Error through misidentification: Some varieties

Journal of Philosophy

Volumn 103, Issue 8, 2006, Pages 403-425

  Coliva, Annalisa ^a  

^a UNIVERSITY OF MODENA AND REGGIO EMILIA   (Italy)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 60950037321     PISSN: 0022362X     EISSN: 19398549     Source Type: Journal    
DOI: 10.5840/jphil2006103824     Document Type: Article

References (37)

1. The label "(immunity to) error through misidentification" wasfirst introduced in the philosophical literature by Sydney Shoemaker, "Self-reference and Self-awareness, " this JOURNAL, LXV, 19 (October 3, 1968): 555-67,
2. reprinted in Quassim Cassam, ed., Self-Knowledge (New York: Oxford, 1994), pp. 19-48
3. (page reference to the latter) in connection with Wittgenstein'sdistinction, between uses of "I" as subject and as object - The Blue Book (Oxford: Blackwell, 1958), pp. 66-67.
4. The notion of quasi-memory is characterized by Gareth Evans as follows:"A subject q-remembers an event e if and only if (i) he has an apparentmemory of such an event, and (ii) that apparent memory in fact embodiesinformation deriving from the perception of that event by a person who is notnecessarily himself" - The Varieties of Reference (New York: Oxford, 1982), pp. 247-48. It is assumed that one could have memory impressions deriving fromsomeone else's past if one had the relevant brain cells transplanted into one'sbrain. Since quasi-memories and ordinary ones are phenomenologicallyindistinguishable, their phenomenological content is best expressed by locutionslike "I seem to remember being in Scotland five years ago, " or as"I have an (apparent) memory as of being in Scotland five years ago."It is then a further issue - an issue which I will not discuss in this paper -whether quasi-memories and ordinary ones should be taken to be intrinsically thesame kind of mental state whose difference resides merely in their differentcausal origins, or whether - in a "disjunctivist" spirit - they shouldbe taken, because of their different causal origin, as distinct and mutuallyexclusive mental states. For the purposes of this paper, we can be neutral onthis issue. For once it is allowed that memories and quasi-memories aresubjectively indistinguishable, it should also be allowed that the judgmentsbased on either of them will have the same structure of rational grounds. Therewill be more about the notion of grounds in the sequel. But to help fix ideas, Itake it that both in cases of memory-based self-ascriptions and in cases ofquasi-memory-based ones, subjects requested to justify them would offer thefollowing: "I seem to remember being F. Therefore, I was F."
5. Shoemaker, "Persons and Their Pasts, " in his Identity, Cause, and Mind (New York: Cambridge, 1970), pp. 19-18, here see pp. 23-24, 46.
6. Evans, pp. 235-48, here pp. 241-45.
7. See Parfit, Reasons and Persons (New York: Oxford, 1984), pp. 220, 222;
8. and McDowell, "Reductionism and the First Person, " in Jonathan Dancy, ed., Reading Parfit (Malden, MA: Blackwell, 1997), pp. 230-50. Accordingto Parfit, memories, as such, have an identity-neutral content and can give oneinformation about the life of a specific human being only through theidentification of oneself with the person whose life is responsible for thememories one is having. Such an identification might be wrong but, moreimportantly, it could be made by different future selves. This would be the caseif, with an appropriate brain transplant, one's memories could survive in two(or more) different brains, belonging to two (or more) different human beings.Psychological continuity would then be preserved at the expense of personalidentity.
9. The driving thought behind McDowell's position is that memories haveimmediate and irreducible first-person content. According to him, this is sobecause memories have the specific, biologically-selected function to retain, for each specific human being, information about his own past, and make itimmediately available to him in the first-person mode of thought. Memories, therefore, are situated in the life of a specific human being and this latterfact, contrary to what Locke and Parfit seem to maintain, cannot be abstractedaway in giving an account of persons and their identity. For an illuminatingdiscussion of Parfit's and McDowell's positions, see Carol Rovane, "Personal Identity: Ethical Not Metaphysical, " in Graham F. Mac Donaldand Cynthia Mac Donald, eds., McDowell and His Critics (Malden, MA: Blackwell, 2006), forthcoming.
10. Peacocke, Being Known (New York: Oxford, 1999), p. 263. The task oflocating the exact source of such a challenge in the case of reflections on theself - whether in what Peacocke calls the phenomenon of " representationalindependence" or in something else - falls outside the scope of this paper.
11. Evans, McDowell, Peacocke, and Crispin Wright in his "Self- knowledge: The Wittgensteinian Legacy, " in Wright, Barry Smith, and Cynthia McDonald, eds., Knowing Our Own Minds (New York: Oxford, 1998), pp. 13-45, justto mention the most prominent figures, have all made these points. Thearguments are usually that immunity to error through misidentification is commonto both some uses of "I" and, for example, to perceptualdemonstrative judgments, for which there is no temptation to think they may notrefer; and that immunity is a function of the absence of any identificationcomponent in the rational grounds for a given first person judgment and not ofthe fact that either there is no self to be identified in the first place, orthat the identification is secured by the special kind of entity presented toone for recognition.
