三角測量戴維森的真理、客觀性和意義概念(1/3)

全文

(1)

行政院國家科學委員會專題研究計畫 期中進度報告

三角測量戴維森的真理、客觀性和意義概念(1/3)

計畫類別: 個別型計畫

計畫編號: NSC91-2411-H-004-006-

執行期間: 91 年 08 月 01 日至 92 年 07 月 31 日

執行單位: 國立政治大學哲學系

計畫主持人: 林從一

計畫參與人員: 兼任助理:李稟潔 王建功(2002/8/1-2003/3/31) 簡淑雯

(2003/4/1-)

報告類型: 精簡報告

處理方式: 本計畫可公開查詢

中 華 民 國 92 年 5 月 26 日

(2)

報告內容

(一) 前言

本研究為三年(91-93)期計劃,本(91)年度的計劃主題為戴維森的客觀性概

念,規劃所欲回答的具體問題包括:(優先性依序遞減)

a. 針對的戴維森的客觀性概念在他的語意理論裡扮演什麼樣的角色?

b. 在客觀性乃至於語意的形成中,第二人乃至於社會性因素是否如戴維森所

言必須扮演角色?如果是,扮演什麼角色?

c. 戴維森的三角測量(triangulation)隱喻和其他的三角測量理論(如騅斯機和布

蘭登的理論)之間有何不同?它在戴維森的哲學中扮演何種角色?如何扮

演?

d. 戴維森的三角測量隱喻是否有丐題的問題?

e. 戴維森的三角測量隱喻是否與戴維森其他重要哲學理論不一致?

針對這些問題及其他相關問題,本人撰寫以下四篇會議論文加以探討之

1.

Chung-I Lin (林從一), "Davidson on Objectivity", presented at: Conference on

Analytic Philosophy, Department of Philosophy, National Chengchi University,

Taipei., March 30, 2002.

2.

Chung-I Lin ( 林 從 一 ), "External vs. Internal Interpreters", presented at:

Conference on Davidson's Philosophy, Department of Philosophy, National

Tsing Hua University, Hsinchu, September 19, 2002.

3.

Chung-I Lin ( 林 從 一 ), "Davidson's Triangulation and the Normativity of

Meaning", presented at: Beijing International Symposium On "Wittgenstein and

the 20

th

Century Analytical Philosophy", Peking University, Beijing, October

11-14, 2002.

4.

Chung-I Lin ( 林 從 一 ), "Triangulation: World, Society and Meaning

Normativity", presented at:To be presented at: Conference on Neo-Pragmatism:

Davidson, Institute of European and American Studies, Academia Sinica,

Taipei, December, 21, 2002.

這四篇論文的主題實質上是相同的,它們皆探討戴維森如何以他的三角測量來

說明語言及思想的客觀性和規範性所必須具有一些重要條件,譬如說,一個社

會性的語言溝通脈絡。雖然這四篇論文的主題相近,但是歷次會議的討論,使

得各論文的內容逐次獲得實質改善。於中研院歐美所新實用主義研討會上發表

之"Triangulation: World, Society and Meaning Normativity",比前三篇論文獲得

(3)

更堅實的結論。

(二) 研究的主要發現

本研究的主要發現如下。我們發現三種對戴維森三角測量的既存的或可能

的詮釋皆不恰當或使得三角測量無法達到它的目標。但是,我們發現在一個較

廣的戴維森哲學脈絡中,亦即,輔之以戴維森其他重要哲學主張,將使得三角

測量成功。

戴維森早期關於三角測量的論述容易讓學者陷入以下的詮釋模式。學者一

般認為戴維森藉由三角測量進行以下的論證:所有命題思想或言說都具有特定

的內容,或具有意向性;然而,詮釋者必須決定導致詮釋對象某類行為的導因,

才能決定思想或言說的特定內容;但是,如果某人不與他人在一個分享的世界

中進行語言溝通,那麼詮釋者將毫無根據說詮釋對象的行為的導因為何,例如

我們無法說他的某個想法是關於某個物理對象(類型)或是關於他神經系統上

的某些刺激物(類型)

,因此無法說詮釋對象擁有具有特定內容的思想及言說;

所以具有命題內容的思想或言說要求(require)社會性(人際(interpersonal, I-you))

的溝通。以這個論證為基礎,思想及言說的客觀性和規範性得到適當程度的說

明。以客觀性為例。客觀性之所以能獲得說明是因為,如果三角測量是成功的

話,思想及言說一般而言是關於那些獨立於行動者而存在的物理對象,因此思

想和言說的真假大體而言決定於那些獨立存在的物理對象,也因此我們有合理

的基礎說,任何一個信念皆有錯誤的可能。

這種詮釋方式將三角測量視為一種因果測量,這種因果測量的目僅僅在於

幫助詮釋者去定位他的詮釋對象的反應的導因。在這個理解下,三角測量的立

即論證標的是:除非一個詮釋者的詮釋對象至少包含兩個在同一世界中(並且

針對此世界的事物)進行互動的生物,否則深度的歧義問題,亦即詮釋對象的

導因位於何方的問題,原則上將無法被回答,詮釋對象的意向性狀態的內容為

何的問題,也將無法被回答。但是,在這個解讀將使得三角測量所面對的問題

成為一個不存在的問題,亦即,對於一個面對一個孤立的生物的詮釋者而言,

深度的歧義根本不是一個問題。本研究發現這個以思想或言談的特定內容為切

入點對三角測量的解讀方式是不適當的,它弱化了戴維森的論證,使得三角測

量過於容易被擊敗,它忽略了戴維森論述中的一些重要訊息,使得三角測量的

一些哲學含蘊隱而不彰。舉例來說,如果三角測量的目的在於使得界定特定內

容成為可能,那麼另一個感覺器官便足夠扮演戴維森三角測量中第二人(a second

person)的角色,思想和語言的社會性便無法被論證出。在舉例來說,戴維森不

止一次強調,他以三角測量所要處理的問題是,行動者視某物為某類物時,涉

及了什麼要件。這個問題不是關於,從詮釋者的角度來看,行動者的思想或言

說如何具有特定內容,而是關於,從詮釋者的角度來看,行動者如何能具有將

事物歸類的能力。面對一個孤立的生物時,詮釋者或許可以具有必要的資源回

(4)

