行政院國家科學委員會專題研究計畫 成果報告
存在的投射概念:戴維森式的存在概念(3/3)
研究成果報告(完整版)
計 畫 類 別 ： 個別型 計 畫 編 號 ： NSC 95-2411-H-002-001- 執 行 期 間 ： 95 年 08 月 01 日至 96 年 10 月 31 日 執 行 單 位 ： 國立臺灣大學哲學系暨研究所 計 畫 主 持 人 ： 楊金穆 計畫參與人員： 博士班研究生-兼任助理：傅皓政、李國揚 報 告 附 件 ： 國外研究心得報告 處 理 方 式 ： 本計畫可公開查詢中 華 民 國 97 年 04 月 18 日
Report on “A Projectivist Conception of Existence:
Toward A Davidsonian Theory of Existence”
〈
〈
〈
〈存在的投射概念
存在的投射概念
存在的投射概念
存在的投射概念：
：
：
：戴維森式的存在概念
戴維森式的存在概念〉
戴維森式的存在概念
戴維森式的存在概念
〉
〉
〉結案
結案
結案
結案報告
報告
報告
報告
計畫執行期間 計畫執行期間 計畫執行期間 計畫執行期間: [1/3 – NSC 93-2411-H-002-029 (93/08/01∼∼∼∼94/07/31); 2/3 – NSC 94-2411-H-002-006 (94/08/01∼∼∼∼95/07/31); 3/3 – NSC 95-2411-H-002-001 (95/08/01∼∼∼∼96/10/31)] 計畫主持人計畫主持人計畫主持人: 楊計畫主持人 楊楊 金楊 金金 穆金 穆穆 穆 計畫執行單位 計畫執行單位 計畫執行單位 計畫執行單位: 台灣大學哲學系台灣大學哲學系台灣大學哲學系台灣大學哲學系 2008 年年年年 4 月月月月I. Papers resulting from the proposed research
(A) Papers published:
Yang, Syraya C. M. (2008 – forthcoming),《Quine 論模態性》,《哲學與文化月刊》, 「走到分析的盡頭──蒯因（Willard Van Orman Quine）百年」專題論文 (鄧 育仁主編).
Yang, Syraya C. M., (2007), ‘Proper names as quantifiers: A neo-Fregean account of the sense of names’, EURAMERICA (《歐美研究》) vol. 37, no.2, 183-225.
(B) Contributed/invited papers on conference
Yang, Syraya C. M. (2008a), ‘De Re modality and the merely possible’, to be addressed at
The Sixth European Congress of Analytic Philosophy (ECAP 6), to be held at
Jagiellonian University, (organized by the Institute of Philosophy of the Jagiellonian University and the Polish Association of Logic and Philosophy of Science). Krakow, Poland. 21-26 August 2008.
Yang, Syraya C. M. (2007a), ‘Existence as a primitive concept,’, 2007 Soochow International
Conference on Analytic Philosophy, June 22, 2007, Department of Philosophy, Soochow University, Taipei, Taiwan.
Yang, Syraya C. M. (2007b), ‘Coming into existence through language’, presented at The 4th
Conference on Experience and Truth, November 9-10, 2007, Department of Philosophy, National Taiwan University, Taipei, Taiwan.
Yang, Syraya C.M. (2006a), ‘Truth and existence in fiction’, presented at The 3rd Conference on Experience and Truth, November 24, 2006, Department of Philosophy, Soochow University, Taipei, Taiwan.
Yang, Syraya C.M. (2006b), ‘Existence is a predicate: A Davidsonian account of existence and predication’, presented at Workshop on Body and Meaning, December 9, 2006, Institute of European and American Studies, Academia Sinica, Taipei, Taiwan.
Yang, Syraya C.M. (2005a), ‘The role of truth in theory of existence’, presented at The 1st
Conference on Experience and Truth, March 12, 2005, Department of Philosophy, National Taiwan University, Taipei, Taiwan.
