In this presentation we give a short reminder on the origin of the two main notions of context that are used in philosophy of language in order to draw attention to the main points of discussion common in the meeting on “Contexts in Philosophy”. The use of Context in philosophy begins with Frege's work: in his book Foundations of Mathematics (1884: xxii) he gives the following maxim: “never to ask for the meaning of a word in isolation, but only in the context of a sentence”. This claim is a criticism of Locke who endorsed the Aristotelian view of meaning as an idea connected with a lexical item (for Aristotle words are signs not of objects, but of the movement of the mind and for Locke the meaning of a word is the idea connected with the word). Frege thought that Locke's mistake was to look for the meaning of a word in isolation; but (i) for many words there are no corresponding ideas, and the only way to understand or define the meaning of a word is to study how it works in the context of a sentence; (ii) different people may connect different ideas with the same words. But Frege was worried also for other aspects of the use of language: context de pendence and fiction. While in mathematics we work with expressions (e.g. numerals) that always refer to the same objects (numbers) independently of the context, in ev eryday speech, the reference of expressions like “I”, “today”, “this”, “here” changes depending on the context of utterance. Frege dedicated much of his last writings to discuss this peculiarity of natural language remarking that “The mere wording, as it can be preserved in writing, is not the complete expression of the thought; the knowledge of certain conditions accompanying the utterance, which are used as means of expressing the thought, is needed for us to grasp the thought correctly. Pointing the finger, hand gestures, glances
may belong here too. The same utterance containing ‘I’ in the
mouth of different men, will express different thoughts.”
Especially in poetry and fiction, but not only, there is another problem: in natural lan guage we may use expressions that do not refer to any existing object (e.g. "the winged horse" etc.). Frege suggested that when we use a definite description we presuppose the existence of the object defined by the description: “when we make an assertion we give it for presupposed that the simple or composed proper name has a meaning” (Frege 1892). But this is a shared agreement, which not always is followed by speakers, when they use non-referring expressions.
These three kinds of remarks are at the beginning of three different concepts of
context, two of which became the most relevant in contemporary philosophy: (1)
Linguistic context, (2) Context of utterance (3) Cognitive context. The three different
concepts of context represent the main divide among different – although connected –
logical, philosophical and technical enterprises. Besides the first notion of linguistic
context – that has been developed in different areas of linguistics and computer sci
ences, especially in information retrieval – the notions of context of utterance and of
cognitive context became the most important ones in the philosophical debate. The
notion of context of utterance, together with the notion of circumstances of evalua
tions, is central in Kaplan’s truth conditional semantics: with the work of Kaplan we
have a standard definition of the notion of what is said as ‘the content of a sentence in
a context’. Kaplan provided a shared terminology concerning content, context (of
utterance) and circumstance of evaluation together with a specification of the proper
role of the context of utterance:
“The character of an expression is set by linguistic conventions
and, in turn, determines the content of the expression in every con
text. Because character is what is set by linguistic conventions, it is
natural to think of it as meaning in the sense of what is known by
the competent language user. Just as it was conventional to repre
sent contents by functions from possible circumstances to
extensions (Carnap’s intensions), so it is convenient to represent
characters by functions from possible contexts to contents. This gives the
following picture:
Character: Contexts => Contents
Content: Circumstances => Extensions
Or, in more familiar language,
Meaning + Context => Intensions
Intension + Possible World => Extensions
Indexicals have a context-sensitive character. It is characteristic of
an indexical that its content varies with context. Nonindexicals
have a fixed character. The same content is invoked in all contexts”
Following Kaplan’s framework, John Perry gives a classification of indexicals, together with the distinction between narrow and wide context:
The narrow context consists of the constitutive facts about the utter
ance, which I will take to be the agent, time and position. These
roles are filled with every utterance. The clearest case of an indexi
cal that relies only on the narrow context is ‘I’, whose designation
depends on the agent and nothing else. The wider context consists of
those facts, plus anything else that might be relevant, according to
the workings of a particular indexical”
The distinction between narrow context and wide context is partly relying on the idea
put forward by
An alternative notion of “cognitive” context has been developed by Robert Stalnaker, working on the idea of dialogue: in every dialogue there is a common ground of beliefs and information that we give for granted. How to characterize this common ground and how to understand the relations between this common ground and what is said in the dialogue? Here Stalnaker uses the notion of presupposition as the proposi tion that must be taken for granted in order for an utterance to be appropriate. His idea is to develop the basic intuition behind Grice's theory of speaker’s meaning according to which when speakers mean things, they act with the expectation that their inten tions to communicate are mutually recognized: “This idea leads naturally to a notion of common ground – the mutually recognized shared information in a situation in which an act of trying to communicate takes place. A representation of the common ground helps to clarify both the end of the communicative ac tion by representing the possibilities among which the speaker intends to distinguish, and the means available to the speaker to distinguish between them – the information that must be available in or der that the act of uttering certain noises reasonably be taken as an act of trying to get someone to acquire certain information.” (Stal naker 2002: 704) This new concept of context as common ground of presupposed beliefs brings about a new notion of meaning as context change potential: the meaning of a sentence (typically an assertion) may change the context by simply adding content to the context, or by challenging something accepted in the context or in other ways such as influencing the direction of the subsequent conversation. Update Semantics is just one of the many frameworks aiming at developing a formal semantics grounded on the idea of meaning as context change potential; while these semantics represent an alternative to truth theoretical semantics, they still keep some idea of truth conditions as character izing the result of what is said.
In both frameworks, truth conditional semantics and update semantics, a problem
still rests unanswered: what is said?
While Yang Hu tries to clarify a minimalist model that might be more coherent with Kaplan’s framework, most papers try to challenge aspects of that framework: the two strongest criticisms are by Aldo Frigerio, who claims that demonstratives are not direct referential expressions, and Palle Leth, who criticises the centrality of inten tions in defining the meaning of context dependent expressions. Tamara Dobler gives semantics a space beyond the limits imposed by Kaplan by appealing to the concept of practical goal to help determining the truth conditions of sentences containing oc casion sensitive predicates. Claudia Picazo Jaque develops a similar critique showing how to derive contents that are linked to specific occasions and at the same time con tents that are invariant across contexts. As to the other concept of context inspired by Stalnaker’s views, Carpintero tries to enrich the idea of common ground with the idea of commitments connected with different speech acts. Filippo Domaneschi and Si mona di Paola derive some relevant theoretical ideas from experimental data on Stalnaker’s notion of accommodation, while Salvatore Pistoia Reda works on the concept of existential presupposition for quantified sentences, claiming that background knowledge of possible worlds have a central role in defining the acceptability of the presupposition.
Finally, Sandro Balletta and Augustin Vicente aim at inserting in truth conditional semantics aspects of ontologically driven lexical components, while Ernesto Perini Santos gives a general challenge to the use of formal models for generating or interpreting natural languages: we cannot ask formal models to perform the role of ex plaining natural language and the reference of its expressions, but formal models should be used to represent compositionality once the references of expressions have been defined outside the formal model itself.
We thank all members of the Scientific Committee1 that helped us to select the papers for this workshop and the North Western Philosophy Consortium for sponsoring this meeting under the project of Graduate Conferences in Philosophy, supported by Compagnia San Paolo, and devoted to put together PhD Students and experts in the field. This project was also supported by the National Research Project PRIN 2015L3BC35 on “Indeterminacy Problem: Meaning, Knowledge and Action”. We thank also Sandro Balletta for his help in preparing the CEUR proceedings.