答第一個問題,但不一定具有必要的資源回答第二個問題。

能將事物歸類,就是具有概念能力,就是具有遵循規則的能力,就是具有

掌握正確與不正確之間的區分的能力。戴維森的三角測量事實上直接針對的問

題是,上述能力所必須具有什麼要件。晚近一些戴維森的詮釋者或批評者看出

了這一點,但是他們的詮釋仍舊走岔了路。

許多晚近的戴維森的批評者認為,戴維森的三角測量主要處理的問題是所

謂的範疇的歧義問題,更精確地說,此問題的核心為,詮釋者必須獲得什麼樣

的資源,才能確定被詮釋對象使用某個而非其他的概念,或遵循某個特定的規

則。戴維森的三角測量所欲達成的結論被理解為,除非詮釋對象與其他詮釋對

象在同一個環境中進行語言互動,否則詮釋者沒有任何基礎說,詮釋對象的行

為是一個遵循規則的行為,而非只是一個傾向性、無概念性的行為,因為,詮

釋者無法說他的詮釋對象遵循了什麼規則。在這個解讀下,批評者認為戴維森

的三角測量無法達成它的目的,其主要的理由可以整理歸結為:三角測量的策

略將會面臨以下這個類似兩難的處境。一方面,如果詮釋者所能獲得的詮釋資

料只不過是詮釋對象的外顯行為資料,特別是詮釋對象和外在世界的互動,在

這樣的詮釋角度下,那麼加入數個彼此互動的被詮釋者,並不能根本地脫卸他

們的行為不是意向性行為的疑慮,亦不能限定住他們的行為體現出麼樣的概念

運用。另一方面,由於為了避免丐題,詮釋者似乎不能預先假定詮釋對象具有

概念,那麼在這種情形下,詮釋者必須訴諸他自己的範疇或概念,來說被詮釋

對象是根據什麼範疇來歸類事物,如此便產生了幾個疑慮:第一,如果詮釋者

可以合法地訴諸自己的範疇來回答,被詮釋者根據什麼範疇作出行為這個問

題,那麼我們沒有理由說,面對單一個詮釋對象時,詮釋者無法根據相同資源

回答相同問題,這意謂被詮釋對象無須多於一人;第二,我們如何能確定詮釋

者加諸於被詮釋者行為的範疇是被詮釋者本身所具有的範疇?詮釋者與詮釋者

之間似乎有可能具有不同範疇。第三,範疇既然是詮釋者施加於被詮釋者的行

為上以理解那些行為,那麼這裡總是存在著被詮釋對象的行為只是傾向性的行

為,這個疑慮不會因為多加入些被詮釋者而排除。除此之外,相當重要的是,

範疇歧義問題一開始對戴維森就不該是一個問題,這是因為戴維森拒斥徹底差

異的概念網落的可能性。

第三個詮釋輔以三角測量兩個理論。第一個理論是意向性超過傾向性,而

這個理論蘊含與詮釋相關的事物之間相似性判斷必須是從被詮釋者的角度作

出來的,亦即詮釋者眼中的相似性。這點事實上蘊含,被詮釋者具有錯誤的概

念。另一個想法是,被詮釋者必須感知到相似性。以這兩個理論為背景,我們

將三角測量理解為對以下想法的一個論證:某個生物除非與另一個生物針對共

同的對象進行互動,否則將缺乏足夠的必要規律性讓它去感知以產生一個歧出

的行動的概念。但是,這個解讀亦不能使得三角測量盡其功。為省篇幅,在此

不陳述我們的理由,評審者可見附件一。

(5)

雖然這些解讀都不能使得三角測量達到它的目的,但是,我們對它們的探

究仍顯現一些與解讀戴維森三角測量相關的要點。第一,如第一個解讀所強調

的,三角測量所要解決的一個問題,是關於如何定位不同反應所共有的導因的

問題。我們可以合理地設想,共同導因這個概念在戴維森的三角測量和他關於

規範性的理論中扮演一個重要角色。第二點是意向性超過傾向性,特別是這點

所蘊含的這個要點:除非一個生物具有歸類事物的能力,否則它不能被詮釋為

一個具有意向性狀態的生物。這兩個觀察加上一些戴維森對意義的看法,我認

為將能讓我們構作一個脈絡,在這脈絡中三角測量將能說明意義的規範性和它

的社會面向。

在我說明我的提案之前,必須先說明兩點。第一,我將在一個戴維森對意

義的一般性看法下解讀三角測量。戴維森以三角測量所關心的主要問題是,意

義的什麼構成要素使得意義的規範面向成為可能?為什麼一個分享世界的語

言互動者,對於任何一個人之具有意向性是必要的?第二,明顯的,分類的能

力和具有錯誤的概念是相互依憑的。這使得我們得以將三角測量解讀成一個處

理一般性客觀性問題的方式。所謂一般性客觀性問題是,為何任何一個基本的

性念都有可能錯誤?更具體地說,三角測量將被理解為一個方法,去說明意義

的什麼構成要素使得一個人理解他的任何一個信念都有錯誤的可能。簡言之,

三角測量將被理解為回答以下問題的方式:意義的什麼構成要素,使得一個人

能理解他的任何信念都有錯誤的可能。這也就是說,三角測量是一個方法,讓

我們理解,為何錯誤的概念和其他相關的概念,如真理的概念,可以運用到我

們任何一個信念上。

如此一來,我們可以說,三角測量所關心的具體問題是,意義的什麼面向

使得一個孤立的生物,一個沒有任何其他處於同世界的生物與之中進行語言溝

通的生物,是不可能理解他的任何一個信念都有可能錯誤?

關於內容,戴維森有一個著名的主張,他認為一個信念不會具有任何內容,

除非它內建於一個一般而言一致且融貫的信念網絡中。在這個想法中,一個人

如果是理性的,亦即具有任何意向性狀態,那麼他的信念彼此之間一般而言就

必須具有邏輯和認知的關係。這個限制是一個先驗的限制,因此任何一個信念

系統都大體上是一致而融貫的。此外,身為一種外部論者,戴維森也主張,我

們「最平常及最基本」(1983:317)的信念的內容與真實環境中導致那些信念的

事件彼此之間具有構成性的連結。這個外部論的想法加上上述關於語意整體論

的想法,對於信念產生一個構成的原則,它要求我們大部分的信念不僅彼此之

間大體為一致的,而且它們大體上皆為真的。語意整體論加上外部論,排除了

任何一個系統具有廣泛錯誤的可能性。這是一個有力的論證,而我要進一步宣

稱的是,正是由於這個原則,使得一個孤立的個體無法理解它的任何一個信念

都有錯誤的可能。

從一個孤立的信念系統之內來看,將無法產生以下這個理解,理解這個信

(6)

