The Spectrum We Split: Science as Philosophy, Religion, Politics, and Method

We didn’t find the categories “science,” “religion,” “philosophy,” and “politics” lying around in nature, helpfully labelled by a cosmic filing clerk. We drew them. The way we carve white light into seven named colours, a continuous thing sliced into useful pieces, we split the human project of figuring out the world into departments, and then forgot that the lines were ours. Boundaries, as I keep insisting, are as real as money: invented, agreed upon, and absolutely consequential. You cannot pay rent with the observation that money is a fiction, though it would be interesting to watch someone try.

So here is the experiment. Take one enterprise, call it the organised human attempt to understand and to act, and turn it slowly under four lights. Watch what each one reveals. Notice, inconveniently, that it is the same object the whole time.

As philosophy

Before science can measure anything, it has to assume things no measurement can defend. That the universe is regular, that tomorrow’s physics matches today’s. That our senses, sharpened by instruments, report something real. That the cosmos is, suspiciously, comprehensible to a primate that evolved to find fruit and avoid being eaten. None of this is provable. It is a wager that keeps paying out.

David Hume noticed three centuries ago that we have no non-circular reason to expect the sun to rise: we believe it because it always has, which is precisely the thing in question. Karl Popper tried to rescue the enterprise not by proving theories true, because you can’t, but by demanding they be the kind of thing reality could prove false. Science, on this view, is the game where you write down in advance what would make you wrong. Then Thomas Kuhn complicated even that, showing that working scientists mostly don’t falsify; they labour inside a paradigm until it cracks under its own anomalies, and then convert to the next one almost the way one converts to a faith. Strip the enterprise to its frame and you find philosophy holding it up: a set of load-bearing assumptions, chosen because they work, not because they are certain. A few first principles, as I like to say, are better than none.

As religion

Now the same object under a different light, and the resemblance is uncomfortable. There is faith here: faith that the regularity holds, that the mathematics means something. There is awe, the genuine article. Anyone who has grasped, even for a held breath, that they are made of the ash of dead stars has felt the thing cathedrals were built to produce. There is a priesthood, credentialed and robed at graduation. There are sacred texts, the journals whose names we lower our voices around. There is liturgy: the Methods section, the ritual of reproducibility. There are saints (Newton, Darwin, Curie), relics, schisms, heresies, and the occasional excommunication. Humans do love a robe.

“Trust the science” is a creed, and like most creeds it is usually recited by people who have not read the scripture. The honest scientist finds this lens disturbing precisely because it fits so well. But the fit is not an accusation. The difference that matters is small and decisive, and I will come back to it.

As politics

Turn it again. Science is also a way of distributing power and money, and of deciding whose curiosity gets fed. Every experiment that happens is a thousand experiments that didn’t, because someone with a budget chose. Consensus is real, and it is also manufactured: through peer review, which is gatekeeping by the already-anointed; through citation, where the rich get richer; through the quiet gravity that pulls researchers toward the publishable question instead of the important one. States fund the science that builds their weapons and their prestige. “Whose questions get asked” may be the most political sentence in the whole enterprise, and it is answered with chequebooks. To insist that science floats serenely above the city is itself a move in the city’s game, usually one that flatters whoever currently holds the microphone.

As method

And yet. Under the fourth light there is one thing here that the other three never had. Not certainty. Science is the least certain of the four, which is its glory, not its embarrassment. The new thing is institutionalised doubt: a procedure for being wrong on purpose and noticing in time.

Religion asks you to believe. Politics asks you to win. Philosophy asks you to argue. The method asks you to predict, then go and look, then, this is the radical part, to lose gracefully when the universe disagrees. Feynman put the whole of it into one sentence: it does not matter how beautiful your guess is, or how smart you are, or what your name is; if it disagrees with experiment, it is wrong. That is the only one of the four traditions with a built-in mechanism for changing its mind that does not require anyone to die first.

The mechanism is leaky. It is slow, and corrupted by every human thing visible in the first three lenses. But it exists, and over centuries it ratchets, the way a democracy can be venal and venal and venal and still, slowly, correct itself. The priesthood is real, and the method works in spite of the priesthood. Sometimes, when the light catches it, because of the priesthood arguing with itself.

The seam that isn’t there

So which is it: philosophy, religion, politics, or method? The question is the mistake. It is the spectrum problem again. You can name the colours, but you cannot find the seams, because there are no seams. Science is a human animal’s drive to understand, wearing four coats, and the coats are ours.

The reason the religion lens fits is not that science is secretly a faith waiting to be debunked. It is that faith, politics, philosophy, and science are all downstream of the same source: a creature that woke up inside a universe it did not ask for and could not stop trying to understand. Everything is connected because it all comes from one place.

I find this clarifying rather than deflating. If the method really is the one new tool the species has invented for being a little less wrong each century, then the work is not to purify it of its philosophy, its awe, and its politics. You can’t; they are the body it walks around in. The work is to keep the small, decisive thing alive inside all of it: write down what would prove you wrong, and mean it.

Everything else — the cathedrals and the committees — is what humans always do. Stay curious. Stay honest. Stay hydrated.

Further reading

• The Problem of Induction (Stanford Encyclopedia of Philosophy) — Hume’s quiet bombshell: why “it always has” can’t justify “it always will.”
• Karl Popper (SEP) — falsifiability, and science as the discipline of writing down in advance what would make you wrong.
• Thomas Kuhn (SEP) — paradigms, normal science, and revolutions that look unsettlingly like conversions.
• Paul Feyerabend (SEP) — Against Method, the gadfly who argued there is no single recipe called “the scientific method.”
• Cargo Cult Science — Richard Feynman’s 1974 Caltech address on the kind of integrity that distinguishes the method from its imitations.
• (video) The Key to Science — Feynman, in one minute: guess, compute, compare; if it disagrees with experiment, it is wrong.
• (video) How to Detect Baloney the Carl Sagan Way — Michael Shermer on Big Think, walking through Sagan’s “baloney detection kit” for thinking past the priesthood.