. There is nothing partial, inconclusive, or provisional about that fact. What this example indicates, and such examples could be multiplied at will, is that causes can be quite sufficient for given effects even if we do not or cannot know the cause of the cause.

It appears indeed that the adequacy of a causal explanation is relative to what we want to explain. The blow to the head is a sufficient explanation for the death of the victim, whether the blow was delivered by a man, a woman, or (as in the Sherlock Holmes story “Silver Blaze”) a horse. We push our inquiry further, not because the causes we have adduced up to now have been insufficient for their effects, but because our interests extend further than merely identifying the proximate cause of death. Our demand for justice prompts us to inquire into the cause of that cause, i.e., to regard the cause of the death as an effect that requires a cause—the one who struck the blow.

Often indeed, when we seek the cause of a cause C, it is not to address any weakness, failure, deficiency, or inadequacy of C; on the contrary, we are frequently seeking to explain the efficacy or the sufficiency of C in producing its effect. We know that bread nourishes. That has been known since the first loaf was consumed. No question about it. We cite mitochondria, ATP, ADP, and the Krebs cycle not to justify or ground our belief that bread nourishes, but to show just why and how it serves our nutritional needs as it does. In a foundationalist epistemology, which Herrick seems to presuppose, non-basic beliefs remain unjustified unless ultimately grounded in foundational, properly basic beliefs, i.e., beliefs that may be accepted as rationally justified even though they are grounded in no further beliefs. A justificatory chain that does not terminate in such a properly basic belief will lend no justification to any belief in that chain, even if the chain is infinitely long. Without such ultimate grounding, the purported justification will amount to a summing up of zeroes, and even an infinite sum of zeroes equals zero.

Building up causal accounts is very different. Instead of having zero value until it is ultimately grounded in a putative uncaused cause, a chain of well-confirmed causal explanations is always good as far as it goes. Adding links to the chain of unjustified beliefs merely extends our ignorance; adding links to the chain of true causal explanations extends our knowledge with every link. If the universe is infinitely old and scientifically explicable throughout, as we have supposed, then it is knowable at each stage and the knowability goes back forever. This is a good thing.

Really, then, Herrick is comparing apples to pipe wrenches here. His purported analogy between infinite chains of causal explanation and infinite chains of epistemic justification is baseless and adds nothing to his case.

But isn’t an endless series of contingent causes really no better than the turtle lady’s account in the famous story about Bertrand Russell? Isn’t such a scheme really no better than saying that the earth rests on the back of a turtle, and that turtle on another, and so on, “all the way down?” Isn’t this just silly?

Let’s change the example slightly and ask how silly this new example is. 18th century deist philosopher William Wollaston asked us to suppose that one day we come across a chain dangling from the sky and terminating a few feet from the ground. Would we not have to suppose, Wollaston asks, that the other end of the chain, though invisible to us, must be attached to something, some sort of skyhook, perhaps, that is holding the whole thing up? Would it be reasonable to suppose that the chain could just extend infinitely without anything to hang upon? Wollaston took the absurdity of such a notion as obvious (quoted in Everitt, 2004, pp. 66-67).

Is it? Why not say that the chain needs no cosmic hitch precisely because it does extend infinitely? Every link of the chain is completely supported by the link above it, and that one by the link above it, and so on forever. At no point does any part of the chain lack adequate support. Besides, the chain example is a poor analogy. Nicholas Everitt comments:

Wollaston’s chain is a misleading analogy. In all our experience, chains are always of finite length, and if they are suspended vertically, they are always supported by something that is not itself a chain or another link in the chain. A chain might be hanging from a nail in the wall, it might be held aloft by a person, and so on. By contrast, in all our experience of series of events, we have never experienced an event which did not depend on a predecessor, and if we try to think of what an event might depend on other than a predecessor, no answer comes to mind. Our background expectations about the nature of physical chains of links are therefore very different from our background assumptions about causes in general, and we cannot assume that what seems obvious with the former will also apply to the latter. (Everitt, 2004, p. 67 [emphasis mine])

What applies to chains applies to turtles. We have never experienced chains that were not suspended from something not a chain and we have generally experienced turtles supported by nonturtles. Yet with causal chains, we have only experienced events caused by preceding events, and, as Everitt notes, it is not clear what besides a prior event we should have in mind when asked to account for events.

Everitt is exactly right, and, generalizing, we might say that stories like the one about the turtle lady or Wollaston’s chain often serve an obscurantist function in philosophical discourse. Such stories are bad analogies, but they are terrific intuition pumps, and philosophers can hardly ever keep their hands off a good intuition pump. But our everyday intuitions are shaped by everyday experiences of things like chains and turtles and are often quite inappropriate when we are trying to think clearly about things that are not of an everyday nature, like metaphysics or cosmology.

It seems to me an odd mental quirk to regard no question as really answered until every possible question is answered. That is, we don’t really know anything until we know everything. I regard this as far sillier than any story about turtles.

8.2.1 Preview

8.3 So What is “Odd” About the Gap Question?

Herrick asks three questions and then remarks about the history of science:

What if there is a way to provide a rationally conclusive explanation for the existence of a contingent universe? On purely rational grounds, wouldn’t this be preferable to an infinite regression of explanation? It would seem so; isn’t it always reasonable to prefer a complete over a partial explanation? The history of science makes no sense otherwise. (p. 14)

The history of science makes no sense apart from the demand for “complete” explanations? In its entire history science has never asked for nor offered complete explanations of the sort Herrick wants. On the contrary, science has thrived specifically by seeking focused, local explanations and happily deferring the deeper questions until later or dismissing them completely. One of Thomas Kuhn’s insights was that our theories determine which questions we regard as salient or negligible. Science therefore has often regarded some phenomena as givens that require no explanation. That is how Newton viewed constant rectilinear motion. That is how Bohr, as opposed to Einstein, viewed the probabilistic nature of quantum events (God does play dice with the universe!). So, the history and practice of science offer zero support to Varghese and Herrick’s project of total explanation.

OK, but—bottom line—isn’t more understanding, when we can get it, always preferable to less? Sure. When we can get it. However, theists and naturalists both agree that there are ultimate terms of explanation, and that when we arrive at these our quest for deeper understanding can reasonably rest; we run out of “why” questions. As naturalists see it, our knowledge will be complete when we have addressed every legitimate mystery. Invoking God after we have reached the limit of physical explanation is pointless because the questions that so engage theists—Why this universe? Why any universe?—appear to them to be pseudoquestions that demand answers to pseudomysteries. Though expressed in the grammatical form of questions, they pose no genuine intellectual problems.

Finally, Herrick says that “the question of existence”—why does “it all” exist instead of nothing—has been asked by philosophers since “the beginning of the discipline.” Actually, no. The Milesians pushed their inquiries to an ultimate arché, which was itself in physis, and then they stopped. The questions “Why anything?” never occurred to Plato or Aristotle. Their “cosmological” arguments (In Laws X and Metaphysics XII) sought an origination of motion, not an answer to Herrick’s “question of existence.” That question became pressing for the medieval Christian, Jewish, and Islamic philosophers, who, of course, were working in theistic traditions. The issue remained salient for some seventeenth century philosophers such as Leibniz, Locke, and Samuel Clarke, but, of course, it was no issue for Hume or Kant, nor, I think it is fair to say, for most major philosophers since. Further, even if certain issues have seemed pressing to the “greats,” there is no reason we cannot now disagree. Sidney Hook’s brilliant essay “The Quest for Being” cogently debunks the age-old quest to define the nature of “being” (Hook, 1991).

Herrick concludes, “Parsons’ deflationary argument therefore gets us nowhere.” No, my argument shows that there is no place to get to.

