I write software for a living, so when I opened the first week of a cognitive behavioural therapy course I reached for the nearest mental model I had: a rule engine. Symptoms in, prescribed response out. The course spends its first week explaining why that model is wrong, and it took me most of the session to understand why it matters.
The Model I Brought With Me
CBT describes a cycle. Something happens, an automatic thought arises, an emotion follows, and the whole thing loops until a belief changes. Sketched out, it looks like this:
- Situation arrives
- Automatic thought fires
- Emotion and bodily response follow
- Behaviour follows that
- Consequence reinforces the thought
It is a tidy loop, and I had been quietly treating it as a mechanism. If a person’s thoughts produce distress, then changing the thought should reduce the distress, in the same way that changing one line of code changes a function’s behaviour. That felt obviously true, and it took the tutor’s first question to puncture it.
The Question That Broke It
If it were a mechanism, the tutor asked, how do you debug it? What do you see when you run it and get an unexpected result?
There was no answer. That is the whole problem, and it is worth sitting with, because the failure is not a detail. A mechanism with no observable output is not a mechanism — it is a diagram someone drew because it made sense on paper. The loop is a model of a conversation about what happens, not a description of a process that runs on its own.
Why It Matters for Practice
This is not a philosophical quibble, because the distinction changes what you do with a person in front of you.
If thoughts are a mechanism, you troubleshoot: identify the faulty input, replace it, expect a predictable result. That is what makes CBT attractive to a technical mind, and it is also why it goes wrong. The therapist is not debugging a system. They are asking questions, and the answer that comes back is the only real data available.
The one piece I am carrying out of week one is this: you cannot observe the mechanism, only the conversation about it, which means the dialogue is the treatment. Not a preamble to the treatment. The treatment.
What I Will Watch For
Now that I know to look for it, the same pattern shows up elsewhere. The self-awareness exercise from this week asks me to rate 32 statements about myself on two columns — how true each one is of me, and how true I think someone who knows me would say it is. My first instinct was to look for the column with the bigger gap, as though the gap were a defect to be found and fixed.
It is not a defect. Two people looking at the same person will not converge, and the amount by which they fail to converge is the interesting part. That is not a mechanism with a wrong output. It is two people, and a difference between them worth understanding.
Reference
Week one materials from the course, read alongside:
- Haarhoff, B., Thwaites, R. and Bennett-Levy, J. (2016). ‘Reflection in CBT: becoming better therapists, supervisors and trainers’. In Reflection in CBT. SAGE.
- Dalal, F. (2018). CBT: The Cognitive Behavioural Tsunami. London: Routledge.
Note
Written from personal study notes. These are my own impressions of a first week, not an account of the course or its teaching, and nothing here is clinical guidance.
