Source
Block (2024), “Précis of The Border between Seeing and Thinking”, LIT-649, read in NOTE-498. The book (LIT-636) is filed unread. This account rests on the précis, which states the thesis and updates it; it does not contain the arguments for the main thesis.
What was actually shown
The thesis (p. 273). Perception is constitutively iconic, nonconceptual and nonpropositional. Cognition is paradigmatically discursive, conceptual and propositional, and can use perceptual material inside a cognitive envelope (NOTE-498, C1: stated, argued in the book).
The envelope (pp. 274–275, Figure 1). Percepts compete in sensory areas; winners trigger frontoparietal “ignition”; the reverberating front-and-back coalition is the cognitive envelope, in which perceptual content is held in an abstracted, partly conceptual form for reasoning. The global neuronal workspace is reread as the mechanism of conceptualisation, not of consciousness (C2, asserted here).
Working memory (pp. 277–279). Nine lines of evidence that working-memory representations differ from perceptual ones and are task-dependent abstractions, including Duan and Curtis (2024). Since 2022 this has increased Block’s doubt that “raw” percepts are used in cognition at all (C3, moderate). Active working memory is taken to be conscious and maintained by broadcasting; activity-silent retention for up to 4 s is set aside as a different phenomenon (pp. 275–276).
Iconicity (pp. 279–280). Three conditions, after Shepard: representations respond to environmental differences by degree; changes in them change what they represent by degree; relations among properties are mirrored by relations among representations.
What could have come out otherwise. Working-memory representations could have matched perceptual ones. On the evidence cited they do not; whether the difference is one of format is left open.
What this does not say
- That the recoding changes format. The nine differences “do not show a
change in format”, and a change could be within iconic representation
(p. 279). That is the open question in
promote_when. - That thought is never iconic. Carey’s parallel-individuation system (about 3 items in infants, 4 in adults) and entropy-based same/different matching are iconic, though enclosed in an envelope. Block concedes he “might have scanted the extent of iconic representation in thought” (C4).
- That the workspace is not a mechanism of consciousness. The rereading is asserted here, not argued against the consciousness reading (NOTE-498, Limitations).
- That infants reason in working memory. Hamlin’s helper and hinderer results give “some reason” to credit 6–8-month-olds with it, and leaner iconic accounts are not ruled out (p. 276).
- That the main thesis is established. The arguments are in the book, unread here.
Connections
- Ventral and prefrontal content. THEORY-102’s finding of sustained ventral content and an onset-only prefrontal event is compatible with, and on Block’s reading predicted by, the cognitive-envelope view. A THEORY combining the two would need the book’s arguments (NOTE-498, Bearing on the record).
- The workspace as consciousness. THEORY-096 holds that perceptual consciousness is global availability through ignition. This account rereads the same ignition as conceptualisation. The two are compatible only if conceptualisation and conscious access coincide, which the précis does not discuss. THEORY-103, that perceptual consciousness overflows access, is the position from Block’s own earlier work that sits behind the rereading; the précis does not restate it.
- High-level perception. Chalmers, French and Hofstadter (THEORY-541) put concepts inside perception at the level where analogy works. Block puts the border at format, with conceptualisation on the cognitive side.
- Framing without penetration. Ivy and Mroczko-Wąsowicz (LIT-099, NOTE-150) cite the book for a border without cognitive penetration.
Note of 2026-10-29: what holds it back
Kept Proposed by the sixteenth status review. NOTE-498 read the précis in full, but the main thesis’s arguments are in the book (LIT-636), unread. The promote_when’s result is not in the record: no test of whether working- memory contents fail the tracking and mirroring conditions for iconicity that the corresponding percepts meet. Block grants that his nine lines of evidence “do not show a change in format”.