12. With respect to the former issue, we will see in the following thatmatters are far from being settled. With respect to the latter, nowadays nobodywould agree with Wittgenstein's position that all and only psychologicalself-ascriptions are immune to error though misidentification (but for adifferent interpretation of Wittgenstein's position, see Brian Garrett, "Wittgenstein and the First Person, " Australasian Journal of Philosophy, LXXIII (1995): 347-55). Since Shoemaker and Evans, it has becomecommonplace to hold that also nonpsychological self-ascriptions, such as thosebased on memory, perception, and somatic proprioception, as well asperception-based demonstrative judgments are immune to error throughmisidentification. Moreover, since Evans (see pp. 219-20), it has been widelyheld that inferential psychological self-ascriptions are liable to error throughmisidentification.
13. Finally, John Campbell by appealing to the phenomenon of "thoughtinsertion", whereby schizophrenic patients claim that they have a giventhought which is not theirs, has gone so far as to argue that alsononinferential ones can be liable to such an error - see "Schizophrenia:The Space of Reasons and Thinking as a Motor Process, " The Monist, LXXXII(1999): 609-25.
14. (For a discussion critical of Campbell's position, see my "Thought Insertion and Immunity to Error through Misidentification, " and "What There Really Is to Our Notion of the Ownership of a Thought, " Philosophy, Psychology and Psychiatry, IX, 1 (2002): 27-34, 41-46;
15. and Campbell's reply "The Ownership of Thought, " Philosophy, Psychology and Psychiatry, IX, 1 (2002): 35-39.)
16. Pryor, "Immunity to Error through Misidentification, "Philosophical Topics, XXVI, 1 (1998): 271-304.
17. See Evans, pp. 235-48;
18. McDowell, pp. 239-41;
19. Wright, p. 20.
20. and McDowell. It should be noticed, however, that while Shoemaker himselfheld the view that memory-based self-ascriptions are grounded in first-personalmemory impressions (Shoemaker, Self-Knowledge and Self-Identity (Ithaca:Cornell, 1963), pp. 33-34)
21. and yet, in the abnormal conditions envisaged, may be liable to errorthrough misidentification (Shoemaker, "Introspection and the Self, " inhis The First Person Perspective and Other Essays (New York: Cambridge, 1986), pp. 3-24 , see p. 15), it is in fact Parfit (pp. 220, 222), who has contributedto spreading the view that memory-based self-ascriptions are always based onidentity-neutral memory impressions and arrived at by holding in place therelevant identity beliefs.
22. Evans, chapters 6, 7;
23. and Peacocke, Sense and Content (New York: Oxford, 1983), pp. 143-14.
24. Pryor, p. 281.
25. Pryor, p. 283.
26. See, for instance, Evans, chapters 6, 7; Peacocke, Sense and Content, pp.143-44.
27. The point is not taken universally. In my "Error through Misidentification, the Split between Speaker's and Semantic Reference and the Real Guarantee, " this JOURNAL, C, 8 (August 2003): 216-31,
28. discuss extensively those interpretations of error throughmisidentification which take it to be a referential mistake, such as Elisabeth Anscombe, "The First Person, " in Simon Guttenplan, ed., Mind and Language: Wolfson College Lectures 1974 (New York: Oxford, 1975), pp. 45-64;
29. Andrea Christofidou, "First Person: The Demand for Identification-free Self-Reference, " this JOURNAL, XCII, 4 (April 1995):223-34;
30. and Rovane, The Bounds of Agency (Princeton: University Press, 1998). Onthese latter views, when that error occurs, in using "a" one would infact refer to the person visually presented to one, rather than to its semanticreferent. So, in the Aunt Miriam example, when I am in the car park and Imistake a red-hatted woman approaching for Aunt Miriam, by saying "Aunt Miriam is wearing a red hat" I would be (speaker-) referring to the womanwho is actually approaching and not to my aunt. Since they do not find thisoutcome plausible in the case of the use of "I, " they conclude thatself-ascriptions are (almost) always immune to error through misidentification.In the paper I put forward the view that such an interpretation might in fact bebased on a conflation between (immunity to) error through misidentification, the (im)possibility of a split between speaker's and semantic reference and what I label the "real guarantee": namely, the fact that any competent useof "I" is such that the speaker knows which object is its semanticreferent, namely, oneself.
31. A somewhat similar point is made by Peacocke, Sense and Content, pp.143-44.
32. and his "(Anti)-sceptics, Simple and Subtle: Moore and McDowell, " Philosophy and Phenomenological Research, LXV (2002): 330-48;
33. and Pryor's "The Skeptic and the Dogmatist, " Noûs, XXIV(2000): 517-49 - here pp. 537-38, respectively. So, in my view, althoughthey do not explicitly distinguish between the subject's own grounds for a givenjudgment and the background presuppositions of it, the best way of interpretingtheir dispute is to read it as bearing on whether background presuppositionsneed (or need not) be warranted.
34. I take it that Evans and McDowell do both at least implicitly challengethe de facto side when they deny that a mere difference in metaphysicalconditions would make memory-based self-ascriptions liable to error throughmisidentification. Campbell, "Schizophrenia, the Space of Reasons and Thinking as a Motor Process, " and Rovane, The Bounds of Agency, pp. 220-32, openly challenge the logical side of it. While I have voiced my perplexitiesabout Campbell's and Rovane's proposals elsewhere (see my "Thought Insertion and Immunity to Error through Misidentification, " "What There Really Is to Our Own Notion of the Ownership of a Thought, " and"Immunity to Error through Misidentification, the Split between Speaker'sand Semantic Reference, and the Real Guarantee"), in the following I willargue that Evans and McDowell can, at most, claim for a revision of that notionin a more epistemologically-oriented direction, whose proper explanationcrucially depends on allowing for the existence of identification components aspart of a judgment's background presuppositions, contrary to Evans's officialposition (cf. footnote 11).
35. This example is in Wright, "Self-knowledge: The Wittgensteinian Legacy, " p. 20.
36. and Being Known, p. 275;
37. also Rovane, The Bounds of Agency, pp. 220-32.