Peter Pagin

This article focuses on the relevance of computational complexity for cognition. The syntactic items may be expressions that are surface strings. But in general, strings are syntactically ambiguous in that they can be generated in more than one way from atomic expressions and operations. The semantic function must take disambiguated items as arguments. When expressions are ambiguous, expressions cannot be the arguments. Instead, it is common to take the arguments to be terms, whose surface syntax reflects the derivation of the string. The semantic function differs in one other important respect from an arithmetic function, since it maps entities between domains, from a syntactic to an ontic or conceptual domain of meanings. Compositionality helps to explain the rate of success in linguistic communication when the sentence used or the content communicated is new.

In his paper ‘Why assertion may yet be social’ (Pegan 2009), Philip Pegan directs two main criticisms against my earlier paper ‘Is assertion social?’ (Pagin 2004). I argued that what I called “social theories”, are inadequate, and I suggested a method for generating counterexamples to them: types of utterance which are not assertions by intuitive standards, but which are assertion by the standards of those theories. Pegan’s first criticism is that I haven’t given an acceptable characterization of the class of social theories. His second criticism is that I have overlooked some alternatives, and that there are social theories that are not affected by my argument. In section 1 I discuss the first, and in section 2 the second.

An assertion is a speech act in which something is claimed to hold, e.g. that there are infinitely many prime numbers, or, with respect to some time t, that there is a traffic congestion on Brooklyn Bridge at t, or, of some person x with respect to some time t, that x has a tooth ache at t. The concept of assertion has often occupied a central place in the philosophy of language, since it is often thought that making assertions is the use of language most crucial to linguistic meaning, and since assertions are the natural expressions of cognitive attitudes, and hence of importance for theories of knowledge and belief.

The term ‘meaning holism’ has been used for a number of more or less closely interrelated ideas. According to one common view, meaning holism is the thesis that what a linguistic expression means depends on its relations to many or all other expressions within the same totality. Sometimes these relations are called ‘conceptual’ or ‘inferential’. A related idea is that what an expression means depends, mutually, on the meaning of the other expressions in the totality, or alternatively on some semantic property of this totality itself. The totality in question may be the language to which the expressions belong, or a theory formulation in that language.

Treating the principle of charity as a non-empirical, foundational principle leads to insoluble problems of justification. I suggest instead treating semantic properties realistically, and semantic terms as theoretical terms. This allows us to apply ordinary scientific reasoning in meta-semantics. In particular, we can appeal to widespread verbal agreement as an empirical phenomenon, and we can make use of probabilistic reasoning as well as appeal to theoretical simplicity for reaching the conclusion that there is a high rate of agreement in belief between speakers who have a high rate of verbal agreement. Although this does not by itself imply that the beliefs agreed upon are generally true, it does imply that any single speaker who is party to such a verbal agreement is justified in taking the other speakers to have mostly true beliefs. She is so justified because of the fact that it is incoherent to take her own beliefs not to be mostly true. Indirectly, this justifies a version of the principle of charity as an empirically correct principle.

Can there be rules of language which serve both to determine meaning and to guide speakers in ordinary linguistic usage, i.e., in the production of speech acts? We argue that the answer is no. We take the guiding function of rules to be the function of serving as reasons for actions, and the question of guidance is then considered within the framework of practical reasoning. It turns out that those rules that can serve as reasons for linguistic utterances cannot be considered as normative or meaning determining. Acceptance of such a rule is simply equivalent to a belief about meaning, and does not even presuppose that meaning is determined by rules. Rules that can determine meaning, on the other hand, i.e., rules that can be regarded as constitutive of meaning, are not capable of guiding speakers in the ordinary performance of speech acts

This is the first part of a two-part article on semantic compositionality, that is, the principle that the meaning of a complex expression is determined by the meanings of its parts and the way they are put together. Here we provide a brief historical background, a formal framework for syntax and semantics, precise definitions, and a survey of variants of compositionality. Stronger and weaker forms are distinguished, as well as generalized forms that cover extra-linguistic context dependence as well as linguistic context dependence. In the second article, we survey arguments for and arguments against the claim that natural languages are compositional, and consider some problem cases. It will be referred to as Part II.

