Niraj ChaurasiyaBuilding systems under uncertainty

The Illusion of Learning

How do we know that learning occurred rather than merely experiencing familiarity, fluency, engagement, or completion?

This research and communication project examines learning as a latent process. It asks what observable evidence can support the claim that capability changed and distinguishes feelings associated with learning from demonstrations of recall, performance, transfer, teaching, creation, and reflection.

Publication note

Developing TEDx talk, essay, and public video series. It is not currently a formal experimental study.

Central claim

The subjective feeling that learning occurred is not sufficient evidence that capability changed.

Method and evidence

How the argument is currently supported

Method

Current approach

Conceptual analysis of learning as a latent process.

Comparison of learning feelings with observable performance.

Development of the Evidence of Learning framework.

Translation of research ideas into a public talk and video series.

Evidence

Supporting observations

Attention demonstrates exposure but not necessarily durable learning.

Fluency can make information feel understood because processing feels easy.

Recognition can occur without independent recall.

Independent performance and transfer provide stronger evidence than passive familiarity.

Teaching and creation can reveal relationships that memorization alone may not demonstrate.

Current structure

The developing argument

01

Learning cannot be observed directly

Learning is inferred from changes in knowledge, capability, behavior, or performance. The process itself is not directly visible.

This makes learning an epistemic problem: we must decide what evidence justifies the claim that learning occurred.

02

The feeling of learning

Attention, completion, enjoyment, note-taking, and familiarity can accompany learning, but none independently establishes it.

The problem is not that these experiences are useless. The problem is treating them as conclusive.

03

Evidence of changed capability

  • Recall without the original material.
  • Imitation with guidance.
  • Independent performance.
  • Transfer to a changed context.
  • Teaching relationships to another person.
  • Creating something that depends on the knowledge.
  • Reflecting on limitations, mistakes, and uncertainty.

Epistemic boundaries

What this work does not yet establish

Limitations

Current limitations

The current model is a conceptual framework rather than a validated diagnostic instrument.

Different learning types may require different forms of evidence.

Performance can be influenced by anxiety, context, prior knowledge, and available tools.

Failure to perform once does not prove that no learning occurred.

Open questions

What remains unresolved

How should evidence differ across conceptual, procedural, perceptual, and motor learning?

How much evidence is sufficient before claiming that learning occurred?

How should delayed retention be incorporated?

Can reflection be evaluated without rewarding polished language over genuine understanding?

Connected work

Research inside a larger system