Hostname: page-component-78c5997874-ndw9j Total loading time: 0 Render date: 2024-11-14T22:43:49.965Z Has data issue: false hasContentIssue false
  • English
  • Français

Fodor's Very Deep Thought

Published online by Cambridge University Press:  01 January 2020

Martin Montminy
Martin Montminy*
Affiliation:
University of Ottawa, Ottawa, ON, CanadaK1N 6N5
Get access Rights & Permissions [Opens in a new window]

Extract

Pooh rubbed his nose with his paw, and said that the Heffalump might be walking along, humming a little song, and looking up at the sky, wondering if it would rain, and so he wouldn't see the Very Deep Pit until he was half-way down, when it would be too late. (A.A. Milne, Winnie-the-Pooh)

Jerry Fodor is loath to have content be constituted, even in part, by inferential relations. This loathing, I will argue, gets him into trouble. In his latest book, Concepts, Fodor contrasts informational atomism, his view of concepts and their content, with inferential role theory. The latter, he argues, is almost certainly false. This strikes cognitive science at its core, he adds, since the major theories of concepts currently in vogue in cognitive science are all variants on inferential role theory.

Type
Research Article
Information
Canadian Journal of Philosophy , Volume 29 , Issue 4 , December 1999 , pp. 595 - 618
Copyright
Copyright © The Authors 1999

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Fodor, J. Concepts: Where Cognitive Science Went Wrong (Oxford: Clarendon Press 1998)CrossRefGoogle Scholar

2 In this paper, names of concepts will appear in capital letters, and names of properties will be written in italics.

3 Fodor holds that to have the complex concept BROWN COW, one has not only to exhibit a certain nomic relation with the property of being a brown cow, but also to have the concepts BROWN and COW. Let us note that informational semantics is compatible with the view that most lexical concepts are like the concept BROWN COW, i.e., they are complex, and in order to have them, one needs to have their constituents.

4 The cognitivist model would also be suitable for someone who holds a prototype theory or a theory theory of concepts. The latter, which can be seen as a version of inferential role theory, says that having a concept is having a mini-theory that describes facts about the things falling under that concept.

5 It should be noted that even though the metaphysical condition for content doesn't require a cognitivist account of concept possession, it could be the case that for many primitive concepts, our minds are such that as a matter of fact, a certain knowing-that is needed in order to sustain the condition. I'll come back to this point in Sections IV and V.

6 From now on, all references will be to Fodor, Concepts, unless otherwise specified.

7 Two points are worth making here. First, the concept DOORKNOB is assumed to be primitive. As Fodor points out (122, n.3), DOORKNOB isn't a good example, since it appears to be composed of DOOR and KNOB. However, we can ignore this problem for the sake of the discussion. Second, the reader should be aware that Fodor's use of ‘acquisition’ is ambiguous. He sometimes uses it interchangeably with ‘learning,’ opposing ‘acquired’ and ‘innate’ (see for example 126), and sometimes uses ‘concept acquisition’ for any mental process which eventuates in the possession of a concept by an organism, leaving open the possibility that certain concepts might be both acquired and innate. The latter use is in accordance with the terminology that he proposes in ‘The Present Status of the Innateness Controversy,’ in Representations (Cambridge, MA: MIT Press 1981) 257-316. Fortunately, this ambiguity doesn't really affect the issues under discussion.

8 A theory theory of concepts, for example, would easily dispose of the d/D problem, since encounters with x's give one the evidence one needs to form the mini-theory that is associated with X.

9 See Fodor, ‘The Present Status,’ 272–3.Google Scholar

10 Fodor (135-6) explains that a non-sensory appearance property is a property that is constituted by the mental states that the things having that property give rise to in us, while a sensory property is one that is constituted by the sensory states that the things that have it produce in us. A sensory property sness can be sensed and be the intentional object of an experience even though one hasn't got the concept S. On the other hand, the perceptual detection of a non-sensory appearance property xness is always inferential, and only a mind that has the concept X can have experiences whose intentional objects are x's. As we will see shortly, the fact that the perception of doorknobs (as doorknobs) is inferential will turn out to be highly problematic for Fodor.

11 Other passages (see for example 140, n. 16) suggest that for Fodor, stereotype is a psychological notion: stereotypic doorknobs would be doorknobs that have the properties that people generally associate with doorknobs, or, as the proponents of the prototype theory would put it, the prototypical features of doorknobs. The psychological and statistical notions of stereotype are different since prototypical features of doorknobs are not necessarily statistically reliable features of doorknobs.

12 Consider the following story: ‘Danny Hillis, the creator of the Connection Machine, once told me a story about some computer scientists who designed an electronic component for a military application (I think it was part of a guidance system in airplanes). Their prototype had two circuit boards, and the top one kept sagging, so, casting about for a quick fix, they spotted a brass doorknob in the lab which had just the right thickness. They took it off its door and jammed it into place between the two circuit boards on the prototype. Sometime later, one of these engineers was called in to look at a problem the military was having with the actual manufactured systems, and found to his amazement that between the circuit boards in each unit was a very precisely milled brass duplicate of the original doorknob’ (Dennett, D. Darwin's Dangerous Idea [New York: Touchstone 1995], 199, n.5)Google Scholar.

13 Hence, readers who are not convinced by my telephone example should consider a similar story involving a computer instead. Let me remind these readers that we sometimes talk of inventors who tried, but failed, to create the first artefact of a certain type (say, a television or an aeroplane).

