Does Therapy Need a Better Theory, or a Better Menu?

Every so often, two smart people take the same problem and land in genuinely different places, and reading both is more useful than either one alone. That's what happened recently in a small corner of the therapy world, and it's worth sitting with.

It started with a New York Times essay by therapist Daniel Duane, arguing that psychotherapy has an identity problem. Dozens of methods all claim to work, yet decades of research keep pointing to the same humbling truth: the specific technique matters less than the relationship between therapist and client. Duane points to a newer body of research, including work out of the University of Cape Town, where our own director happened to complete her undergraduate studies, on something called predictive processing, the idea that our brains constantly build an internal model of the world and experience their own predictions as reality itself. If that's true, patterns like chronic anxiety or self-doubt might really be outdated predictions still quietly running the show, and therapy's job becomes helping someone update a model that stopped serving them long ago. For the first time, Duane suggests, the field might have something close to a unifying explanation for why good therapy works at all.

Not everyone is convinced that a grand theory is what the field actually needs. Writer Michael Goldstein, responding to Duane, agrees with the diagnosis but not the prescription. He raises two concerns. First, a better theory doesn't guarantee better practice, he points to Behavioral Activation Therapy, a method with strong evidence behind it that many therapists still avoid simply because it's tedious to deliver. Second, and more pointedly, he argues there's no real appetite among therapists, policymakers, or funders for a major overhaul in the first place. Their energy tends to go toward access, getting more people into therapy, rather than improving the quality of the therapy already happening. His alternative isn't a new theory at all. It's a wider menu. Instead of asking which therapist someone should see, start by asking what problem they actually want to work on, then let them choose from real options, a sleep coach, a tutor, a friend group, a therapist, and go from there.

Neither piece cancels the other out. One makes the case that better science could finally explain what's been working all along. The other makes the case that explaining it won't matter much if nobody's actually looking to change. Read together, they say something neither says alone: that progress in mental health care might require both a deeper understanding of why therapy works and a genuine willingness to rethink who gets offered what, and when.

Both are worth reading in full, the summaries here can only carry so much of the argument:

• 📄 Therapy Needs Therapy. Neuroscience Can Help. — Daniel Duane, The NYT

• 📄 Therapy Needs to Get Help — Michael Goldstein, CTF WTF