AP News recently ran a story on the interface of religious inspiration and the search for dark matter. I thought about commenting on it, then thought better of it, but now here I am again. I know some but not all of the people who are quoted, all good scientists. The article ends with a nice quote from Jennifer Wiseman:
“Studying the deep universe may make us feel insignificant,” Wiseman said. “But it also gives us a sense of unity that we’re all on the same planet. … The hope is we get a sense of joy, humility and love from these contemplations.”
I’ve known Dr. Wiseman since undergrad days, before either of us were Ph.Ds. We don’t agree on much that is specific to religion, but somehow manage to be friends anyway, and I completely agree with the sentiments she expresses here.
There is a tendency to portray religion as being at odds with science, and that can certainly happen. But they share more in common than this trope implies. Both arise from a deep wellspring in the human spirit, the same wellspring: the desire to know. How does the universe work? How did it come to be so? Why? and on and on. Where they differ is in approach: when encountering an unknown, especially things that are fundamentally unknowable (e.g., does God exist?), religion asserts an answer and asks that we accept it on faith. Faith is anathema to the scientific process; one is, in principle, to search for the truth through observation and experimentation.
Religion and science come into conflict when scientific knowledge encroaches on religion’s known unknowns. There was a time when the structure of the heavens must have seemed unknowable; the realm of the sacred, safe from access by mundane human knowledge. So it is easy to see how Galileo came into conflict with the Church despite being a faithful member:
Church: Scripture teaches us that the Earth is the unmoving center of mortal corruption; the heavens above us are perfect and unchanging.
Galileo: I can see spots on the sun and mountains on the moon.
Church: We are the center about which the heavens rotate every day, clearly the center of all rotation.
Galileo: I see satellites circling Jupiter.
Church: God is great; He could do one thing one day and an entirely different thing another day*.
Galileo: It appears that the Earth revolves around the Sun, never the other way around.
Church: Do you see these instruments of torture?
This is a tongue-in-cheek portrayal of a serious historical incident, but my point is that conflict arises when scientific knowledge encroaches onto turf that had formerly been the exclusive province of religion. Addressing our spiritual need to know can be inspirational, but it can also be a let down. Giving specific answers to formerly unknowable questions can be a bit of a mood killer. Is that all there is?
Another downside of scientific knowledge is that it can never be complete. There is always something left unknown, and we are not supposed to fill that void with something we take on faith. Yet we are profoundly uncomfortable with not knowing. A little extrapolation at and beyond the fringes of current knowledge is natural, and can sometimes provides a useful way forward in driving new discoveries. But it can get carried away, e.g., string theory. So extraordinary caution is also warranted: the temptation to fill in the blank is where empiricism transitions to theology.
Understanding that we know a lot – as we do at this juncture in the history of science – and yet still don’t know something important, like what most of the universe is made of, is hard to accept. It becomes even harder when admitting something we thought we understood is wrong, or is less complete than we thought. We know most of the mass in the universe is made of non-baryonic cold dark matter&. Don’t we?
The dynamical evidence for acceleration discrepancies is abundant. That these must be caused by non-baryonic cold dark matter is less clear. This is where cosmology intrudes. Cosmology has always been the nexus where science and religion meet, with philosophical imperatives often obscuring essential observational facts. Before Kepler, orbits had to be circular. After Inflation, the density parameter had to be one. (We meant Ωm = 1, not the modern weak-sauce version Ωm + ΩΛ = 1.) The mass density is larger than the baryon density from BBN (Ωm > Ωb), so there has to be non-baryonic dark matter. That such a substance is also required to fit the acoustic power spectrum of the CMB amplifies our faith in the existence of such stuff.
We think we’ve solved cosmology, and that solution requires non-baryonic cold dark matter. So to admit that maybe we were wrong about dark matter just because it persistently remains undetected and provides unsatisfactory explanations of many astronomical observations and is consistently outperformed in predictive capacity by an alternative theory is to admit that we don’t understand as much about the universe as we thought. That’s really, really hard. Our unwillingness to admit that maybe we have been wrong about such an important issue is where human nature kicks in to blur the line between scientific knowledge and religious faith. It’s much easier to ignore those nagging doubts% and have faith that we were right all along.
Cosmology works so well that dark matter has to exist. I’ve head this sentiment expressed over and over by many different scientists, yet this is an assertion of faith. A more conservative statement is that cosmology as we currently conceive it works if, and only if, an appropriate form of non-baryonic dark matter exists with the required cosmic density. If not, then we need a new model – something that presumably^ stems from a more general underlying theory.
Since both religion and science arise from the same desire to know, it is easy to have faith that we know more than we actually do.
