Abstract
Noam Chomsky and Frances Egan argue that David Marr’s computational theory of vision is not intentional, claiming that the formal scientific theory does not include description of visual content. They also argue that the theory is internalist in the sense of not describing things physically external to the perceiver. They argue that these claims hold for computational theories of vision in general. Beyond theories of vision, they argue that representational content does not figure as a topic within formal computational theories in cognitive science. I demonstrate that Chomsky’s and Egan’s claims about Marr’s theory are false. Marr’s computational theory contains a mathematical theory of visual content, based on empirical psychophysical evidence. It also contains mathematical descriptions of distal physical surfaces and objects, and of their optic projection to the perceiver. Much computational research on vision contains these types of intentional and externalist components within the formal, mathematical, theories. Chomsky’s and Egan’s claims demonstrate inadequate study and understanding of Marr’s work and other research in this area. Computational theories of vision, by containing empirically based mathematical theories of visual content, to this extent present naturalizations of semantics.
Similar content being viewed by others
Notes
The terms ‘externalism’ and ‘internalism’ have different senses in the philosophical literature, involving different issues. In one sense a theory of vision is externalist just in case it contains descriptions of things physically external to the perceiver, otherwise it is internalist. In the present article I will be using these terms primarily in this sense, addressing the question of whether Marr’s and other computational theories of vision contain descriptions of external physical things. In another sense of the terms, a theory is externalist if and only if it individuates visual content in terms of things external to the perceiver, internalist if and only if it individuates visual content in terms only of things internal to the perceiver, and neither externalist nor internalist if and only if it does neither. In the present essay, after introducing the issue of whether computational theories of vision are externalist in this sense, which has motivated much of the literature, I return to it only at the very end of the article for a brief remark. I discuss this issue at length in: Silverberg (in preparation).
Shapiro’s description of ‘chief’ and ‘service’ tasks presented at the topmost level (Shapiro, 1997, p. 134) should have alerted him that Egan might claim that descriptions of internal processes are presented at this level of the theory.
Egan’s statement that Marr says that primitives of the raw primal sketch, like ‘edges’, “considered individually” lack physical reality is due to a further misunderstanding of his theory on her part. She is confusing Marr’s use of the spatial coincidence assumption, according to which ‘edges’ are located at coinciding zero-crossing segments that appear in at least two ∇2 G channels of nearest sizes, with his discussion of grouping processes by which the full primal sketch is formed from the raw primal sketch. This can be seen in her attribution to Marr of the statement, “[These primitives are] descriptions of an image at different scales” (Egan, 1991, p. 198). The different scales Marr is discussing in the source from which Egan constructs her quotation are not the different sizes of channels. He is, rather, talking about the formation of the full primal sketch by grouping processes, which results in different scales in the analysis of the retinal image. The sentence from which Egan forms her quotation, which is part of the annotation to a diagram, reads, “A diagrammatic representation of the descriptions of an image at different scales which together constitute the primal sketch” (Marr, 1982, p. 53). Marr’s terms “primal sketch” and “full primal sketch” are equivalent.
An anonymous referee for this journal informed me of Losonsky (1993). Losonsky cites Marr’s discussion of the computation of distance from disparity, and notes that it contains mathematical descriptions of distal points, and ‘depth’, which latter Losonsky notes concerns “the subjective impression of distance”. In my earlier manuscript, which the referee saw, I had presented detailed discussion of Marr’s mathematical descriptions of external physical phenomena and of visual content, especially of the 2.5-D sketch and of 3-D representations.
I disagree with some significant details of Losonsky’s interpretation of the passage by Marr which he cites, and with his claims of the import of this passage for the question of whether Marr’s theory of visual content is externalist in the sense that Burge and Egan discuss. I present my arguments on these in Silverberg (in preparation).
Patricia Kitcher has claimed that Marr’s computational theory contains descriptions of things external to the perceiver, and of visual content (Kitcher, 1988). Egan has eluded Kitcher’s claims, due to Kitcher’s not documenting them (Egan, 1992). Kitcher says that “a Computational theory will include mathematical results, in particular, theorems about the geometry of three-dimensional space (Marr, 1982, p. 188)” Kitcher (1988, p. 13), but Egan could deal with this the way she deals with Ullman’s rigidity assumption.
This is not a criticism of Kitcher’s paper, since her article appeared before Egan’s and Chomsky’s writings on the subject, and does not address the types of claims about Marr’s theory that they present.
In Silverberg (in preparation), I also discuss in greater detail the topics of the present article. In Silverberg (1998), I argued against Chomsky’s general claims that cognitive science is internalist and not intentional.
References
Biederman, I. (1995). Visual object recognition. In S. M. Kosslyn & D. N. Osherson (Eds.), Visual cognition (pp. 121–165). Cambridge, MA: The MIT Press.
Burge, T. (1979). Individualism and the Mental. Midwest Studies in Philosophy, 4, 73–122.
Burge, T. (1986). Individualism and psychology. Philosophical Review, XCV, 3–45.
