About Oneself: De Se Thought and Communication – GARCÍA-CARPINTERO (M)

GARCÍA-CARPINTERO, Manuel; TORRE, Stephan (eds.). About Oneself: De Se Thought and Communication. Oxford: Oxford University Press, 2016, 368p. Resenha de: VALENTE, Matheus. Manuscrito, Campinas, v.40 n.2 Apr./June 2017.

About Oneself: De Se Thought and Communication is a recently published collection of eleven papers on de se thought, i.e. thoughts about oneself as oneself, and its implications for a theory of communication. Edited by Manuel García-Carpintero and Stephan Torre, this volume contains contributions from many distinguished experts and presents the state-of-the-art discussions on this important field.

The issue of de se thought dates, at least, as far back as the late 1960s1, when some noticed that, in order to characterize the information encoded in a de se thought, one must reject some traditional assumptions about propositional attitudes. Take, for example, Perry’s famous story of when he tried to find the person making a mess in the supermarket by following a trail of spilled sugar only to subsequently realize that he himself was the culprit. It seems that, whatever the content of Perry’s epiphany is – the one he could express by uttering “I am making a mess” – it resists characterization by traditional conceptions of propositions (i.e. as sets of possible worlds or structured Russellian propositions).

Most of the chapters in this volume are primarily concerned to defend one or another theory of de se thought and to examine their implications for an account of communicative success. Due to shortage of space, this review will focus on the papers that are directly involved with these questions. Our main objective will be weaving these papers together such as to make explicit their points of agreement and disagreement.

What, after all, is so special about de se content? This is not an easy question to answer. Indeed, some philosophers even got as far as declaring that there is no real problem of de se content over and above the typical issues that singular thought (in general) brings about, such as Frege’s Puzzle2. Dilip Ninan’s paper – What is the Problem of De Se Attitudes? – intents to reach a verdict about the extent to which some content is essentially indexical. According to Ninan, only de se attitudes (as opposed to other de re attitudes) are such that they give rise to cases where two subjects agree about all the objective properties of a situation while still diverging in their behavior. This happens, for example, in Perry’s bear scenario (1979), in which I am being chased by a bear while you are watching from a safe distance. Even though we may agree on our objective beliefs about the situation (e.g. that I am being chased by a bear and you are not) as well as in our desires (e.g. we both desire that I don’t get killed), it is still true that we will be motivated to behave differently (e.g. I will curl up into a ball and you will run to get help). That is, only when de se attitudes are concerned, there can be full belief-desire agreement concomitant with divergent behavior. That conclusion leads to an outright denial of Explanation:

(Explanation) If a subject’s behaving in a certain way is explained by his set of belief-desire pairs, then any other subject possessing the same set of belief-desire pairs will be disposed to behave identically.

Ninan’s paper help us see how most theorists of de se thought are trying to hold onto Explanation in the face of conflicting evidences by rejecting one of the following two theses:

(Absoluteness) The contents of attitudes are absolute, i.e. contents do not vary in truth-value across individuals or times.

(Publicity) The contents of attitudes are public or shareable, i.e. if an agent x can entertain a content p, then so can any other agent y. (p. 111)

If Explanation and Publicity are true, Ninan claims, one must agree that someone could possess one of my de se beliefs, e.g. that I am being chased by a bear. Per Explanation, we would then be disposed to behave identically – we would both be disposed to, say, curl up into a ball. However, since it is possible that our beliefs diverge in truth-value – one of us could just be overly paranoid – we would then have to dispose of Absoluteness. This is the path (Lewis, 1979) famously took by defending that the content of de se attitudes are properties (or, equivalently, centred propositions), entities which vary in truth-value relative to non-worldly parameters.

Clas Weber – in Being at the Centre: Self-location in Thought and Language – explicitly sets out to defend a Lewisian theory of propositional attitudes and to show how the communication of de se thoughts would be possible inside that framework. In order to do this, Weber advances the Transform-and-Recenter Model of communication, according to which there is no single content being replicated from speaker to hearer in successful instances of de se communication. Since assertions present their contents as being essentially from a particular perspective, Weber claims, we must perform a series of transformations on other people’s asserted contents, so that a piece of information that was originally presented from the speaker’s perspective becomes a piece of information relative to the hearer’s.

In his contribution, De Se Communication: Centered or Uncentered?, Peter Pagin argues that no Lewisian theory positing centered contents can give a suitable account of de se communication, since “the connection between a [centered] content and its thinker is not representational” (p. 275). In a Lewisian theory, the content of de se thoughts are impersonal properties, e.g. the property of being chased by a bear. Only when a thinker believes that property, i.e. self-ascribes it, a connection between her and the content of her thought is drawn. Pagin argues that this entails that a thinker may never merely entertain a thought, e.g. that she is being chased by a bear; what she entertains is just the property of being chased by a bear, not that she is being chased by a bear. After considering various recent incarnations of the Lewisian theory, Pagin goes on to advance his own Fregean-inspired view, maintaining Absoluteness in exchange for Publicity, and thus, giving rise to the famous issue of ‘limited accessibility’. Pagin claims the denial of Publicity should be seen as non-problematic since (i) it is an independently motivated thesis that subjects rarely have the same conception of the concepts they employ and that (ii) this should not harm communicative success in the least.