This is the sticking point we’ve hit in the dark matter debate. We’ve been calling the acceleration discrepancy the dark matter problem for so long that this linguistic mistake has morphed into an absolute certainty that invisible mass exists. It has become a matter of faith.
*Of course an omnipotent deity could do that. Apparently He chooses not to do so, sparking the schism between theists, for whom God actively intervenes in miraculous ways in real world affairs, and deists, who view God more as the great watchmaker, setting creation in motion but not interfering with its operation. Deism is conducive to science, as the search to identify the rules by which the universe works is to gain insight – however remote – into the mind of God. I suspect this attitude informed Einstein’s complaint against quantum mechanics: “God does not play dice with the universe.”
&I certainly thought so before I didn’t.
%The social pressure to conform to the preferred cosmology is enormous. I know I’m only making it worse for myself. But an explanation that omits MOND is a lie of omission. Here science and religion certainly overlap in the sentiments expressed by the seventeenth century cleric Paul Gerhardt:
“When a man lies, he murders some part of the world.”
Assertions that somehow “feedback” explains MOND are simply a numerical form of magical thinking; an excuse to not have to explain the inexplicable. By eliding MOND, we murder a part of the world.
^Here I am extrapolating at the fringes of knowledge.
Strong point about how easily inference turns into belief. The distinction between acceleration discrepancies and dark matter is especially important, since the observation itself does not uniquely select the explanation. The same tension shows up in cosmology where model consistency is often treated as confirmation. MOND’s predictive success at galaxy scales also highlights that alternative frameworks cannot just be dismissed on sociological grounds. At minimum it forces the question of whether we are fitting a paradigm rather than testing it.
Thanks for the great article.
“scientific knowledge … can never be complete. There is always something left unknown.”
Exactly. But I think it’s a lot more than that. The more we know, the more we discover that there is to be known. Every new piece of knowledge we gain combines with every other piece of knowledge to increase the apparent dimensionality, the parameter space of reality in a hyper-exponential manner.
Biological life explores this space all the time, and becomes more and more incredibly, incomprehensibly diverse, but visits only a minute fraction of it because it is so vast.
Human minds are far more limited.
In other words, we’re really returning, it seems, to the state you describe in this:
“There was a time when the structure of the heavens must have seemed unknowable; the realm of the sacred, safe from access by mundane human knowledge.”
It’s not “sacred” in any way, but it is unknowable because it’s just too complex. There is too much of it. It doesn’t reduce to a simple set of laws. We find that we amplify the complexity of reality by exploring it.
So perhaps it really is unknowable, and we should temper our expectations as we continue to explore. We can continue to discover and explain things, but what we are explaining becomes a smaller part of the new whole yet to be explained.
I encourage everyone to go read Peter Woit’s comments on string theory linked above: https://www.math.columbia.edu/~woit/wordpress/?p=15647 It was an offhand comment for me, but as it happens there is a very timely example there of what happens when science turns into faith.
Strong point with the Woit piece. Once a framework becomes the only accepted reading of the data, inference starts to feel like necessity. Similar issue with expansion. It is often treated as proof when in reality it is an integration. I myself am not a fan of approaches that lean on handwaving over logic.
There is an irony worth naming here. MOND proponents correctly diagnose that ΛCDM has been pushed past its validity range, but the proposed remedy accepts the same foundational assumption: that there exists a single compact gravitational description adequate across all scales and complexity regimes. The dispute is over which law occupies that role, not whether such a role should be assumed to exist. That is naive reductionism one level up.
The more fundamental point is that scientific theories are compression schemes. This is not a philosophical gloss, it has precise technical content through algorithmic information theory. AIT results establish that the compact effective description of a low-complexity system generically cannot be the optimal description of a high-complexity one. Extrapolating it will produce apparent anomalies, tensions.
The “missing mass” problem looks different through this lens: not a signal that mass is missing from the ontology, but a signal that the compression scheme is failing, that we are operating outside the complexity regime in which its parameters were calibrated.
G was measured in effectively two-body systems. MOND’s a₀ was fitted from galactic dynamics. Both are regime-contextual parameters fingerprinting the complexity scale of their measurement context. Neither should be expected to govern dynamics cleanly outside that context, and the search for the one true gravitational law that unifies them may be the wrong research program entirely.
The regularities McGaugh has documented empirically are real and important. But the theoretical interpretation placed on them remains inside the paradigm being critiqued.
Expecting gravity to have a “clean” compact description at every scale is a vestige of 19th-century clockwork determinism.
Lovely post, as always!
A few observations on the history and philosophy of science and religion, as they may relate:
.
“God is great; He could do one thing one day and an entirely different thing another day.”
