Search

As we saw in the last chapter, certain arguments in support of (event-) realism make explicit reliance upon the possibility of ‘identity-statements’ involving grammatical singular terms, such as ‘the murder of Caesar’. Thus, Davidson’s positive argument from causation, and his argument from intention, both appeal to sentences like ‘The killing of the old man was the killing of Oedipus’ father’. These sentences might uncontroversially be called ‘grammatical identity-statements’, in so far as they involve grammatical (definite) singular terms flanking an inflection of the verb ‘to be’. The realist who appeals to such sentences must maintain that the grammatical singular terms in question are on a par, logically, with their grammatical brethren. Whether this is so is a question treatable separately: we have discussed it separately in the foregoing chapters. But the very fact that singular terms for properties or events occur in grammatical identity-statements may be taken as evidence for their being logically on a par with their grammatical brethren. For should we not assume, of a given sub-class of grammatical identity-statements, that they are logically on a par with their grammatical brethren?

In the previous chapter, we argued that actions are objects which explain agency-causings. In order to fulfill this purpose the informational content of the object ascribed with a true action ascription must be determined by the practical and motivational content of its intentional cause. Where the agent’s intention fails to be realized, the object ascribed must contain additional extra-intentional information which explains why the (mis-)intention had the effect it did.{Si1} All of the distinctions which we find useful to make among the objects ascribed with action discourse will have their source, therefore, in the articulations of the relevant representations. Thus, in order to specify the structures of the various act-objects, one needs an account of Practical Rational Agency. The act-object types action, mistake, accident, etc., will mirror, in their structures, generic structural features of agency and certain corresponding standard deviations therefrom. In this chapter we begin our analysis of the processes of agency, which we complete in the sequel. Ours is a causal cum functional analysis of natural agency2 in which will occur states of a process which have the logical and functional properties of our psychological terms such as ‘desire ’ ‘belief’ and ‘intention’, as well as other states wh ich our basic terms and relations allow.

If words have meanings, then how is it that these meanings unite to form the meaning of a sentence? This question will naturally arise for anyone who recognizes the composition principle--the principle that the meaning of a sentence is a function of the meanings of its constituent words--and also recognizes that the sentence is the basic unit of discourse. Given the first principle plus some fairly intuitive notions of word meaning, sentence meaning, and sentence, it may seem surprising that we can do with sentences what we cannot do with their constituent words: assert propositions, utter commands, ask questions, etc. Crudely put, the question amounts to something like this: why is the meaning of a sentence not just the meaning of a list of words?

The first few years of my philosophical life were so entangled with learning from Karel Lambert that I can scarcely separate the two. In the fall of 1959 I entered my first philosophy class at the University of Alberta, a class of about seventy students with Lambert as instructor. He told us firmly that he had no intention of discussing ethics or morals, told us to read something about the pre-Socratics for next time, and dismissed the class. It became clear quite soon that he expected us to learn philosophy by doing: despite the size of the class, there was a great deal of discussion and we were constantly challenged. His response was always measured. People who were struggling found sympathy, and he would turn their questions into something significant for discussion. But students who were catching on would immediately find themselves made to face greater difficulties. It was also, I can think of no other way to put it, a lot of fun. Twice, I remember, he put me down, to everyone’s amusement. The first time I wouldn’t back down from a point when I was clearly losing, and Lambert ended the discussion with “Van Fraassen, you are logical, but you are not reasonable.” The second time we had gotten into that subject he did not want to do at all, and I brought up Sartre’s famous example of the moral dilemma of the young Frenchman, who had to choose between care for his aging mother and joining the Free French. Lambert listened patiently, then retorted: “Van Fraassen, there comes a time when a boy has to leave his mother.”

The three-level theory of language processing is almost a direct consequence of the theory of partitioned representations. It makes three basic claims: (1) In discourse understanding communicated information is distributed over spaces as the internal representation evolves. This means that at various points in language understanding a decision is made that some particular intermediate structure (or part thereof) will project into some particular space. (2) There are two kinds of processes responsible for these decisions. One implants an intermediate structure in some space in which further processing can begin. The other distributes an intermediate structure between a parent and a child space. I will call these contextualization and distribution respectively. (3) Most linguistic processing is completed with reference to the contents of a particular space. This includes most of the familiar processes. I will call such processes parochial.