9. Parsons’ Second Critique of the Gap Argument: Deflationary Argument #2

9.1 The “Principle of Parsony”

10. Why the Principle of Parsony is Unacceptable

As Herrick notes, much of my argument can be summarized in this quote from my original essay:

Why should it surprise us that there is a universe? Why should it surprise us that we have this universe? What else should we expect? Of course, we can imagine that there might (i.e., conceivably could) have been nothing at all or that all sorts of other universes might have existed instead of ours, but this need not create any mystery. There are always innumerable imaginable possibilities whose failure to be realized creates no mystery at all. The moon could conceivably have been made of cheese, but it is no mystery that it isn’t. In general, it is no mystery why something does not exist unless, given our background knowledge, its existence was expected, or at least no more unexpected, than what does exist. Nothing in our knowledge base supports the slightest expectation that the moon would be made of cheese. Nor do we have any basis for thinking that some other (ex hypothesi eternal) universe should have existed all along instead of ours. Therefore, it is hard to see how asking “Why doesn’t some other universe exist?” is very different from asking “Why don’t we have a moon made out of cheese?” (p. 1)

Herrick amusingly calls the sentence italicized above “The Principle of Parsony” (PoP):

In general, it is no mystery why something does not exist unless, given our background knowledge, its existence was expected, or at least no more unexpected, than what does exist.

He then proceeds to argue that the PoP is unacceptable:

Is the principle of Parsony true, or even plausible? Why should we accept it? What is the reason for thinking it true? Is there a noncircular argument for it? Does the principle reflect our actual explanatory practices? First off, the principle of Parsony is certainly not necessarily true: its denial is not a contradiction or an impossibility. Neither is it self-evident. (p. 14)

In my original essay, I regarded the principle as obvious. I still do. When someone does not find obvious a principle that you do, all you can do is try to unpack it or elaborate on it in a way that, one hopes, its obviousness will be revealed. The key word of the principle is “mystery.” A mystery, as I have used the term, is something that cries out for explanation. It presents us with an urgent need for understanding. It would be intellectually irresponsible for a serious inquirer to ignore or dismiss that demand for explanation, that is, there is an epistemic (and perhaps a moral) duty to account for the difference between reality and expectation.

Since I regard the PoP as obvious, I will put the burden on Herrick to show that it is wrong. He offers four purported counterexamples. Before turning to these, I will consider three cases of genuine mysteries that conform to and illustrate my principle:

In the terrible 1849 cholera outbreak in London, infections clustered in the poor neighborhood of St. James Parish around the communal Broad Street Pump. However, the residents of a workhouse and a brewery in that area were largely unaffected. (See Goldstein and Goldstein, 1978).
Based solely on natural factors, the climate would be expected to have cooled since 1900. However, from the 1970s on, every decade has been hotter than the one before, with the ten hottest years all occurring since 2005. (See Romm, 2016 and Dessler, 2022)
When Xerxes, the Great King of Persia, invaded mainland Greece in 480 BCE, he came with an army and fleet much larger than anything the Greeks could match. Persia was vast, stretching from Anatolia to India, unified, and fabulously wealthy. The Greeks were few, disunited, and poor. Yet, at the Battle of Salamis, the Persian fleet was smashed, and at the Battle of Plataea the next year the Persian army was slaughtered.

Each of these cases presents a mystery: Why was infection absent from the workhouse and brewery? Why did climatic cooling not occur and dramatic warming instead? Why did the wealthy, powerful, and united Persians not easily defeat the fractious Greek city states? In each case, something is absent that, on the basis of background knowledge, would be strongly expected. Their nonoccurrence therefore required explanation. In his inquiry into the causes of cholera, John Snow found that the workhouse had its own water supply and that the workers at the brewery drank beer instead of the infected water from the pump. Climate scientists explain the expected absence of cooling and the dramatic warming that has occurred instead in terms of anthropogenic factors, such as the massive injection of carbon into the atmosphere due to the burning of fossil fuels. The triumph of the Greeks was explained in detail in Herodotus’ Histories.

Innumerable other real and imaginable cases conform to the PoP by the nonoccurrence of something strongly expected which therefore demands an explanation. On the other hand, we can imagine many circumstances where something imaginable does not occur, but its absence creates no mystery: Why is the Moon not made of cheese? Why (presumably) is there no teapot in orbit between Earth and Mars? Why did Godzilla not attack Tokyo last year? These things are not mysteries. These scenarios are possible in the sense that they can be coherently supposed, but there is no urgent (or any) demand to explain their nonoccurrence. They do not cry out for explanation. Not addressing these does not make us irresponsible, dishonest, or in violation of epistemic duties. There is nothing, absolutely nothing in our background knowledge that would make such “possibilities” the least bit likely, expected, or even barely plausible. Therefore, their absence is no mystery. So, as the PoP indicates, some questions are urgent, and others are nugatory.

These and innumerable other examples could be cited as conforming to the PoP, but let’s turn to Herrick’s alleged counterexamples:

Susan, a physics undergraduate, hears some grad students mention that the charm quark, c, has a 2/3 charge. Though she has no reason for thinking that the charm quark should have this charge rather than some other, or no charge at all, she wonders why the charm quark would have a 2/3 charge.
A philosopher is pondering the free will problem. She wonders why we have this feeling that the future is open to us. She has no reason for thinking that we should or should not have this feeling, or that this feeling is not more unexpected than the feeling that the future is not open. Yet she still wonders why we have such an intuition and seeks an answer.
Thales sitting on the dock at Miletus ponders the question of the one and the many. Nothing in his experience tells him that there is any expectation that there would or would not be one ultimate arché behind the apparent plurality of things. Nevertheless, he asks the question.
A student in a class on medieval philosophy hears that Thomas Aquinas walked from Italy to Paris to study under Albert the Great, but, learning that Albert had taken a position in Germany, walked to Cologne. The student, though having no knowledge of how people generally traveled in those days, still asks why Aquinas walked rather than riding a horse or donkey.

After offering these supposedly crushing counterexamples Herrick comments:

In these and other realistic case of explanatory practice, the underlying explanatory principle is not the Principle of Parsony (PoP), but instead seems to be what I shall name the daring inquiry principle (DIP):

When confronted with the existence of some unexplained phenomenon X, it is reasonable to seek an explanation for X if we can coherently conceive of a state of affairs in which it would not be the case that X exists. (p. 15)

Far from being crushing, these purported counterexamples are glaringly irrelevant. There is nothing, absolutely nothing in the PoP or in any aspect of my argument that forbids or makes it the least unreasonable to ask such questions. The PoP is not a stricture against curiosity, not even the idlest curiosity. Indeed, simple curiosity is one of humanity’s most appealing traits. Wonder is wonderful. I have said nothing whatsoever to imply that such curiosity is the least dubious, and by implying that I have Herrick argues against a straw man.

No, the PoP addresses what we must ask, not what we may ask. It does not judge which questions are possible, but which are compelling. For Varghese and Herrick, the questions “Why this universe? Why any universe?” are deep—actually, the deepest—mysteries. For them, these questions are exigent, pressing, compelling and cry out (loudly) for explanation. They are not simply questions that may be asked, but questions that philosophers, at least, should ask. The whole purpose of my essay was to debunk this demand, to show that, once all scientific questions are answered, the further metaphysical question of why this or any universe exists is not a pressing issue, indeed, not an issue at all. Thales stopped his inquiry when he reached what he took to be the basic stuff of nature. I think he was right.

Perhaps Herrick would deny the possibility of brute facts. Perhaps he is assuming a strong version of the Principle of Sufficient Reason (see below), a principle much stronger than his DIP, that insists that everything must have a sufficient reason why it is so and not otherwise. In this case, any brute fact would count as a “mystery” for Herrick.

Herrick nowhere states or defends such a strong principle. He suggests that there could be a total explanation in terms of a self-explaining God. He also makes it very clear that he would prefer such a total explanation to any account that leaves some facts unexplained. However, such possibilities and preferences are irrelevant because reality is under no obligation to respect them. Unless Herrick shows that there cannot be brute facts, he cannot deny their possibility, and the fact that he does not like them is neither here nor there.

As for the DIP, it would seem to lead to precisely the kind of infinite regress of explanation that Herrick wants to avoid. Would not every explanation leave us with something that we can coherently conceive not to be the case? Even if we invoke God, is it not coherently conceivable that God might not exist? In that case, according to the DIP, it would be reasonable to ask for an explanation why God exists. Of course, the only way to stop this regress would be to claim that God’s nonexistence is logically impossible and cannot be coherently conceived. We see below if Herrick makes this case.

11. Another Try: The Physicalist Interpretation of POP

Herrick next says that I have claimed that only physical antecedents and the laws of nature can give us expectations about what we will or will not find in the world, and he calls that alleged claim “the physicalist interpretation of PoP.” He then offers an argument that physical antecedents and laws of nature cannot be the only reasons governing our expectations about the world:

(a) We obviously do have knowledge of the physical antecedents and laws of nature required by PoP and this knowledge is certainly empirical in nature (empirical or a posteriori knowledge is knowledge justified on the basis of our sense experience of the world, in other words on the basis of data provided by our physical senses—sight, touch, taste, etc.).

