Overview of the project
From September 2019 to December 2021, I was a Marie Skłodowska-Curie Fellow in philosophy of language and philosophy of mind at the École normale supérieure (Paris). My research project ESSINDEX (for "Essential Indexicality") proposed a renewed assessment of the puzzling phenomenon of mental indexicality. This project received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie Grant Agreement No. 799162.
Indexicals are words like ‘I’, ‘here’, ‘now’, ‘that’, etc. whose linguistic meaning encodes a context-dependency: what they designate (their referent) varies according to the context of their use. For instance, my uses of ‘I’ refer to me, while yours uses of the same word refer to you — same word, same conventional meaning, different contexts of use, different referents.
Pioneers of modern semantics (Gottlob Frege, Bertrand Russell, Rudolf Carnap) regarded indexicality (and other sorts of context-dependency such as ambiguity or vagueness) as a defect of natural languages, to be eliminated from ideal, logical languages suited to the purposes of science. They assumed that, for any indexical sentence in a natural language L expressing p in a context c, there is always some nonindexical sentence in L that expresses p in all contexts of use. Thus they regarded indexicality as a dispensable feature of the words whereby we express thoughts.
But around 1970, some philosophers (including John Perry and David Lewis) noticed that some thoughts cannot be expressed without the help of indexicals. The so-called “problem of the essential indexical” suggests that indexicality is, more fundamentally, an essential feature of certain thoughts playing key roles in perception, self-knowledge, and action. Such indexical thoughts pose a challenge to otherwise plausible doctrines about the semantics of mental states like beliefs, given that they seem to represent the world from an irreducibly first-personal, egocentric perspective.
Mental indexicality is now widely regarded as a nexus between classical themes tied to subjectivity in speech, thought, knowledge, action, perception, or consciousness. Nevertheless, debates are still raging about its nature, its extent, its varieties, and even its semantics. The project ESSINDEX contributes to those contemporary debates by developing an original account of what is distinctive about indexical thoughts. In particular, it explores the novel hypothesis that the reason why the problem of the essential indexical has sounded so intractable is that two different problems were conflated from the beginning, which would each call for a separate solution.
Hypotheses of the project
The following case, imagined by John Perry (1977), provides a famous example of the problem of the essential indexical. Lingens is an amnesiac lost in the Stanford Library, who reads a biography of himself. By postulation, he learns all the objective information about Lingens. Yet there is something that the nonindexical sentences in the book do not teach him. As he would put it: ‘I am Lingens, this place here is the Stanford Library.’ The knowledge he lacks seems essentially indexical.
This suggests that thoughts themselves can be indexical. A nonindexical thought represents the world objectively, from outside, while an indexical thought represents the world from some local perspective inside the world, like a map but with an arrow on it indicating, ‘You are here (now)’.
This problem threatens a doctrine of propositions, according to which the contents of mental states like beliefs are propositions. Formally, a proposition has at most one truth-value (true or false) relative to any 'possible world' (i.e. conceivable history of the world). w. It cannot be true in some part of a possible world w (a subject, time, place, in w) and false in another. It describes the world as a whole. The point of the Lingens case is to highlight that a subject like Lingens could know all the propositions about Lingens without knowing everything about Lingens.
David Lewis (1979) concluded that propositions are not the only kinds of contents one can know or believe. On his view, a proposition corresponds to the set of possible worlds in which it is true, and informativeness is explained in terms of excluded possibilities: to learn a new proposition is to rule out the possible worlds in which it is false. As Lingens knew all the propositions about Lingens, the possibility excluded by his new belief cannot be a possible world. Thus Lewis replaces propositions by properties. A property is the set of possible individuals of which it is true. Some properties distinguish individuals in a same world, by being true of some individuals and false of others. Upon learning who he is, Lingens learns such a property: the set A of possible amnesiacs called ‘Lingens’. To this property picture, Lewis adds Egocentrism: all beliefs are self-ascriptions of properties. Lingens’s new belief is true if and only if Lingens belongs to A. Against the doctrine of propositions, here what plays the role of being true/false is not just a (propositional) content: it is a (property) content as ascribed to the subject (at the time and world of belief).
But no consensus emerged about the correct solution to the problem. And lately, a growing number of sceptics have argued that indexicals pose no special problem at all. Consider a variant case inspired by Robert Stalnaker (2008). Daniels, who is not amnesiac, sits in the same room as Lingens and reads a copy of the same book. By postulation, he learns all the propositions he could know about Lingens. Yet he could still conceivably learn something about Lingens if someone pointed to Lingens and said: ‘He is Lingens’. In an important sense, what he learns is just what Lingens learns upon realising: ‘I am Lingens’. But Daniels knows well who and where he (himself) is!
Here the sceptical reasoning will run as follows: if all indexicality is egocentric, if Daniels knows who he is, and if he learns what Lingens learns, then what Lingens learns was not essentially indexical after all.