Bertrand Russell [8] and John Bacon [1] argue that definite descriptions are not what they seem — referring terms on a par with names. Instead, they propose contextual definitions which serve to eliminate them. Russell analyses the sentence ‘the F Gs’ as ‘there is at most one x such that Fx and there is at least one xsuch that Fx and Gx’. Bacon analyses it as ‘for all x if x uniquely Fs, then Gx. ’ In either case all sentences containing definite descriptions are true or false even when there fails to be a unique entity described. For Russell ‘The present King of France is bald’ is false, for Bacon true. Bacon’s analysis, in particular, meshes nicely with a feature of generalizations: ‘All present Kings of France are bald’ is also true when there are no present Kings of France.

I hope in this paper to establish that, in an important sense of ‘empirical’, the concept of personal identity is not wholly an empirical concept. It cannot therefore be equated, as some philosophers have maintained, with the concept of some relation or set of relations whose existence could only be discovered empirically. I also hold that the concept of identity is not the concept of a relation at all. If, therefore, it is implicit in thoughts that we have of ourselves as remembering, intending, fearing, etc., which are crucial to our thought of ourselves as persons, it is difficult to see how it could be replaced, as has also been envisaged, by some relational concept or concepts. I do not believe that the concept of personal identity can either be reduced to some other concept, or allowed to fall into disuse as a relic of outworn, metaphysical beliefs no longer important to us.

The purpose of this paper is to try to see what the logical gist of Bertrand Russell’s doctrine of knowledge by acquaintance is. I take the outlines of what Russell says on this topic as being familiar enough.1 What is not equally clear is how Russell’s doctrine is to be understood and evaluated. At least this was not clear to myself before I came upon some ideas in the semantics of propositional attitudes which suddenly made the logic of Russell’s views— if not their correctness — perspicuously clear. Hoping that this clarity is not an illusion, I shall try to capture and preserve it by conveying it to others.2

In Meaning and Necessity, Carnap briefly and critically assessed the approaches to definite descriptions associated with the names of Frege, Russell and Hubert and Bernays.2 Ultimately he adopted a version of the Frege approach, now widely known as the chosen object method, in the development of his influential doctrine of meaning and modality. Since Carnap’s discussion, however, another approach to definite descriptions, called free description theory, has emerged. Moreover, Carnap did not concern himself with the question whether there exists a way of understanding the various approaches to definite descriptions such that they can be construed as different options to a common technical problem. Such a way exists in set theory where various approaches can be construed as different options to the paradoxical axiom of set abstraction in naive set theory. That is the task of this essay; to provide a way of understanding both the traditional and free logic approaches to the treatment of definite descriptions such that they all can be seen as different ways out of a common technical problem, a problem manifested in what will be called The Naive Theory of Definite Descriptions.

This paper presents a revised view of the syntax of pronominal binding. It argues that in (1) Every man that owns a donkey beats it. a donkey is a quantifier that binds it, in the same way that every donkey binds it in (2) Every donkey hates the man that beats it.

In this chapter I undertake to look closer at parochial processing, at distribution, and at the interaction of the two.

Theoretical linguists have long grappled with what is known as the projection problem for presupposition. In this short chapter I demonstrate that what has long been considered a formidable problem requiring a complex solution is in fact a direct consequence of the use of partitioned representations in language understanding.

There are some interesting correlations between the semantic notions of “role,” “value,” “connector,” and the grammatical peculiarities of the so-called “copulative” sentences in French noted by Moreau (1970) and Ruwet(1972).

Goudge introduces two types of explanations that challenge the standard models of explanation. These are “integrating explanations” and “narrative explanations.” The first two sections of this chapter are devoted to discussing these two types of explanation in turn to see if they establish that the positivist account of explanation is inadequate to account for what biological science does. The third and fourth sections deal with some further aspects of the axiomatic model and its applicability to biology.

Kuhn’s Structure of Scientific Revolutions presents a picture of science that is often thought to be at odds with the philosophy of science of the positiviste — even though it was published in the International Encyclopedia of Unified Science series which had been founded by Otto Neurath to stand as a monument to positivism. It is certainly true that Kuhn’s picture of scientific research does not fit the probabilistic picture that Carnap defended. But in fact Carnap’s programme was supported only by the narrow group of positivists whose loyalty he had garnered. It was never accepted by Neurath, for example, nor by several other members of the Vienna Circle, including Gustav Bergmann. What is not so clear is that the account of the logic of theory structure that Kuhn’s picture requires is incompatible with the logic defended by the positivists.