Yang, Syraya C.M. (2005b), ‘Truth in virtual reality’, presented at The 2nd Conference on Experience and Truth, June 24, 2005, Department of Philosophy, National Taiwan University, Taipei, Taiwan.
Yang, Syraya C.M. (2004), ‘Towards a naturalistic conception of existence’, presented at
2004 Symposium on Neo-pragmatism, October 2, 2004, held at Institute of European and American Studies, Academia Sinica, Taipei, Taiwan.
II. Previous works related the proposed project
Yang, Syraya C. M., (1998), 《一個投射論者的存在概念》(A Projectivist conception of existence),Taiwan Philosophical Studies, l (〈台灣哲學研究台灣哲學研究台灣哲學研究台灣哲學研究 第一期第一期第一期第一期: : 思想: : 思想思想思想、、、、語言語言語言語言 與真理
與真理與真理
與真理〉)，63-95.
III. Summary of the report
The primary concern of this research is to propose a theory of existence which, as it stands, not only can be applied to ordinary things (i.e., what there are) in the world but also can be applied to some seemingly mysterious and problematic discourses, the ontology of which seems rather disputable, such as modal discourse, fictional discourse and virtue reality.
The theory of existence I propose is essentially a Davidsonian conception of existence. My strategy is, first of all, to follow Davidson’s (1990) approach in taking the existence-predicate as primitive and then to offer an axiomatized theory of existence by applying Tarski’s definition of truth to existential sentences. More specifically, I argue (2007a, 2006b) that in dealing with the concept of existence, we may mimic the way Tarski deals with the concept of truth, namely, to propose a characterization of the semantic conception of existence by showing how the existence predicate (typically, ‘x exists’) is used, rather than attempting to provide an analysis of the concept of existence. That is, we may apply Tarski’s T-scheme to existence sentences and then take as a partial definition of existence any equivalence of the following form
(E) ‘a exists’ is true iff a exists.
For simplicity, let us call an equivalence of this form an E-sentence, and a sentence of the form ‘a exists’ or any of its cognates an existential sentence.
As is well-known, Davidson has long argued that we should take the truth-predicate as primitive and construct an axiomatized theory of truth to serve as the required theory of meaning for a given natural language. It is then sensible to construct an axiomatized theory of existence based on all E-sentences. Of course, the gist of this approach is essentially an application of Davidson’s truth-conditional based theory of meaning to existential sentences. That is, we should read E-sentence from right hand side to the left hand side. For it is clear that based on Davidson’s radical interpretation, to know the meaning of an existence sentence
s is to know its truth-conditions when it is held as true. Therefore, at the core of the programme is to show (2005a, 2007b) how existential sentences are to be so interpreted, namely, to show how the truth-conditions in question are to be obtained, such that the desired collection of true existential sentences would suffice to show what there are, as well as how things are, in the external world.
In doing so, I shall (1998) introduce a projectivist account based on both Blackburn’s ((1984), (1986)) projectivism and Nelson Goodman’s ((1976), (1978), (1983)) notion of projectibility, to indicate how truth-conditions of an existential assertion are to be characterized. But it is striking that neither of Blackburn’s projectivism nor Goodman’s conception of projection can be straightforwardly applied to characterization of truth-conditions for existential sentences. To offer a satisfactory projectivist treatment by virtue of which the truth-conditions of an existential sentence can be characterized, I shall once again appeal to Davidson’s (1997) idea ‘seeing through language’. More specifically, I shall argue that it is the language in use, which supplies a setting for the characterization of the truth-conditions for existential sentences. Without the appeal to the language in use we are in no position to talk about the truth-conditions of existential sentences. I hope this approach can offer a better way of constructing a theory of existence by virtue of which our use of the existence predicate is closer to our intuition of the concept of existence.
It strikes me that if Davidson’s truth-conditional programme offers a satisfactory theory of meaning, it will also provide a promising approach to a theory of existence. The primary concern of this project is to specify how to apply Davidson’s approach to the very concept of existence.