(b) But if the physical interpretation of PoP were strictly true, we could never attain knowledge of physical antecedents in the first place. Consider the exact moment before the first instance of coming to know a physical antecedent or a law of nature. At that moment, we would not know any physical antecedents or laws of nature. But then (according to PoP) we could not ask questions such as “Why does this rather than that happen? And “Why does this rather than that exist? (p. 15)

Herrick continues:

Yet questions of this type are surely necessary for any scientific investigation of the world, including the discovery of the very physical laws of nature and physical antecedents required by PoP. In other words, at the beginning of all empirical investigation, lacking the basis of expectations required by the physicalist interpretation of PoP—knowledge of physical antecedents and laws of nature—we could not legitimately ask the sorts of questions that must be asked if we are to pursue and attain the empirical knowledge presupposed by the physicalist interpretation of Pop Thus the physicalist interpretation of PoP is self-referentially incoherent: it presupposes that we possess a particular type of knowledge, while at the same time making it impossible that we ever attain that type of knowledge in the first place. The physicalist interpretation of PoP cannot be right if we have any empirical knowledge at all. (p. 16)

Put simply, Herrick’s argument here seems to be that it cannot be right that we base all our expectations about what can or cannot be on physical antecedents and the laws of nature because, if this were so, we could never have discovered those physical antecedents or laws of nature. Hence, the claim is incoherent because it is self-refuting. QED, supposedly. Herrick adds a self-congratulatory note: “So much for the PoP.”

I have searched diligently through my original essay, and I cannot find any proposal of what Herrick calls “the physicalist interpretation of PoP.” The closest statement I could find was this:

What possible grounds could we have for thinking that it is a puzzle or mystery why our universe exists rather than some other imaginable universe, or even none at all? In scientific contexts, when only one out of a range of relevant alternatives (what Bas van Fraassen calls the “contrast class”; van Fraassen, 1980) is realized, we naturally and rightly assume that there is some reason why this happened rather than that. We rightly assume that there were physical antecedents and relevant physical laws that determined, at least probabilistically, the occurrence of one event out of a contrast class. But when the “occurrence” we are interested in is the existence of the universe as a whole, by definition, there are no physical antecedents or applicable natural laws that justify the expectation of an answer to our question “Why this rather than that?” What could make the universe as a whole probable or improbable, expected, or unexpected? (p. 1)

I can see no way to torture this passage and make it say what Herrick claims I have asserted as the “physicalist interpretation of PoP.” The passage explicitly cites a scientific context, in which it is surely true that we explain in terms of physical antecedents and the laws of nature. Nor can the passage be made to yield a categorical claim about such scientific explanation being the only kind of explanation. At most, the passage puts the burden of proof on those who would offer another kind of explanation to account for the existence of the universe “as a whole.”

So, Herrick has proposed and refuted his own principle.

11.1 Preview

Conclusion to Part I

This pretty much concludes Herrick’s critique of my original essay. The remainder of his paper is devoted chiefly to the development of his own theistic argument, and I will address it below. To this point, Herrick’s critique has chiefly been a farrago of misrepresentation and irrelevance. I therefore see no need to retract, qualify, or modify any of my original claims.

Part II

I am sure he would not, but Herrick could concede everything I have said to this point and still insist that nevertheless he does have an argument that shows that there is an intellectually compelling explanation of why this universe, or any universe, exists. So, let’s see.

12. Does the Principle of Sufficient Reason Justify asking Varghese’s Question?

12.1 “No” Argues Parsons

Theists have frequently supported arguments such as Varghese’s by appeal to the Principle of Sufficient Reason (PSR). Most memorably stated by Leibniz, that principle holds that nothing can be so unless there is a sufficient reason why it is so. Further, as Simon Blackburn notes, for Leibniz, there has to be an ultimate reason why all things are so, and this reason must be of a special sort:

But the reason has to be of a particularly potent kind: eventually it has to ground contingent facts in necessities, and in particular in the reason an omnipotent and perfect being would have for actualizing one possibility rather than another. (Blackburn, 2005, p. 355)

The PSR therefore makes two claims: (1) Every fact has a reason sufficient to explain why it obtains and not something else. (2) Every contingent fact must ultimately be explained in terms of something necessary. Something is contingent if it is possible for it to be or not be. Something is necessary if it is not possible for it not to be. As philosophers often put it, whatever exists contingently exists in some possible worlds but not others, while what exists necessarily exists in all possible worlds.

J. L. Mackie in The Miracle of Theism (1981) denies the first claim of the PSR, that is, he denies that every fact has a sufficient reason why it is so. Actually, he says that this claim must be false. I go over this argument in my original essay and Herrick outlines it well, so I will here only offer a summary. Mackie appeals to the logic of explanation. He notes that in every explanation there is something explained (the explanandum) and something that explains it (the explanans). The explanans can then be regarded as a new explanandum, and a new explanans will be sought to explain it. This process can be repeated indefinitely, but at every stage the given explanans will itself be unexplained, and this is so however far our explanations go. This is a necessary feature of the logic of explanations, and, qua necessary, it means that the PSR is necessarily false. The demand for an explanation of everything cannot, in principle, be satisfied. Necessarily, something will always remain unexplained.

The importance of Mackie’s critique of the PSR for my argument is that if the PSR is false, then there is no reason to be dissatisfied with an ultimate, brute unexplained explainer. The basic features of the universe, as discoverable by science, can be that ultimate brute fact, and there will be no rational demand to push our search for explanations any further.

Herrick, however, defends the possibility of total explanation.

13. A Reply to This Argument

13.1 Self-Subsuming Explanation

13.2 Stage One: The Elemental Modal Distinction: Necessary and Contingent Truth

13.3 Stage Two: From Truth to Being

13.3.1 Contingent Being

13.3.2 Necessary Being?

13.4 Stage Three: Reflection

13.5 Stage Four: Solution

I will summarize Herrick’s argument in these sections, pausing along the way to ask questions or to make observations. I will then offer a preliminary evaluation to see what Herrick has accomplished to this point and to see what remains to be shown.

Herrick says that scientific explanations will indeed always end with that, at least for the time being, is brute. However, philosophical explanations might answer everything:

Consider one way in which the universal claim that all explanations must leave something completely brute in their wake might be false. Suppose there were an existentially self-subsuming explainer—an explainer of existences whose own explanatory field curves around to account for its own existence (and other existents explained in the explanandum). It that case, the hypothesized existence of the existentially self-subsuming explainer could serve as the explanation for the existence of a whole realm of thing without leaving its own existence brute (and thus without leaving any irreducibly brute facts of existence in its wake). This idea may be summed up by saying that such an explainer would be a “self-explaining explainer.” (pp. 17-18)

Herrick proposes a self-explaining explainer (SEE) as a theoretical possibility and asks whether such an explanation of existence is possible. That it is not possible is not obvious, but neither is it obvious that it is. Note that saying that something is self-explaining is not to say that it or any of its parts or properties need no explanation. That could be true of a brute existent. Rather, it is to say that something needs an explanation and somehow provides it for itself. It is both explanandum and explanans. What kind of explanation would self-explanation be? None of our standard models or paradigms of scientific or mundane explanation would apply, so the SEE’s mode of self-explanation would presumably be sui generis. And, of course, suggesting an SEE in no way indicates that such a thing is real or even possible.

The SEE would have to be unique in another sense as well. To do the job Herrick wants, there will necessarily have to be only one SEE. If there could have been others, then there could have been indefinitely many other universes, each with its own SEE. In that case, it would be unexplained why this universe with its SEE exists rather than some other universe with its SEE. In other words, the question of why this universe exists rather than some other coherently conceivable universe would still not be answered.