In opposition to both Lewisian and Fregean theories, some philosophers argue, following (Perry, 1979), that sameness of behavior between A and B should not be explained by them believing a common content, but by them believing possibly distinct contents under the same guise. François Recanati and Manuel García-Carpintero both identify themselves as developing their own Perrian accounts of de se thought. These two authors argue that, in order to clarify these issues, one has to take into account two distinct semantic levels about which one’s attitudes and assertions are accountable: the presuppositional content that one’s representations carry and the content that they properly expresses. The former accounts for the cognitive significance of a thought, whereas the latter accounts for our intuitions on (dis)agreement and sameness of subject matter. Recanati, in Indexical Thought: The Communication Problem, presents his version of a Perrian two-factor theory by means of his independently motivated framework of Mental Files, i.e. psychological guises by means of which we produce and retain singular thoughts. Recanati considers multiple accounts of de se communication before settling for an improved variant of Weber’s Transform-and-Recenter model. Differently from Weber, who frames his model by means of metarepresentations, Recanati fleshes it out in terms of Mental Files. For this reason, he claims to be able to overcome the aforementioned objections from Pagin. One interesting consequence of Recanati’s account of de se communication is that the idea of the thought expressed by an utterance, over and above the thoughts of the speaker and her interlocutor, comes out as otiose. As the author emphasizes, as long as we have our hands on the thought of the speaker, the thought of the hearer, and a suitable relation of coordination between them, there remains no theoretical role to be played by a neutral notion of the thought expressed by the utterance.

Manuel García-Carpintero’s paper, Token-Reflexive Presuppositions and the De Se, agrees with Recanati’s in that both argue that the cognitive state of a subject undergoing a de se thought must be characterized not only by (i) that subject (mentally) asserting a certain content but also by (ii) her thought triggering certain reference-fixing presuppositions. Thus, even if the content of de se attitudes are to be fleshed out as familiar singular propositions, “when I judge I am hungry I [also] presuppose that the person of whom I am predicating being hungry is the thinker of this very judgement” (p. 191). García-Carpintero points out the importance of distinguishing the attitude a subject holds towards an asserted content from the attitude held towards a content that she presupposes in virtue of having made that assertion. The author claims that the attitude towards a presuppositional content cannot be that of belief, since anyone else could believe a certain presuppositional content without being motivated to act in the special way de se attitudes motivate us to act. Because of these reasons, García-Carpintero notes, even a Perrian theory is bound to posit some kind of limited accessibility to de se thoughts. In his own framework, the limited accessibility arises from the fact that, while anyone can have a belief about the owner of a certain thought of mine (e.g. that the owner of that thought is hungry) only I can have a thought about myself by correctly presupposing that the owner of that thought is hungry. One shortcoming of that paper, as the author himself admits, is that it does not provide a deeper development of the attitude of presupposition. More particularly, it would be enlightening to know more about a subject’s understanding of her own presuppositions besides the fact that it should be merely implicit. Should it, for example, be characterized as dispositional knowledge that one would be able to manifest given sufficient time and reflection or is it something even less substantial, such as, perhaps, a matter of knowledge-how? These questions remain open for further inquiry.

Robert Stalnaker’s positions about de se thought have been usually understood as being in opposition to a Perrian two-factor theory. Be that as it may, in his contribution – Modeling a Perspective on the World – we see him getting closer and closer to that tradition. Stalnaker’s most immediate concern in that paper is proving that one does not need to add centers to possible worlds in order to model attitudinal content. According to him, the formal apparatus of centred possible worlds is theoretically useful, not for modeling contents, but for modeling belief states, i.e. the relation between particular thinkers and the set of doxastic alternatives accessible to them. More specifically, belief states are modeled by Stalnaker as pairs consisting of a base world (the world and time in which the subject is in that state) and a set of doxastic alternatives available to that subject (the worlds that might, for all that subject believe, be the actual one). As more than one author in this volume has noticed3, Stalnaker’s resulting theory resembles a typically Perrian theory with two levels of content, one playing the internal role of psychological rationalization (the doxastic alternatives) and the other, playing the external role of providing absolute truth-conditions for the relevant attitude (the base world). One might wonder, as García-Carpintero (p. 188 ff. 21) suggests, whether this means that Stalnaker’s theory is not concerned, as it used to be in earlier works, with providing a holistic individuation of a subject’s total belief state, but with characterizing specific parts of a subject’s belief states. Unfortunately, Stalnaker does not comment on that, so we are left to our own speculations.