念系統內的任何一個信念都有錯誤的可能。這是因為,根據上述原則,任何一

個系統事實上都大體為真。從一個信念系統之內來看,一個信念只能從它是否

與同系統內的其他為數極多的信念相互一致,來判斷它是否為真,而那些為數

及多的信念的真是不能被質疑的。我們必須將我們大部分的信念毫不質疑的視

為真,這個一般性的事實顯示,從一個信念系統內部而言,我們將缺乏一個一

般性的根基作為對比,讓我們能說這系統中的任何一個信念都有錯誤的可能。

或許此時有人會認為,外在世界能從信念系統之外幫助我們解決難題。有

人或許會論證說,既然我們大部分的基本信念是導致於外在事件,這些事件的

堅實性質一但成為信念的導因,它們將持續決定我們基本信念乃至於衍生的信

念的真假。最後這點或許是真的,但是就算它是真的,它也不蘊含,外在世界

可以成為一個孤立的生物產生客觀性概念的一個一般性的對比基礎。這是因

為,外在世界可以扮演那個一班性的對比角色,只有當外在世界是從一個所謂

的上帝之眼的角度(God’s-eyes-view)被觀察,而這個觀點當然是戴維森所不允

許的。對戴維森而言,我們不可能獨立於我們自己的信念系統,以無涉信念的

方式描述世界,來產生形成客觀性概念的一般性對比基礎。我們所描述的只是

我們所相信的,我們只不過再度掉入我們適才所描述的處境中罷了。

這裡是社會性因素必須進入意義的地方。上述的討論迫使我們處在一個位

置上,讓我們只能說,信念之間一般性的對比基礎,只能存在於分屬不同信念

集合的信念之間。換言之,一個個體要能具有客觀性的一般概念,亦即理解他

的任何一個信念皆有錯誤的可能性,必須至少有一個不同的信念系統,讓他能

拿之與他自己的系統相比較。但是,那種對比只有在他能理解其他個體的言談

才是可能的。“We grasp the concept of truth [and the related concepts, such as the

concept of error] only when we can communicate the contents - the propositional

contents - of the shared experience, and this requires language.” (Davidson, 1997:

27) As early as 1975, Davidson has already said: (170)

We have the idea of belief only from the role of belief in the interpretation of

language, for as a private attitude it is not intelligible except as an

adjustment to the public norm provided by language. It follows that a

creature must be a member of a speech community if it is to have the

concept of a belief. And given the dependence of other attitudes on belief,

we can say more generally that only a creature that can interpret speech can

have the concept of a thought.”

要理解他人的言談,我們必須理解,那些言談與我們於類似處境種所作出

的言談,具有相同的主題。在語意外部論的基本想法下,要理解他人的言談與

我們的言談具有相同的主題,我們必須理解那些言談〈至少那些基本的言談〉

是導致於一些相同的事物。而要理解他人的某些言談語我們的某些言談具有相

同的導因,我們必須理解我們是處於一個人際之間的三角測量處境中。

(7)

現在應該很清楚了,並不是所有的三角測量皆有助於我們說明客觀性概念

的產生。只有那種有助於語言溝通者決定他們思想或言談的共同導因的三角測

量,才有助於說明客觀性概念的產生。人際之間的三角測量幫助語言溝通者將

他們的言談連結到外在世界中的相同面向上,而這些相同的面向成為他們溝通

的共同主題。很明顯的,共同導因和相同主題這個概念在戴維森對客觀性的說

明中,扮演關鍵角色。承認他人與你所談論的對象是相同對象,就算是以不同

的字詞來言說,同時也就承認他人處在一個可以反對你的言談的位置上。易言

之,每當你認為他人所談論的與你所談論的是相同事物的時候,你都至少隱含

的承認,在那個時候你相信為真都有可能事實上是假的。

一個徹底詮釋者一開始所能獲得的詮釋證據,僅僅是詮釋者所顯示出來的

傾向性反應。詮釋者當然能感知並判斷這些反應彼此之間的相似性關係,根據

這些判斷,他能說被詮釋對象的某些反應是正確的某些反應是錯誤的。然而,

當詮釋者理解只要那些反應彼此之間的相似性全然是根據他自己的相似性標

準作出的,他將會理解他的詮釋釋過度的,因為有可能那些反應不是概念性的

反應。我們的詮釋者必須將那些使得他能做出相似性判斷的能力讓渡給他的詮

釋對象,而無論這個讓度所需的條件是什麼,我們至少清楚的是,要使的詮釋

持續進行,詮釋者必須一開始就預設,如果他的詮釋對象具有任何意向性狀

態,它就必須具有感知何判斷相似性的能力。

此外,藉由反省意義的基本性質,我們的詮釋者也應瞭解以下關於詮釋的

事實。一個生物具有客觀性的一般性概念時,它必須理解它所具有的任何一個

信念都有錯誤的可能,而如果它從未與其他生物進行溝通,它將不能產生那樣

的理解。它必須至少與另一個生物進行過語言溝通,至少必須存在著另一個生

物以不同於它的方式擇出相同的對象作為溝通的共同主題,它才能理解它任一

個信念都有錯誤的可能。而對外部論者而言,說兩個信念關於相同事物,等於

說它們可靠地導致於類似的對象,而一個個體要知曉另一個個體同它一樣針對

相同的對象作出反應,它必須能做出如戴維森所描述的那種三角測量。事實

上,如戴維森所言,那個個體必須能從那第二人、對象何他自己之間的那三角

聯繫中獲得「認知的利益」(cognitive advantage)。獲得那樣的「認知的利益」

是理解它與它的語言溝通者談論的是相同的對象,而那樣的理解涉及了另一個

理解,亦即,理解它的任一個信念皆有錯誤的可能。因此,我們的詮釋者最後

理解了,他的詮釋具有一個不可避免的預設,他的詮釋對象必須最少包含兩個

進行語言溝通的生物。

當然,要能從對三角測量獲得「認知的利益」

,或甚至要能進行語言溝通,或

甚至要能三角測定自己的和溝通者的言談的共同導因,已經預設了具有概念的

能力。戴維森的說明的確如他自己所言存在著循環的問題(“move in a circle”)。

但是,循環的問題所必定阻礙的是化約的企圖,它不一定對戴維森造成傷害,

因為戴維森相信,理解哲學基礎想法的唯一方法,或許只能透過將基本概念之

(8)

間的連結顯示出來。

在我們對戴維森關於客觀性想法的說明中,意義的規範性和社會性因素的

必要性,事實上是獨立於三角測量以一個先驗的方式從意義的性質導出來的,

但是,我們必須透過三角測量,才得以理解那社會性因素如何具體成為意義規

範 性 的 構 成 要 素 。 此 外 , 我 們 的 解 讀 , 並 不 預 設 語 意 的 外 部 論 (semantic

externalism),事實上它可以說提供了語意外部論一個獨立的說明基礎。這一點

將是相當有趣的,因為訴諸語意外部論來說明語言的客觀性,難免被攻擊為丐

題之舉。

(三) 計劃成果及進度自評

本研究本年度規劃所欲回答的具體問題包括:

a. 針對的戴維森的客觀性概念在他的語意理論裡扮演什麼樣的角色?

b. 在客觀性乃至於語意的形成中,第二人乃至於社會性因素是否如戴維

森所言必須扮演角色?如果是,扮演什麼角色?

c. 戴維森的三角測量隱喻和其他的三角測量理論(如騅斯機和布蘭登的

理論)之間有何不同?它在戴維森的哲學中扮演何種角色?如何扮

演?

d. 戴維森的三角測量隱喻是否有丐題的問題?

e. 戴維森的三角測量隱喻是否與戴維森其他重要哲學理論不一致?