Semantics is the theory of meaning; in particular, natural language semantics is the theory of meaning in natural languages. But what is meaning? This question is as old as philosophy, and it has, like every other central philosophical question, led to an unending string of different, often incompatible, answers. In fact, one of the principal tasks of semantics is to define, or at least to clarify, what meaning is. And so we find, corresponding to many of the different conceptions of meaning which reflection on the original question has spawned, different semantical theories — theories which often differ from each other so much in presuppositions or purpose that there is barely any sense in trying to compare them. Indeed, we find such divergence and incommensurability not only when we survey the various approaches to the theory of meaning generally, but also when we confine our view to theories that are concerned specifically with the semantics of natural language. Even within this more restricted domain many approaches are separated by underlying pre-theoretical assumptions so different as to make any attempt at comparative evaluation well-neigh pointless.

‘Devadatta, the son of a barren woman, does not speak—but neither is he silent.’ Most non-philosophers would be inclined to accept the Buddhists’ claim that this sentence is true, and that it is about Devadatta. There are widely shared pre-theoretical intuitions to the effect that it is possible to make true assertions about non-existent entities. But many philosophers see this as yet another instance in which our intuitions lead us astray. In the West there is a tradition beginning with Parmenides that denies that anything meaningful can be said about the nonexistent. In response to the challenge that at least this much may be said about a non-existent, that it does not exist, this tradition has given us various analyses of negative existentials all of which are designed to show that such sentences are actually about something else. Only a relative handful of Western philosophers-Meinong being the best known example-defend common sense on this point. But the Meinongian defense comes at a price--a bloated ontology replete with the golden mountain, unicorns, and the son of a barren woman-that many find too steep to pay.

According to the thesis of the primacy of the intentional, the reference of language is to be explicated in terms of the intentionality of thought. The word “Pferd,” for example, refers to horses in so far as it is used to express thoughts that are directed upon horses. But most contemporary philosophers of language, until recently at least, have held that the intentionality of thought is to be explicated in terms of the reference of language. But no such explication is at hand.

While the work of the contemporary Polish logicians is fairly well known by their colleagues in western Europe and America, little is known in these countries about recent Polish philosophy. This is perhaps due to the fact that the logicians generally published in universal languages while the philosophers published their works mostly in Polish. This is unfortunate because western philosophers, especially logical empiricists, might have profited from a knowledge of the work of such men as Tadeusz Kotarbinski and Kazimierz Ajdukiewicz, who are probably to be considered as Poland’s outstanding contemporary philosophers. Both have formulated philosophical views influenced by modern logic: Ajdukiewicz has created radical conventionalism, and Kotarbinski has developed reism.

In this paper, I want to return to an attempt I made a few years ago in my Frege’s Conception of Numbers as Objects (henceforward FCNO)1 to bring general ideas about the nature of singular terms and singular reference to bear on the issue between mathematical platonism and its opponents. The ideas in question are at least implicit in Frege’s way of proceeding in Grundlagen, and are also canvassed in various of Dummett’s writings.2 They have been deepened and strengthened in Hale’s recent book3 and provide, it seems to me, by far the best hope for a straightforward platonistic construal of large portions of classical mathematics. My principal purpose here is to add to the reasons for confidence in them by responding to some criticisms canvassed by Field, in his critical notice of FCNO.4

The hierarchy of salience of the items of the knowledge assumed by the speaker to be shared by him and by the hearer constitutes one aspect of a dynamic account of discourse (Sect. 1). It is claimed that a representation of this hierarchy is a good support for discourse analysis (reference assignement, Sect. 2) and for discourse production (pronominalization, definite description, Sect. 3).

In Marty’s Untersuchungen zur Grundlegung der allgemeinen Grammatik und Sprachphilosophie (Marty 1908, 438f., henceforth ‘U’) there is a long footnote about the semantic status of proper names. This footnote occasioned a remarkable defence of the thesis that proper names have meaning. The discussion, found in L. Landgrebe’s work on Marty (Landgrebe 1934, § 16), presents strong arguments against the currently popular causal theory of proper names. These arguments, which have an obvious relevance to Kripke’s views, will be discussed in what follows. I will not go into Landgrebe’s critique of Marty’s views. Instead, I will emphasize the shared strand of systematic argument. But before we come to the relevant passages, a few basic reflections about Kripke’s work on proper names are in order.

I owe much to my thesis director Nuel Belnap, but I wonder if I can blame him for my recent idealist tendencies. The novelist Colin Wilson (1967) has a theory of ideas as infectious agents, much like viruses and needing hosts. It is interesting to speculate how the idealist infection was transmitted from Blanshard to Anderson and Belnap at Yale, and from them to me as their graduate student at Pittsburgh. Perhaps rather than being an infectious disease it is a hereditary one, linked in some complicated way with the intellectual genes for relevance logic, Platonism, and logical humor.