No theory of de se thought has an easy way out of the problem of communication. In Varieties of Centering and De Se Communication, Dirk Kindermann claims that no variant of the Lewisian and Perrian accounts of de se thought4 allows us to maintain a simple picture of communication as the replication of thought from speaker to hearer. He takes that and related facts to motivate a neutral position on the issue of de se thought: “everything that can be done by one view can also be done by the others; the views cover exactly the same empirical data and do so in equally simple ways; the choice between the views is a matter of (theoretical) taste and prior commitments” (p. 309). Kindermann’s conclusion implies that, at least some disagreements between philosophers working on de se thought are not fundamental, such as those about which theory provides a simpler account of de se communication, propositional agreement/disagreement or of samesaying. Coming last in a volume about de se communication, Kindermann’s paper has a particularly anti-climactic feel. If, as this author suggests, there are no knock-down arguments waiting to be discovered in favor of this or that theory, philosophers might need to put the issue of de se thought into a new perspective in order to avoid reaching an argumentative dead-end.

There are four remaining papers. Isidora Stojanovic’s Speaking About Oneself investigates the concept of samesaying in relation to de se utterances and argues that it neither tracks the character nor the Kaplanian content of an utterance. In Why My I Is Your You, Emar Maier presents a formal model of de se communication using the apparatus of Discourse Representation Theory. Aidan McGlynn’s Immunity to Error Through Misidentification and the Epistemology of De Se Thought claims that de se thoughts are not epistemically special and that the phenomenon of IETM should be characterized as a matter of degree. Finally, Kathrin Glüer’s Constancy in Variation tackles the issue of perceptual content and defends the thesis that it does not need to be modeled by centered contents.

One may ask, what is the upshot of the discussions and arguments contained in this volume? Here’s a tentative answer: to account for de se intentionality, we need to explain (i) why two people who have identical de se beliefs are (ceteris paribus) disposed to behave identically while (ii) two people who seem to agree on all the objective properties of a scenario may nonetheless go on to act differently. This was, as we have seen, one of the lessons of Ninan’s contribution to that volume. Most authors seem to agree that, in order to explain (i), we need fine-grained representations of one’s beliefs and/or belief states; however, it is crucial to distinguish the attitude one has towards those fine-grained objects from the attitude one has towards one’s beliefs. On the other hand, there is not much consensus nor positive suggestions about how to explain (ii). Again, most authors point out that merely having beliefs with the same objective truth-conditions is not enough for two people (or two temporal stages of the same person) to count as being in agreement with each other. However, it is not clear what else is necessary. One interesting theoretical possibility is suggested by Recanati’s talk of “coordination” among different thoughts, but the idea is arguably underdeveloped as it stands. Another possibility is to see Weber’s Transform-and-Recenter model as providing a constraint on the agreement relations between two thoughts, e.g. two thoughts A and B agree with each other if and only if B would be the output of the Transform-and-Recenter operation on A and vice-versa. However, that route seems to lead one to conclude that communication and belief retention are highly intellectual inferential processes, whereas intuition has it that they are just the opposite.

All in all, the issue of de se thought is by now a firmly established area of philosophical research and this volume points the way future discussions should take. We recommend it to any reader who is interested in the latest discussions in the philosophy of language and their ramifications into the philosophy of mind and epistemology.


Castañeda, Hector-Neri ‘He’: A study in the logic of self-consciousness. Ratio 8 (December):130-157, 1966. [ Links ]

Perry, J. The problem of the essential indexical. Noûs 13 (December):3-21, 1979. [ Links ]

Lewis, D. Attitudes de dicto and de se. Philosophical Review 88 (4):513-543, 1979. [ Links ]

Prior, A. N. Thank Goodness That’s over. Philosophy 34 (128):12 – 17, 1959. [ Links ]

Frege, G. The thought: A logical inquiry. Mind 65 (259):289-311, 1956. [ Links ]

Cappelen, H. & Dever, J. The Inessential Indexical: On the Philosophical Insignificance of Perspective and the First Person. Oxford University Press, 2013. [ Links ]

Magidor, O. The Myth of the De Se. Philosophical Perspectives 29 (1):249-283, 2015. [ Links ]


1(Castañeda, 1966), (Perry, 1979) and (Lewis, 1979) are usually referred as the first authors to address the issue of de se thought. However, the origins of the argument in favor of essentially de se thoughts can be traced to even earlier works, such as (Prior, 1959) and (Frege, 1956).

2For some recent de se eliminativists, see (Cappelen and Dever, 2013) and (Magidor, 2015).

3García-Carpintero, p. 188 ff. 21; Recanati, p. 144 ff. 5; Weber, p. 249 ff. 5.

4lt seems that Kindermann could very well extend his argument to Fregean theories, although he does not go in that direction.

Matheus Valente – Universitat de Barcelona – Department of Logic, History and Philosophy of Science, Carrer de Montalegre 6 Barcelona 08001, Spain, [email protected]

Acessar publicação original


Deixe um Comentário

Você precisa fazer login para publicar um comentário.