While the first half of this research is mainly devoted to the presentation of a Davidsonian theory of existence, the second half intends to show the application of my proposed theory of existence can be applied to some seemingly disputable discourse, noticeably, existence of individuals in modal discourse (typically, in possible worlds), the existence of fictional characters in fictional discourse, and the concept of existence in virtual reality. I (2008, 2008a) show that based on my proposed theory the existence of possible individuals, the merely possible, or individuals in possible worlds can be satisfactorily explained. Similarly, the issue with regarding the ontology of characters in the works of fiction can be dealt with (2006a). And moreover, we can have a more appealing account of the ontology of virtue reality (2005b).
Keyword: Existence; truth; projection/projectivism; Davidson; truth-conditional
theory of meaning; predicate/predication
IV. Abstracts
Abstract (2008): Quine on modality
Quine 一向以反對模態性著名，對 Quine 而言，對於事物能滿足所謂的模態條件(modal conditions)的宣稱是難以理解的(unintelligible)。因為，宣稱同一個事物在不同的可能世 界中得以滿足同一個條件，這種同一性是無法成立的，因而 Leibniz’s Law 無法成立。 所謂的模態語句最多只是用來表示我們談論事物的方式(way of speaking of things)。因 而，只有關於語句的模態性(de dicto modality)是可被接受的，而關於事物的模態性(de re modality)則是不合法的(illegitimate)。這個反模態性的論旨在二十世紀後半葉引起相當 廣泛的討論。反對者認為，所有的「必然性」與「可能性」這種模態性，如果有任何的 形上學或知識論意涵的話，當然是歸屬於事物本身。我們一般會認為，當有人宣稱「蘇 格拉底可能是個音樂家」為真時，他是在斷言：是蘇格拉底本人可能成一個音樂家，而 不是「蘇格拉底是個音樂家」這個語句或命題，有所謂的可能性。然而 Quine 的論證似
乎是沒有挑戰的空間。晩近 Timothy Williamson 嘗試使用條件句(conditionals)來說明模 態性，以取代以往訴諸可能世界語義學(possible world semantics)的進路，其好處在於， 可以避開同一性的問題。然而，條件句本身似乎在某種意義下預設可能性的概念。
本文主旨在於說明 Quine 對於關於事物的模態性之批評基本上是正確的，如果模態 性是在所謂實在論的觀點下。但是我將指出，問題的癥結不在於同一律能否適用模態論 境(modal discourse)，而在於沒有所謂的模態條件(modal conditions)，也沒有所謂的實際 存在的可能事物(mere possibles)。因而，如果關於事物的模態性語句之為真係取決於某 一個可能事物能否滿足某一性質，或者某一個事物能否滿足某一模態性質，則這種關於 事物的模態性之宣稱自然是難以理解的。Quine 的論證當然無可置喙。然而，從反實在 論的觀點來看，我們仍然有堅實的理由可以宣稱關於事物的模態語句為真，而無需訴諸 於「模態條件」或是「可能事物」這種實在論的概念。我將進一步論證模態語句之真假 取決於一組適當的真值條件。重點在於，有些真值條件在實際世界中並不成立，但與其 他的真值條件可以一致地同時成立，作為該模態語句的真值條件。
我將先概略說明 Quine 的論證，並檢討若干批評，特別集中於 Kit Fine、Dagfinn Føllesdal。其次檢討 Williamson 的化約主義（把模態語句化約為條件句）。最後，我將 說明如何把模態性當作一種原初概念(primitive concept)來看待，並利用真值條件來證成 模態語句的真。在此，我將論證，訴諸於「真值條件」與「一致性」的概念不致於有訴 諸循環或無限後退的困境。我希望此進路不只可以用來說明關於事物的模態語句之合法 性，也可用來解決條件句的一些爭端。
Abstract (2008a): De Re Modality and the Merely Possible
In this talk I propose an anti-realist account of metaphysical modality. I assume that if the notions of necessity and possibility have any metaphysical significance, then modality of this sort can be, and should be, ascribed to individuals, rather than to sentences or statements. I then take as the starting point that we do have de re modal truths and legitimate de re constructions. And at the core is to address how de re sentences to be interpreted intelligibly without positing possible objects and modal properties. In particular, I show that de re modal truths concerning the merely possible can be fully comprehended without being committed to alleged entities of this sort.