Starting from the familiar observation that no straightforward treatment of pure quotation can be compositional in the standard (homomorphism) sense, we introduce general compositionality, which can be described as compositionality that takes linguistic context into account. A formal notion of linguistic context type is developed, allowing the context type of a complex expression to be distinct from those of its constituents. We formulate natural conditions under which an ordinary meaning assignment can be non-trivially extended to one that is sensitive to context types and satisfies general compositionality. As our main example we work out a Fregean treatment of pure quotation, but we also indicate that the method applies to other kinds of context, e.g. intensional contexts

This paper is an attack on the Dummett-Prawitz view that the principle of bivalence has a crucial double significance, metaphysical and meaning theoretical. On the one hand it is said that holding bivalence valid is what characterizes a realistic view, i.e. a view in metaphysics, and on the other hand it is said that there are meaning theoretical arguments against its acceptability. I argue that these two aspects are incompatible. If the failure of validity of bivalence depends on properties of linguistic meaning, then there are no metaphysical consequences to be drawn. The case for this view is straightforward as long as we are discussing a language different from our own. But it seems that the distinction between failure because of meaning and failure because of reality cannot be applied to our own language, simply because our own language is just what we use to represent reality. I argue that this impression is illusory. In order to draw a conclusion about reality, meaning must be connected with truth in a non-trivial way, and precisely this cannot be done in the language for which the meaning theory itself is correct.

In his recent paper ‘Analyticity: An Unfinished Business in Possible-World Semantics’ (Rabinowicz 2006), Wlodek Rabinowicz takes on the task of providing a satisfactory definition of analyticity in the framework of possible-worlds semantics. As usual, what Wlodek proposes is technically well-motivated and very elegant. Moreover, his proposal does deliver an interesting analytic/synthetic distinction when applied to sentences with natural kind terms. However, the longer we thought and talked about it, the more questions we had, questions of both philosophical and technical nature. Hence the idea of this little paper – for how better to honor a philosopher than by trying very hard to criticize him? After quickly running over some background in possible worlds semantics and setting out Wlodek's proposal against that background, we shall bring up and discuss our questions in sections 3 – 5. In the final section, we shall also make a stab at a different solution to the problem, making use of our own earlier idea of relational modality.

In earlier work (Glüer, K. and P. Pagin. 2006. Proper names and relational modality. Linguistics & Philosophy 29: 507–35; Glüer, K. and P. Pagin. 2008. Relational modality. Journal of Logic, Language and Information 17: 307–22), we developed a semantics for (metaphysical) modal operators that accommodates Kripkean intuitions about proper names in modal contexts even if names are not rigid designators. Graeme Forbes (2011. The problem of factives for sense theories. Analysis 71: 654–62.) criticizes our proposal. He argues that our semantics predicts readings for certain natural language sentences which these simply do not have. These sentences contain mixed contexts involving factive attitude verbs. We argue that the readings our semantics predicts do indeed exist, even if it might take a little work to bring them out. Moreover, denying their existence would have some rather unattractive consequences

Suppose you are stranded on an island and you want to get over to the nearby mainland. Your only option is to swim. But is the other shore close enough? If you embark and it isn’t, you drown. So you prefer to know before taking off. Happily, you are well equipped. You have not only a yardstick, but also a theodolite for measuring angles, and a good knowledge of trigonometry. You then determine the distance to the other shore by means of triangulation. You fix two points, A and B, on your own shore, and identify a third point, C, in sight on the opposite shore. You then measure the distance between A and B with your yardstick, and the two angles ABC and CAB with your theodolite. With this information you can calculate the distances AC and BC to the other shore.

However, if Wittgenstein’s so called rule-following considerations are correct, then this reason for believing in the validity of (C), is mistaken. The conclusion of those considerations is that we must reject the idea that rules are things which determine possible cases of application before those cases are actually encountered and decided by speakers. If this is right, then there is no rule which determines the meanings of new sentences, i.e. before those sentences have actually been used. Therefore, it might seem that (C) is not valid for natural languages.

In the recent debate on the semantic/pragmatic divide, Herman Cappelen and Ernie Lepore (2005) on the one hand, and Fran¸cois Recanati (2004) on the other, occupy almost diametrically opposed positions as regards the role of semantics for communication, while largely agreeing on important features of pragmatics. According to Cappelen and Lepore (CL), semantic context sensitivity of natural language sentences is restricted to what is determined by a particular minimal set of canonically context sensitive expressions. If you try to go beyond that set, as has often been done in recent semantic theories, to reach a position of moderate contextualism, your reasons will force you to the much more extreme position of radical contextualism. That is CL’s instability thesis. They argue for it by means of a number of examples intended to illustrate how you are off on a slippery slope if you admit any context sensitivity beyond the basic one. If radical contextualism is true, systematic semantics is not possible, since, according to CL, there cannot be any systematic theory of speech act content. The one exception is that whatever is said by the utterance of a sentence, its minimally context dependent content is part of it. Precisely this is denied by Recanati. Not only is this “minimal proposition” not part of speech act content. The minimal proposition plays no..