Phil Halper, a science communicator and commentator who debated Stephen Meyer and got a number of things wrong in doing so, has released another video, this time attacking cosmologist Brian Keating over Keating’s commentary on two surveys that Halper co-authored with physicist Niayesh Afshordi. The first survey was a poll of 85 specialists at the 2024 Black Holes Inside and Out conference in Copenhagen, and the second was the much larger Big Mysteries Survey conducted through the American Physical Society’s Physics Magazine (see also here). Readers may recognize Keating as the distinguished experimental cosmologist at UC San Diego who was a generous and open-minded participant in the documentary The Story of Everything. Keating needs no help in the fight that Halper has picked with him over string theory’s standing, sigma levels, and survey percentages. In my estimation, Halper’s hostility in the video considerably outruns its substance.
What does merit a response is the use to which Halper puts these surveys on a different question. He contends that the results are “bad news” for Meyer, whose case, he says, “depends on fine-tuning demanding a metaphysical explanation.” This ill-considered gibe fails at every step.
Seeing Why Is Instructive
Halper rests his spurious claim on three findings: (1) among the survey respondents, the most frequently offered “explanation” of fine-tuning was that the constants of nature are brute facts that do not need an explanation at all — only 9 percent of the APS sample and 3 percent of the Copenhagen specialists took fine-tuning as evidence for God; (2) even when support for a multiverse as an explanation of fine-tuning was added to those judging it to have theistic import, the combined total was still less than a third of those surveyed; and (3) the strongest consensus in the surveys, at 68 percent, was that the Big Bang only gives evidence for a hot dense early state of our universe while saying nothing either way about a beginning of time.
First, let’s consider the form that Halper’s argument takes, because the terrain on which he has chosen to build his fort is pure quicksand. The very nature of the surveys to which Halper appeals is that they measure what physicists believe. They do not and cannot measure what the evidence actually supports. The fine-tuning argument, developed with care, is a claim about comparative likelihoods. If FT = fine-tuning, T = theism, N = naturalism, and B = our background knowledge, the ratio of the likelihoods P(FT|T·B) over P(FT|N·B) measures the extent to which fine-tuning confirms theism over naturalism, and the numbers show unequivocally that a life-permitting universe is far more probable given theism than it is given naturalism. The truth of this judgment depends on physics and probability theory, not on a show of hands.
The Pool of Voters
Beyond this, the pool of voters is not well-suited to answer the question. Most working physicists have no reason to study confirmation theory and their extemporaneous opinions on the significance of fine-tuning are about as useful as the average philosopher’s opinion on what a scattering amplitude shows. Afshordi himself made this point directly when the results were initially released, commenting that “Scientific truth is not decided by a vote… [but] lack of consensus can be a clue. It marks places where better data, sharper theory, or new connections between subfields may be needed.” An ironic observation, sociologically telling, comes from watching such surveys being used two ways — as sociologically descriptive when the survey data is released, then as epistemically probative when aimed at Meyer or intelligent design. They cannot be both, and the second use is simply a mistake. Argumentum ad populum is a fallacy independent of the perpetrator, and it would have been equally inappropriate to claim explanatory superiority for design had the majority of respondents declared in its favor.
Halper makes a big deal out of the popularity of the “brute fact” response holding that the constants just happen to be what they are and their values have no need of explanation. He casts this as a tertium quid that obviates any need to decide between theism and the multiverse. It is, of course, nothing of the kind. It is rather a bald denial of explanatory demand and, as such, is no refutation of any substantive explanation. Where the design inference is concerned, it leaves the likelihood ratio exactly where it was and simply refuses to compute it or engage with the issue. Aside from its complacency, denying the need for an explanation carries an epistemic price unacknowledged by many who obliviously pay it.
Let Me Elaborate
The assertion of brute factuality requires abandonment of the principle of sufficient reason (PSR), at least in the restricted form requiring contingent states of affairs to have explanations. Denying that contingent states of affairs have explanations comes at a steep price. If it is possible for something to be, or to happen, for no reason at all, then this possibility becomes a competitor for every explanation we think we have for why something that could be otherwise is the case, and what is more, the probability of that state of affairs obtaining, in the absence of any reason for it, is completely inscrutable. We cannot even say it is unlikely that a given state of affairs has no explanation. And, of course, this applies as much to our current state of conscious awareness as anything else. If the restricted PSR is false, then we cannot say that it is improbable that our current perception of our environment and our occurrent beliefs are what they are for no reason at all. Arguably, then, denying the restricted PSR is epistemically self-defeating, in which case its affirmation is epistemically necessary (see also here and here). Stated boldly, physicists who think that the constants of nature are brute facts without need of explanation simply have an incoherent opinion.
On a more pedestrian level, the enterprise of physics itself runs on the engine of explanation. If the physics community were to permit declarations of brute-factuality whenever certain of their members felt metaphysically uncomfortable, then abandonment of many lines of research would become acceptable. The whole community could disavow decades of theoretical labor over why the early universe has such low entropy or why the cosmological constant is so fine-tuned, and the door would be left open to embracing unexplained occurrences anywhere in nature. This would undermine the very practice of science in which the respondents of the survey are supposed to be engaged. The multiverse literature itself is a monument to the community’s conviction that fine-tuning demands explanation. One does not simply build an infinite ensemble of unobservable universes to explain what needs no explaining! Read through the eyes of sociological explanation, then, the startling preference for abstention over multiverse or theistic explanation says less about the actual evidence than it does about the manifest discomfort many physicists have with its implications, and with the paucity of metaphysical naturalism’s explanatory resources in relation to what needs to be explained.
In the Beginning
This brings us, at long last, to the beginning of the universe. No fewer than 68 percent of respondents agreed that the Big Bang describes the universe’s evolution from a hot dense state rather than implying a beginning of time, a result that Halper vainly tries to wield against cosmological arguments in general. It is difficult to suppress a yawn. I agree with the 68 percent, and so does Meyer. None of us think that the hot Big Bang model, taken by itself, says anything about t = 0. The more interesting questions gain their foothold elsewhere. For example, under what conditions does the Borde-Guth-Vilenkin (BGV) theorem show spacetime to be past-incomplete, and are these conditions, in fact, satisfied? What would their satisfaction or failure signify about cosmological origins? These questions are the focus of ongoing research by many, and central to one of my current research projects evaluating the explanatory adequacy of different families of cosmological models. It is, however, worth noting that the surveys asked their respondents nothing about geodesic incompleteness, and a poll that never poses the question can hardly return a verdict on the theorem that answers it. The bottom line is that moving from “the Big Bang model does not itself assert a beginning” to “physics gives no reason to think the universe began” changes the subject between premise and conclusion, and no percentage, however large, could license that inference.
What do the surveys actually show, then? No more and no less than a discipline whose foundations are contested, whose standard cosmological model is under genuine strain, and for which no rival theoretical framework has commanded anything close to a consensus.
We know this picture. It is not only the one to which Afshordi alluded earlier; it is what our ongoing examination of cosmological origins reveals every day: model choice at the frontier of science is underdetermined by the data and thoroughly shaped by philosophical commitments, whether they are acknowledged or not. Read rightly, the Big Mysteries Survey is not “bad news” for intelligent design research at all. What it constitutes is a candid admission, from 1,600 respondents, that the deepest questions remain open. So let’s keep pressing those questions, and follow the evidence where it leads.