In response, I explore another tack in the project: maybe someone like Daniels lacks indexical knowledge which is not fundamentally egocentric. Arguably, Daniels is propositionally omniscient about Lingens, and egocentrically omniscient about his own identity and location inside the world (‘de se’), but not indexically omniscient about the identity and location of Lingens in the world (‘de re’). At the very least, here, Lewis’s Egocentric account misses what Daniels learns: his de re belief can rule out neither a possible world nor another possible individual he might (for all he believes) himself be in his world.
Thus the project ESSINDEX has explored the idea that two problems of the essential indexical were conflated. In the present instance, the first would be to specify a new truth that Lingens and Daniels both learn, while the second would be to explain a further cognitive contrast between their respective de se and de re beliefs.
The solution to the first (and overlooked) problem would be to give up Egocentrism: an indexical belief B of a subject s about an object o (at time t in world w) to the effect that o is F ascribes a property directly to the object o. Its content is the property F (i.e. the set of possible objects that are F), and B is true if and only if o (at t in w) belongs to F. In this sense, Lingens and Daniels would acquire the same new indexical belief. The solution to the second (and more familiar) problem would be this: while for de re beliefs and concepts, the object o can be anything to which the subject s is de facto (a posteriori) appropriately related in the context of belief (pragmatic freedom), in the special case of de se beliefs and concepts, o=s de jure (a priori), in virtue of a reflexive rule requiring that the object be identical to the subject of belief (semantic constraint).
Results of the project
(1) PUBLICATIONS.
The project has given rise to the publication of three articles and a book:
Bochner, G. (2020) “A Puzzle about Assertion,” in S. Biggs & H. Geirsson (Eds.), The Routledge Handbook on Linguistic Reference. New York: Taylor & Francis, 268-280.
In this paper I argue that some indexical (demonstrative) assertions pose a special problem (of the essential indexical) in Stalnaker's two-dimensional pragmatics. The problem shows, I suggest, that possible world propositions in that framework ought to be replaced by properties ascribed directly to the object of reference.
Bochner, G. (2021) Naming and Indexicality. Cambridge University Press (Collection: Key Topics in Semantics and Pragmatics). ISBN: 9781108428453. (See the review here: https://ndpr.nd.edu/reviews/naming-and-indexicality/)
This book offers a detailed and critical survey of the central debates concerning linguistic reference in the twentieth century. It then uses the survey to identify and argue for a novel version of current 'two-dimensional' theories of meaning, which generalise the context-dependency of indexical expressions.
Bochner, G. (2022) “Contextual Analyticity,” Analytic Philosophy 63(4), pp. 268-276.
In this paper I argue that, when they are formed in certain ways, judgements expressed by “I am here now” are both (a) apriori (although not solely in virtue of their meaning), and (b) necessary (albeit trivially so). I try to show that, beyond conventionally-analytic sentences (known apriori to be) true just in virtue of (knowledge of their) meaning, there are also contextually analytic judgements (known apriori to be) true in virtue of (apriori knowledge of the) reflexive relation between the context of the judgement and the circumstance of evaluation which it targets.
Bochner, G. (2022) “Externalism, Transparency, and Diagonal Propositions,” Synthese 200(3), pp. 1-23.
(2) CONFERENCES:
Here is a selection of talks I gave to present results of the project:
- November 28, 2019 (invited): International workshop Monsters, Demonstratives and Token-Reflexive Concepts New Thoughts on the Semantics of Indexicals, organized by Shimon Mercer-Wood, École Normale Supérieure, Paris. Title: “A Puzzle about Demonstrative Reference in Stalnaker’s Two-Dimensional Pragmatics.”
- February 27, 2020 (invited): Seminar of LINGUAE (research team in formal semantics), Institut Jean Nicod, Paris. Title: “A Puzzle about Demonstrative Reference in Stalnaker’s Two-Dimensional Pragmatics.”
- March 2020 (invited): Series of talks over one week about centred world semantics, invited by Prof. T. Ede Zimmerman, Linguistics Department, Goethe Universität Frankfurt. Cancelled due to the pandemic.
- August 24-28, 2020: ECAP 10 (10th European Conference of Analytic Philosophy), Utrecht University. Title: “Stalnaker on Externalism and Transparency.”
- February 17, 2021 (invited): « Grandes Conférences », Laboratoire d'Histoire des Sciences et de Philosophie, Archives Henri Poincaré, Université de Nancy 2, France. Title: “Pragmatic Internalism.”
- November 3, 2021 (invited): EXRE Philosophy Colloquium, University of Fribourg, Switzerland. Title: "Pragmatic Internalism."
(3) PUBLIC EVENTS.
Two public lectures organised at the Cité Internationale Universitaire de Paris in 2020 had to be cancelled due to the pandemic: