Extensionalism

Under extensionalism is understood in the philosophy of language , logic and semantics of the view that all languages (or weaker: all scientific languages) are extensional ultimately (or should be). It means that

the specification of the extension of the atomic expressions of a language is sufficient to characterize their meaning;
the extension of a complex expression depends solely on the extension of its sub-expressions and the way in which they are composed.

Extensionalism can be viewed as a normative tightening of the extensionality principle, according to which the extensionality only applies (or should apply) to some languages or language segments, but not necessarily to all. Most modern semantics accept intension as the central concept of semantics and thus contradict extensionalism. For advocates of extensionalism, the intensionality of some expressions in natural languages is a surface phenomenon that can be made to disappear through in-depth logical analysis.

Extension and Intension

Since the logic of Port-Royal (1662), it has been customary to use linguistic expressions to differentiate between their subject-matter (reference or extension) and their content (meaning or intention). In modern semantics, which is based on Gottlob Frege's essay On Meaning and Meaning (1892) and essentially goes back to Alfred Tarski and Rudolf Carnap , the following classification has established itself as the standard:

Expression type	Extension	Intension
Proper names	Bearer of the name	Individual term
single-digit predicates	Sets of individuals	Terms
multi-digit predicates	Sets of n- tuples	Relations
sentences	Truth values	Propositions

Willard Van Orman Quine , a lifelong advocate of extensionalism, has already asserted in his essay Two Dogmas of Empiricism (1951) that the intensive vocabulary belongs to a family of terms that can only be defined in terms of each other, so that one ultimately moves in a circle of definitions when trying to explain these terms. In connection with this, the identity conditions of intensional objects such as propositions are often not unequivocally clear; for example, one can be in doubt whether the sentences "The first man on the moon was American" and "Neil Armstrong was American" express the same or different propositions.

Intensionality in natural languages and their extensional reinterpretation

Definition of terms: dimensional and extensional contexts

Under a extensional context is since Bertrand Russell and Alfred North Whitehead's Principia Mathematica understand a linguistic structure (1910), whose extension depends only on the extension of its sub-expressions, so for example, an ordinary, non-contextual sentence like

(1) It is raining or snowing on November 19, 2014 at 6:07 p.m. in Bingen am Rhein.

The criterion for extensionality is the substitutability of extension-equivalent partial expressions salva extensione , i.e. H. in the case of statements, salva veritate . The truth value (extension) of the example sentence just mentioned does not change if “Bingen am Rhein” is replaced by an expression with the same extension: The sentence “It is raining or it is snowing on the day following November 18, 2014 , at 6:07 pm in the city in which the Rhineland-Palatinate State Garden Show took place in 2008 “is true if and only if (1) is true.

The expression "intensional context" (also: "referential opaque context") is mostly used simply as a contradictory contrast to "extensional context", for example in Carnap's Logical Syntax of Language :

“Intensional” should mean the same as “not extensional” (in the various contexts). ("Intensional" means nothing more for us, especially not something like "related to meaning" or the like [...].)

If you follow this terminology, then every linguistic expression is either extensional or intensional. Occasionally, e.g. B. by Carnap himself in meaning and necessity , the term "intensional" is used in a narrower sense and restricted to the cases "in which the condition of extensionality is not fulfilled, but the analogous condition is fulfilled with regard to intension." Examples of this would be modal contexts. Linguistic contexts that are neither intensional in this latter sense nor extensional are often referred to as “hyperintensional”. These include sentences about propositional attitudes , that is, sentences that begin with "A believes that ...", "A fears that ..." etc. In such contexts, intensional equivalent expressions not salva veritate can also be substituted for one another, because subject A of the propositional attitude does not necessarily know about the intentional equivalence. (If A were an omniscient, logically perfect being, a sentence of the form "A believes that ..." would, however, be an extensional context.)

Extension reinterpretation of intensional speech

It seems obvious that in everyday language there are often intensive (or non-extensional) contexts. There are typically three to four groups:

Citations,
Reports on propositional attitudes and indirect speech,
modal contexts,
non-truth-functional sentence operators like “because”.

Examples and reinterpretation:

(2) " Giorgione " is three syllable

It is obvious that salva veritate cannot be substituted within quotations , because reference is often made to the linguistic form of what is between the quotation marks: the sentence

(3) "Giorgio Barbarelli da Castelfranco" is three-syllable

is obviously false while (2) is true. The case is a little more complicated

(4) Giorgione was so named because of his size,

because (4) is first about the painter Giorgione and not about his name. Still is

(5) Barbarelli was so named because of his size

wrong and even incomprehensible: The ending “-one” is an augmentative suffix in Italian , and the anaphoric “so” in sentence (4) refers to the linguistic object. If the anaphora is dissolved, the true (stylistically somewhat awkward) sentence emerges

(6) Giorgione was called "Giorgione" because of its size,

in which the first, referentially transparent occurrence of "Giorgione" salva veritate can be replaced by an extension-like expression:

(7) The Renaissance painter Giorgio Barbarelli was called "Giorgione" because of his size.

Generally speaking, you should never substitute within quotation marks because terms enclosed in quotation marks stand for themselves and lose their ordinary meaning. Tarski and Quine have therefore proposed using functional terms in canonical notation instead of quotation marks , which describe expressions as concatenations (concatenations) of letters or other characters:

(8) is three-syllable.