Frege’s extensional approach to reference affected his understanding of ontology in two ways: 1) it minimized the importance of his distinction between the complete being of objects and the incomplete being of concepts, and 2) it obscured the role of senses as “modes of presentation,” by making it impossible to regard concepts as senses. Let us first situate the analytic interpretation of ontology within its historical context, next review Frege’s comments on the being of concepts and objects more in detail, then discuss the levelling effect of his extensional approach to reference, and finally consider how that approach encouraged reductive interpretations of the meaning of ‘being’.

‘Bantu nomenclature’ should not only feature on the academic menus of anthropologists, sociologists, sociolinguists, psychologists or the so-called African Studies scholars, as history reveals, but should also be, and doubtless it is, but for the infancy in which African philosophy finds itself, of interest to philosophers in general, and to philosophers of language in particular. The urgency of such endeavours is justified by the live and burning issues in contemporary philosophy of language. Needless to say that the academic arena is replete with anthropological, sociological and even linguistical literature on Bantu cultures in general, and Bantu nomenclature in particular in so far as they serve the anthropological, sociological or the linguistical purposes, but, lamentably, no comparable philosophical literature has so far been realised. Hence the attempt to open a preliminary, but by no means exhaustive, philosophical perspective towards Bantu nomenclature in the hope of provoking a philosophical debate on issues and problems raised by contemporary philosophers of language, particularly those relating to names and naming, and their implications when considered with Bantu proper names within the purview.

This paper describes the treatment of some anaphoric expressions in certain referentially opaque contexts by the experimental discourse understanding system LODUS. Noun phrases occurring in potentially opaque contexts are given one of two possible interpretations: (1) The noun phrase describes the way the agent refers to the entity in question, or (2) the noun phrase acts solely as a means of identifying the referent, regardless of the linguistic expression used for referring to it. The first case, which is the opaque case, is handled by complex variables called roles. Here a further distinction is made between value-loaded and value-free roles, indicating whether the expression is referential or not. The second, and transparent, case is handled like other non-opaque cases. Some examples of the treatment in LODUS of opaque contexts are given, among these Quine's Ralph and Ortcutt story.

This article has two main parts. In the first, the subject pronoun in a cleft sentence together with the cleft clause is shown to function pragmatically as a discontinuous definite description. Applying the GIVENNESS HIERARCHY (Gundel et al. 1993) makes it possible to explain the distribution of this-clefts and that-clefts in discourse, and predicts the more frequent occurrence of it-clefts. Clefts also semantically share existential and exhaustiveness conditions with definite descriptions. The second part presents a new syntactic analysis of clefts, which treats the cleft clause as an extraposed complement of the cleft subject pronoun, adjoined to the clefted constituent.

Theories of personal identity try to explain what the identity of a person necessarily consists in, but frequently leave open what kind of necessity is at issue. This paper is concerned with conceptual necessity. It proposes an analysis of the concept of personal identity in terms of a definite description. The analysis coheres with out judgments about clear cases and explains why cases of division seem indeterminate. The apparent indeterminacy results from attempting to apply a definite description to a situation in which more than one object would satisfy the description. The definite description analysis also explains the strengths of the influential no-branching theory, while avoiding the problems with that view. The no-branching theory is in effect a second-order analysis, i.e., a combination of the definite description analysis of personal identity plus a Russellian analysis of the definite description.

Preface Part I. Situation Theory: 1. Constraints, channels, and the flow of information Jon Barwise 2. Extended Kamp notation: a graphical notation for situation theory Jon Barwise and Robin Cooper 3. States of affairs without parameters Mark Crimmins Part II. Logical Applications: 4. Labelled deductive systems and situation theory D. M. Gabbay 5. Events and processes in situation semantics Michael Georgeff, David Morley and Anand Rao 6. Nonmonotonic projection, causation and induction Robert C. Koons 7. Modal situation theory Stephen M. Schulz Part III. Linguistic Applications: 8. Generalized quantifiers and resource situations Robin Cooper 9. Situation theory and cooperative action Keith Devlin and Duska Rosenberg 10. Propositional and non-propositional attitudes Jonathan Ginzburg 11. Episodic logic: a situational logic for natural language processing Chung Hee Hwang and Lenhart K. Schubert 12. A situation-theoric formalization of definite description interpretation in plan elaboration dialogues Massimo Poessio 13. A situation-theoretic representation of text meaning anaphora, quantification, and negation Dag Westerstahl, Bjorn Haglund and Torbjorn Lager Name index Subject index.

