Mind, Matter, and Meaning

Vehicle and Content

The vehicle of a representation is what carries it: ink on a page, pixels on a screen, a spoken sound, or a pattern of neural activity. Its content is what it represents. A photograph on a desk can depict the Eiffel Tower in Paris; neither the tower’s height nor its location is a property of the paper. The distinction separates questions about a representation’s physical realization from questions about what it says or depicts.

Vehicles, Contents, and Objects

Three terms need to be distinguished. The vehicle carries the representation. The content specifies how things are represented. The object, where there is one, is what the representation concerns. A photograph may depict the Eiffel Tower from below. The tower is its object; the way the tower is depicted belongs to its content. A drawing of an imaginary tower has a vehicle and content even if no such tower exists.

Philosophers disagree about the nature of content. Propositional theories treat the content of a belief as something that can be true or false. Other theories describe perceptual content through accuracy conditions, spatial structure, or nonconceptual presentations of properties. Some views make particular objects constituents of content; others distinguish an object’s contribution to reference from a more general descriptive content. The vehicle/content distinction does not settle these disputes.

The same content can often be carried by different vehicles: a spoken sentence, its written version, and a translated sentence may express the same claim. Conversely, an identical mark can carry different contents in different languages or contexts. This does not make the vehicle irrelevant. Its structure, history, and use can help explain why it carries one content rather than another.

Physical Properties and Represented Properties

A map represents a winding road without its paper winding through the countryside. A neural state can represent a red apple without itself being red or apple-shaped. Moving a map to another room need not change the place it represents. These examples show why an inference from the location or properties of a vehicle to the location or properties of its object requires further premises.

Fred Dretske develops the distinction in Naturalizing the Mind (1995). For his representational theory, the relevant sensory qualities are properties represented as belonging to objects, rather than qualities that the neural vehicle literally acquires. This helps explain how physical processes can concern things outside the nervous system.

It does not follow that every indirect theory of perception confuses a vehicle with an object. A sense-datum theorist may deliberately posit immediately perceived objects on independent grounds, such as arguments from illusion or hallucination. Nor does every defender of qualia hold that neural tissue literally turns red. Their claim may concern properties of experience that neither resemble the neural vehicle’s ordinary physical properties nor reduce to its representational content. Those positions require substantive evaluation, not just a reminder of terminology.

Perception, Transparency, and Hallucination

Representational theories of perception distinguish the processes enabling vision from what vision presents. Looking at an apple ordinarily directs attention toward its color, shape, and position, not toward the neurons enabling the experience. The transparency thesis generalizes this observation: introspection of perceptual experience seems to disclose represented features rather than a further inner object.

Different transparency claims have different strengths. The inability to attend to one’s neural vehicle does not establish that every phenomenal feature reduces to content. Tim Crane emphasizes the intentional mode under which content is presented, while Ned Block argues that qualitative character can exceed representational content. Seeing and imagining the same scene, or seeing it sharply and blurrily, provide tests for accounts of how content, mode, and phenomenal character relate.

On a representational account, hallucination can involve content whose accuracy conditions are not met. An experience as of a red apple need not involve an actual apple or an inner red object. Relational and disjunctivist theories question whether veridical perception and an indistinguishable hallucination share the same fundamental nature. The vehicle/content distinction remains useful in articulating the alternatives, but does not select between them.

Accuracy and Perceptual Contact

Accurate content is not always enough for perception of an object. Imagine an actual hand hidden behind a screen. Brain stimulation produces an experience as of a hand at exactly that location, independently of the hidden hand. The experience matches the scene, yet the match alone does not establish that the subject sees that hand.

Causal theories require an appropriate connection between the object and the perceptual process, not merely correspondence between content and reality. They must also explain which causal chains count as appropriate: an object might cause an experience through an unusual route that still fails to constitute seeing it. Relational theories instead make the perceptual relation central to the nature of the experience. The example separates accurate representation from successful perceptual contact.

Cognitive Science and Artificial Systems

In cognitive science, identifying a neural or computational vehicle differs from determining its content. Researchers may ask which pattern encodes a location, how downstream systems use it, and whether intervention changes behavior in ways predicted by that interpretation. Causal, informational, teleosemantic, and interpretation-based theories offer different accounts of what makes these assignments correct.

A model’s token vectors or weight matrices are candidate vehicles; their contents cannot simply be read off their numerical shape. Training data, learning objectives, environmental relations, and subsequent use may all be relevant. Establishing content at one level does not by itself settle whether a whole system understands language or has conscious experience. Equally, the distinction supplies no general argument that artificial vehicles cannot carry genuine content.

Further Reading

  • Dretske, Fred. Naturalizing the Mind. MIT Press, 1995.
  • Crane, Tim. Elements of Mind: An Introduction to the Philosophy of Mind. Oxford University Press, 2001.
  • Block, Ned. “Mental Paint.” In Reflections and Replies: Essays on the Philosophy of Tyler Burge, edited by Martin Hahn and Bjørn Ramberg, 165–200. MIT Press, 2003.
  • Mind, Matter, and Meaning — Gordon Swobe’s book-length argument about mind, consciousness, and meaning; an author position, not an endorsement by this reference page or its subjects.