Theists do generally believe in an ordered cosmos. The question about miracles is generally framed as pertaining to ~rare exceptions to established expectations (although these expectations don’t always correspond obviously to natural laws as such — there’s not a law per se that no one can walk on water, even if doing so might prove noteworthy).
In a sense, the contention between deists and theists is about marginal, non-systematic phenomena. Cosmology, and science generally (necessarily at least a little more systematic), don’t always come into it.
.
I might suggest that there’s no such thing as an identifiable “religious” epistemic. Early Christian texts for instance attempt to present (features at least of) the religion as grounded in empirical observation, which, true or false, is a different frame of mind than the “take it on faith” stuff you encounter in many Christian traditions. Along a different axis of the spectrum, you have Buddhists for instance who try not to believe in anything (doors, dogs, dads), let alone in miracles.
However, I do think the analogy you offer remains useful, whatever terms one attaches to it.
.
In this vein, religious epistemics in the West underwent a number of changes leading up to and along with the scientific revolution. Ironically, ancient and canonical Jewish religion is famously unengaged in what we might call speculative philosophy, including as it pertains to nature. Jewish cosmology (such as it is) appears in canonical texts fleetingly when at all, generally in service to some other point, and ambiguously related to other appearances of Jewish cosmology. “God conquered the primordial beast Rahab (Oh let us praise him for establishing order in the cosmos etc etc.).”
The Ptolemeic model came to have the importance it did for Christians in late antiquity as Christian scholars attempted to synthesize their religion with Aristotelian philosophy, which was extremely well-regarded at the time, and remained so until the scientific revolution.
The scientific revolution is mirrored in a similar, though less clearly successful, revolution in religious epistemics at that time, which arose from similar conditions. The arrival of the printing press, better access to ancient manuscripts and foreign interlocutors, the rise of the academy, mass literacy, the refinement of text critical techniques. If, during this time, natural philosophy went from largely doubtful and speculative to basically reliable and conservative, religious philosophy at least got very good at making legible the texts and councils it treated as authoritative and developing interpretive apparatuses and research programs for engaging with those things. It had a harder time demonstrating which, if any, of those things *should* be considered authoritative, when it even tried, but in many other ways the revolution was transformative, politically and epistemically. Modern biblical studies is a different and larger beast than it was when Galileo had his trials. (Incidentally, I’m not certain many modern academic scholars would agree with the Catholic Church of his time that the Christian scriptures meaningfully suggest a geocentric view.)
.
Galileo’s story I find weirdly and annoyingly complicated. It’s unfortunate that the thing he got imprisoned for was teaching heliocentrism, when the better grounds for broad epistemic lessons lie in his earlier telescope work — there were well-respected astronomers who refused, point-blank, to even look through his device! As I understand it, his advocacy of heliocentrism is complicated by the (at the time) unresolved parallax issue. I’ve encountered claims, which I haven’t looked into, that the authorities persecuted him not for exploring it per se but for teaching it as truth when the matter hadn’t been settled, whether scientifically or ecclesiastically (there was some back and forth between the two modes of thought, iirc the hierarchy at the time wasn’t universally and unambiguously opposed to heliocentrism, though obviously it was later). In any case, I don’t think they had any business persecuting him, and he remains a personal hero of mine.
.
Are you familiar with William Harvey, who discovered the circulatory system? Great parallels there too. “I’d rather be wrong with Galen than right with Harvey.” His seminal publication, De Muto Cordis, is available in translation online and is a short and fascinating read for anyone who’s interested. Another personal hero of mine. (Also, I was surprised to learn the circulatory system *needed* to be discovered in the first place, but it really stumped people for a long time! Some great parallels there for current debates over the acceleration discrepancy, imo. Harvey was above all a persistent and careful observer.)
It’s pretty easy to make the case that Lambda-CDM as a whole is a faith based model, given that none of its defining elements are observed phenomena. Modern cosmologists are not allowed to question the standard model. they can only accept it on faith or suffer excommunication from what can be reasonably called the cult of modern cosmology, an academy-based community of belief that is completely untethered from the basic scientific principle of empiricism.
Hi Stacy.
Thank you for this post on “The well of inspiration”.
I think the church gets a lot of bad press. For centuries the church was the main provider of education (teaching and learning). Education has been taken away from the church and is now carried out by the state. And for centuries the church was the main provider of health (medicine and healing). Health has been taken away from the church and is now carried out by the state. It could be argued that the church/religion did a better job than the state. And nowadays it seems that populations are expected to blindly accept the outpourings of the state rather than those of religion. We are lucky that most research institutions (universities) are separate from both state and religion.