基於本研究所作出的四篇會議論文,給於後四項問題相當實質的回答,並

對戴維森相關哲學思想有實質的新發現。本研究計劃提及,本人將於三年

計劃完成時,將研究成果集結成書,書名現暫定為『三角測量戴維森的客

觀性、真理和意義概念』

。此書現已完成大部草稿,其中的導論部份,對於

問題(a)多所著墨。整體而言,本研究達成甚至超過原計劃預期進度。為取

信評審者,特於報告後附上(1) "Triangulation: World, Society and Meaning

Normativity"一文及(2)上述書籍導論草稿。

附錄一

(9)

Chung-I Lin (林從一)

Department of Philosophy National Chengchi University

Taiwan

To be presented at:

Conference on Neo-Pragmatism: Davidson Institute of European and American Studies

Academia Sinica, Taipei December, 21, 2002

Direct Correspondence to: Lin, Chung-I (Assistant Professor) Department of Philosophy National Chengchi University No. 64, Sec. 2, Chih-nan Rd. Taipei, 115, Taiwan.

Tel: 886-2-29393091 ext. 88146 Fax: 886-2-29390514

Email: cyberlin@nccu.edu.tw

Triangulation: World, Society and Meaning Normativity

Chung-I Lin (林從一)

Department of Philosophy, National Chengchi University

1. Introduction

Thinkers create and entertain meaning; but thoughts and whatever they issue or conspire are bound by meaning. The bounding is normative, in the sense that meaning creates a space where thought and action can be said to be correct or incorrect. What is it about meaning that makes the normative dimension of meaning possible?

It is well known that Wittgenstein dismisses three accounts of the normativity of meaning (see, e.g.,1958: §§ 139-42; §§193-4; §§ 218-9): the first bases it in private mental state of individual

(10)

speaker; the second has it homed in a use-transcendent, platonic realm; the third trades it in by regularity described in purely behavioralistic-naturalistic terms.The first two accounts conceive the mental state or the platonic entity as essentially possessing ultimate content, only in connecting to which can things acquire the contents they have. However, any of such ultimate contents is liable to further various interpretations for its applicability, since such a content is supposedly existent in isolation from its particular application and, thus, does not carry within itself its applicability to special cases which is infinite in number. The two accounts would then either lunch a regress or end up with a termination of the very distinction between correct and incorrect linguistic practices they seek to explain. The futile of the first two accounts naturally makes the last account tempting. However, it also faces similar paradox and regress. Any set of verbal behaviors in this conception can be found regular according to some classifications, but irregular according to others. Seeking an answer to the pending question of which classification is correct spoils the spirit of the account and will further hang the normativity of meaning in the air. It does not fundamentally change the situation by appealing to regularity of external events causing those behaviors, or regularity of crowed identical behaviors, if the regularity is still couched in purely behavioralistic-naturalistic terms.

In fact, we may say that the behavioralistic-naturalistic account misplaces the subject matter at the very beginning. Explaining the normativity of meaning in terms of regularity is on the right track, but we have a good intuition that mere regularity does not manifest normativity. Daily sun rises and sets do not enter the domain of normative evaluation. It is certainly true that picking out any regularity involves classification and classification is essentially a matter of conceptualization, nonetheless we do not intend to discover anything conceptual on the part of the system depicted solely in behavioralistic-naturalistic terms. A concept user not only manifests in its behaviors reliable dispositional responses to stimuli, but also essentially embodies in its regular behavior application of concept to stimuli, namely, attitude of taking some stimuli as similar with respect to a concept. The regularity that matters here should be normative regularity, regularity constrained by the meaning of the term, or more specifically, by the agent’s own understanding of the meaning. To release off this constraint is to down grade conceptual behaviors to mere dispositional responses.

The three Wittgensteinean critiques suggest respectively the following three morals in regard to the nature of meaning. The first is the externalist thought that meaning must be anchored to something beyond individual speakers. The second is the pragmatism spirit that ongoing linguistic practices must be there from the start for meaning to emerge. The third is normative regularism, as we might call it, that regularity of linguistic practice, and any regularity internally connected to it, is essential to an explanation of meaning, especially its normativity dimension, but the relevant regularity must be made from the concepts in the eyes of their beholders.1 We can expect that an account of meaning incorporating the three morals, or at least part of them, might render a satisfactory explanation of the normativity of meaning.

Some philosophers attempt to explain the normativity of meaning on the basis of the first two morals -- semantic externalism and pragmatism. They, however, carry out the attempt in different ways -- usually, they have various conceptions on the first moral, and therein tend to assign different theoretical weights to the second moral. Some externalists (Burge (1988) may be a case in point), whose position might, following Davidson, be dubbed perceptual externalism (2001b: 2), regard event or object in the external world as essential for the constitution of meaning and hence for its normativity. It is thought that the when causal correlation between some external events (of the same kind) and the uses of a word (of the same kind) is nailed down, the content of the word is determined and the standard of its further uses is set up. In Pears' words, when we set up a language "the criteria

1 The ideas of normativity of meaning and normative regularity are interrelated. One’s taking some stimuli as similar with respect to a concept involves at least the understanding or awareness that one's so taking is under the constraint of the concept. It is in effect an understanding that what one takes to be correct might be not. Thus, the understanding must be involved with the distinction between

correctness and incorrectness, which is just the main feature of the normativity of meaning. On the other hand, as the notion of meaning entity and the idea of truth as correspondence are rejected, the correctness and incorrectness distinction has its genuine application only in the case where an action is judged to be similar or not similar to some previous actions regarded as similar in accord to a rule or concept.

(11)

of identity of objects and their types are simply imposed on us" (1987: 30); "when a name is attached to a thing, the nature of the thing takes over and dictates its subsequent use" (1987: 65).2 In this conception, linguistic practice of speaker is mere a complementary item needed for the completion of a causal correlation with external thing. On the other hand, for those externalists, whose position might be called in Davdison's term as social externalism (2001b: 20, emphasizing on social aspect of meaning, group behavior plays the crucial explanatory role on normativity. They take uniformity of linguistic behavior among speakers to be all that are needed for the normativity of meaning; causality of events or objects in the external world fade into or even fade out from the background of discussion. Dummett (1993) and a version tentatively promoted in Kripke's [1982] and are representatives of this view. However, both theoretical orientations take the third moral very lightly, if not completely ignored. This might due to their proscribing anything intensional to be a basis in their explanations.3 With the notion of normative similarity missing from the scene, an explanation of the normativity of meaning cannot be satisfactory. The crucial defect of an account ignoring the moral is that it will fail to distinguish clearly, in Davidson’s terms, the difference between “having a concept and simply having a disposition.” (2001b: 4).4

However, it is not easy, if not of formidable difficulty, to come by a coherent picture weaving up the three morals in whatever version they are conceived. There are no jump-out conceptual connections among them -- externalism and the notion of normative similarity look even incompatible with each other. Davidson agrees in general, not necessarily in the same terms and for same reasons, with the three Wittgensteinean critiques and the respective lessons (see, e.g., 2001a:37; 2001b: 2-3). But, in a series of papers from 1982 on,5 Davidson endeavors to develop an externalistic picture aiming to make it clear the connections among the three morals and,thereby, to make intelligible the notion of the normativity of meaning.6 The crucial element of the picture Davidson gives us is his famous triangulation.