{\ displaystyle G \ circ i \ circ o \ circ r \ circ g \ circ i \ circ o \ circ n \ circ e}

The concatenation functor is extensional.

An example of the attribution of a propositional attitude is

(9) Chiara believes that Giorgione was a Renaissance painter.

However, it does not follow from (9) that Chiara believes that Giorgio Barbarelli was a Renaissance painter, because if she has all her knowledge about Giorgione from a book about the Italian Renaissance in which he is exclusively called "Giorgione", she can draw this conclusion dont pull. Here, too, the substitutivity fails, which, as I said, is the criterion of extensionality. A reference to propositions (sentence contents) as objects of belief, doubt, etc. is also generally not open to extensionalists, because propositions are intensional.

A trivial return to the case of citation according to the model

(10) Chiara believes: "Giorgione was a Renaissance painter"

Among other things, it is forbidden because Chiara, if she does not understand German, can hardly have such a relationship with a German sentence.

In his essay On Saying That , Quine's student Donald Davidson tried to subject indirect speech and propositional attitudes to a paratactic analysis, which can be viewed as extensionalistic in its basic approach.

Prominent representatives and formulations of extensionalism

One of the main exponents of extensionalism was the early Rudolf Carnap, one of the leading members of the Vienna Circle . In his first major work The Logical Structure of the World (1928) he wrote (p. 63):

All statements are extensional. In each sentence, the sign of the subject matter assessed by the statement […] may be replaced by any sign of the same meaning, even if it has a different meaning.

It should be noted here that Carnap is based on Frege's terminology from On Sense and Meaning ; d. H. "Meaning" here means something like "extension", "sense" means something like "intension".

In Carnap's second major work Logical Syntax of Language from 1934 it says (p. 188):

A universal language of science can be extensional, more precisely: for every given intensional language S1, an extensional language S2 can be constituted in such a way that S1 can be translated into S2.

In his third major work, Meaning and Necessity (1947), Carnap gave up extensionalism and took the view that extensional and dimensional systems stand side by side on an equal footing.

Quine, heavily influenced by Carnap, at least in his early years, wrote:

I doubt I have ever fully understood anything that I could not explain in an extensional language.

Individual evidence

^ Wuttich, Contexts called Intensional
^ Carnap, Logical Syntax of Language , p. 185
^ Carnap, Meaning and Necessity , p. 61
^ Quine, Reference and Modality
↑ See Cappelen / Lepore, Varieties of quotation
↑ Schreiber, citation , chap. 2.2
↑ See Quine, Reference and Modality
↑ See Church, On Carnap's Analysis of Statements of Assertion and Belief
↑ Carnap, The logical structure of the world
^ Carnap, Logical Syntax of Language
↑ Quine, Confessions of a Confirmed Extensionalist , p. 500, cited above. Bar-Am, Extensionalism in Context , p. 13 [Orig .: "I doubt that I have ever fully understood anything that I could not explain in extensional language."]

literature

Ajdukiewicz, Kazimierz: “A Method of Eliminating Intensional Sentences and Sentential Formulas”, in: Atti de XII Congresso Internazionale di Filosofia , Vol. 5. Florence 1960, 17-24
Bar-Am, Nimrod: Extensionalism. The Revolution in Logic . Berlin 2008
Bar-Am, Nimrod: "Extensionalism in Context", in: Philosophy of the Social Sciences 42 (2012), 543-560
Cappelen, Herman / Lepore, Ernest: "Varieties of quotation", in: Mind 106 (1997), 429-450
Carnap, Rudolf: The logical structure of the world . Berlin-Schlachtensee 1928. New edition Hamburg 1998. ISBN 978-3-7873-1464-5
Carnap, Rudolf: Logical Syntax of Language. Vienna 1934
Carnap, Rudolf: Importance and Necessity. A study of semantics and modal logic . Vienna / New York 1972 (Original edition: Meaning and Necessity. A Study in Semantics and Modal Logic . Chicago / Toronto / London 1947)
Church, Alonzo: " On Carnap's Analysis of Statements of Assertion and Belief ", in: Analysis 10 (1950), 97-99
Davidson, Donald: "On Saying That", in: Synthesis 19 (1968/69), 130-146
Quine, Willard Van Orman: "Three Grades of Modal Involvement", in: Proceedings of the 11th International Congress of Philosophy , Vol. 14. Amsterdam 1953, 65-81
Quine, Willard Van Orman: "Reference and Modality", in: ders., From a Logical Point of View . Cambridge, Massachusetts 1953
Quine, Willard Van Orman: Word and Object . Cambridge (Massachusetts) 1960. ISBN 0-262-67001-1
Quine, Willard Van Orman: Confessions of a Confirmed Extensionalist . Cambridge, Massachusetts 2008
Schreiber, Jan: citation. Linguistic philosophical considerations for the nomination of linguistic entities . Saarbrücken 2008. ISBN 978-3639022049 ( available online )
Wessel, Horst: “Against the myth of intensional contexts”, in: Zeitschrift für Semiotik 17 (1995), 369–378
Wuttich, Klaus: “Intensionally named contexts”, in: G. Meggle (ed.): Analyomen 2. Proceedings of the 2nd Conference “Perspectives in Analytical Philosophy” , Vol. I. Berlin / New York 1997, 174–182