Separately, with the Artemis II space capsule travelling around the Moon, I am minded to reflect on how unfortunate it is that the Moon’s rotation is locked to the Earth and that we see only one face. If the Moon’s rotation were different, then early people on Earth would have seen different features on the Moon and concluded that the Moon was a ball in space. The whole of astronomy and cosmology would have followed a very different path.
“He could do one thing one day and an entirely different thing another day* – a good description of President 47.
Good point about the tidal locking of the moon’s rotation. Similarly, if our eyes were a little better, or Venus were a bit bigger, we could see it go through phases and a geocentric cosmology would never have been viable.
“Religion and science come into conflict when scientist knowledge encroaches on religion’s known unknowns.” Think about 2 ideas: The Prophet X is the greatest, because the Sacred Book Y tells the greatest truths. The scientist X is the greatest, because empirical evidence A, B, C supports X’s research. In the development of plate tectonics, there was a conflict between sunken land bridges versus continental drift — in some respects, the controversy resembled many religious controversies — there were belief systems with fervent disciples. Now, we have dark matter particles versus MOND — we need FUNDAMOND so that Professor Milgrom can win. There is always some doubt about the correct experts versus the incorrect experts. One might divide knowledge into 4 categories: (1) known knowns (by almost everyone); (2) unknown knowns (known to a few experts but nobody else); (3) known unknowns (ignorance agreed upon by the experts and/or almost everyone); (4) unknown unknowns (anybody’s guess — everybody’s ignorance).
tritonstation wrote (April 8, 2026):
> […] We think we’ve solved cosmology, and that solution requires non-baryonic cold dark matter. […] % […] But an explanation that omits MOND is a lie of omission.
This remark inspired me once again to “Search this/your blog” for the terms “conformal” and/or “Mannheim”; as I had done occasionally already in past years.
There are (still only) two instances of P. D. Mannheim being mentioned (rather briefly).
Would you please comment (or if appropriate: blog) a bit more on your opinion about conformal gravity (of which Mannheim still seems the principal exponent) and relations to your work. Thanks.
Good question. I got enthusiastic about conformal gravity when I first read the work of Mannheim & Kazanas on the subject in 1989 (https://scixplorer.org/abs/1989ApJ…342..635M/abstract). This was before I had considered MOND, or even thought that dark matter was problematic. Conformal gravity just seemed like an intriguing idea. I even had a section on it in my series of 1998 papers with de Blok. However, I cut it out at Mannheim’s request because he wasn’t confident that the theory was ready for testing at that level of detail yet. At that time, it did not work. There is a Newtonian term and a term that predicts a minimum, constant acceleration. That in turn predicts rising rotation curves, not flat ones. Flatness occurs over a finite range of radii due to the competition between the falling Newtonian term and this rising one. That’s OK for many galaxies but not all. Some (e.g., UGC 128) probe to low enough acceleration that we should see a rising RC. We do not.
There is a third, “cosmological” term that we all agreed should be ignored at the time. It turns out this was a bad assumption; it starts to matter in the regime of low accelerations. Mannheim and O’Brien have written several papers (e.g., https://scixplorer.org/abs/2012PhRvD..85l4020M/abstract) making fits that include this term. It seems to work to fit the data, but things get rather contrived. UGC 128, for example, has a sharp cut off in V(R) just slightly beyond the last measured point. We had already extended that RC once hoping to see the edge of the DM halo, which we didn’t, so it’d be a heck of a thing if we just missed this strong signal.
Mannheim and O’Brien were doing the RC fitting, so I didn’t feel the need to join in. More generally, the use of three terms where one suffices I find unappealing, so I haven’t worked on it further. It seems like supersymmetry: a good idea that Nature declined to implement.
IMHO bootstrapping string theory to basic assumptions, or having faith instead of realism, is not such a big problem, although the hidden assumption that gravity must be quantized on a background should be properly mentioned. It contributes to knowledge, and people are stubborn enough to ignore it anyway. I do find it a huge problem if “faith” in String theory makes you ignore MOND successes.
A problem I do see is the amount of effort and imagination required to study one tree of string theory (one theory) and calculate predictions, while at the same time there’s so much trees that it’s impossible to see the forest. The predictive power lacks in many respects, and the effort sinks into a kind of black hole. Also the need for a background is a small concern. In physics, theory and experiment ought to have closer contact than this to remain healthy. Still I do find some useful things in string theory, such as holographic QCD predicting with only small errors the masses of several observables.
One of the allegedly beautiful things about string theory is that gravity “falls out of it” naturally. From that starting premise, one naturally ignores MOND – that’s the wrong kind of gravity! I do wonder if they might have made more progress if they used MOND as an input. For all we know, they’ve derived the correct theory but thrown it away for doing this crazy thing.