14 I should once again specify that this is what I have been assuming all along. The difference between the previous reading and this one is that on the former, being nomologically locked to doorknobs guarantees, as a matter of fact, being able to recognize doorknobs.

15 In this Section, for the sake of the argument, I'm assuming Fodor's account of concept possession. I'm also assuming (for now) that, for us, responding selectively to doorknobs amounts to being able to recognize doorknobs when confronted with them.

16 I should add that Fodor's view is probably even less plausible with respect to superordinate categories like furniture, vehicle and weapon, whose members show great perceptual variation. Certain types of weapons (missiles, chemical weapons) are very different from stereotypic ones (knives, guns), and it is extremely unlikely that someone could, without specific instructions, come to group them under the same category.

17 See for example Landau, B. ‘Object Shape, Object Name, and Object Kind: Representation and Development,’ in The Psychology of Learning and Motivation 31, Medin, D.L. ed. (San Diego: Academic Press 1994) 253–304Google Scholar; Landau, B. Smith, L. and Jones, S. ‘Object Shape, Object Function, and Object Name,’ Journal of Memory and Language 38 (1998): 1–27CrossRefGoogle Scholar; and Merriman, W.E. Scott, P.D. and Marazita, J. ‘An Appearance-Function Shift in Children's Object Naming,’ Journal of Child Language 20 (1993) 101–18CrossRefGoogle ScholarPubMed. However, see Nelson, D.G. Kemler ‘Principle-Based Inferences in Young Children's Categorization: Revisiting the Impact of Function on the Naming of Artifacts,’ Cognitive Development 10 (1995) 347–80CrossRefGoogle Scholar, for different results.

18 I'm told that one punt notes were withdrawn from circulation a few years ago, and were replaced by coins.

19 This, of course, doesn't mean that we are exempt from explaining the fact that experiences with doorknobs play a crucial role in the acquisition of DOORKNOB.

20 Malt, B.C. Sloman, S.A. Gennari, S. Shi, M. and Wang, Y. ‘Knowing versus Naming: Similarity and the Linguistic Categorization of Artifacts,’ Journal of Memory and Language 40 (1999) 230–62CrossRefGoogle Scholar

21 See Kronenfeld, D.B. Armstrong, J.D. and Wilmoth, S. ‘Exploring the Internal Structure of Linguistic Categories: An Extensionist Semantic View,’ in Directions in Cognitive Anthropology, Dougherty, J.W.D. ed. (Urbana, IL: University of Illinois Press 1985) 91–110Google Scholar, for cases involving English, Hebrew, and Japanese categories.

22 The same point would apply to a variant of this view that claims that ‘striking us as being a doorknob’ means ‘striking us, when told by an informed observer, as being a doorknob.’ This, by the way, should also be taken as a serious problem for the view defended by Fodor in The Elm and the Expert (Cambridge, MA: MIT Press 1994). In that book, Fodor claims that one could have the concept ELM if one could detect elms by using an expert. The problem is that on this view, concepts are just too cheap: knowing a person who can refer one to experts and being able to utter ‘What is this?’ would be sufficient to have any concept possessed by these experts.

23 In this Section, to avoid confusion, I will use the words ‘symbol’ to designate a mental representation that is individuated syntactically and ‘concept’ for a type that is individuated in terms of its content.

24 In his ‘A Theory of Content, II: The Theory,’ in A Theory of Content and Other Essays (Cambridge, MA: MIT Press 1989) 89-136, Fodor explains that if the nomic relation between y's and tokens of X is asymmetrically dependent on the nomic relation between x's and tokens of X, then, ceteris paribus, breaking the relation between x's and tokens of X would break the relation between y's and tokens of X, but not the other way around.

25 I take it that this problem would persist even if Fodor opted for the purely nomic view.

26 Let's assume that if Bonnie were presented with a watch battery and asked ‘What is it?’ she would answer ‘I don't know,’ and if asked ‘Is this a battery?’ she would answer ‘No.'

27 See Fodor, ‘Reply to Stalnaker,’ in Meaning In Mind, ed. Loewer, B. and Rey, G. (Oxford: Blackwell 1991), 302–303Google Scholar, for similar remarks in the context of a thought experiment that bears some similarities to the one under discussion.

28 See for instance Fodor, ‘A Theory of Content,’ 96Google Scholar.

29 In his ‘A Theory of Content,’ 107-108, Fodor makes a similar remark, though in dealing with a slightly different issue.

30 Note that I'm still assuming, for simplicity's sake, that non-x's don't cause tokenings of X.

31 Note that Fodor would not run into this problem if he held that Bonnie's symbol BA ITERY means whatever it covaries with (say, under normal viewing conditions and careful examination) and that being a doorknob consists in being the kind of thing that covaries with Bonnie's symbol BAITERY. But then, instead of being circular, his view would be false, since it is not the case that all instances of the property of being a battery covary with tokens of Bonnie's BAITERY.

32 I am grateful to Paulo Faria, Sherri Irvin, Ernie Lepore, Brian Loar, Barry Loewer, Brian McLaughlin, Mark Moyer, John Sarnecki, Jonathan Schaffer, Susan Schneider, Barry Ward, Jonathan Weinberg and Julie Yoo for useful discussions about the issues discussed in this paper. I would also like to thank the participants in the symposium ‘Fodor's anatomic Bêtes Noires,’ which took place at the CPA meeting in St. Catharines in 1996 and in which an old version of this paper was read. Finally, I owe thanks to the anonymous referees for this journal for useful suggestions.