Davidson’s remarks are often succinct; crucial points sometimes are hidden between lines from commentators. It may, therefore, be a subject worthy of curiosity to inquire the nature of the triangulation to see whether and how it may shade lights on, if it may, the interconnections among semantics externalism, pragmatism, and the notion of normative similarity. There are two major parts in my paper. The first is both an exposition and an evaluation. Several existent and possible readings on the triangulation are put on the table to examine. It is found that the triangulation, in whatever version considered in our path, stands by itself suffers from various defects and fails to its immediate purpose, which is to vindicate the daring idea that a world-shared linguistic interlocutor is necessary for any one to have contentful state. But the discussion also paves a way, together with the resource available from some of Davidson’s general ideas of meaning, allowing me to venture, in the second part of this paper, an account in which the triangulation would gain a position adequate to its purpose.

2 Pears' remark is attributed to Wittgenstein's Tractatus. I borrow Pears' words from Child (2001: 30). 3 In fact, both accounts presuppose the unexplained notion of normative similarity. Any set of events will have endless properties in common. Talks of similarity make sense only by relating to concepts. 4 The basic points of what I have said so far can be found in Davidson’s “Externalisms” (2001b: 1-13) He criticizes two externalistic approaches on two points. First, they all fail to explain the notion of normative similarity, and, therefore, fail to make a clear distinction between “having a concept and simply having a disposition” (2001b: 4). Second, without an answer to the question of normative (relevant) similarity, the notion of causality or regularity alone does not suffice for an account of content.

5 To mention only a few: [1982], [1983], [1989], [1991a], [1991b], [1991c], [1992], [1993], [1997], and [2001b].

6 It should become clear as our discussion unfolds that the Davidson’s externalism is different

significantly from the foregoing kinds. The account, like perceptual externalism, takes external cause to be essential to meaning constitution, but, unlike perceptual externalism, it proposes that, without a social mechanism of linguistic interaction, external cause will fail to fulfill its constitutive role of meaning. On the other hand, Davidson’s externalism differs from social externalism in that the former requires social interaction, but the latter requires only group behaviors, for a satisfactory explanation of meaning.

(12)

2. Basic strategy

Davidson's triangulation admits different conceptions. One reason is that the triangulation is utilized by Davidson to advocate various basic but interrelated ideas, including, in part, the interdependence of language and thought, the sociality of intentionality, the interdependence among knowledge of one’s own mind, other minds, and the external world. But, for our purpose, we choose to conceive it as partaking in an argument for the following thesis: (see, e.g., Davidson, 2001a: 129): (E) [1] One cannot have any intentional state if one does not acquire the concept of error, [2] which one cannot have unless one is in linguistic communication with a real second person.

Davidson says very little about thesis [1], but has the triangulation for the plausibility of thesis [2]. To have a germane understanding of the triangulation, one has to take into seriously consideration its purpose, the key concept of which is the concept of error. The concept of error, according to Davidson, comes in a package with the concept of truth, the concept of belief, the concept of objective world, the capacity of judging similarity and mastering the distinction between correct and incorrectness, and some other concepts. It is thus a good way to understand the concept of error by sketching out its connections to the other concepts. The concept of truth is a good place to cut in, since for Davidson there is a close and obvious connection between the concept of truth and meaning.

Davidson’s conception of the concept of truth is meant to capture an essential feature of human cognition. A human being not only has the mental capacity to represent the world, but also has the mental capacity of appreciating that he is doing so. The latter kind of cognitive capacity is not the kind of disposition to directly respond to whatever comes up in the environment, but a kind of capacity to qualify some direct representations by way of comparing them with some concept, norm or principle. The mental capacity of comparing direct representations with some norms requires that the agent knows that the direct representations of the world are as the agent takes them to be, not as the way the world is. In other words, the agent must know that some representations are merely subjective and hence subject to error. Knowing something to be subjective and knowing something to be objective are two faces of a single coin. Thus, One cannot have the ability of appreciating that one is thinking of the world unless one is in command of the distinction between the objective and the subjective. In Davidson’s words, one cannot know one is in a state of believing unless one has the concept of truth.7

Let's put the point from a different angle. Human cognition not only represents things in accord with rules, but also applies rules to representations.8 We apply rules to representations because we know that our representations could be wrong. A rule is a normative conception, by means of which can one distinguish what is correct from what is wrong. One cannot know one has a belief without apprehending that the belief could be wrong. Davidson would say that if I have a belief that it is raining, since I know that it is a belief, I must know that it could be wrong. In its nature, the concept of belief is intrinsically connected to the concept of truth and the concept of error. In short, Davidson connects the concept of truth (and the concept of error) with the concept of belief, and takes them as essential features of thought.9 We have seen that for Davidson, the concept of error, the concept of belief, the concept of truth, the concept of objective and the concept of a norm (and some others) are all intimately connected to each other. One cannot have one without any of the others and each of them is necessary for one to have thoughts and meaningful talk.

One more stroke is needed to the portrait of the purpose of the triangulation for our latter

7 As we have pointed out, Davidson has said little about thesis (1). But, if we consider the fact that human cognition is a process of constantly evaluating the quality of whatever comes to mind, it is hard to imagine that such a kind of cognitive capacity would be possible if we were not aware of the fact that we are thinking and believing and that we are consequently subject to error. As for linguistic practice, it is hard to imagine that one could have a language without the ability to apply rules that would distinguish the correct applications of words from the incorrect ones.

8 The formulation of the two concepts is directly borrowed from R. Brandom’s [1997].

9

A qualification is needed here. In attributing Davidson with this view, we do not thereby force upon him the thesis that there are rules, norms, or the like, existing prior to successful communication which impose standards dictating what is correct and what is not.

(13)

discussion. Comparing direct representation to concepts can be understood as applying concepts to stimuli. And applying a concept to stimuli involves judgement on the similarity among them according to the concept. Correspondingly, (E) is, in effect, equivalent to the following thesis:

(N) One cannot have any intentional state if one does not have the capacity of judging similarity, which one cannot acquire unless one is in linguistic communication with a real second person in a shared world.

With the purpose of the triangulation set, I choose, for two reasons, to probe the triangulation by conceiving it as a continuance of Davidson’s project of radical interpretation. The first is a general reason that both of them are intended to reveal constitutive elements of meaning. It is well known that Davidson employs radical interpretation is to reveal some crucial elements involved in the constitution of meaning. And, as (N) indicates, the triangulation is meant to vindicate the idea that a world-shared linguistic interlocutor is necessary for the constitution of anyone’s capacity of judging similarity and hence intentionality.