Introduction. Tense functions in scientific texts (S.-G. Andersson). German tenses, tense logics, context change, and the temporal interpretation of texts (J. Ballweg). Negation, probability, and temporal boundedness: discourse functions of negative tenses in Swahili narrative (E. Contini-Morava). Frame and reference time in complex sentences (C. Fabricius-Hansen). Toward a theory of tense-aspect in narrative discourse (S. Fleischman). Deixis, relevance, and tense/aspect (A. Fuchs). Meaning and interpretation of tense (J. Gvozdanovic). Meaning and use of the Dutch perfect (S. de Haan). Preterit as definite description (Th.A.J.M. Janssen). Multiple meanings in context: the French present tense (M. Monville-Burston, L.R. Waugh). Les valeurs temporelles des formes verbales en grec ancien (C.J. Ruijgh). The proximal/distal temporal axis: the meaning and use of then in discourse (D. Schiffrin). The analysis and reconstruction of the temporal structure of narrative texts (A. Schopf). Invariant meanings and context-bound functions of tense in Spanish (C. Silva-Corvalan). The 'future tense' in modern Hebrew (Y. Tobin).

Now that you have settled into your new post, I am required to evaluate your teaching performance. The semester is well under way, so the sooner this is done, the better. If we were being scientific about sampling your teaching, the class observation would be a surprise inspection. Thus reason dictates that the visit should occur on the first day that you do not believe it will occur. But on reflection, this is an impossible demand. Could I give the inspection next Monday? No, because you would realize that Monday is the first available unexpected day. What about the next class on Tuesday? Well the previous elimination would make Tuesday the first available unexpected day. Hence, it falls prey to the previous reasoning. And indeed, parallel reasoning would eliminate all of the remaining days. Hence, the earliest unexpected inspection is impossible. Since we hired you in a buyer's market, you have obviously noted the resemblance to the surprise examination paradox. But you may have not noticed that the elimination proceeds in the reverse direction. It is odd that such a reversal could be effected by substituting a definite description ('The event will occur on the first unexpected day') for an existential generalization ('The event will occur on an unexpected day'). But that's logic for you. Thus I am compelled to take refuge in philosophy department's custom of asking you to propose a day for the class visit.

In "On Denoting," Bertrand Russell maintains that the key to an understanding of denoting ordinary things is an analysis of the logical syntax of sentences containing the definite article 'the'. In somewhat the same vein, I think that the key to an analysis of denoting propositions and facts must be an analysis of the logical syntax of sentences containing the complementizer 'that', the word that forms that-clauses, such as 'that snow is white'. I But my conclusions are the opposite of Russell's. His insight was to reject the natural idea that combining 'the' with a common noun phrase yields a phrase, a definite description, that acts as a syntactic-semantic unit and refers to a thing. I defend the natural idea that combining 'that' with a sentence yields a phrase, a thatclause, that acts as a syntactic-semantic unit and refers to a proposition or fact. Treating that-clauses as units flies in the face of conventional practice of the last few decades of work in modal, epistemic, and deontic logic. In these enterprises, the role of the word 'that' is to adhere to a verb or adjective to produce an operator, such as 'It is possible that' or 'x knows that', the whole of which combines with sentences in non-truth functional ways. Russell, had he considered this view, would have said "This leaves that-clauses, by themselves, wholly devoid of meaning, but gives a meaning to every proposition in whose verbal expression they occur" (paraphrased from On Denoting). This practice sees 'Mary believes that snow is white' as having the structure

Gottlob Frege observed that a sentence of the form 'a = b' has a different cognitive value (Erkenntniswert) from a sentence of the form 'a = a,' where 'a' and 'b' are co-referential proper names; 'a = b' is informative, one might say, and 'a = a' is not. From this he inferred that (proper) names have senses (Sinne) and that the sense of a name may be different from the sense of another co-referential name.1 It follows from this that the sense of a name is not determined by its referent. Frege thought that the sense of a name is analogous to the sense of a definite description on three counts: (i) it is the descriptive content of the term; (ii) it determines the reference of the term; and (iii) it contributes to the cognitive value of any expression in which the term occurs. Although the exact nature of Fregean senses might be open to endless debate, it is clear that the postulation for each name of a sense satisfying (i), (ii), and especially (iii) would provide an explanation of the informativeness of 'a = b.' This is the strength of a Fregean theory of names. However, Saul A. Kripke has convinced many philosophers that names do not have Fregean senses, or any descriptive contents, in terms of which their reference is determined.2 Kripke resurrected John Stuart Mill's idea that names are "connotationless" tags. Kripke's arguments were novel and ingenious. He taught us a great deal about names and modality. And yet, Kripke has not completely dispelled the Fregean reason for postulating descriptive contents for names. Or so it appears. If, as Kripke has us believe, names are thoroughly nondescriptive, then we seem to have a problem; the informativeness of 'a = b' seems to defy explanation. If names have no descriptive contents, co-referential names seem not to differ from each other in any way semantic. But if so, it