I start with an examination of Quine’s argument against the intelligibility of de re constructions. In general, a de re modal sentence, say ∃x□ϕ(x), holds only when a given object assigned as the value of the bound variable in question satisfies the very modal condition, i.e.□ϕ(x). However, for Quine, talk of the satisfaction of modal conditions embedded in de re sentences with regard to individuals in modal discourse makes no sense owing to the lack of identity of objects in modal discourse. We can at most have de dicto modal truths which can be at best construed as expressing certain ways of speaking of things. I shall argue that Quine’s argument seems impeccable for those who would adhere to the realist’s view of modality. Nonetheless, a further examination of Quine’s argument would reveal that talk of de dicto modality also makes no sense. As a matter of fact, there is no part of reality, upon which de dicto modal truths hinge. It is then more sensible to give up the realist’s view and instead, maintaining that we can talk of modal truths without presupposing the existence of possible individuals and the satisfaction of so-called modal conditions should be no longer a threat to talk of de re modality.
Recently Williamson proposes that modality can be explained in terms of conditionals. But it seems more likely that conditionals and modality presuppose each other. Surprisingly, if this observation is right, it is appealing to argue that at the root modality and conditionals share something in common – a certain type of logical consequence: just like the consequent of a given conditional can be treated as a logical consequence of its antecedent, we may take a modal truth as logical consequence of some other statements which may be held as true but not actually true. That is, a modal sentence can be held as true only when a certain set of sentences – its truth conditions – can be held as true though some of them are not actually true.
Based on this interpretation of modal sentences, we can even legitimately talk about de re modal truths concerning the merely possible. Notice that we must distinguish the merely possible from the fictional – the characters in works of fiction. I shall argue that talk of the merely possible makes sense only when they are logical consequences of certain truths concerning actual things. I shall further argue that appropriate characterization of truth conditions of a given modal statement concerning the merely possible would suggest that they can have names, and a fortiori, identity as well. And these considerations would indicate that an appropriate modal system for metaphysical modality needs a free logic with identity as its underlying system.
Abstract (2007): Proper Names as Quantifiers A Neo-Fregean Account of the Sense of
Names
The main burden of this paper is to present a quantificational treatment of names, by construing the sense of a name in a sentence as indicating a special type of quantification (in fact, a constant quantification) in character, which is supposed to impinge upon the scope of application of the associated predicate(s). In brief, a name occurring in a sentence will be treated as a constant quantifier. That is, to treat the standard formula of the form “Fa” (where “F” is a predicate and “a” a proper name) as “axFx” (where a is a constant quantifier, to which an object a in the given domain will be assigned as its reference, if there is any), and the variable x always takes the object a, if there is any, as its semantic value whenever it is bounded by the constant quantifier a. This account substantially follows Frege’s guidelines for his semantic theory in general which he lays out at the very beginning of his The
Foundations of Arithmetic. I shall start with a brief analysis of how his guidelines would carry weight with an adequate account of the sense of names. Then I propose that the sense of a name in a sentence should be construed as a special type of quantification (in fact, a
constant quantification) in character. I shall further justify the formal adequacy of this quantificational treatment of names by constructing a first order language, in which the symbols ordinarily used as name letters or individual constants will be treated as constant quantifiers, together with appropriate semantic rules for these constant quantifiers. Finally, I show how this treatment could help us to deal with some persisting problems that the use of names may give rise to.