Herrick begins his search for a SEE by noting the distinction between necessary and contingent truth:

[I]n logical theory we say that a proposition is necessarily true if and only if it is true in all possible circumstances, and thus false in absolutely none. A necessary truth like the Pythagorean Theorem is therefore steadfast across all possible circumstances…. A proposition is contingently true if and only if it is true in some possible circumstances and false in others. (p. 19)

Actually, as Herrick notes, the Pythagorean theorem is not true in all circumstances, but only in Euclidean non-curved space. In fact, many truths are necessary only given certain axioms or stipulations. Further, at least one “logical truth,” the law of the excluded middle—p ∨ ~p—even though it is a tautology of sentential logic, is rejected by some philosophers. As the joke goes, there are three kinds of philosophers, those who accept the law of the excluded middle and those who do not (quoted in Henle, Garfield, and Tymoczko, 2012, p. 167). Finally, W. V. O. Quine famously argued that there are no genuinely necessary truths and that, as Simon Blackburn puts it, necessity is really disguised contingency (Blackburn, 2005, p. 248). Perhaps, though, these are quibbles, so, for the sake of argument, let’s allow Herrick’s statement to pass.

Herrick next makes a rather breathtaking claim:

It is not that much of a leap from a recognition of two modes of truth—the contingent and the necessary—to the thought that there might be two corresponding modes of existence or being as well. (p. 19)

Many philosophers have held that logical necessity only applies to propositions and that it is a category mistake to speak of necessary beings. Ronald Hepburn puts it like this:

Necessity in the logical sense is a characteristic not of beings but of propositions. To speak of a necessary being is as crude a category mistake as to speak of a “contradictory sheep” or a “self-evident frying pan”—and for exactly the same reasons. (Hepburn, 1967, p. 234)

Whether Herrick means necessity in a logical or some other sense remains to be seen.

I previously said that a contingent being is one that can be or not be. There is a minimalist meaning of “can not be” such that an existent entity x is contingent if its nonexistence is coherently conceivable, that is, if we can suppose its nonexistence without self-contradiction. Mt. Everest exists, but, though factually incorrect, there seems to be no contradiction in the assertion that it does not exist. For Herrick, here at least, “contingent” apparently has a much stronger meaning. He regards contingency as a real condition or property of things—a kind of existential defect afflicting every physical object of our experience, including ourselves.

For Herrick, for something x to be contingent means that, considered both synchronically and diachronically, x’s existence is wholly dependent on conditions, causes, or circumstances external to itself. For instance, a human being’s continued existence is dependent upon any number of factors, such as the presence of oxygen, food, water, a tolerable ambient temperature, and many other factors (p. 19). Further, that human’s beginning was dependent upon a long string of antecedents, such as having certain parents, grandparents, and so on back through the whole course of evolution, and, indeed, the whole history of the universe.

Herrick thus defines “contingent being” as follows:

Thus we get the strict definition of a contingent entity or being: there are possible circumstances in which it would not exist. To which it seems logical to add: If certain conditions fail to obtain, its existence is toast, snuffed out like a candle in the rain. This is the miserable contingency, the “might not have been-ness,” that is our lot, and that is the lot of all the mundane or material things around us from atoms to molecules, chemicals, rocks, lakes, mountains planets, stars, galaxies, and so on. (p. 20)

The sense of “possibility” Herrick is using here does not appear to be logical possibility—coherent conceivability—but physical possibility. As he says, humans require a tolerable ambient temperature. If climate change raises the average summer high temperatures in Pakistan and India to 50 degrees Celsius, those areas will not be habitable by human beings. The physical possibility of our existence has depended, does depend, and will depend upon innumerable such circumstances.

The opposite of contingent being is non-contingent being:

A completely noncontingent being, if such were to exist, would logically have to be a thing whose existence is not dependent on any external circumstance (nor ever could have been). It would seem to follow that its existence could not possibly ever have been threatened by any possible circumstances or entity, nor ever could be. Moreover, no external circumstances or entity could possibly have caused its existence (or prevented or aborted it). That is the only way such a being could exist in a manner completely independent of all possible circumstances, that is, independently of all contingencies. This is a matter of pure logic. In theory, such a being would have exited no matter what, and would continue to exit no matter what. That is, it would exist in all possible circumstances, and be nonexistent in absolutely none. (p. 20)

It is appropriate to call such a noncontingent being a “necessary” being.

There seems to be an ambiguity here; Herrick’s use of “possible” and “could” might be construed in at least two different ways. When he says that the necessary being “would exist in all possible circumstances,” does he mean in all logically possible circumstances, i.e., in every possible world, or in every physically achievable circumstance? The definition of “contingent being,” as noted above seems to be in terms of physically realizable circumstances, that is, circumstances possible given the laws of physics in our universe—circumstances like the presence of absence of oxygen or having or not having certain parents.

If “necessary” is defined in opposition to “contingent,” which seems to be Herrick’s aim here, then a necessary being would be one that would exist in every physical circumstance realizable in our universe but not in all possible worlds. In this case, however, there seems to be no reason that an ultimate brute physical existent could not be noncontingent in this sense. There seems to be no reason why an ultimate, brute physical reality could not be such that, in Herrick’s words, “it has always existed and always will, as there never was a time when it did not exist, nor will there ever be a time when it does not exist…” (p. 20).

Perhaps, as scientists have hypothesized since the pre-Socratics, there is a permanent stock of basic “stuff” and laws governing it that always has been and always will be. There will be possible worlds, worlds with different laws of nature, in which these do not exist, but in our universe, the ultimate archai will have always existed and always will exist. Note that a beginning in a Big Bang would not contradict such noncontingency. Since time began with the Big Bang, it will still be true of whatever starts then that “there never was a time when it did not exist.”

Further, if we take contingency to mean what Herrick seems to intend—existent in some physical circumstances but not others—then it is obviously absurd, a category mistake, to speak of the universe as a whole as contingent in this sense. The universe is the sum total of all physical circumstances, space-time and whatever its contents, and so its existence logically cannot be dependent upon any particularities of physical circumstance. If physical circumstances had been different, if, for instance, the asteroid that ended the Cretaceous had not hit, then that would be how the universe is. If, then, the universe is contingent, it must be so in some other sense.

I attempt later to sort out and pin down Herrick’s precise sense of “contingent” and “necessary.”

Herrick quotes Sartre to confirm the alleged awfulness of contingency, the nausea felt when you contemplate that everything is “gratuitous.” I am reminded of a cartoon I once saw in the New Yorker. A man is staring disconsolately out the window and his wife inquires, “Why did you read a book called Being and Nothingness anyway?” Really, though, such sentiments are a kind of self-pitying Weltschmerz, a comprehensive disappointment with reality that can be indulged only by one physically comfortable and with angst and time to spare. My answer to such pooped world-weariness is “Get a life!”

Finally, Herrick makes another small leap from necessary being to God:

Now imagine a necessary being with certain features: sufficient power, knowledge, and will to create a universe; free will in the normal, everyday-language sense of the term “free” (i.e., the power to initiate a course of action, to make a choice that is not necessitated, caused or predetermined by the sum total of all antecedent conditions and external circumstances or forces); and will to freely give the gift of existence to a world of creatures, either directly by divine fiat, or through an independent evolutionary process. It seems crystal clear that hypothetically, that sum total—a free act of creation by a sufficiently knowledgeable and powerful necessary being—would all by itself constitute a logically sufficient causal condition for the existence of a contingent universe. (p. 21)

And in virtue of providing such a logically sufficient causal condition, the clear superiority of theism over atheism is demonstrated:

So let’s take stock. Philosophical theism provides us with a sufficient cause, sufficient reason, and sufficient explanation for why a contingent universe exits. On the other hand, atheism, on Parsons’ own admission, does not; rather, it leaves the existence of the material universe an unexplained, purely brute fact our contingent universe has no explanation claims Parsons: it just is. But in science and philosophy, and in many explanatory contexts, providing a logically sufficient reason for something is normally considered a very respectable form of explanation. The philosophical theist adds: It sure beats the only alternative—leaving it all unexplained. (p. 22)

Preliminary Evaluation: The Duty of Nonexplanation

It is easy to see why Herrick’s scenario is attractive to many. We crave closure. Who would read the adventures of Hercule Poirot or Sherlock Holmes if they did not discover “whodunnit?” What could be more intellectually satisfying than an explanation of everything? Yet a sometimes painful lesson of both philosophy and science is that the rush to closure must often be resisted. Humanity has never lacked sufficient reasons for every puzzle: The Milky Way is milk from the breast of Hera. Lightning strikes when Zeus hurls the thunderbolt. Plague is caused by arrows from Apollo’s bow. Spring comes when Persephone returns from the underworld. Myth, lore, and legend are full of sufficient reasons.