The second and a more specific reason is that both are intended, partly, to show crucial elements of meaning by way of manifesting what must be added to disposition to make it into conceptualization. Radical interpretation reveals crucial elements of meaning by revealing some elements that a radical interpreter has to attribute to her creature to make any sense of it. It is a kind of enterprise of constructing a theory of meaning the evidence for the justification of which must be “described in non-semantical terms” (Davidson, 1984b:142). By a way of finding out what must be added into the behavioral evidence to make sense of a creature, the radical interpreter reveals corresponding constitutive elements of meaning. On the other hand, Davidson, in “Externalisms” (2001b), places the triangulation in a picture to "make sense of the transition from mere disposition to conceptualization." (2001b: 8) His way of making sense of the transition from mere disposition to conceptualization is built on the difference between "an external commentator slipping in his categories to make sense of an isolated creature" and "a participant observing another participant doing his or its thing," (2001b: 8) (or, "seeing the relevant similarity of one animal's responses through the eyes of another animal.") (2001b: 12) The reason why an external commentator has to slip in her categories, standards of judging similarities, in her interpretation is that what her creature provides her is nothing but dispositional response as resource of interpretation, but dispositional response is too thin to conceptualization. The transition from disposition to conceptualization is a process, like radical interpretation, of finding out what must be added into the initial evidence to make sense of the creature. Eventually, Davidson arrives at the conclusion that, for interpretation and hence intentionality to be possible, a social setting must be added into disposition as its embodiment. I shall spell out the detail of the transition later.

An external commentator and a radical interpreter share with each other the same initial position and the general goal of interpretation – both have to begin with nothing but dispositional responses of their creatures to arrive at some general claims on the nature of meaning. But the former approach, with the triangulation as an essential part of it, reveals social elements of meaning that are not explicitly told in the project of radical interpretation. On this very general reason, we may expect that triangulation is an extended part of the project of radical interpretation.

3. The ambiguity of the concept of cause

So, let’s set the first scene of our survey be a special kind of radical interpretation, which consists of an interpreter and an isolated creature to be interpreted. The plan, to repeat, is that if it is shown that this kind of radical interpretation founders, then it is shown that something crucial for meaning is absent from the isolated creature case – something essential for the constitution of meaning must lay besides individual speakers. And if it is shown that a second creature needs to be add into the picture as a world-shared linguistic interlocutor to the first creature for the interpretation to get going, then sociality is shown to be an constitutive part of meaning.

A radical interpreter is an externalist who believes that what determines the content of an intentional state, utterance or thought, is partially but essentially involved with the external things

(14)

causing the intentional state.10 Thus, to make sense of a creature, our interpreter is required to be able to determine the typical external causes (hereafter, causes) of the creature’s intentional states. That is, she must be able to correlate the creature's prompted responses, especially linguistic ones, with the circumstance that elicit them. To recognize the correlation, she is further required being capable of not only perceiving similarity of his creature's verbal responses and similarity of the stimuli, but also making judgments on the similarities.

Davidson suggests that here our interpreter faces the problem that he calls “the ambiguity of the concept of cause” (2001a: 129), which, he thinks, will block the interpretation. According to Davidson, in the present case, the cause of a response of the isolated creature is doubly ambiguous. “The first ambiguity concerns how much of the total cause of a belief is relevant to content”; the second has to do with the ambiguity with respect to the location, whether it is distal or proximal, of the relevant stimulus. (2001a: 129) Without the problem of double ambiguity being solved, “there is no answer to the question what it is in the world to which we are responding” (2001a: 128) and hence there is no answer to the question what specific content of our intentional state is. Let’s call the first ambiguity the ambiguity of width, and the second, the ambiguity of depth.

The failure of resolving any of the two ambiguities will indeed render the content of an utterance indeterminate. They are nonetheless distinct -- one may persist when the other is removed. One might determine the relevant part or aspect of the total cause of a verbal response without being able to determine the whereabouts of the cause. On the other hand, one might find a way of determining the location of the cause of a verbal response without the part of the cause relevant to the response being delineated. Furthermore, the two ambiguities have bearings on different aspects of the issue of objectivity. Our beliefs can be said to be about things in the world external to us, and hence their truths to be determined by things independent of us, unless the location of the causes of our beliefs can be determined in the distal. Thus, an important sense of objectivity of content would be buried in the ambiguity of depth, even when we assume that the ambiguity of width is resolved. On the other hand, unless the relevant range of the total cause of a verbal response can be fixed, there is no way to tell whether a further response is to a similar cause and hence the distinction between correctness and incorrectness will be demolished. This is because that anything is similar with another thing in some respect. A different aspect, and perhaps a more fundamental one, of objectivity (and normativity) will be lost in the ambiguity of width even with the assumption that the ambiguity of depth is removed.

This is where Davidson explicitly wedges his triangulation in his explanation of the normativity of meaning. He introduces it to show why either ambiguity cannot be removed unless our interpreter has in her interpretation at least a world-shared, real, interlocutor to her previously isolated creature. To see why, for Davidson, the isolated creature case dooms at the two ambiguities, we have to see what, for Davidson, needed resource for the removal of the two ambiguities is available in the interpretations of multiple object in interaction, but not in the case of isolated creature.

Let’s consider the ambiguity of depth first, since it is the problem Davidson deals with when he first employed the triangulation. In fact, the triangulation is often formulated explicitly as a solution to the problem of depth, but not obviously to the problem of width (1991b: 159, emphasis mine; see also 1993: 263):

What seems basic is this: an observer (or teacher) finds (or instills) a regularity in the verbal behavior of the informant (or learner) which he can correlate with events and objects in the environment. This much can take place without developed thought on the part of the observed, of course, but it is a necessary condition for attributing thoughts and meanings to the person observed. For until the triangle is completed connecting two creatures, and each creature with common features of the world, there can be no answer to the question whether a creature, in discriminating between stimuli, is discriminating between stimuli at sensory surfaces or somewhere further out, or further in. Without this sharing of reactions to common stimuli, thought and speech would have no particular content - that is, no content

10 See Davidson, 1985: 478-80; 1989: 195; 1991a: 200; 1996: 161-2. The adoption of radical interpretation as the main strategy of our discussion presupposes the idea that meaning requires something beyond speaker’s skin. But the assumption is very minimal in that it does not involve what those external elements are and how they partake in the constitution.

(15)

at all. It takes two points of view to give a location to the cause of a thought, and thus to define its content.

The idea of the triangulation is this. Our radical interpreter observes that there are two causal lines, each of them consists of two sets of coverying events, stretching from near to far, correlating with some response(s). Then by intersecting the two causal lines the interpreter can pick out an event from both sets of coverying events as the common cause of the response(s) (if those responses have a common cause). The crucial but controversial point Davidson wants to made in the story is that it takes at least two creatures to establish the needed causal lines and the triangulation. If Davidson is right, then our radical interpreter must have as objects of her interpretation at least two creatures interacting with each other in a shared world for the determination of the causes and the references of their utterances, and the attribution of conceptual state, to be possible.