Abstract (2007a): ‘Existence as a primitive concept’
A concept can be defined in terms of some other more fundamental/basic concepts. When such an analysis is available, we may apprehend the given concept in accordance with our understanding of these basic concepts. Alternatively, a concept can be defined extensionally by specifying a class of objects, taken as its extension. When no appropriate definition is available, a concept can be construed by appealing to a certain promising reductionism so as
to paraphrase statements expressing the seemingly troublesome concept into some equivalent statements containing other well-known or more solid concepts.
However, it can be argued that the concept of existence can not be defined in terms of other basic concepts. Intuitively, any statement containing the alleged more basic concepts in terms of which an analysis of the concept of existence can be carried out will be essentially much more informative than existential statements. Nor can the concept of existence be defined extensionally. Any specification of a desired class of objects for the concept of existence either presupposes the existence of some objects already, e.g., Quine’s well-known slogan: ‘to be is to be the value of a (bound) variable’), or results in the commitment to a class of posited objects based on the language in use (e.g., Quine’s ‘semantic ascent for terms’).
I shall next show that no appropriate reductionism with regard to existential statements is available. Two typical examples will be examined, that is, Frege’s identification of identity statements and existential statements, and Dummett’s appeal to the notion of reference in his use theory of meaning.
This line of reasoning leaves us with no further option but to take existence as an undefinable, in fact, a primitive concept. Of course, a primitive concept can be grasped by virtue of some informal heuristic treatment. However, it would be much better to characterize such a primitive concept of existence via an axiomatization of all true existential sentences in a given language, i.e., to take all true existence sentences as axioms. Nonetheless, it would be ideal to have a certain version of axiom-schema, instead of taking all true existential sentences as required axioms. Moreover, to guarantee that true existential sentences do have metaphysical import, the axioms under investigation must be equipped with appropriate interpretation. Application of Davison’s truth-conditional program of meaning to existential sentences then suggests itself to be an appealing approach. On the one hand, Tarski’s T-sentence provides a required axiom-schema sufficiently ranging over all existential sentences, a fortiori, all existing objects. On the other hand, each true existential sentence, taken as axioms, is equipped with intended interpretation based on its truth-conditions. More specifically, we may take a special kind of Tarski’s T-sentences containing only existential sentences as the desired axiom-schema for the required axiomatization:
(ET) ‘τ exist(s)’ is true iff τ exist(s)
where ‘“τ exist(s)”’ in the left-hand-side of (ET) is an affirmative existential sentence of the language in use; while ‘τ exist(s)’ in the right-hand-side of (ET) is an interpretation of the sentence ‘τ exist(s)’. The concept of existence characterized in this way is in essence Davidsonian. I show finally that a Davidsonian conception of existence characterized in this way is conservatively informative in the sense that a true existential statement (i) offers some information, resulting from the truth-conditions for the sentence ‘τ exist(s)’, about the world, but (ii) says nothing more than the existence of the object(s) under investigation.
Abstract (2007b), ‘Coming into existence through language’
The primary concern of this paper is to show an application of two related these that
Davidson proposed in ‘A nice derangement of epitaphs’ to a theory of existence. The first one, known as no-language thesis, states that ‘there is no such thing as a language, not if a
language is anything like what many philosophers and linguists have supposed’; whereas the second, called no language-world boundary thesis, aims at the elimination of ‘the boundary
between knowing a language and knowing our way around in the world generally.’ (Davidson 1986/2005 V-9: 107)
The first part is devoted to a defence of the no-language thesis. In particular, I focus on Professor Ian Hacking and Professor Michael Dummett’s criticisms. I argue that what is invariant on the basis of which a variety of distinct interpretation can be rendered is not the existence of a unique language, in the standard sense, to which a given utterance belongs. I shall show that for Davidson, the interpretation occurs with regard to utterances of sentences; while the indeterminacy of interpretation holds for sentences. And indeterminacy occurs whenever a vocabulary is rich enough to describe a phenomenon in more than one way. And it is the phenomenon under investigation which is invariant. The no-language thesis is therefore not inconsistent with the principle of the indeterminacy of translation. Also, the no language thesis is free from Dummett’s challenge because the alleged norms of the use of language is not a requisition for successful communication. I shall argue that communication requires certain normative elements. But alleged normative elements are in interpretation, rather than in the use of language.