Less innocently, pseudoscientists, paranormalists, and ideologues can always provide sufficient reasons. How do we explain the gigantic and architecturally astounding pyramids, henges and other amazing monuments from antiquity? Ancient astronauts, employing superadvanced technology, must have aided our ancestors in such constructions. How do we explain alleged instances of extrasensory perception, mind reading, second sight, etc.? They must be due to the mysterious but powerful workings of “psi.” I recently asked a biblical literalist how Jonah could have survived undigested in the belly of the “great fish” for three days and three nights. Easy: God could have provided him with a La-Z-Boy recliner, a stocked refrigerator, and a big screen TV if he had wanted to. You can provide a sufficient reason to dismiss almost every difficulty if you are allowed to invoke ad hoc miracles.

The upshot is that human beings have never lacked appealing pseudoexplanations for just about anything. You can think of six before breakfast. Herrick thinks that his presentation of a scenario of a sufficient explanation of the universe in terms of a necessary being (and a scenario is all he has so far given us) shows that his account “sure beats the only alternative—leaving it all unexplained.” Right?

Wrong. Rationality very often requires “leaving it all unexplained.” Proposed explanations often have to be rejected despite many attractive features. Indeed, much obscurity has resulted from the rush to intellectual closure, seizing and refusing to let go of explanations too good to be true. Theism explains more? So do conspiracy and crackpot theories. In short, no explanation is always better than a bogus one.

Even good explanations, ones later universally accepted by scientific communities, were initially rejected for plausible reasons. Alfred Wegener proposed the theory of continental drift to explain many puzzling phenomena: the fit of continents now separated by ocean, the ranges of fossil organisms now divided by seas, and mountains on different continents with identical geology, as if they had once been part of a single range. Though geologists at the time could not explain these phenomena and recognized the inadequacy of their current theories, Wegener’s theory was rejected largely because of methodological considerations (Oreskes, 2019, pp. 80-87). Similarly, evolutionary explanations were rejected, and reasonably so, until Darwin explained it in terms of natural selection.

A hard-bitten skepticism is a necessary component of science and philosophy. The human mind abhors an explanatory vacuum, but sometimes we have to suspend our abhorrence and embrace the vacuum. We have a responsibility to leave phenomena temporarily or even permanently unexplained rather than accept an attractive but poorly supported explanation. On many occasions, both science and philosophy have had to displace received explanations that had become so deeply entrenched that they became obscurantist. Bacon, Galileo, Hobbes, and other pioneers of the new science and philosophy reviled Aristotle as the archobscurantist. Really, though, the problem was that Aristotle’s explanations were so thorough, so beautiful, so satisfying, and so congenial both to religion and common sense, that abandoning them for something less comprehensive seemed like intellectual temerity. Nonetheless, the purge was necessary.

It follows that Herrick’s encomia to the beauties of total explanation are irrelevant or even misleading. He must show that his God scenario is not just attractive, appealing, or would be a nice thing to have. He has to show that it has the credentials of a genuine explanation: that it is (a) true, and (b) really does explain what it is supposed to.

And there are even deeper reasons for resisting the rush to explain.

The first is the Kantian one. In his first Critique, Kant argued that when we succumb to the blandishments of metaphysics, when we try to achieve ultimate answers, our thinking quickly becomes “dialectical.” That is, we employ concepts that are indispensable in science and mathematics but those concepts, deracinated from the only contexts that make them meaningful, now lead us into antinomy and confusion. As Kant saw it, the tragedy of human reason is that we feel compelled to ask questions that can only be answered by pure reason, but which pure reason is not qualified to answer. Maybe all metaphysics can give us are appealing scenarios with no solid answers. I think that many of today’s philosophical theists do not take Kant’s caveats seriously enough.

More basically, we have so far seen no reason to expect that all of our intellectual aspirations will succeed. Reality is under no obligation to conform to our intellectual desiderata. For all that Herrick has so far said, the universe might nonetheless be a brute fact. There might be no total explanation. To our yearning for intellectual closure, reality replies “Tough!” Our prime intellectual duty is not the imperative to insist on explanations for everything, but to understand everything as it is. If there are ultimate brute facts—and Herrick has not yet shown that there are not—then that is how we should want to understand them, and we will misunderstand them if we insist that they must have an explanation.

14. A Critique of This: The Problem of the Infinite Re-Asker

14.1 How to Stop the Infinite Re-Asker

15. A Powerful Objection to All of This: The Dilemma Argument

15.1 The Dilemma for the Theist

15.1.1 Ockham’s Razor

16. A Solution Proposed: The Logic of Personalistic Explanation

16.1 Freedom and (Lack of a) Sufficient Cause

16.2 Explaining Free Choices

16.3 Problem Solved?

Since I am already over 13,000 words, I will not summarize these sections but will refer back to some of these points when I make my final evaluation of Herrick’s argument. I will now proceed to his final sections, his definitive defense of the necessary being hypothesis.

17. Back to the Necessary Being Hypothesis: A Closer Look

17.1 The Philosophical Advantage of a Necessary Being Hypothesis

18. The Argument Continues: Theism vs. Grand Unified Theory

18.1 Analysis of This: The First Inference

18.2 The Second Inference

18.3 Is Theism Explanatory at All?

19. A Further Worry

19.1 Five Models of Explanation

20. One Last Time: In Defense of Theistic Explanation

21. Concluding Thoughts

In these final sections, Herrick provides his exhaustive (and exhausting) defense of his necessary being hypothesis:

NB: A necessary being, with sufficient power, knowledge, and other attributes, freely chose to create a universe.

He notes that NB breaks down into two parts:

There exists a necessary being with sufficient power, knowledge, and other attributes requisite to creating a universe.
This being freely chose to create a universe.

Given my argument in my original essay and given my additional arguments in Part I showing the complete failure of Herrick’s critique of that essay, I am taking as the default hypothesis my claim that the universe exists as the ultimate brute fact, neither having nor needing an explanation. Therefore, I am putting the entire burden of proof on Herrick to show both (a) that a necessary being with sufficient power, knowledge, and other attributes requisite to creating a universe exists, and (b) that the free acts of that being satisfactorily explain the existence of the universe. Therefore, in the remainder of this essay I will address these questions to Herrick’s NB hypothesis: (1) Is it true? and (2) Can it explain the universe? Note that Herrick needs to answer both of these. Establishing the existence of a being with the attributes necessary to create a universe does not mean that his proffered explanation in terms of the free choices of that being is a legitimate one.

Is it True?

Herrick succinctly states his case for the existence of his necessary being:

I will argue that: (a) philosophical theism provides us with a (philosophical) hypothesis that answers the ultimate question of existence, while atheism does not and cannot; and (b) this counts as a pretty good explanatory reason to believe that God exists. After all, if one hypothesis H1 accounts for the data, while a second hypothesis H2 does not, that is a good reason to prefer H1 over H2. (p. 16)

Which data? Theists think that the universe is a datum to be explained. Atheists do not. Herrick often mentions my claim—as if it were a forlorn concession—that the universe is unexplained, but he ignores the second and more important part of my statement, viz., that the universe needs no explanation. Atheists and theists do not agree on what the “data” are. Theism and atheism are not competing explanations for the existence of the universe; they are opposing theses about where explanation satisfactorily terminates. Herrick therefore begs the question against my default hypothesis that the universe is a brute given, neither having nor needing an explanation. It is no argument against atheism that it fails to explain what it regards as needing no explanation. Of course, Varghese and Herrick try very hard to show that atheists ought to seek an explanation of the universe, but I have argued at length and in detail that these arguments fail.

Pretty much Herrick’s whole case for the existence of his NB is that his hypothesis offers a more complete explanation. But when is explanation complete? Explanation is complete when we reach the ultimate terms of explanation, but those ultimate terms are not ours to choose. They are determined by reality, not the questions we want answered. On naturalism, my default hypothesis, the basic features of the universe are those ultimate terms of explanation. Between my claim and Herrick’s here is no neutral criterion of the completeness of explanation, and he cannot appeal to his to disabuse me of mine. If theism were true, it would explain more, but, of course, the truth of theism cannot be assumed by an argument intended to establish the truth of theism.

