Saturday 20 February 2016

Against Cognitive Representation


In a recent conference presentation, William Ramsey (2015) argues that when it comes to the question of representations in the brain, the burden of proof lies with those who deny the existence of these alleged entities. He makes it clear that the consensus amongst “the vast majority of cognitive scientists” is that “representations are realised in the brain” and he agrees with Godfrey-Smith (2006) that the term “representation” is not used in a technical sense but is based on the standard definition in which representations are stand-ins for the things they represent. On these last two points I agree.

Ramsey focusses his attention on what he calls "representational deflationism." By this he means "views that deny that the value of representational theories depends upon the existence of actual representational structures in the brain." Ramsey argues very persuasively that representational deflationism either falls into anti-representationalism or what he calls "representational realism." On this point too, I think Ramsey's analysis is correct.

Even though I disagree that the burden of proof falls upon the shoulders of those who deny that brains produce representations—and I will explain why—I hereby take up Ramsey’s challenge.
  1. Anyone who claims the existence of an entity, whether it be a fairy, a god, a ghoul in the attic or a ghost in the machine, is under an obligation to support their claims with evidence, preferably empirical evidence. If the burden of proof lies with those who deny the existence of nonentities, then there would be no end to the nonentities they would have to disprove.
  2. Representations are actions or artefacts produced by creatures as part of communicative transactions, usually involving two or more agents. Brains do not interact or make transactions, only creatures do.
  3. Brains have no medium of representation.
  4. Unlike embrained creatures, brains do not have any manipulative agency or resources that can be mobilised for representation use or production.
  5. All representations that are not manifested as behaviours, are manifested as artefacts produced by tool manufacturing creatures. Brains cannot fashion artefacts of any kind.
  6. Representation has evolved in response to social pressures and opportunities. There  are no social pressures or opportunities for speculative exploitation of available resources in the brain.
  7. Representations either resemble or symbolise the things they represent. If there are any resembling representations in the brain, then they must be observable. Causal correlations between world and brain are commonly observed, but cognitive representations remain elusive.
  8. Symbolic representation has an evolutionary reliance on more rudimentary skills in the use of resembling representations (imitation for example). Not even the most simple brains contain any evidence of mimetic representation or any other rudimentary representation, let alone more sophisticated forms.
  9. Representations are purposefully produced by intelligent creatures. Brains are inextricable parts of the central nervous systems of many creatures, but they do not have any intentions that are not the intentions of those creatures as whole organisms.
  10. Cognitive representations are often invoked to explain intentionally directed action, but this explanatory strategy is circular because all representations presuppose intention. There are no unintentional representations, only occasional artefacts and patterns that happen to resemble representations to varying degrees. Cognitive representations are therefore explanatorily vacuous in this important respect.
  11. If brains act on their own intentions, then we are host to forms of agency which need to be explained without recourse to an infinite regress of further intention initiating representations. Thus, all cognitive representations must necessarily be unintentionally produced and cannot therefore contribute to an explanation of intentional behaviour.
  12. Representations are necessary for communication. Brains do not need to communicate because they have direct causal influence over their own processes and those of the organism as a whole. Unlike the brain/body relation of influence, the influence of representations is always indirect (see note 3 above).
  13. If brains were sufficiently intelligent to develop as yet undetectable representational systems during their evolution without leaving evidence of more rudimentary systems observable in simple brains, then there is no reason to believe that they would need such systems in the first place.
  14. The only entities in need of encoded representations are creatures with secrets and those they wish to share them with.
  15. Representations are used to mediate between intelligent agents capable of exploiting opportunities and of making mistakes. Brains do not exploit anything that the organism as a whole does not exploit and they do not make any mistakes that the organism as a whole does not make.
  16. Ramsey claims that all representations have content. Representational content is culturally determined. There is no culture in the brain. Isomorphic representations (copies, replicas etc.) do not depend on content, but there is no unequivocal evidence of such isomorphic representation in the brain.
  17. Symbolic content (what representations are about as opposed to what they are of) is not fixed. Disputes over the content of representations are arbitrated according to social norms and discursive practices. There are no social norms or discursive practices in the brain.
  18. If there are multiple agents in the brain capable of producing and interpreting representations, then there is no possibility of clearly explaining agency because there is no way of explaining how final arbitration is reached in cases of conflicting representational interpretations.
  19. Ramsey suggests (pace Putnam 1975) that it would be a miracle if there are no representations in the brain. There are two replies to this claim: A: The real miracle would be if there are representations in the brain. B: If a creature has a capacity to produce representations, then this capacity alone is sufficient to fit the explanatory job description. The capacity to make a pot is not an inner pot.
  20. Ramsey argues that the widespread scientific consensus on cognitive representation is a reason to be a realist on the subject. This amounts to nothing more than the observation that cognitive representation is a widespread article of faith amongst scientists who believe that they have better things to do than to determine whether their philosophical foundations are firmly rooted or not. It is the job of philosophers to point out the significant conceptual weaknesses in this theoretical position.
  21. To see or otherwise experience a representation of X is not to experience X. Thus, an experience of a representation cannot be an experience of the thing/s represented.
"Suppose someone were to coat the occipital lobes of the brain with a special photographic emulsion which, when developed, yielded a reasonable copy of a current visual stimulus. In many quarters this would be regarded as a triumph in the physiology of vision. Yet nothing could be more disastrous, for we should have to start all over again and ask how the organism sees a picture in its occipital cortex, and we should now have much less of the brain available in which to seek an answer." (Skinner 1969, 232)

If you have any suggestions of further objections to representationalism then please let me know.