I. What scientism actually is

Scientism is not science, and it is not evidence-based reasoning. It is a philosophical position — usually unstated — that the natural sciences are the only legitimate source of knowledge, and that claims falling outside their reach are either reducible to them or meaningless.

The critiques of this position are serious, and worth stating plainly.

It is self-undermining. The claim "only scientifically verifiable statements are meaningful" is not itself scientifically verifiable. Scientism smuggles metaphysics in through the back door while insisting the front door is the only one.

It makes category errors. Ethical obligation, aesthetic response, the meaning a person makes of a loss, the insight that arrives on the cushion — these are not failed empirical claims awaiting better instrumentation. They belong to different orders of description. Treating them as deficient science is a mistake about what kind of thing they are.

It has a colonial history. In mental health specifically, the record is not ambiguous. Diagnostic categories developed in a narrow cultural context have been exported as universals. The bulk of our outcome literature rests on WEIRD samples — Western, educated, industrialized, rich, democratic — while being spoken of as though it described human beings as such. Healing traditions with long records of use have been dismissed as superstition without ever being studied. Research funding has followed the interests of those who hold it.

It confuses method with authority. The prestige of scientific language has repeatedly been used to settle questions that were never empirical to begin with — who counts as disordered, what a good life looks like, whose distress is legitimate.

All of this is fair criticism. None of it is controversial among careful people.

II. The reaction, and where it goes wrong

Here is where the argument tends to break.

Having established that evidence-based practice has been shaped by power, a further step gets taken: therefore the evidential standard itself is a colonial artifact, and clinicians should be free of it.

This is a conflation, and it is worth naming precisely, because two genuinely different objects are being collapsed into one.

The first is a critique of the sociology of knowledge production — who gets funded, who gets sampled, whose experience counts as data, which questions get asked, which get ignored. This critique is correct and urgent.

The second would be a critique of the logical structure of evidential reasoning — the idea that beliefs about causation in the observable world should be answerable to observation, that claims which cannot fail should be held more loosely than claims which could have failed and didn't, that the strength of evidence should scale with the magnitude of what is being asserted.

The first critique does not establish the second. Bias in how a standard has been applied is not evidence that the standard is invalid. If anything, the demonstration of bias depends on the standard — you can only show that samples were unrepresentative, that findings failed to replicate across populations, that funding skewed the questions, by appealing to exactly the kind of reasoning the argument then proposes to discard.

The strongest decolonial arguments in our field are empirical arguments. They are made with data. They require the very standards they are sometimes said to overturn.

There is also a straw man in circulation. Evidence-based practice in psychotherapy is the integration of best available research with clinical expertise and client values and preferences. It was never "run the manual." Critiquing the narrow, manualized, insurer-friendly distortion of EBP is entirely legitimate. Treating that distortion as the whole concept is not.

And the practical cost lands where it always lands. Without any shared way to distinguish what helps from what harms, the vacuum is filled by charisma, confidence, seniority, and market forces. Clients — often the ones with the least power to walk away — carry the risk. The history of psychotherapy contains a good number of confidently delivered interventions that turned out to hurt people. Sincerity was never the safeguard.

III. A way forward

The path through is not a compromise between the two positions. It is a clearer account of what is being claimed, and by whom.

Distinguish types of claims. Not everything said in a therapy room is an empirical assertion about physical causation, and the ones that aren't shouldn't be judged as though they were. "This practice has helped my people carry grief for generations" is a claim about lived tradition and meaning. "This intervention reduces PTSD symptoms relative to no treatment" is a testable causal claim. Both can be spoken with integrity. They answer to different standards, and clarity about which is which prevents a great deal of pointless argument.

Expand who meets the standard rather than suspending it. The response to a narrow evidence base is a wider one. That means samples that reflect the populations actually being served. Outcomes defined by communities rather than only by researchers. Practice-based evidence gathered in real settings alongside efficacy trials. Indigenous research methodologies taken seriously as methodologies — with their own criteria of rigour, relationality, and accountability — rather than as culture to be studied by someone else's method.

Be honest about what our tools can and cannot answer. Randomized trials answer certain questions well and others poorly. They are good at "does this outperform that, on average, on this measure." They are weak at "what does recovery mean to this person," "what makes an alliance hold," "what happens over fifteen years." Recognizing the limits of an instrument is not the same as abandoning measurement. It is part of using it properly.

Keep epistemic humility on both sides of the ledger. The evidence base is provisional, partial, and shaped by the conditions of its making. That is a reason to hold it lightly, revise it often, and widen it deliberately. It is not a reason to stop asking whether what we do actually helps.

Closing

The two positions being argued over are mirror images. Scientism claims a single method exhausts knowledge. Its reaction claims that because the method has been misused, no method is owed. Both foreclose the harder and more useful work.

The position worth holding is unglamorous and, I think, correct on both counts:

The knowledge base of our field has been shaped by power, and needs to be widened, corrected, and shared.

And we still need some way to tell whether what we are doing is helping the person in front of us.

These are not in tension. Only the caricatures are.