But is not a necessary being automatically a more satisfactory endpoint for our explanations than a brute fact? In what sense is God’s existence necessary? Kenneth Pearce succinctly states three senses in which a claim can be necessary:

A logically necessary claim is one whose denial is a contradiction, like “all humans are human.”

A physically necessary claim is one that, according to the laws of nature, must be true.

A metaphysically necessary claim is one that would still have to be true even if the laws of nature were different (Pearce, 2022, p. 26).

It is easy to state, mutatis mutandis, what a necessary being will be in the corresponding senses:

A logically necessary being will be one whose existence cannot be denied without contradiction.

A physically necessary being is one that, according to the laws of nature, must exist.

A metaphysically necessary being is one that would still have to exist even if the laws of nature were different.

Frankly, and I do not think I am being unfair or obtuse, I can find no passage where Herrick unambiguously affirms the claim that God is necessary in any of these senses. As noted previously, when he develops his idea of contingent being (p. 19), he appears to define it in terms of physical dependency. Human life depends upon the availability of oxygen, food, water, tolerable temperatures, etc. If noncontingency is defined in parallel with such physical contingency, then a non-contingent being would be one such that it could not be threatened by any possible physical circumstance realizable in our universe.

In other places, Herrick appears to endorse a metaphysical sense of necessity. On pages 20 and 21, he approvingly considers the notion of divine necessity in terms of the medieval concept of “self-existence.” Self-existence, aseity in the scholastic Latin, means that a being has that property if, in John Hick’s words, “it exists eternally and independently as an uncreated and indestructible entity” (Hick, 2004, p. 76). Hick notes that this was the concept of an ontological or factual necessity, not logical necessity (Hick, 2004, p. 76).

On a later page Herrick says that “philosophical theism” invokes logical necessity:

Philosophical theism traces explanation back and back until it finally grounds its account of the material universe in an explanatorily ultimate basis of existence, an explanatory ground that is itself logically necessary and explanatorily self-subsuming (or self-explanatory) with respect to its own existence And from the beginning of philosophy, the logically necessary (unlike the contingent) has been recognized as a rationally satisfying stopping point for explanatory regress. (p. 31)

Yet on p. 33 he specifically recommends a metaphysical notion of necessity for Varghese to employ against my criticisms and repeats that recommendation on p. 36. If Herrick does endorse metaphysical necessity, he never explicitly spells out its nature. Perhaps the clearest hint is in his eighteenth footnote where he discusses necessity and contingency in modal logic and endorses Alvin Plantinga’s discussion in The Nature of Necessity. I am therefore going to interpret Herrick as endorsing a logical notion of necessity and contingency, where a contingent being is one that exists in some but not all possible worlds, and a necessary being is one that exists in all possible worlds.

Usually, those who defend the idea that God is a logically necessary being support that claim with some version of the ontological argument, but Herrick does not offer or even hint at such an argument. Does God exist in every possible world? It certainly seems that possible worlds can be imagined in which God does not exist. For instance, God would not exist in any possible world containing gratuitous evil. Gratuitous evil is evil that a perfectly good being would prevent, and an all-powerful being could prevent. A world containing gratuitous evil will therefore not contain a being both perfectly good and all-powerful. That is, such a world will not contain God. Consider, for instance, a world in which the wicked go to heaven and the good go to hell. God could not exist in such a world because, being all-powerful and perfectly good, he would not permit such an atrocious situation. Further, if God is defined as existing in all possible worlds, but we find a possible world where God does not exist, then this will show that God so defined cannot exist.

A defender of the claim that God is a logically necessary being should at least address such qualms, but nothing of this sort is found in Herrick’s extended essay. As I say, pretty much his whole case for the alleged superiority of theism is the claim that it explains more. So, really, part 1 of his NB hypothesis is dependent upon his success in part 2, i.e., showing that God’s free choice to create is a legitimate explanation of the existence of the universe. So, we turn to that claim.

Personalistic Explanation

How could God explain the universe? Herrick proposes “personalistic explanation”:

Personalistic explanation (also called “intentionalistic explanation”) occurs when we explain something by saying that a person freely brought it about for a reason. In this type of explanation, the explanans refers to the powers, knowledge, and purposes or reasons of a rational agent, including the power to freely bring something about. Thus we might explain a freshly mown lawn by referring to the homeowner’s power to mow it, his knowledge of how to mow it, his desire that lawn be mowed before company comes over for dinner and so on. (p. 35)

Normally, of course, we attribute intentions to humans and other animals, but Herrick says that they are also rightly ascribed to God:

There is no good reason why the explanatory model of personalistic explanation cannot be applied to the (hypothesized) divine choice to create a world. The capacity part of the explanation will be obvious: According to philosophical theism, God (by hypothesis) possesses unlimited power, knowledge, and love. The reason part the explanation is only slightly less obvious. Philosophical theists propose that the creation of the world was an act of pure love, undertaken to share being (existence or be-ing) with creatures, and an act of love is by its very nature a free act. In short: God freely created the world out of love. Now, love is a pretty good reason to do something, a reason that is intrinsically intelligible apart from all else. And with that, it seems that we have a rationally satisfying explanation for the divine choice, one that is conclusive in the sense that no further explanation is called for, and we reach explanatory finality (with respect to the choice to create). (p. 29)

Can we speak meaningfully of God as choosing, intending, loving, etc., in a sense that is the same as, or closely like, the sense in which humans choose, intend, love, etc.? Isn’t this way of speaking a kind of “folk theology,” employed by unsophisticated believers but rejected by the most serious thinkers, such as Aquinas (Dawes, 2009, p. 46)? Assuming that Herrick could clarify and support such attributions, it is still not evident that we could ever explain occurrences in such terms. Identifying the intentions of a disembodied being is not as straightforward as attributing them to your neighbor cutting his grass. You have seen him mow his lawn on numerous occasions, and so there is no question that he has the power and knowledge to do it. You may even know that he is expecting company and infer that he wants them to be greeted by a neat lawn. You could be wrong, of course. Actions underdetermine intentions. Attributions of intention are often dicey, even with respect to human actions.

The first problem with explanations in terms of the intentions of a putative disembodied, invisible agent would be to identify which occurrences are due to that agent’s actions. Maybe we can imagine instances where we could. Suppose that troubling mysterious events begin to happen in your home such as objects spontaneously flying across the room, heavy items such as bookcases tumbling over with no one around, and your pets reacting as if to a threatening invisible presence. You might conclude that you were being haunted by a poltergeist, but you would still not know the identity of the troublesome spirit or what motivates its mischief. And just where are the public, observable phenomena, analogous to the effects of the poltergeist, that would indicate the activity of a creator? If we saw the dead being raised, or seas parted, or thousands fed with a single Happy Meal, then maybe we would conclude the existence of a god or God. But where are the miracles?

If we admit to acts by disembodied agents, how do we determine which agent caused any particular effect? With physical agents, it is usually straightforward to connect the action with the actor. It is our neighbor Bob who is mowing his grass. However, if the grass in Bob’s yard were suddenly and without physical instrumentality to become perfectly trimmed and neat, to what supernatural agent would we attribute this miracle? God? The Devil? The HOA fairy? How could we know? Given that the universe had a creator, how do we identify that being with the theistic God? Advocates of “intelligent design theory” rightly admit that the creator could have been some other highly advanced and powerful entity rather than the theistic God. (Though one suspects that their admission has more to do with evading the establishment clause of the first amendment than an honest concession.)

What basis could we have for attributing the creation of the universe to an act of pure love? If we cannot be sure of the identity of the creator, then how can we know his (Her? Its?) motivations for creating? However, for the sake of argument, let’s focus on Herrick’s hypothesis that the universe was created by the God of theism, motivated by pure love.

Many years ago, I came across a tract that claimed to present the “Four Spiritual Laws.” The first law was “God loves you and has a wonderful plan for your life.” What ought to convince you that God loves you and has a wonderful plan for your life? God’s plan for some people is that they die in infancy in house fires. In another case—an actual one—God’s plan for a brilliant young graduate student—a devout Christian, by the way—was that he die of melanoma leaving behind a devoted wife and two beautiful young girls. It is vacuous to claim that there is a wonderful plan if the purported plan is indistinguishable from no plan or a bad one.