Burge, T. (1988). Reply: Authoritative self-knowledge and perceptual individualism. In R. H. Grimm & D. D. Merrill (Eds.), Contents of thought (pp. 86–98, 195–196). Tucson, AZ: The University of Arizona Press.
Butler, K. (1998). Internal affairs: Making room for psychosemantic internalism. Dordrecht: Kluwer Academic Publishers.
Chomsky, N. (1992). ‘Explaining language use’, reprinted in New horizons in the study of language and mind (2000) (pp. 19–45, 195–197). Cambridge, UK: Cambridge University Press.
Chomsky, N. (1995). Language and nature. Mind, 104, 1–61.
Chomsky, N. (2003). Reply to Egan. In L. M. Antony & N. Hornstein (Eds.), Chomsky and his critics (pp. 268–274). Oxford, UK: Blackwell Publishing.
Cooper, L. A., & Lang, J. M. (1996). Imagery and visual-spatial representations. In E. L. Bjork & R. A. Bjork (Eds.), Memory (pp. 129–164). San Diego, CA: Academic Press.
Davies, M. (1991). Individualism and perceptual content. Mind, C, 461–484.
Egan, F. (1991). Must psychology be individualistic? Philosophical Review, C, 179–203.
Egan, F. (1992). Individualism, computation, and perceptual content. Mind, CI, 443–459.
Egan, F. (1994). Individualism and vision theory. Analysis, 54, 258–264.
Egan, F. (1995). Computation and content. Philosophical Review, CIV, 181–203.
Egan, F. (1996). Intentionality and the theory of vision. In K. Atkins (Ed.), Perception (pp. 232–247). New York, NY: Oxford University Press.
Egan, F. (1999). In defense of narrow mindedness. Mind and Language, 14, 177–194.
Egan, F. (2003). Naturalistic inquiry: Where does mental representation fit in? In L. M. Antony & N. Hornstein (Eds.), Chomsky and his critics (pp. 89–104). Oxford, UK: Blackwell Publishing.
Faugeras, O. (1993). Three-dimensional computer vision, a geometric viewpoint. Cambridge, MA: MIT Press.
Kitcher, P. (1988). Marr’s computational theory of vision. Philosophy of Science, 55, 1–24.
Kripke, S. (1980). Naming and necessity. Cambridge, MA: Harvard University Press.
Losonsky, M. (1993). The cognitive unity of external and internal states. In R. Casati & G. White (Eds.), Philosophy and the cognitive sciences: Papers of the 16th international Wittgenstein conference (pp. 313–318). Kirchberg am Wechsel: Austrian Ludwig Wittgenstein Society.
Low, A. (1991). Introductory computer vision and image processing. London, UK: McGraw-Hill.
Marr, D. (1977). Analysis of occluding contour. Proceedings of the Royal Society of London, B, 197, 441–475.
Marr, D. (1979). Representing visual information. In A. R. Hanson & E. M. Riseman (Eds.), Computer vision systems (pp. 61–80). New York, NY: Academic Press.
Marr, D. (1982). Vision. New York, NY: W.H. Freeman and Company.
Marr, D., & Hildreth, E. (1980). Theory of edge detection. Proceedings of the Royal Society of London B, 207, 187–217.
Marr, D., & Poggio, T. (1977). From understanding computation to understanding neural circuitry. Neurosciences Research Progress Bulletin, 15, 470–488.
Marr, D., Poggio, T., & Hildreth, E. (1980). Smallest channel in early human vision. Journal of the Optical Society of America, 70, 868–870.
Marr, D., Ullman, S., & Poggio, T. (1979). Bandpass channels, zero-crossings, and early visual information processing. Journal of the Optical Society of America, 69, 914–916.
Nalwa, V. S. (1993). A guided tour of computer vision. Reading, UK: Addison-Wesley.
Peacocke, C. (1995). Content, computation and externalism. Philosophical Issues, 6, 227–264. (Also in Mind and Language, 9, (1994), pp. 303–335).
Peacocke, C. (1999). Computation as involving content: A response to Egan. Mind and Language, 14, 195–202.
Putnam, H. (1975). The meaning of ‘meaning’. In Mind, language, and reality (pp. 215–271). Cambridge, UK: Cambridge University Press.
Shapiro, L. A. (1997). A clearer vision. Philosophy of Science, 64, 131–153.
Silverberg, A. (1998). Semantic externalism: A response to Chomsky. ProtoSociology, 11, 216–244.
Silverberg, A. (in preparation) Externalism and Internalism in the Science of Vision (book manuscript).
Stevens, K. A. (1981). The interpretation of surface contours. Artificial Intelligence, 17, 47–73.
Ullman, S. (1979). The interpretation of visual motion. Cambridge, MA: MIT Press.
Acknowledgements
I thank James Fetzer, James Moor, and an anonymous referee for this journal for comments on longer manuscripts from which the material presented in this article is drawn.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Silverberg, A. Chomsky and Egan on computational theories of vision. Minds & Machines 16, 495–524 (2006). https://doi.org/10.1007/s11023-006-9050-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11023-006-9050-2