In the second part, I offer an exposition of the no language-world boundary thesis. Special attention will be paid to the relationship between the two theses. I will first point out the key double role that the passing theory plays in communication. On the one hand, a passing theory paves a setting to successful communication, on the other hand, it also serves as a theory of the external world. Then I show that at the core of this double role is a Quinean conception of truth – truth hinges on reality. Thus the passing theory involved in a successful communication provides the hearer’s understanding of the very sentence uttered by the speaker; hence it also provides her knowledge of the required truth-conditions of the very sentence. Now, based on a Quinean conception of truth, to know the truth conditions of a sentence is to know the way things are as what these truth conditions show.
Finally, in the third part, I apply the two theses to existential sentences, showing that our understanding of true existential sentences suffices to justify the existence of the objects that the given true sentences are about. Following this line of thought we may claim that things come into existence through language.
Abstract (2006a): Truth and Existence in Fiction
In this paper, I propose a Davidsonian account of fictional truths and the existence of fictional objects. I argue that it helps little for our understanding of fictional truths to characterize fictional objects, or to analyze the nature of fictional objects,. Instead, we shall start with the naïve assumption that we do have fictional truths and then ask: how would we have knowledge of fictional truths. And I argue that our understanding, or our having knowledge, of fictional truths mainly results from our proper interpretation of fictional sentences, which depends partly upon our knowledge of the world and partly upon the descriptions that the author states in the story. Accordingly, our interpretation of fictional sentences is in a way similar to Davidson’s having knowledge of other mind (the author’s mind). The well-known triangulation program thus is applicable to our interpretation of fictional sentences. I shall firstly show that both the object-theorist’s and anti-realists’ approaches offer no satisfactory account of fictional objects and fictional truth. For the former, I examine Peter van Inwagen’s existing-object theory and a variety of Meinongian-object accounts. For the later, I focus on Lewis’s possible world account and the well-known Pretense theory.
Abstract (2006b): Existence Is a Predicate −−−− A Davidsonian Account of Existence and
Predication
The main burden of this paper is to present a Davidsonian conception of existence. I argue that it helps little for our understanding of the very concept of existence to characterize what there is, or to analyze the nature of existence. Instead, we shall start with the naïve assumption that ‘existence’ can be treated as a primitive concept and we do have some true existential sentences. The gist of the issue then becomes to ask: how would we know the truth-conditions of a given existential sentence. And I propose that our understanding, or our having knowledge of the truth conditions, of a given existential sentence mainly results from proper interpretation of that sentence, which depends partly upon our knowledge of the world and partly upon the language in use. The well-known triangulation program thus is applicable to our interpretation of existential sentences. I shall firstly offer a historical survey of the debate over whether existence is a predicate or not, and show that both sides have rooted on a misleading realistic view of predicate, that is, to show what kind of entity should be associated to predicates, such as universals, properties, relations, or sets, . . . , etc. A variety of different account of the concept of existence will be examined, including Owen’s epistemic account, Frege-Russell-Quine’s taking existence as a second-level concept, and Redmon’s pragmatic account. Then I show that Davidson’s theory of predication offers a promising treatment to the problem of predication, especially for the predicate ‘exists’. According to Davidson, predication has to be related to truth, for having a truth value is the simplest mark of the unity of sentences; so in any satisfactory account of predication we need to explain how predicates contribute to the truth value of the sentences in which they occur, And finally, I show that a Davidsonian conception of existence can be characterized by applying Tarski’s T-scheme to existential sentences.
