Reverse Causality

Reverse causality, as this page uses it, is a fault in how a chunk gets built: you justify why one idea matters by pointing at a second idea, and the second is itself something you would have to memorize. The two now hold each other up, and nothing outside the pair explains either one, so a move meant to reduce what you carry has added an item to it. The name is borrowed rather than shared — this is not the statistical case where an outcome turns out to cause its predictor. The repair is to anchor the first idea in things you already hold: what else it influences, what produces it, what mechanism it shares. Then the link to the second can be rebuilt from more than one direction.

The loop you can see

A chunk is a group of ideas held as one unit instead of many isolated facts. Importance-based grouping works when the reason A matters can be generated from something already known: a broader relationship, a prior fact that does not itself need to be learned tonight, a logical step. Elaborative interrogation and self-explanation have the same bound. The why has to be answerable from what the learner already holds.

The failure is narrower. A is marked important because it creates B. The only way to know that A creates B is to memorize the arrow. Nothing outside A and B can reconstruct it. The justification has not grounded A. It has added a second item — B, and the A→B link — that must be stored on its own.

Isolation
  A  →  B
  A matters because it creates B; B is itself a memorized fact.

Wider network
  E produces A.
  A influences C and D.
  A shared mechanism sits under all of them.
  B can now attach without the A→B arrow having to be stored alone.

Once B sits in that network, the A→B relationship can be reconstructed from more than one cue instead of recalled as a standalone fact. Inferred is the hope. Reconstructable from more than one cue is the test.

After a chunk is built, two sentences catch the loop:

A is important because B.

Would I only know that if I had memorized that B is due to A?

If the answer is yes, the justification is circular. The chunk is not finished. It has to be reconnected to the wider network — the Schema the surrounding topic is trying to become — before the grouping will reduce load instead of adding it.

Why it happens

The loop is most likely when chunks are formed in the same order as the source, rather than by independently evaluating relationships. Aim stops that textbook-order encoding: write what the material must answer before opening it. Shoot then works the source against those questions into a rough map, so the first grouping is not the author’s next sentence.

It is also most likely when importance questions are answered too quickly, with the first plausible connection rather than the most meaningful one, and when the chunk structure is committed before the wider topic has been surveyed. A first shallow pass should precede a committed structure.

Two operators raise the risk because they reward fast, locally coherent justifications. Importance checklisting is running a fixed list of “why this matters” prompts and accepting the first hit. Simple Q&A is a local question-and-answer pair that never leaves the sentence it came from.

The repair is not a second name for the same check. Intuitive chunking — finding the root reason something matters across the whole topic — is the move that Importance-Based Chunking owns. Ground importance in the whole topic, not the next sentence.

Reverse causality and spiderwebbing often appear together. Circular justifications produce shallow, locally determined arrows; the map then looks chaotic because no wider logic constrains them. Mindmaps owns that chaotic-arrow failure. This page is the upstream cause. Fixing the loop at the chunking stage usually cleans up the spiderwebbing downstream.

The same loop, now caught, is what the next pass reconnects. A yes on the check means A gets attached to something that does not itself have to be memorized, so the relationship can be rebuilt from more than one cue. That is what makes the next pass cheaper.

Related

• Importance-Based Chunking — the repair: ground importance in the whole topic, not the next sentence
• Mindmaps — owns the chaotic-arrow failure; this page is the upstream cause
• Schema — the wider network the reconnection is trying to build
• Aim — write what the material must answer before opening it, which stops textbook-order chunking
• Shoot — work the source against those questions into a rough map
• The Shortcut Problem — the sibling failure: a locally satisfying move that does not do the work

Open Questions

• How often the check fires on notes already marked done.

Sources

• House encoding diagnostic. The check and the A–B isolation are this system’s own; they run on any notes already in hand.
• Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest. Elaborative interrogation has moderate utility when the why is answerable from prior knowledge.
• Pressley, M., McDaniel, M. A., Turnure, J. E., Wood, E., & Ahmad, M. (1987). Generation and precision of elaboration: effects on memory. Elaborative interrogation helps when prior knowledge can generate the elaboration.