We may generalize: A world containing pediatric cancer, flesh-eating bacteria, brain-eating amoebae, the rabies virus, fawns burned in forest fires, mental illnesses such as schizophrenia, birth defects, sickle cell anemia, EF-5 tornadoes, and such human specimens as Adolf Hitler, Josef Stalin, and Donald Trump does not appear, prima facie, to have been a creation of pure love, especially if the loving agent was a being of unlimited power and perfect foresight. If an intelligent but inexperienced being—perhaps one behind Rawls’ veil of ignorance—were informed before experiencing the world that it is the creation of an all-powerful being motivated by pure love, would that being have expected anything like the actual world? Put another way, a good explanation is one which, if it were true, the explanandum would follow as a matter of course (Dawes, 2009, p. 21). I do not think that any honest (or maybe even sane) person could say that if the world were the creation of omnipotence and pure love that it would follow as a matter of course that the world would resemble our vale of tears. The moral and logical contortions of theodicists over the centuries show that it does not obviously follow but requires herculean labors of attempted justification.

Of course, the world contains many wonderful things as well, but, as David Hume cautioned long ago, we cannot infer a pure cause from mixed phenomena. What does “love” even mean if it allegedly created—freely and with perfect foreknowledge—a world in which innocent creatures are tortured to death? Again, it is vacuous to claim that the world is a product of love when such “love” is indistinguishable from indifference. Is Herrick’s hypothesis testable? Does it have any empirical content? What does it rule out? What would have to happen in the world to count against the hypothesis that it was the creation of pure love? Herrick gives no indication how he would answer these questions.

Moreover, even if love could explain God’s motivation for creating a world, it cannot explain his whole motive for creating this one. God, being free and omnipotent could have created indefinitely many other worlds upon which to lavish his love, but for some reason he chose this one. Herrick assures us that God’s choice to create this world was rational. Therefore, his choice to create this world could not have been arbitrary. Some other factor or factors must have motivated (without determining) his choice to make this specific one. Unfortunately, we do not know what those factors might have been. So, Herrick still leaves us largely in the dark with respect to God’s putative intentions, and so his NB hypothesis remains nebulous.

With humans, intentional explanations in terms of love are often sufficient. Herrick gives the example of an artist who expresses his love for his wife by painting a beautiful picture of a rose, his wife’s favorite flower. Surely, this explains the artist’s actions. If such explanations are good enough for understanding the motives of the artist, why not of God? With human agents, like the artist, we have a very large set of background beliefs and expectations that inform the explanation in terms of love. Why did the artist express his love with a painting rather than, say, buying her a diamond ring? Because his own artistic creation would be a far deeper and richer expression of love than an expensive purchase. With God, however, we are much more ignorant of crucial information. We have no idea of the kinds of alternate worlds omnipotence might have chosen to create or what kinds of opportunities they would offer their creator for the expression of love. There is no analogy between Herrick’s artist and God.

There are deeper problems with explaining the creation as God’s free act motivated by pure love. Philosophers Patrick Francken and Heimir Geirsson note that if every element of the explanans of an explanandum is necessary, then the explanandum must be necessary also because only necessary truths can follow from necessary premises (Francken and Geirsson, 1999, p. 287). This is not controversial; it is an accepted modal principle that necessary truths can only entail necessary truths. However, arguments such as Varghese’s and Herrick’s are attempts to explain a contingent universe in terms of a necessary being. This is impossible if the explanans of the universe cites only necessary truths. In this case, as Spinoza argued in the 17th century, the universe must be necessary also (Francken and Geirsson, 1999, p. 287).

What, then, are we to make of God’s alleged choice to create the universe? Was that act necessary or contingent? In the former case, the universe would also have to be necessary since its existence is entailed by necessary premises. In the latter case, if the choice is contingent, it must lack a sufficient cause, and thus will seemingly be unexplained, i.e., a brute fact. How, then could the theist claim any advantage over atheism? Why would it be better to end our chain of explanations with an unexplained choice rather than an unexplained universe? Herrick recognizes the cogency of this argument (p. 25). He notes that if God’s choice was just a brute fact, then theism and atheism both end with brute facts, and Ockham’s Razor would shave off theism:

In this case, then, Parsons is certainly right: theism is a logically superfluous hypothesis, an unneeded explanation. God is just an unnecessary explanatory epicycle, a spinning wheel doing no explanatory work, an unnecessary cog in the cosmic clock work—indeed, as Parsons argues, “a Creator with nothing to do.” (p. 26)

Herrick’s answer is that God’s choice is contingent and not necessary. God’s creative act was free, and therefore it had no logically sufficient cause (p. 28). Yet this does not mean that the choice was unexplained. Free choices are explained in terms of their motivations and reasons; it is these factors that make the choices intelligible, and not merely brute (p. 29). God’s act was free, but was motivated by pure love, and this is what makes God’s creative act understandable. Unlike the atheist’s brute universe, God’s creative act is quite intelligible, and this gives the explanatory edge to theism over atheism.

Francken and Geirsson reply that if God’s choice was contingent, that is, if God’s motivations and reasons did not determine his choice, then it must have been a real possibility that he could have freely created some other world or none at all:

If it is contingent that God created this world for his reasons—if it is contingent, in other words, that God’s reasons resulted in God’s choice to create the universe—then it is possible for God, on the basis of those same reasons, to have created some other world or perhaps no world at all…. It is a real possibility, that is, that some other choice, and hence some other act of creation and some other world, should have been the outcome of God’s reasons. (Otherwise those reasons will have been sufficient and the choice will not have been contingent or “free” in the required way.) (Francken and Geirsson, 1999, p. 291)

So, even if we allow Herrick to claim that God’s free choices are explained in terms of his motives and reasons, we still have no explanation why God created this world and not some other.

In an earlier paper written in response to Richard Carrier, Herrick offers a reply to this line of argument. He appeals to our ordinary, everyday explanatory practices, as when we explain why Ann and Bob got married:

In the end, when all is said and done, that Ann and Bob ultimately chose to marry, that their choice was fully informed and morally autonomous, and that they married out of love, ends the regress of questions by grounding it in an intellectually satisfying and reasonable stopping point. Granted, it is a contingent fact that they chose to marry, but it is not, on that account, a brute fact. A brute contingent fact has no explanation, whereas a free, morally autonomous act of the type under consideration does have an explanation, albeit one that is not deterministic. Love, as a motive, can make an action intelligible without making it metaphysically necessary. The more general point is that reason-explanations that cite motives are sometimes perfectly acceptable stoppers of explanatory regresses. I am appealing here to our everyday explanatory practices, which, I would argue, should guide us unless and until they are shown to be based in error or confusion. (Herrick, 2006)

Since Ann and Bob’s decision to marry is a contingent fact, it is indeed conceivable that they might have chosen to marry someone else or not to marry at all. However, we do not hesitate to accept the explanation that Ann and Bob married out of love as an explanation for this marriage. Why, then, would such an explanation not suffice for God’s choice to create this universe? However, the comparison is not apt. Ann and Bob love each other, not anybody else, and that is why they married each other. Prior to the creation, God was not in a loving relationship with any specific universe. As noted previously, God could have chosen any of indefinitely many possible universes to have made the object of his love. So, we are still left with a blank if we want to know why God created this universe.

Actually, in the end, it is hard to see God’s choice to create as anything other than a brute fact. If God’s choices are free—truly free in the libertarian sense—then God must have had the power to act against his motivations. In that case, God’s choice to act in accordance with his motivations cannot be explained by those motivations. On the libertarian view, the moment of choice is a go/no go node where the outcome is determined purely by the agent’s power to choose, a power that supersedes all motivations, desires, purposes, etc.—as Herrick admits (p. 28). Hence, contrary to Herrick’s claim, invoking God’s love does not bring us to finality of explanation. We would still need to know why he chose to act out of love instead of against it, yet the libertarian notion of freedom seems to preclude ever having such information. Indeed, the choice would appear to be a brute fact, and in that case, we must ask why a brutally factual divine act is a better terminus for our explanatory quest than a brutally factual universe.

