Skip to main content
Log in

Why it doesn’t matter to metaphysics what Mary learns

  • Published:
Philosophical Studies Aims and scope Submit manuscript

Abstract

The Knowledge Argument of Frank Jackson has not persuaded physicalists, but their replies have not dispelled the intuition that someone raised in a black and white environment gains genuinely new knowledge when she sees colors for the first time. In what follows, we propose an explanation of this particular kind of knowledge gain that displays it as genuinely new, but orthogonal to both physicalism and phenomenology. We argue that Mary’s case is an instance of a common phenomenon in which something new is learned as the result of exploiting representational resources that were not previously exploited, and that this results in gaining genuinely new information.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. For a survey of such replies, see Alter (2005). As we are going to use ‘knowledge,’ knowledge involves states with representational content. While we recognize that there may be accounts of Mary’s knowledge gain that do not characterize the gain in terms of states with content—versions of the abilities reply (Lewis 1983, 1988; Nemirow 1990) may be examples—we are not going to discuss such views here. While we think there are compelling reasons to accept that Mary’s genuinely new knowledge involves states with representational content, our aim isn’t so much to defend this claim as to show why characterizing her genuinely new knowledge in terms of states with content (a) has no interesting metaphysical implications and (b) does not turn on features that are peculiar to first-person phenomenology.

  2. See Haugeland (1991), Fodor (2007), Heck and Richard (2007), and Cummins (2010) for a sustained discussion and defense of theses along these lines.

  3. The words ‘form’ and ‘format’ are used here as a generalization of what is generally called syntax. Syntax is what we call the form or format of a structured propositional representation—e.g., a sentence.

  4. We also take it to be obvious that there is a good deal of diversity of format between perceptual modalities, and the science seems unequivocal that there is a good deal of diversity within a given modality as well.

  5. According to both Fregean and Russellian accounts of propositions, propositions are significantly structured entities that have either senses (Fregean) or objects and properties (Russellian) as constituents. Each account reveals intimate parallels between propositional structure and sentential structure, and some philosophers (e.g., King 2007) hold that the structure of propositions is inherited directly from the structure of sentences. We think such parallels should come as no surprise.

  6. Stanley and Williamson (2001), while not using our terminology, posit content-demonstrating speech acts in their account of knowing how. On their view, knowing how to X entails knowing a proposition that contains a way of X-ing. They allow that a way of X-ing may be introduced into a proposition by way of demonstratives. In one of their examples, Hannah comes to know that some way is a way for her to ride a bicycle when Susan points to John and says, “That is a way for you (Hannah) to ride a bicycle.” Contrary to our view, Stanley and Williamson do not think this commits them to positing non-propositional knowledge [cf. Stanley (2011)].

  7. Accuracy can be graded, of course, and we will want to allow for (A) being true even if the picture introduced is not perfectly accurate. The point is tangential to the main argument, however, so we will ignore this complication here.

  8. This is a kind of search by gradient descent in error space analogous to having someone find a hidden object in response to “warmer” or “colder” signals from someone who knows the location.

  9. Due to production limitaions, the square is not colored in the print version of the paper. The square is colored red in the online version of the paper, and should be understood as being red by the reader. Mary, of course, knows it is the color—not the shape—that is at issue.

  10. Lingua can get a good verbal description, but there would still be a substantial cognitive gain when the picture becomes available. This is why you want the police to have a picture of the suspect rather than a verbal description.

  11. Michael Tye puts the point this way:

    What Mary thinks is not new when she leaves her room. What is new is the way she is thinking what she is thinking. That isn’t enough. What Mary knows before time t (the time of her release) is exactly the same as what she knows after time t. But if what she knows before and after her release is the same, she does not make a discovery in a really robust sense. This is counter-intuitive. Surely if anyone ever made a significant discovery, Mary does here. The proposal, in the end, is not convincing (2009, p. 55)

    Notice, by the way, that if you (a) want to be a physicalist, (b) assume that limiting non-propositional representational resources (or opportunities to exploit them) does not limit epistemological access to the physical, (c) accept that Mary gains genuinely new knowledge after her release, but (d) deny that the new knowledge/old fact reply explains how Mary’s knowledge can be genuinely new, then the abilities reply will look rather compelling. Because we hold that Mary’s gain should be characterized in cognitive-cum-representational terms, with think the new knowledge/old fact reply is superior to the abilities reply. The important point here, however, is that the replies are equally flawed insofar as each assumes (b).

  12. We don’t wish to legislate concerning what counts as metaphysics. Our point is simply that the fact of Mary’s cognitive gain provides, by itself, no better reason to accept dualism than is provided by Lingua’s cognitive gain.

  13. If information about what it’s like to see red depends on information about what red things look like, as it surely does, then won’t it trivially follow that the information gained about what it’s like to see red is robust if the information about what red things look like is robust? Yes, but we can avoid trivializing the issue by reframing it as follows: is the information about what it’s like to see red robust in a way that is not entirely parasitic on the robust gain involved in acquiring information about what red things look like? If the answer is no, then the information gained about what it’s like to see red is robust in a merely derivative sense. We think this would be enough to show that information about what red things look like is doing the important work in this debate, not what it’s like information. Of course, if what it’s like aspects associated with perceptual states just are the representational contents of such states (e.g., Dretske 1995; Tye 1995), it is necessary that gains concerning the former are robust if gains concerning the latter are robust. While we have some sympathy for this position, we don’t assume it here.