Saint Anselm of Canterbury offered several arguments for the existence of God. We examine the famous ontological argument in Proslogium II. Many recent authors have interpreted this argument as a modal one.' But we believe that Jonathan Barnes has argued persuasively that Anselm's argument is not modal.2 Even if one were to construe the word 'can' in the definite description 'that than which none greater can be conceived' in terms of metaphysical possibility, the logic of the ontological argument itself doesn't include inferences based on this modality. In this paper, we develop a reading of Anselm's Proslogium that contains no modal inferences. Rather, the argument turns on the difference between saying that there is such a thing as x and saying that x has the property of existence. We formally represent the claim that there is such a thing as x by '3y(y=x)' and the claim that x has the property of existence by 'E!x'. That is, we represent the difference between the two claims by exploiting the distinction between quantifying over x and predicating existence of x. We shall sometimes refer to this as the distinction between the being of x and the existence of x. Thus, instead of reading Anselm as having discovered a way of inferring God's actuality from His mere possibility, we read him as having discovered a way of inferring God's existence from His mere being. Another important feature of our reading concerns the fact that we take the phrase "that than which none greater can be conceived" seriously. Certain inferences in the ontological argument are intimately linked to the logical behavior of this phrase, which is best

The present paper treats the problems of reference and especially the single definite reference. Some examples from the filmic discourse (which is considered as a possible world) illustrate the designation of individuals through definite descriptions and particularly through proper names : proper names are considered as rigid designators, which at the same time connote and function as a predicate of denomination to be called /N/ for an x., Il s'agit dans cet article de traiter de la référence, sous l'angle particulier de la référence définie unique. Des exemples pris dans le discours filmique, envisagé comme monde possible, illustrent la désignation d'individus par des descriptions définies et surtout des noms propres, considérés comme des "désignateurs" à la fois "rigides" et connotatifs, ayant le sens d'un prédicat de dénomination être appelé / N / pour un x., de Mourgues Nicole. La référence définie unique dans le discours filmique. In: Documentation et recherche en linguistique allemande contemporain - Vincennes, n°34-35, 1986. Paroles inachevées. Etudes linguistiques pragmatiques et processuelles de conversations, brouillons et textes. pp. 337-357.

Analyse du probleme de la reference des noms propres a la lumiere des Theories de l'A. sur l'intentionnalite. Apres avoir defendu, contre les partisans de la theorie de la reference "directe" des noms une version renouvelee du descriptivisme (noms et descriptions n'ont pas un fonctionnement semantique et pragmatique substantiellement different), l'A. articule tout acte de reference a l'expression d'un contenu intentionnel.

within a language whose only modal operators are the box and the diamond; other modal idioms cannot be expressed within such a language at all. Nonetheless, quantified modal logic has enjoyed considerable success in uncovering and explaining ambiguities in modal sentences and fallacies in modal reasoning. A prime example of this success is the now standard analysis of the distinction between modality de dido and modality de re. The analysis has been applied first and foremost to modal sentences containing definite descriptions. Such sentences are often ambiguous between an interpretation de dicto, according to which a modal property is attributed to a proposition (or, on some views, a sentence), and an interpretation de re, according to which a modal property is attributed to an individual. When these sentences are translated into the language of quantified modal logic, the de dicfo/de re ambiguity turns out to involve an ambiguity of scope. If the definite description is within the scope of the modal operator, then the operator attaches to a complete sentence, and the resulting sentence is de dicto. If the definite description is outside the scope of the modal operator, then the operator attaches to a predicate to form a modal predicate, and the resulting sentence is de re. Quantified modal logic has the resources to clarify and disambiguate English modal sentences containing definite descriptions. In this paper, I explore to what extent the analysis in terms of scope can be applied to modal sentences containing denoting phrases other than definite descriptions, phrases such as ‘some F’ and ‘every F.1 I will focus upon categorical modal sentences of the following two forms: mo s a1 idioms must be artificially restructured if they are to be expressed

The immediacy of interpretation of anaphoric expressions occurring in written discourse was examined in two experiments, using a technique for measuring the time to detect misspellings on subsequent words in the sentence. The critical contrast was between spelling error detection latency for verbs which were either consistent with a full contextual interpretation of the prior anaphor or inconsistent. In the first experiment a substantial consistency effect was observed for verbs which immediately followed either an anaphoric proper name or definite description. This result indicates that the full contextual interpretation of the anaphors occurs as soon as they are encountered. In the second experiment anaphoric pronouns were used which could refer to either the main character in the story or a subsidiary character. Consistency effects were only observed for verbs which immediately followed the main character pronouns. This result indicates that unambiguous pronouns may not be given a full context...

