Jackendoff’s Intermediate-Level Cognitive Representations
Linguist/philosopher Ray Jackendoff presents an “intermediate-level cognitive representations” theory of consciousness as a cognitively and computationally grounded, representational theory that accounts for consciousness, not by locating it in specific brain structure (as in the neural correlates of consciousness), but by identifying which types of mental representations are accessible to consciousness and why. There is no “central observer,” no “Cartesian Theater.”
Ray Jackendoff
Professor of Linguistics & Philosophy
Ray Jackendoff is an American professor of linguistics and philosophy, and co-director of the Center for Cognitive Studies at Tufts University. He bestrides the boundary between generative linguistics and cognitive linguistics, committed to both the existence of an innate universal grammar (an important thesis of generative linguistics) and to giving an account of language that is consistent with the current understanding of the human mind and cognition (the main purpose of cognitive linguistics).
Key Takeaways
Core Claim
Consciousness happens at a middle layer of mental processing—not raw data or abstract thoughts.
How It Works
We experience representations like shapes, sounds, and spatial layouts—not the brain’s inputs or concepts.
Distinguishing Idea
Consciousness is tied to how information is formatted, not where it happens in the brain.
Jackendoff’s Intermediate-Level Cognitive Representations
Linguist/philosopher Ray Jackendoff presents an “intermediate-level cognitive representations” theory of consciousness as a cognitively and computationally grounded, representational theory that accounts for consciousness, not by locating it in specific brain structure (as in the neural correlates of consciousness), but by identifying which types of mental representations are accessible to consciousness and why. He claims that conscious experience occurs at a specific level of mental representation, which he calls the “intermediate level.” It is situated between low-level sensory processing (raw data) and high-level cognitive representations (like abstract ideas). There is no “central observer,” no “Cartesian Theater” (Jackendoff, 1987).
Intermediate-Level Representations
Take the visual system. In Jackendoff’s model, low-level processing features light intensity and edge detection; intermediate-level: shapes, object contours, spatial relationships; high-level: object recognition, labeling, memory associations. According to Jackendoff, we are conscious of intermediate-level representations, not the raw sensory input or the high-level symbolic interpretations. He states, “We are not conscious of low-level sensory processing, nor of high-level conceptual structures, but of the representations in between.”
Jackendoff's theory is modality-specific, meaning that each sensory system (vision, auditory, tactile, etc.) has its own intermediate representations that enter consciousness. This explains why we are not typically conscious of language syntax or phonology directly, he says, but we are aware of its perceptual output (e.g., sound or meaning).
Consciousness as Representation
Jackendoff has phenomenal consciousness as representational, not mystical. Given that he works within the computational theory of mind framework, he posits consciousness as the result of information processing, operating at the computational-cognitive level. Thus, he has consciousness arising from how information is formatted in the mind, in addition to that which it is about (Jackendoff, 1987).
How does Jackendoff’s theory work? He supports a distributed, representational account, where consciousness is a result of specific kinds of neural representations. He is neutral regarding the underlying theory, but he believes that his theory—intermediate-level cognitive representations—articulates well with several of the leading theories of consciousness, including those globally accessible and those integrated.
Tools
Cite