Abstract (2005a): ‘The role of truth in theory of existence’
For decades, Quine’s well-known slogan, ‘No entity without identity’, has been widely construed as providing a specification of the existent ─ a key to the question of what there are ─ only those, to which identity applies, count as objects in our ontology. In particular, the slogan appears irreproachable, when it goes hand in hand with another metaphysical maxim, the criterion of ontological commitment ─ ‘to be is to be the value of a (bound) variable’. Intuitively, if both criteria function well, we would have a satisfactory characterization of the concept of existence. As Quine has summarized, the combination of these two slogans serve as an explication of the notion of existence: ‘We may indeed take ‘∃xx=a’ as explicating ‘‘a exists”.’ (Quine 1969: 94)
In this paper I shall firstly draw a distinction between two aspects of the characterization of a concept, that is, the normative aspect which stipulates what objects there are, which fall under the very concept so that a class of object can be specified as the extension of the very concept; and the descriptive aspect which describes the conditions under which the application of the very concept to an object holds. Then I argue that to characterize the concept of existence by specifying an associated class of object on the basis of Quine’s two criteria can be at best treated as offering a normative aspect of the concept. And as the concept of existence is so basic and fundamental such that such a normative approach could serve little help. By contrast, I shall instead propose a descriptive approach to the speciation of the class of the existent by describing the conditions of application of the predicate ‘exists’ and its cognates to certain objects. The proposed approach is an application of Davidson’s
truth-conditional theory of meaning to the class of existential statements. The main idea is to characterise the concept of existence by a specification of the class of the existent, which will obtain by applying Tarski’s T-scheme to all existential statements,
(ET) ‘a exists’ is true if and only if a exists.
I shall further argue that a theory of existence can be constructed by such an application of T-scheme to all existential statements. Accordingly, the concept of truth will play a central role in the theory of existence.
Abstract (2005b): ‘Truth in virtual reality’
What sort of reality is virtual reality? And what kind of knowledge would we have when we claim that we know of something in virtual reality? In this paper I shall investigate into the ontology of virtual reality to see what sort of entity it should be. I shall argue that these issues can be settled only in a satisfactory theory of truth which can be applied to virtual reality as well. I shall firstly examine Heim’s (1993; 1998) virtual realism, and Stanovsky’s (2004) phenomenalist account of the ontology of virtual reality, respectively. Of course, some other versions will be mentioned in due course. For instance, it seems to me quite clear that physicalism should be excluded. For no one would take objects in virtual reality as merely phantom of bits and bytes. And of course what is involved in virtual reality can never be only the circuits and wires that actually produce virtual reality. Then I propose an antirealist view of the ontology of virtual reality, a Davidsonian approach in character. Accrodingly, our knowledge of virtual reality is quite different from our knowledge of the real world. As a matter of fact, it seems to me that the former is more like our knowledge of other minds. If this line of thought is on the right track, then the role of truth would play a significant role not only in our distinction between virtual reality and the actual reality but also in our knowledge about virtual reality. Accordingly, we need a satisfactory account of truth. And it can be further argued that a Davidsonian account of truth is in order. For obviously, none of the traditional accounts of truth serves as an adequate theory of truth in virtual reality, including correspondence theory, coherence theory, semantic theory, redundant theory, identity theory, disquotational theory, and deflationary theory of truth.
Abstract (2004), ‘Towards a naturalistic conception of existence’
Yang (2007a) is a substantial refinement of this paper.
