Egocentric thought and self-representation

CPER « Représentationnalisme et Fondements des sciences cognitives : nouveaux problèmes, nouveaux défis »

Programme :

- James Higginbotham ; Departments of Linguistics and Philosophy, University of Southern California : Rules of Use
- Joëlle Proust, Institut Jean Nicod, CNRS, Paris ; Max Planck Institute for Evolutionary, Anthropology, Leipzig

Three ways of thinking de se

- François Recanati, Institut Jean Nicod, CNRS, Paris

Implicit and explicit de se thoughts

- Jérôme Dokic, Institut Jean Nicod, Paris,EHESS, Paris Metacognition of the first person
- Philippe Schlenker, Institut Jean-Nicod, CNRS, Paris ; Department of Linguistics, University of California, Los Angeles

To be announced

- Alexandre Billon, Institut Jean-Nicod, CNRS, Paris/ CREA, Ecole Polytechnique, Paris ; Marie Guillot, Institut Jean-Nicod, Paris, EHESS, Paris : Could first-per­so­nal thoughts be self-refe­ren­tial ?

General Argument

The aim of the confe­rence is to shed light on the com­plex rela­tion­ship bet­ween the salient epis­te­mic pro­per­ties of ego­logy - or what we may call, after Lewis 1979, ’de se thoughts’ - and the puta­tive exis­tence and nature of an under­lying self-repre­sen­ta­tio­nal struc­ture.

Egocentric thoughts, like those expres­sed by a sub­ject when she decla­res « I wish a woman could win the elec­tion » or « I see a canary », can be conve­niently sin­gled out in terms of their spe­ci­fic epis­te­mic pro­per­ties. These thoughts are repu­ted to be invul­ne­ra­ble to cer­tain forms of error : they dis­play what Shoemaker has famously called « immu­nity to error through misi­den­ti­fi­ca­tion », as well as, in core cases, immu­nity to error through misas­crip­tion of pro­per­ties. Other remar­ka­ble epis­te­mic fea­tu­res of I-thoughts include their appa­rent ground­less­ness, or exemp­tion from the need of an epis­te­mic jus­ti­fi­ca­tion, the pri­vi­lege of first-per­so­nal access to such thoughts, and the sub­se­quent autho­rity atta­ched to first-per­so­nal reports on them.

Traditionally, the phi­lo­so­phy of mind has explo­red two anta­go­nist paths to account for these pecu­lia­ri­ties. The diver­gence bet­ween these two options rests on the stance taken up on the ques­tion as to whe­ther such de se contents involve self-repre­sen­ta­tion, in the two-fold sense of self-refe­rence (a repre­sen­ta­tion of the sub­ject in her own thoughts) and token-reflexi­vity (a repre­sen­ta­tion of occur­rent thoughts, in part or in tota­lity, within their own content).

The first option stems from a posi­tive answer to this ques­tion. According to a number of authors, who appeal to a plau­si­ble prin­ci­ple of iso­mor­phism bet­ween I-thoughts and first-per­so­nal utte­ran­ces, the well-docu­men­ted asym­me­try bet­ween first- and third-per­so­nal mental states is due to the pecu­liar way in which a sub­ject is repre­sen­ted in her own thoughts in the first case. In view of avoi­ding posi­ting a direct acquain­tance with one’s own self, or alter­na­ti­vely a sui gene­ris « mode of pre­sen­ta­tion » through which the thin­ker, and she only, could refer to her­self, a number of phi­lo­so­phers have recently pro­po­sed to cons­true type-ego­cen­tric thoughts as invol­ving a token-reflexive struc­ture, whe­reby the thin­ker is indi­rectly repre­sen­ted in her own (occur­ring) thoughts through the spe­ci­fic rela­tion she has with these token-thoughts, whose repre­sen­ta­tion contri­bute a cons­ti­tuent to their own content. Contemporary accounts per­tai­ning to this stra­tegy include J. Perry’s arti­cles from the 1990s, recent sug­ges­tions from M. Garcia-Carpintero, as well as J. Higginbotham’s 2003 arti­cle « Remembering, Imagining, and the First Person », which argues that a token-, or rather event-reflexive struc­ture, gover­ned by com­plex ana­pho­ric rela­tions, may under­lie and explain the pecu­liar seman­tic and epis­te­mic beha­vior of I-thoughts, at least in cases where they are embed­ded within the scope of higher-order atti­tu­des.

