Why Factory Education Fails to Teach Thinking

The complaint that schools do not teach thinking is old enough to have become furniture. It is usually made loosely, and loosely made it is easy to dismiss: students plainly learn a great deal, examinations are plainly difficult, and the people who pass them plainly go on to build bridges and treat patients. So the claim has to be narrowed until it can be wrong.

Here is the narrow version. Thinking, in the sense this book cares about, is not the recall of approved content. It is the capacity to generate a question of one's own, to test it, and to revise a position when the test comes back badly. Factory methods measure the first thing with considerable precision and leave the second unmeasured. What is not measured is not practised, and what is not practised does not become a habit. The result is not ignorance. The result is a large population of educated people who can perform intelligence on demand and who have very little practice at examining a position they already hold.

Four things a scored curriculum quietly teaches

None of these appear in a syllabus. All of them are learned, because the structure teaches them whether or not anyone intends it.

• Every question has a settled answer somewhere. The answer exists before the student arrives and is held by someone else. A question is therefore a request for retrieval rather than an opening of an investigation.
• Uncertainty is a symptom of insufficient preparation.Saying "I do not know, and here is what would tell me" is not a scoreable move. Saying something confident and wrong often scores better than saying something tentative and accurate.
• The efficient path to the reward is the correct path.This is the same optimisation that shows up in a trained model with a badly specified objective. A student who finds a shortcut through the proxy has not cheated the system; they have read it correctly.
• Revision is an admission of failure. A changed answer costs marks. Nothing in the structure rewards a student for saying which piece of evidence moved them.

Set beside those four, the standard reform proposals — more critical thinking modules, more project work, better rubrics — are patches on the proxy rather than changes to it. A critical thinking module with an answer key still teaches the four propositions above. It simply teaches them using the vocabulary of examination.

What the missing practice actually is

The practice that goes missing is small and specific, which is what makes its absence measurable. It has three parts.

1. Commitment before consultation. The learner states a position, in writing, before any authority — teacher, textbook, or model — supplies one. Without a recorded prior position there is nothing for a correction to act on.
2. Named falsifier. The learner states what observation would make the position wrong. This is the operational test of inner direction used throughout this book: can the person say, unprompted, what would change their mind?
3. Recorded reason for revision. When the position changes, the learner states which specific thing changed it. A revision with no stated reason is compliance wearing the clothes of learning.

A classroom that runs those three moves has not abandoned content. Content is what the moves are performed on. What it has abandoned is the permanent visible presence of the answer key during the interval when the student is forming a view.

Why credentialed competence is not a counter-example

The bridges do get built. This is the strongest objection to the chapter and it deserves to be stated at full strength rather than deflected: factory schooling demonstrably transmits difficult technical competence, at scale, to people who did not have it. Nothing here disputes that.

The claim is narrower. Technical competence and the habit of examination come apart, and they come apart in a direction that is easy to observe. A person may be entirely reliable inside a specification and entirely unpractised at asking whether the specification is the right one. That gap is not a character defect. It is the predictable output of years of training in which the specification was never once in question, because the specification was the thing being graded.

The failure this argument would suffer

Two outcomes would put the chapter out of business, and both are available to a modest study.

• If students in conventionally scored classrooms produce unprompted falsifiers for their own positions at the same rate as students in classrooms running commitment-before-consultation, then the answer key is not doing the damage attributed to it here, and the mechanism in this chapter is wrong.
• If students trained in the three moves show the capacity only on the tasks they were trained on, and no transfer to an unfamiliar domain, then what was taught is another performance and not a habit — the precise failure this chapter accuses the factory of.

There is also the objection that runs against the whole series and is not resolved: measured across countries, more years of formal education correlates with lower support for authoritarian government, not higher. If factory schooling trains outer direction as claimed, that correlation is awkward and no amount of restatement makes it less so. The most that can honestly be said is that years of schooling and the practice of examination are different variables, and that no study yet separates them cleanly. Until one does, the argument in this chapter remains a mechanism with a live objection attached.