Abstract (1997) 《一個投射論者的存在概念》(A Projectivist Conception of Existence)
以歷史的角度觀之，有兩組不同但密切相關的進路環繞存在概念的爭論。第一條進路集 中於存有什麼（what there are）的問題。這樣的進路已有悠久的歷史，特別可追溯到共 相問題的爭論。當然，如果我們有一個決定性的論證足以顯示「存有什麼」，所有的存 在語句能有一個確定的真值。拜 Quine 之賜，因他有名的 ontological commitment 之 判準──「存在即是一個（約束）變元的值」─為這個進路帶來新的高峰。第二條處理 有關存在爭端的普遍進路，著眼於存在語句的分析，從而希望能澄清「存在」（exists 及 其同源字）一字所扮演的角色。自從康德宣稱存在不是一個述詞，這條進路已在哲學研 究的舞台上佔得一席之地。近這幾十年來，這條途徑上哲學家們所付出的心力已蔚為主 流。存在語句的充分分析依賴於─在語言研究之中對名字（包括專名和普通名詞）的使
用作出令人滿意的考慮。然而，這兩條進路，就我所知，似乎都不夠充分地提供一個成 功的方法來處理有關存在真正概念的爭論。這篇論文裡，我試著提供一個觀點從這裡能 夠解釋存在語句的意義，雖然我們尚不能決定客體語句是否關乎存在或不存在。我將呈 現的觀點在本質上是投射主義式 (projectivistic)的。而我先從前述進路的簡短討論開 始，繼而說明存在的投射主義概念是什麼。 V. Expected development
At the present, I am revising some conference papers and hope that some of them can be published in some journals. However, all these efforts can be seen as preparation for the publication of a book. For the details, the content of the proposed book can be sketched as below:
Introduction
Chapter One Towards a naturalistic account of existence (based on 2007a)
Chapter Two Existence as a predicate: An application of Davidson’s theory of predication to the concept of existence (based on 2005a, 2006b)
Chapter Three Existence through language: To characterize the existence of individuals in terms of the projection of certain related truth conditions of existential sentences (based on 2007, 2007b, 1998)
Chapter Four Existence in modal discourse (based on 2008, 2008a) Chapter Five Existence in fictional discourse (based on 2006a) Chapter Six Existence in virtual reality (based on 2005b) Conclusion
赴國外研究心得報告
計畫編號 NSC 95-2411-H-002-001 計畫名稱 存在的投射概念:戴維森式的存在概念(3/3) 出國人員姓名 服務機關及職稱 楊金穆 國立臺灣大學哲學系暨研究所副教授 出國時間地點 96 年 7 月 5 日至 96 年 9 月 13 日國外研究機構 英國牛津大學歐雷學院(Oriel College, Oxford)
工作記要：
本人於 96 年 7 月 5 日至 96 年 9 月 13 日至英國牛津大學進行研究訪問。本人抵達英國 牛津大學後，隨即至歐雷學院(Oriel College)報到，並拜會歐雷學院副院長 David Charles 教授。 本人與 Charles 教授除討論本人 ‘Coming into existence through language’(2007b)一文之構想 外，並討論邀請 Charles 教授來台訪問之可能性，因本系擬於下學年度籌辦維根斯坦哲學會 議，因 Charles 為知名維根斯坦專家，故擬敦請 Charles 來台為講者。Charles 教授對此一邀訪 亦表示欣然接受。
隨後本人前往拜會 David Bostock 先生。Bostock 先生為牛津大學哲學系退休教授，本人 與 Bostock 討論本人關於此計劃之一些文章，包括 ‘Existence as a primitive concept’(2007a)， ‘Coming into existence through language’(2007b)， ‘Truth and existence in fiction’(2006a)， ‘Existence is a predicate: A Davidsonian account of existence and predication’(2006b)等文章。此 外，Bostock 亦與本人討論其新書 Philosophy of Mathematics: An Introduction 中的想法。
本人於訪問期間，亦拜會 Timothy Williamson 教授，並討論本人‘Truth in virtual reality’(2005b)，‘Theroleoftruth in theory ofexistence’(2005a)，‘Truth and existence in fiction’(2006a)，‘Existence as a primitive concept’(2007a)，‘Coming into existence through language’(2007b)等文章中的想法。此外，本人亦與 Williamson 教授討論關於 de re modality 與 conditionals 等哲學與邏輯問題。
於訪問期間，本人亦拜會歐雷學院其餘資深教授，並參訪其學院之圖書館與哲學系圖書 館等。並於 9 月 13 日返台結束此一訪問行程。