  14. Though we don’t think the case has been made for showing that what it’s like information gains are especially relevant to the Knowledge Argument, we think an “inner sense” account (e.g., Armstrong 1968; Lycan 1987), suitably adapted to accommodate the points we make about representation and knowledge, could provide a plausible model of such gains. Just as you can take a photo of a photo, so you can have a perception of a perception—this is what inner-sense is. The important points are this (1): on the inner sense account, we just have another case of an opportunity to apply a representational resource to a target not previously available, namely seeing red; (2) that resource is not (purely) propositional any more than vision is purely propositional. If it were purely propositional, Mary could just be told what it is like to see red. See Churchland and Churchland (1998) for an example of what an inner sense view developed along these lines might look like.

References

  • Alter, T. (2005). The knowledge argument against physicalism. Internet encyclopedia of philosophy. http://www.iep.utm.edu/know-arg/.

  • Armstrong, D. (1968). A materialist theory of the mind. London: Routledge.

    Google Scholar 

  • Brooks, R. (1991). Intelligence without representation. Artificial Intelligence, 47(1–3), 139–559.

    Article  Google Scholar 

  • Churchland, P., & Churchland, P. (1998). On the contrary. Cambridge, MA: MIT Press.

    Google Scholar 

  • Cummins, R. (2010). Representational specialization: The synthetic a priori revisited. In The World in the Head (pp. 194–209). Oxford: Oxford University Press.

  • Dretske, F. (1995). Naturalizing the mind. Cambridge, MA: The MIT Press.

    Google Scholar 

  • Fodor, J. (2007). The revenge of the given. In B. McLaughlin & J. Cohen (Eds.), Contemporary debates in philosophy of mind (pp. 105–116). Malden, MA: Blackwell.

    Google Scholar 

  • Haugeland, J. (1991). Representational genera. In W. Ramsey, S. Stich, & D. Rumelhart (Eds.), Philosophy and connectionist theory (pp. 61–89). Hillsdale, NJ: Lawrence Erlbaum Associates.

    Google Scholar 

  • Heck, R. (2007). Are there different kinds of content?” In B. McLaughlin & J. Cohen (Eds.), Contemporary debates in philosophy of mind (pp. 117–138). Malden, MA: Blackwell.

  • Horgan, T. (1984). Jackson on physical information and qualia. Philosophical Quarterly, 34, 147–152.

    Article  Google Scholar 

  • Jackson, F. (1982). Epiphenomenal qualia. Philosophical Quarterly, 32, 127–136.

    Article  Google Scholar 

  • Jackson, F. (1986). What Mary didn’t know. Journal of Philosophy, 83, 291–295.

    Article  Google Scholar 

  • King, J. (2007). The nature and structure of content. Oxford: Oxford University Press.

    Book  Google Scholar 

  • Lewis, D. (1983). Postscript to “mad pain and martian pain.” In Philosophical papers (Vol. 1). Oxford: Oxford University Press.

  • Lewis, D. (1988). What experience teaches. Proceedings of the Russellian Society, 13, 29–57.

    Google Scholar 

  • Loar, B. (1990). Phenomenal states. Philosophical Perspectives, 4, 81–108.

    Article  Google Scholar 

  • Lycan, W. G. (1987). Consciousness. Cambridge, MA: MIT Press.

    Google Scholar 

  • Lycan, W. G. (2003). Perspectival representation and the knowledge argument. In Q. Smith & A. Jokic (Eds.), Consciousness: New philosophical perspectives (pp. 384–395). Oxford: Oxford University Press.

  • Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 84(4), 435–450.

    Article  Google Scholar 

  • Nemirow, L. (1990). Physicalism and the cognitive role of acquaintance. In W. G. Lycan (Ed.), Mind and cognition (pp. 490–499). Cambridge, MA: Blackwell.

    Google Scholar 

  • Stanley, J. (2011). Knowing (how). Nous, 45(2), 207–238.

    Article  Google Scholar 

  • Stanley, J., & Williamson, T. (2001). Knowing how. The Journal of Philosophy, 98(8), 411–444.

    Article  Google Scholar 

  • Tye, M. (1986). The subjective qualities of experience. Mind, 95, 1–17.

    Article  Google Scholar 

  • Tye, M. (1995). Ten problems of consciousness. Cambridge, MA: The MIT Press.

    Google Scholar 

  • Tye, M. (2009). Consciousness revisisted: Materialism without phenomenal concepts. Cambridge, MA: MIT Press.

    Google Scholar 

  • Van Gelder, T. (1996). Dynamics and cognition. In J. Haugeland (Ed.), Mind design II. Cambridge, MA: MIT Press.

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Ian Harmon.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Cummins, R., Roth, M. & Harmon, I. Why it doesn’t matter to metaphysics what Mary learns. Philos Stud 167, 541–555 (2014). https://doi.org/10.1007/s11098-013-0110-1

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11098-013-0110-1

Keywords

Navigation