But why, in the first place, does the ambiguity of depth arise in the interpretation of an isolated creature? The answer indicted by the triangulation seems to be that it is the number of creature to be interpreted that somehow introduces the ambiguity. However, a careful inspection of the triangulation shows otherwise. The triangulation is a causal triangulation which, for our interpreter, consists of two causal lines from the two distinctive responses, issued at a given time, belonging respectively to her two creatures to a common cause. But, since it is a casual triangulation, what crucial for the determination of the common cause is the two distinctive causal lines, not the two different creatures. The second creature, in Davidson’s picture, functions only as a response generator, a point in a causal line. The second creature is substitutable by some other response generator, say, a sense organ. That is to say, what matters for the triangulation is the amount of different causal connections, not the amount of different creatures.

To put the point differently, the resource of the ambiguity of depth is not the singularity of creature but the singularity of causal connection between response and stimulus. Given a single causal chain type between a type of responses and a type of stimuli, there are various regularities, covarying within the whole causal chain type, that qualify as candidates for the stimuli. The regular stimuli that prompt the response might be the similarity of things at a distance, or the similarity of the stimulation of its nerve endings, or any of various regularities in between. That is to say, a causal chain type by itself is insufficient for the determination of the common cause of those response instances. The insufficiency indicates the ambiguity of depth.

Thus, to say that an isolated creature does not offer enough resource for the triangulation of common cause of responses is presupposing that the creature at a given time can generate at most only one casual response to a given object. But this presumption is simply a fantasy. A single creature could, of course, generate simultaneously two distinctive responses from two different sense organs to a common object. Material things are accessible to more than one of the senses at a given time. You can see a table, but you can also touch, taste, smell or even hear it at the same time. In fact, you can generate two distinct responses to an object with just one organ. You can touch it with both hands and see it through both eyes.

The ambiguity of depth so understood then will not be a real problem for our radical interpreter as an observer. The interpreter observes the causal connections between responses of two creatures and stimuli and correlates them to determine the location of the stimuli, but the causal connections of the same kind are available to, and similar triangulation can be done by the interpreter even when she faces only one creature. The latter kind of triangle is often small in size or short in one side, but they are solid triangles nonetheless. As Dretske observes (1981: 160), if a system has two or more different sensory devices, then the system can single out the cause of the responses of different sensory devices by the “intersection” of the different “flows of information,” causal lines, of the responses of the devices. Dretske here offers a kind of triangulation similar to Davidson’s, when the latter is conceived as a kind of causal triangulation, and hence it is possible to understand it in a similar way. If this is the case, then the role of Davidson’s second creature is the same as Dretske’s second device within a system. And our interpreter of course can observe those causal connections and correlates them to locate the common causes of the responses.

Another possible answer to the question why the ambiguity of depth is a problem for the interpretation of an isolated creature is suggested in an analogy to the problem of depth from Davidson’s [1982]. The analogy goes as follows. Suppose that there is a miserable creature (or, one-eye creature, for the case to be more plausible), say I, bolted in the earth without possibility of

(16)

any moving, it would be impossible for me to change my previous visual perspective to acquire the concept of a three-dimensional space. Likewise, in order to say that my mental state is about something in the external world, I must be able to abandon my perceptual perspective. But how can I abandon my perceptual perspective? No matter how I may move my body, what I perceive could be just something on my nerve endings. The change of my perceptual content does not provide me sufficient resource to determine the locus of the things I think of, just like the change of my visual content, when I am bolted to the earth, cannot determine the distance of the objects for me. But now this could only mean, as it seems, that I become another person who is related to the same content and has a different perspective on it. Since this is impossible, we need, it is argued, a real second person, someone whose “innate similarity responses are sufficiently like [his] own to provide an answer to the question of what the stimulus is to which the speaker is responding” (Davidson, 1992: 264; cf. Davidson, 1990: 77-8). “If we consider a single creature by itself, its responses, no matter how complex, cannot show that it is reacting to, or thinking about, events a certain distance away rather than, say, on its skin.” (Davidson, 1992: 263) A solitary person may be experiencing a feeling of sameness or discrepancy, but there is no answer to the question of what he takes this sameness or discrepancy to concern. Davidson puts this point with his customary verve and dash: “The solipsist’s world can be any size; which is to say, from the solipsist’s point of view it has no size, it is not a world.” (1992: 263) Davidson suggests that in addition to the discriminative mechanisms, we need some mechanism to fix right cause (in fact, fix it in the external world), and therefore content of speech. He argues, via the triangulation, that the mechanism must be a social interaction.

Here our interpreter imagines that she was in the position of an isolated creature with access to the world only through what happen in her nerve endings. Since what happen in her nerve endings, no matter how dynamic and complex, might just be there, she thinks that she alone cannot determine the locus of the causes of those nerve states. In a sense, she thinks she might be bolted in her own mind.

It should be clear that the triangulation by itself renders no solution at all to the ambiguity of depth so conceived. The problem now is not generated from the interpreter’s external perspective, but from the isolated creature’s internal perspective. But from the internal perspective, providing more creatures, or any thing, for the creature in question would be just give it more to discriminate. That is to say, the second creature and whatever it does, for the first creature, could be just some stimulation on its nerve endings.

The foregoing criticism will make it more difficulty to see how the triangulation works, but it does not defeat the triangulation, since what Davidson tries to demonstrate is only that a social setting is necessary, but not sufficient, for the content. The lesson the triangulation tells us should be that if content determination is based solely on what is available from perceptual discriminative mechanism, there is no way to solve the ambiguity of depth. It is said that, in addition to the perceptual discriminative mechanisms, we need some other mechanism to fix right cause and fix it in the distal. This is fine, but why must the additional mechanism be a social interaction as suggested by the triangulation?

It amounts to ask why do we need the social triangulation to save our isolated creature out of his mind? The answer should be that unless the social triangulation is available to the creature, it is unable to acquire a different perspective to contrast with its own and thus to realize that the whereabouts of the causes of its mental contents transcend its perceptual discriminative mechanism. It is not clear what the “different perspective” means. But if a different perspective means a perspective, whatever it is, of a different creature, then the argument simply begs the question at issue. One must have already realized that there is something outside of his perceptual discriminative mechanism before he realizes that there is a real someone other than him. The different perspective would thus be more plausibly be conceived as nothing but a different causal line, especially when we conceive the social triangulation as a causal triangulation served as a means to determine the common cause of various responses. However, if the social triangulation is merely a causal triangulation available to the creature, there is less reason to assume that the creature’s own “inner triangulation” is not available to itself. “Inner triangulation” means the triangulation of two causal lines issued from two organs of one’s own to a common cause. But we are simply in a position to have a better observation of our own responses than others’ responses. Even though it is too good a position on which we sometimes do not aware of that we have it.