Critiquant les theories des descriptions definies qui traitent ces expressions comme des pronoms dont la reference est anaphorique ou semblable a celle de variables liees d'un calcul des predicats logique, l'A. propose qu'on les traite comme des parametres semblables a ceux qu'utilisent les systemes logiques de deduction naturelle de type Gentzen ou Fitch.

Pierre Lerat: descriptive study of utterances like Ça, les féministes protestent The structure considered (NP, S) is a particuliar (oral) case of dislocation (see J. Deulofeu 1977 and 1979). This article draws attention to two semantic properties of utterances such as Ça, les féministes protestent: 1) NP refers to a definite referential object (deictic, anaphoric, proper name, definite description); 2) the typical interpretation of the structure is "concerning NP".

A teacher has graded exam papers which he will return. Just before meeting his class, a colleague tells him that a spy entering his office the previous night has already revealed the grades to his students. Actually, no such thing happened; the colleague is playing a hoax on his friend, who now sincerely believes him. Later in class, before distributing the papers, our teacher says: 'The student who wrote the best paper got, of course, full marks', with the intention to refer to the particular student who indeed wrote the best paper, and who, he mistakenly believes, is known by the others. The question is: has the definite description been used referentially? No doubt the students took it in the attributive way.' They do not yet know which particular student the teacher was talking about. They do not even realize that the teacher intends his description referentially: there is nothing in the situation allowing the recognition of such an intention. But can we say that the teacher used it referentially? This answered in the affirmative will entail the admission that at least under the circumstances specified, 'what the speaker means by his utterance X' and 'how X is understood' have some difference in semantic content. Accordingly, the speaker will be viewed as saying something true about a particular person (whether or not this person bears the features mentioned in the description), while the audience interprets him as saying something about anyone satisfying a particular description. What is crucially interesting in this situation is that we cannot explain this discrepancy as we usually do, i.e. in terms of the audience's misinterpreting or misunderstanding the utterance. There is no way in which we can characterize the audience as unjustified; no mistake or fault can be attributed to him. If so, do we have before us a situation in which the speaker's meaning diverges from

As is well-known, the reference of singular terms in natural languages often depends on the context of utterance. A token of'I', for example, refers to the person who utters that token, a token of 'now' refers to a period of time which includes the moment when this token is uttered, and so on. In the fifties, some philosophers thought that reference was always a pragmatic, contextual matter; and this view is again current among contemporary linguists and philosophers working on the pragmatics of natural languages. Does this mean that all referring expressions are 'indexical', like 'now' or 'I'? Obviously they are not. A definite description like 'the president of the United States of America in 1986' is surely not indexical in the ordinary sense. But then, how are we to understand the claim that reference is always contextdependent? Can we make that claim consistent with due recognition of the fact that not all singular terms are indexical? This is the question I will try to answer in this paper. I hasten to add that I will not be concerned with trivial forms of contextdependence, such as the dependence of reference on the language being spoken. There might exist a language, say Lambdese, where the words 'the president of the USA in 1986' mean the same as the English description 'the inventor of the zip'; supposing Julius to be the inventor of the zip, the description 'the president of the USA in 1986', uttered in a context where Lambdese and not English is spoken, would refer to Julius and not to Ronald Reagan. This form of contextdependence is trivial and will be set aside in this paper; I will always consider the language spoken as fixed. Contextdependence will be deemed non-trivial and worthy of consideration only to the extent that a given expression, with a fixed meaning, makes in different contexts different contributions to

In this paper I discuss the late Gareth Evans's theory of descriptive names-names that are explicitly introduced into language by means of a definite description. Evans has an argument to show that such names belong in the same semantic category as ordinary (i.e. non-descriptive) names. 1 I do not think that this argument works: and I believe the conclusion to be unacceptable, especially given Evans's initial assumptions. No one has grappled with the really deep issues concerning reference as much and as successfully as Evans: and whether I am right or wrong about the present issue, the effort is worthwhile just because of the illumination to be had from further discussion of his views.