Finally, in ordinary cases of intentionalistic explanation the instrumentality of the accomplishment of those intentions is well understood. If we know that our neighbor intended to cut his grass, we also know that he used a lawnmower. If someone inquiries about how lawnmowers work, we could pursue explanations all the way down to basic physics if we wanted. With God’s creative acts, there is no instrumentality. God’s acts are basic actions. Yet even with basic actions we can distinguish between the will to do something and how it was done, the powers wielded in performing that action. Raising my arm is a basic action, but very complex neurological, anatomical, and physiological information is needed to explain how I raise my arm. By contrast, the powers whereby God’s will is realized are totally and in principle occult, like magic or psychokinesis. Unlike natural powers and energies, God’s creative powers are supernatural and beyond our understanding. God supposedly has the power to create and rearrange matter as he wills. What is the nature of this power? We can never know.

Explanations in terms of God’s direct acts (and not in terms of “secondary” causes) therefore introduce a large measure of ineradicable inexplicability into our understanding. As in the first chapter of the Book of Genesis, God says, “let there be,” and there it is! End of story. That is why Darwin’s account was such a boon for human understanding.

Theists often reply to such criticisms with a tu quoque: Scientific explanations end with bare attributions of causal powers that have no deeper explanation, supposedly just like theistic explanations. Fundamental particles like electrons just have a certain rest mass, charge, and spin with no deeper account of why they do, just as certain basic powers are attributed to God. Three points: (1) We know what mass, charge, and spin are, but have no idea of the nature of God’s powers. (2) The properties of electrons were ascertained by rigorous experiments. The powers attributed to God are dictated by a theological imperative (p. 3). The comparison is not apt. Scientific explanations take us very deep indeed before hitting explanatory “rock bottom.” For instance, we have a very detailed account that tells us just how and why thermonuclear devices explode. It was this knowledge that allowed Stanislaw Ulam and Edward Teller to design the two-stage thermonuclear bomb. With God’s actions, however, we hit a wall right away. Saying “God did it” precludes any genuine understanding, and this, by the way, is why such explanations have had such an obscurantist history.

Another way to attempt to avoid this objection would be to define God’s causation in terms that make no mention of his powers. Perhaps a counterfactual explication could be offered: “God causes e” means that had God not willed e, e would not have occurred. However, since anything God wills must occur, we can equally say that had e not occurred, then God would not have willed it, and this creates a spurious dependence of God’s act of willing upon the occurrence of e (Dawes, 2009, p. 54). It appears, then, hard to explicate God’s causal efficacy without mentioning his creative powers, which are in principle mysterious.

Saying “God did it” is like the notation “Here be dragons” on old maps; it is a marker for our ignorance. Maybe, then, as John Hick suggests, it is reasonable to rest with a brute universe than to posit invisible mysteries. (Hick, 2004, p. 111) Better an unexplained universe than an “explanation” in terms of an invisible supernatural being who wields occult powers to create a universe for purposes that we may not understand and cannot confirm.

The Demand for Complete Explanation

Herrick repeatedly claims that theism explains more than naturalism, and, ipso facto, is superior. We can now see that this claim to more complete knowledge has rhetorical bark but no logical bite. Naturalism claims that, with respect to ultimate explanations, less is more: The universe is all that there is, and our knowledge is complete when our scientific knowledge of the universe is complete. There is no problem with regarding something as a brute reality if that is what it is. The fact that theists can ask for more does not mean that there is more, and the fact that they offer an explanation does not mean that it is a good one. The point at which knowledge is complete is determined by reality, not by the questions we ask, and naturalists think that they acknowledge reality, and that theism is a fantasy. Herrick has offered no reason for them to think otherwise.

Why do theists so insistently ask for more? I think that John Hick has the answer:

As consciousnesses, we can rest in the idea of an ultimate consciousness as the source of the existence and character of the universe, whereas the thought that the universe itself is ultimate leaves us unsatisfied: we cannot help wondering why it exists and why it exhibits the particular basic regularities in virtue of which it is ordered. Thus the idea of a creative divine mind possesses to our human minds greater intrinsic intelligibility than that of a realm of purely material forces and movements. (Hick, 2004, p. 80)

Naturalists will reply: “Speak for yourself. We are quite satisfied with a ‘realm of purely material forces and movements.’ Ultimate brute facts do not dismay us. You have not shown that we are wrong. The universe is under no obligation to respect your plaintive “whys.” Further, those “whys” appear to arise from questions that we see no reason to ask, such as why our universe exists rather than none, or why our universe exists rather than some other we can imagine. If you do not ask pointless questions, you will not expect unneeded answers.”

Conclusion

Herrick’s essay covered much ground at considerable length, so this reply has had to cover much ground also and at a comparable length. However, the conclusion may be very briefly stated. Herrick had two aims, to critique my essay “No Creator Need Apply” and to defend his own theistic thesis. Both efforts fail. My original critique of Varghese remains unscathed. His theistic thesis is nebulous in content, weakly argued, and his personalistic explanation model is far too rife with problems to serve as an adequate explanatory scheme. My naturalistic default hypothesis remains untouched.

References

Blackburn, Simon. (2005). Dictionary of Philosophy. 2nd ed. Oxford, UK: Oxford University Press.

Dawes, Gregory W. (2009). Theism and Explanation. New York, NY: Routledge.

Dessler, Andrew (2022). Introduction to Modern Climate Change. 3rd ed. Cambridge, UK: Cambridge University Press.

Everitt, Nicholas. (2004). The Non-Existence of God. London, UK: Routledge.

Francken, Patrick and Heimir Geirsson. (1999). “Regresses, Sufficient Reasons, and Cosmological Arguments.” Journal of Philosophical Research, Vol. 24: 285-304.

Hawking, Stephen. (1988). A Brief History of Time: From the Big Bang to Black Holes. London, UK: Bantam Press.

Henle, James M., Garfield, Jay L., and Thomas Tymoczko (2011). Sweet Reason: A Field Guide to Modern Logic. 2nd ed. Malden, MA: Wiley-Blackwell.

Hepburn, Ronald W. (1967). “Cosmological Argument for the Existence of God.” In The Encyclopedia of Philosophy, Vol. 2 (pp. 232-237) ed. Paul Edwards. New York, NY: Macmillan.

Herrick, Paul. (2006, June 14). “Contra Carrier: Why Theism is Needed to Make Sense of Everything.” The Secular Web. <https://infidels.org/library/modern/paul-herrick-contra/>.

Herrick, Paul. (2009, May 2). “Job Opening: Creator of the Universe—A Reply to Keith Parsons.” The Secular Web. <https://infidels.org/library/modern/paul-herrick-parsons/>.

Hick, John. (2004). An Interpretation of Religion: Human Responses to the Transcendent. 2nd ed. New Haven, CT: Yale University Press.

Hook, Sidney. (1991). The Quest for Being. Amherst, NY: Prometheus Books.

Mackie, J. L. (1982). The Miracle of Theism: Arguments for and Against the Existence of God. Oxford, UK: Clarendon Press.

Margenau, Henry and Roy A. Varghese. (1992). Cosmos, Bios, Theos: Scientists Reflect on Science, God, and the Origins of the Universe, Life, and Homo Sapiens. Chicago, IL: Open Court.

Oreskes, Naomi. (2019). Why Trust Science? Princeton, NJ: Princeton University Press.

Parsons, Keith. (2006, August 8). “No Creator Need Apply: A Reply to Roy Abraham Varghese.” The Secular Web. <https://infidels.org/library/modern/keith-parsons-varghese/>

Pearce, Kenny. (2022). “Classical Theism: An Exposition and a Defense.” In Is There a God? A Debate (pp. 11-81) ed. Graham Oppy and Kenneth L. Pearce. New York, NY: Routledge.

Plantinga, Alvin. (2000). Warranted Christian Belief. New York, NY: Oxford University Press.

Romm, Joseph. (2016). Climate Change: What Everyone Needs to Know. Oxford, UK: Oxford University Press.

Salmon, Wesley. (1990). Four Decades of Scientific Explanation. Minneapolis, MN: University of Minnesota Press.

Searle, John R. (1999). Mind, Language, and Society. New York, NY: Basic Books.

Cookie	Duration	Description
cookielawinfo-checkbox-analytics	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".
cookielawinfo-checkbox-functional	11 months	The cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".
cookielawinfo-checkbox-necessary	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".
cookielawinfo-checkbox-others	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.
cookielawinfo-checkbox-performance	11 months	This cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".
viewed_cookie_policy	11 months	The cookie is set by the GDPR Cookie Consent plugin and is used to store whether or not user has consented to the use of cookies. It does not store any personal data.