The second option, whose origin could be traced back to Wittgenstein 1953 and Anscombe 1975, consists in an endea­vor to explain the epis­te­mic pecu­lia­ri­ties of I-thoughts by the absence of self-repre­sen­ta­tion. As Perry 1993/2000 states it, one’s basic know­ledge of one­self can be consi­de­red as « intrin­si­cally sel­fless ». According to accounts that are sym­pa­the­tic to this view, the content of fun­da­men­tal I-thoughts, like those invol­ved in per­cep­tion and action, is a neu­tral pro­po­si­tion which concerns, ins­tead of being about, the thin­ker. The latter’s iden­tity being, so to speak, the default value of the point of origin defi­ning the orien­ta­tion of the per­cep­tual fra­me­work, it impo­ses such a strong archi­tec­tu­ral cons­traint on one’s mental life that repre­sen­ting one­self would be, in many cases, redun­dant and need­lessly costly in terms of cog­ni­tive invest­ment. On the other hand, if there is, in one’s most pri­mi­tive ego­cen­tric mental states at least, no self-repre­sen­ta­tion, hence no need to iden­tify one­self as such in order to pick one­self out cor­rectly, then there is a for­tiori no room for any misi­den­ti­fi­ca­tion. All the other, rela­ted epis­te­mic pri­vi­le­ges of the first-person pers­pec­tive can then be taken to follow from this basic assump­tion. Contemporary uses of this intui­tion include, in the phi­lo­so­phy of cog­ni­tion, contex­tua­list treat­ments of ego­cen­tri­city in the theo­re­ti­cal fra­me­work of embo­died and situa­ted cog­ni­tion, as recently advo­ca­ted, among others, by J. Dokic’s 2003 and for­th­co­ming papers. In the phi­lo­so­phy of mind and lan­guage, simi­lar consi­de­ra­tions could be argua­bly attri­bu­ted to F. Recanati in his for­th­co­ming book on pers­pec­ti­val thought. Setting his account of first-per­so­nal contents within the fra­me­work of a mode­rate seman­tic rela­ti­vism, Recanati ana­ly­zes the content of core cases of I-thoughts in the light of a dis­tinc­tion bet­ween two dis­tinct com­po­nents, a context-sen­si­tive pro­po­si­tio­nal struc­ture, which he calls the ’lekton’, and a wide context which pro­vi­des a clus­ter of points of eva­lua­tion in addi­tion to the world para­me­ter, inclu­ding a sub­ject, rela­tive to which the lekton is truth-eva­lua­ted.

Faced with the alter­na­tive of explai­ning the pecu­lia­ri­ties of de se thoughts by invo­king either the nature and struc­ture of an under­lying self-repre­sen­ta­tion, or the radi­cal absence of such a mecha­nism, one may want to weigh up the advan­ta­ges and dif­fi­culties invol­ved by each side of the dilemma.

Contextualist theo­ries are sup­por­ted by plau­si­ble consi­de­ra­tions of cog­ni­tive eco­nomy, since they dis­pense with posi­ting com­plex repre­sen­ta­tions behind ego­cen­tric mental states. However, one of the chal­len­ges to such theo­ries lies in the ques­tion as to whe­ther they can give a convin­cing account of the rela­tion­ship bet­ween basic, pri­mi­tive ego­cen­tric thoughts, like those invol­ved in per­cep­tion and action, and more com­plex I-thoughts to which we give verbal expres­sion using expli­cit mar­kers of the first person – e.g. first-per­so­nal reports of embed­ded atti­tu­des like (1) « I remem­ber wishing you were dead » - and which unques­tio­na­bly involve self-repre­sen­ta­tion. The latter share with basic ego­cen­tric mental states all the epis­te­mic pecu­lia­ri­ties that single out de se thoughts, but it is dif­fi­cult to see how the contex­tua­list expla­na­tion could be gene­ra­li­zed to their more com­plex seman­tic beha­vior. In par­ti­cu­lar, the ana­pho­ric struc­ture of an exam­ple like (1), in which the elided sub­ject of the gerun­dive clause has to be inter­pre­ted as refer­ring back to the same indi­vi­dual as the sub­ject of the embed­ding verb, seems to require the prior intro­duc­tion of a marker of dis­course cor­res­pon­ding to the said sub­ject, hence a self-repre­sen­ta­tio­nal under­lying struc­ture.