(17)

In fact, even if the problem of depth so conceived is a real philosophical problem, there might be an individualistic account to solve the problem. It is not difficult to construct an evolutionary account to make the point that unless a creature is responding to distal objects rather than others, it will not survive long.11 And, surely, many animals with a generally solitary life have survived long.12

4. The indeterminacy of category

Despite some obvious textual supports (e.g., Davidson, 2001b: 4, 7-8), the ambiguity of depth conceived in either of the foregoing ways might not be what Davidson’s triangulation is really meant to deal with. There are two different questions involved when we ask where the causes of some responses are, and it is important to keep them clear. When a scientist inquires for a general causal law and asks where the cause of a response of a system is, he is trying to locate the cause independent of how the system conceives it. But when our interpreter works on to rationalizes her creature and inquires into where the cause of a verbal response is, she is trying to understand what the location the creature takes to be the location of the cause of his response. Since the triangulation concerns with meaning, it is the latter, not the former, question that matters for Davidson.13 Accordingly, our talks of locating cause of utterance should be completed by the talk of determining the whereabouts that speaker locates the causes of his own utterances. The matter is about how our interpreter understands how her creature classifies things, whether as the distal or the proximal.

The question we should, then, be concerned with is what must be involved in determining the causes taken by a creature to be the causes of its own responses? To see the ambiguity of depth as the problem of locating causes of verbal responses would substantially underestimate the force of Davidson's argument. Davidson might happily grant that it is an individualisticly obtainable fact that our intentional states are about distal objects without yielding the point that there is no answer to the question of what a person in isolation takes to be the objects of his talk. Let me demonstrate this point.

Let our interpreter be P and the isolated creature S. Since the sets of covarying events typically cause the two person’s similar responses are different, it is legitimate to suppose that what causes S’s response R is some event in the covarying event set Sdistal, S1, S2, Sproximal and what cause P’s response R is some event in a different covarying event set Sdistal, P1, P2, Pproximal. Now, suppose that it is indeed a naturalistic fact that S is responding to Sdistal, P, by observing S, can indeed preclude S1, S2, and Sproximal from being the event that cause S’s response R. However, this does not give P any entitlement to identify the remaining event as Sdistal, since its covarying events in set Sdistal, P1, P2, Pproximal, i.e., P1, P2, Pproximal, are not precluded by P. It might be a fact that P responds to Sdistal, but, as a theorist, we, by P’s similarity of responses alone, lack of any ground to say that P is taking the causes as Sdistal but not others, since we know that different events might cause the same similar responses.

Now if what we concern becomes that what must be available to our radical interpreter to determine the whereabouts that her creature locates the causes of his utterances, we are concerned with the question what must be available to our interpreter to determine the way that her creature classifies things. Here, the ambiguity of depth and the ambiguity of width find their common ground. Recall that the ambiguity of width concerns “how much of the total cause of a belief is relevant to content” (Davidson, 2001a: 129). What is the relevant part of the total cause to content is the part that speaker take to be relevant to his own utterance. Both ambiguities are about how a speaker classifies things. Some textual evidences suggest this is the question that Davidson is dealing with,14

11 Davidson seems to express sympathy with evolutionary line of thought by his comment that "evolution had something to do with" (1991a: 200) the fact that we find it natural and easy to class things together in the ways that we do, and that we do classify distal things together.

12 In “Progress on Two Fronts” (1996), Quine has proposed a detailed theory alone with this line. 13 Some critics of Davidson's triangulation are based on a confusion of this distinction. See, for instances, Talmage, 1997: 142-3, and Verheggen, 1997: 363-4.

14 Davidson remarks: “The challenge is to put the lions in a position to distinguish these cases (actions in unison and in deviance). To do this we have to eliminate the dependence on my arbitrary

(18)

and for Davidson the necessary resource mentioned above must include a triangulation situation. Davidson seems to suggest that responses of an isolated creature do not force the interpreter to cite among various possible standards of classification one standard as the standard of the creature’s classification.

However, it should be noticed that, as the above demonstration indicates, the causal triangulation by itself does not provide sufficient resource for our radical interpreter to determine how its creatures classify things. Neither creature might share the same standard with our interpreter. But, again, it is of no harm to Davidson, since he is merely arguing for necessity, not sufficiency.

Is an interpreter facing an isolated creature really doomed at the indeterminacy problem? Let’s see what we can get from Davidson’s answer to the question why an interpreter, after she triangulates the two causal lines issued from her two creatures, chooses distal objects rather than others as the subject matter of the responses? Davidson’s answer to this question is both concise and suggestive: "because it is natural - to us." (1992: 262; cf. 1991a: 200) First thing to be noticed is that it is natural to us, not natural tout court. Second is that, taking into consideration Davidson's conception of citation of cause, the naturalness here must be normative. For Davidson, citing cause is interest related. What cause is cited or how cause is specified depends on what explanatory interest one is after. (cf, 2001b: 216) When semantics is the concern, the explanatory interest is normative, since the purpose of interpreting one's talks is to make one’s talk and thought as intelligible and rational as possible. Accordingly, when Davidson says that it is natural to us to cite distal objects as the subject matter of our responses, the naturalness here must be our normative nature.

Now if this is really what Davidson has in mind about "natural to us", then, it is argued, Davidson's reason for the question why the radical interpreter chooses distal objects as objects of responses is that the choice is dictated by her normative standard, including standard of grouping things. Our interpreter has no choice but to cite distal objects as objects that her interpretee takes to be the causes of their responses. If this is so, then we may wonder why the problem of indeterminacy of standard of similarity arises for the interpreter who counters an isolated creature. The interpreter can do the triangulation, and classify common causes by her own categories, and the problem of indeterminacy is solved. That is to say, if the interpreter can solve the problem of indeterminacy of standards of similarity by appealing to her own standard of similarity, she can solve, by the same way, the problem generally, both in the social and isolated creature cases. Moreover, if our worry is that since the standard of similarity is projected by the interpreter, the creature might have different standard from hers, then the worry will not be dispelled by merely adding more creature into her interpretation.

In fact, some have argued that the worry of indeterminacy of standard of similarity should not even make sense to a philosopher like Davidson who rejects the notion of a radically different conceptual scheme. It is argued that the rejection of the notion of a radically different conceptual scheme implies a rejection of the notion of a radically different standard of similarity (on perceived objects). Thus, it is unintelligible, for Davidson, to say that the interpreter and the speaker might have different standards of grouping things, so different that the interpreter might lose track to the truth and reference of her speaker’s utterance. As Yalowitz criticizes Davidson: “Without the possibility of ‘alternatively’ constituted interpreters [with different standard of similarity], the part of this remark [“because it is natural to us”] loses its power to elevate the importance of some particular interpreter, constituted in some particular way, in the determination of what the speaker is responding to.” (1999: 117)15 In a word, without the worry of the indeterminacy of standard of classification, the triangulation seems to lose its immediate target.

5. Intentionality outruns disposition

I am not sure that the rejection of a radically different conceptual scheme is incompatible with the admitting of different standards of grouping basic things. The rejection concerns more with the

(or interested) choice of relevantly similar responses on the parts of the lionesses.” (emphases mine, 2001b: 7)

數據

Updating...

參考文獻

Updating...