Three difficulties which emerge in Bertrand Russell’s Theory of Descriptions are analysed in this paper. In each case one or several possible solutions of the analysed difficulty are studied together with the consequences which such solutions have with respect to the validity and application of the theory. The problems considered are: first, the one concerning the definition of “definite description”; second, that about the application of the theory to the analysis of proper names which appear in historical discourse and finally the Russell-Strawson controversy with respect to the existential content of sentences whose grammatical subject is a definite description. In the first part, the question of presenting a general account of definite descriptions is discussed. Three formulations which attempt to define these expressions are examined. The first definition specifies a formal criterion: “a definite description is a phrase of the form ‘the-such-and-such’”. There are two arguments against this formulation, showing that this criterion is neither sufficient nor necessary to explain the semantic features of definite descriptions. After this the following formulation is considered: “A description is a phrase of the form, the term that etc., or more explicitly the term x that satisfies Φ x, where Φ x is some function satisfied by one and only one argument.” Notwithstanding that this characterization includes a semantic notion, it has the disadvantage of excluding empty descriptions, that is to say, those constructed with functions satisfied by more than one individual or by none. Finally, an account of the semantic behaviour of definite descriptions is analysed: “A definite description is a linguistic expression by which it is intended to mention an individual by alluding to a property that only he possesses.” The difficulty of this definition is that it excludes descriptions which a particular speaker believes to be empty. As a solution to the difficulties which the aforementioned accounts present, a definition is proposed to define the expression “‘α’ is a definite description for a in t” where ‘α’ and ‘β’ are variables of linguistic expressions and ‘a’ and ‘t’ are variables of language users and time instants respectively. The outstanding feature of such a definition is that in its definiendum the concept of “intension” appears, as it is used by Carnap in his article “Meaning and Synonymy in Natural Languages”. Def; α is a definite description for a in t = def. there exists a β such that it fulfills the following conditions: (a) β is a general term (for a in t) (b) α contains β (syntactically) (c) In virtue of the intension of α for a in t (and not merely for empirical questions) the following will hold: (y) (α denotes y ≡ y is the only individual that has the property associated with β). The author finishes this section criticising the definition and pointing out some difficulties, namely that it does not apply to some cases of ordinary usage and that it relies on the notion of intension. In the second part, problems arising from the relation between definite descriptions and proper names in a logical sense is a symbol that represents an object of which one has knowledge by acquaintance and does it without assigning any property to it. However, with the exception of logical proper names, other expressions considered generally to be names are in fact abbreviated definite descriptions. Hence, occurrences of common proper names are replaceable by descriptions and, in accordance with this, it may be said that the meaning of a proper name is formed by the meaning of an associated definite description. Nevertheless, a direct consequence of this interpretation of the relation between proper names and descriptions, is that the latter can vary according to the context in which the names are used, so the meaning of a name might be varied. In this part of the article problems arising from the preceeding considerations are developed. Making use of the formal properties of descriptions, as presented in Principia Mathematica, it is shown that this interpretation does not give coherent results within the context of historical proper names. The author of the article suggests that in order to consider proper names as abbreviated descriptions, a condition is required: the general terms that the description contains should be specified. It is assumed that the application of this limiting condition would unify the formal and semantic aspects of the theory of descriptions. However, in the article it is shown how in the case of names which appear in historical contexts it is not possible to fulfill this condition. Thus, from this a new difficulty arises in the application of the theory of descriptions; i.e., that it is not possible to extend the results which are specifically obtained for definite descriptions to other singular terms. In the third and last part of the paper the objections which Strawson raises against Russell’s Theory of Descriptions are analysed. Originally, Strawson thought that Russell was mistaken on the following points: (a) in that whichever person utters a sentence in which the grammatical subject is a definite description he would be formulating a truthful assertion or a false one, and (b) that part of which he would be asserting would be that in reality there exists one and only one object which satisfies the function contained in the description. In the article is stressed that Strawson, in a later paper, withdraws his objection to (a) pointing out that the thesis of truth value gaps and the theory of falsity (the thesis accepted by Russell) present advantages which are not mutually exclusive ((b), however, is not discussed in this paper). The author then describes two technical meanings for the expressions “false” and “asserts”. Let p and q be any two statements, Def: p is falsetc if it does not fulfill some of its truth conditions. Taking truth conditions of a statement p, the conditions that ought to be fulfilled in the world for that statement to be true and that appear specified in its meaning, he gives the following definition: Def: Given two statements p and q, p assertstc q if the truth conditions of q are also truth conditions of p. With these definitions an interpretation of (a) and (b) is given.