Self-repre­sen­ta­tio­nal theo­ries may seem, prima facie, more apt to give a uni­fied account of the epis­te­mic pecu­lia­ri­ties atta­ched to de se thoughts, but they can also be addres­sed a number of objec­tions. S. Predelli 2006 has thus poin­ted out that they are expo­sed to the risk of an uncontrol­led onto­lo­gi­cal pro­li­fe­ra­tion. Moreover, self-repre­sen­ta­tio­nal theo­ries must jus­tify the addi­tio­nal cog­ni­tive cost their com­plexity invol­ves, and demons­trate their com­pa­ti­bi­lity with the pre­sence of meta­cog­ni­tio­nal, hence ego­cen­tric mental states in ani­mals and yound chil­dren. As Proust for­th­co­ming convin­cin­gly shows, there are prin­ci­pled rea­sons to think that meta­cog­ni­tion, espe­cially in the most pri­mi­tive cases, does not rest on self-direc­ted meta­re­pre­sen­ta­tio­nal capa­ci­ties. Could self-repre­sen­ta­tio­nal theo­ries of de se thought account for the fact that even sys­tems lacking the capa­city to repre­sent them­sel­ves expli­citly show self-direc­ted meta­cog­ni­tio­nal apti­tu­des ?


- Castañeda, H.-N. 1966, « ‘He’ : A Study in the Logic of Self-Consciousness », Ratio 8, 130-157.
- Dokic, J. for­th­co­ming. « La der­nière per­sonne : sur la cri­ti­que de l’égologie cog­ni­tive dans Le com­plé­ment de sujet de Vincent Descombes ».
- Dokic, J. 2003a. "The Sense of Ownership : An Analogy bet­ween Sensation and Action", in N. Eilan et J. Roessler (eds), Agency and Self-Awareness. Issues in Philosophy and Psychology, Oxford, Clarendon Press, p. 321-344.
- Dokic, J. 2003b. "L’iden­ti­fi­ca­tion de soi, entre savoir-faire et intros­pec­tion", in S. Chauvier (ed.), Gareth Evans, Cahiers de phi­lo­so­phie de l’Université de Caen, n° 40-41 p 45-64.
- Harman, G. 2006,.« Self-Reflexive Thoughts », Philosophical Issues 16 (2006) : 334-45.
- Higginbotham, J. 2003. « Remembering, Imagining, and the First Person », in Alex Barber (ed.), Epistemology of Language. Oxford : Oxford University Press. pp. 496-533.
- Higginbotham, J. 1995, « Tensed Thoughts », Mind and Language vol. 10 n°3, 226-249.
- Kaplan, D. 1990. « Thoughts on Demonstratives », in P. Yourgrau (ed.), Demonstratives. OUP.
- Lewis, D. 1979, « Attitudes, De Dicto and De Se », Philosophical Review 88 : 513-543 ; repris dans Philosophical Papers, vol. I, New York : OUP, 1983.
- Millikan, R. 2001. « The Myth of Mental Indexicals », in Self-Reference and Self-Awareness, Brook, Andrew and Richard C. DeVidi (ed.), 2001, 163–177.
- Perry, J. 1997. « Reflexivity, Indexicality and Names », in Direct Reference, Indexicality and Proposition Attitudes, ed. W. Kunne, M. Anduschus and A. Newen, Stanford : CSLI, CUP, 1997.
- Perry, J. 1998. ”Indexicals, Contexts and Unarticulated Constituents”, in Proceedings of the 1995 CSLI-Amsterdam LLC Conference, Stanford : CSLI Publications.
- Perry, J. 1997. ”Myself and I” in M. Stamm ed., Philosophie in Syntetischer Absicht, Stuttgart : Klett-Cotta, 1998, pp. 83-103.
- Predelli, S. 2006. ”The Problem with Token-Reflexivity”, Synthese 148 (1) 5-29.
- Proust, J. for­th­co­ming. ”Metacognition and Metarepresentation : Is a self-direc­ted theory of the mind a pre­condi­tion for meta­cog­ni­tion ?”, Synthese, 2007.
- Proust, J. 2003. ”Thinking of Oneself as the Same”, Consciousness and Cognition, 2003, 12, 4, 495-509.
- Proust, J. 2000. ”Les condi­tions de la connais­sance de soi”, in Philosophiques 27, 1, 2000, 161-186.
- Recanati, F. for­th­co­ming. Perspectival Thought. A Plea for (Moderate) Relativism. Prepublication on the web­site www.ins­ti­tu­ni­
- Recanati, F. 2005. ”Loana dans le métro. Remarques sur l’indexi­ca­lité men­tale”, in S. Bourgeois-Gironde, ed., Les formes de l’indexi­ca­lité, Paris, Editions rue d’Ulm, pp. 19-34.
- Schlenker, P. for­th­co­ming. ”The Elimination of Self-Reference (Generalized Yablo-Series and the Theory of Truth)”, accep­ted for publi­ca­tion in the Journal of Philosophical Logic .
- Schlenker, P. 2003. « A Plea for Monsters », Linguistics and Philosophy 26 : 29-120.