Feeds:
Posts
Comments

Archive for September, 2012

Magic Move

yanjianghk vs chessospar2012   Live Chess at Chess.com

Read Full Post »

Week 2: Montague Semantics

[For easy reference only. Not meant to be read wholly before our meeting this week.]

1. Partee: Reflections of a Formal Semanticist as of Feb 2005

2. Partee: Richard Montague

3. Montague: Formal Philosophy (Introduction + Chapter 8 PTQ)

4. Formal Semantics: the Essential Readings

5. Partee & Hendriks: Montague Grammar

 

Links to my Dropbox folder containing the above e-files can be provided upon request. No anonymous requests please. Please send me an SMS with a self-intro. and an email address.

Read Full Post »

CBS 303, Lec. 2.1

CBS303 Linguistic Meaning and Contextual Use, Lecture Note 2.1

[Cf. Löbner (2002, Chapter 2)]

 

Outline of this lecture:

1. Descriptive Meaning; 2. Social Meaning; 3. Expressive Meaning 4. Denotation and Connotation

 

1. Descriptive Meaning

a. Word meaning is related to concepts

Words are recorded in dictionaries, but dictionaries are no more than artifacts, which vary in size of vocabulary. A language user’s knowledge about the vocabulary of a language [whether it is his mother tongue or a second language he has learnt] resides in his mind. It is a mental lexicon, of which a printed dictionary is only an imperfect copy [even though a dictionary could include a larger vocabulary]. In this mental lexicon, the meaning of content words relates to concepts, which is a psychological notion that forms the basic unit of thinking. One word may be related to several concepts, and several words may be linked to one and the same concept. Concepts help people to categorize or group entites. Lobner (2002: 20), "the concept is a mental description of a potential referent…. [A] word can only be considered established if its form and meaning are linked in the minds of a great number of language users." So words should be public, not private.

The descriptive meaning of a content word is a concept for its potential referents. [Lobner 2002: 23]

We can interpret the above definition in two ways:

i. Internally related to the semantic system of a language: word meaning relates to concepts. Such concepts can be defined in the form of descriptions, or in the form of a bundle of features.

ii. Externally: word meaning relates to concepts, which can identify entities in the world. Therefore, words can refer to human individuals or objects. But at an abstract level, we say words denote, and have denotation.

Definition in Lobner [p. 25]: the denotation of a content word is the category, or set, of all its potential referents.

The difference between denotation and reference: denotion is about all the potential references of a linguistic unit.

When is denotation and reference collapsed ? [when we consider names] Why?

 

b. Sentence meaning is related to propositions

The descriptive meaning of a sentence relates to a proposition.

For statement sentences [i.e. assertions] it refers to a proposition.

For other types of sentences[such as questions or imperative sentences] it is defined in terms of a related assertion.

The meaning of a statement sentence refers to a situation it describes. The situation thus described can either be a genuine reflection of the world or a distorted one. Hence we can say that a sentence has a true or false referential meaning. Taking it a step further, we can say that a sentence refers to the true value of a situation. There are altogether two truth values: true and false. So the referential meaning of a sentence is 0 or 1, where 0 means false and 1 means true. Or we can say that the meaning of a sentence is taken from {0, 1}. Or the meaning of a sentence S refers to {0, 1}.

Definition [Lobner p. 24]: The descriptive meaning of a sentence, its proposition, is a concept that provides a mental description of the kind of situations it potentially refers to.

But do we need to check the truth value of a sentence in order to study meaning? Such a kind of "field work" is not necessary, because we only need to define the truth conditions of a sentence.

Definition [ibid]: the truth conditions of a sentence are the conditions under which it is true.

 

2. Social Meaning

 

a. Terms of Address

b. The Honorific System

c. Politeness and estrangement, social distance and distancing

d. behaviourial verbs

e. small talk

f. phatic communion

 

3. Expressive Meaning [not to be confused with expression meaning]

Interjections, exclamatives [four-letter words, mild oath, amplifiers]

 

4. Connotations: taken in contrast to referential meaning/denotations, they are associations one recalls when using a word or an expression.

examples: dog, pig, revolution, counter-revolutionary

Read Full Post »

Week One: the Generative Paradigm

Readings:

Barsky: the Chomsky Effect

Chomsky: of mind and language – a dialogoue

Collier: the Anti-Chomsky Reader

Cook: Chomsky’s Universal Grammar

Neil Smith: Chomsky – Ideas and Ideals

The Cambridge Companion to Chomsky

The Oxford Handbook of Linguistic Analysis

 

Links to my Dropbox folder containing the above e-files can be provided upon request. No anonymous requests please. Please send me a SMS with a self-intro. and an email address.

Read Full Post »

Mon. : Writing up notes for the past two days, including this morning. Then attending the Ontology subject. Lots of new ideas. Will need to go with a PC next time. Had to leave 45 minutes earlier for my own lectures. Will need to learn more programming. Don’t know if I will be able to keep up with the pace. Then to the CBS303 lectures. Sold out all the textbooks by Lobner. Regretted not having ordered more. Ordered enough for CBS500 but may not be able to sell all. Which is better? Tried to talk up at the lecture theatre because the microphone was not functioning. Plus some students couldn’t help chatting among themselves. That is always the problem with undergraduates at large classrooms. Students should learn to be self-disciplined. Please don’t talk among yourselves while I am talking. Exhausted as a result. Didn’t know if I could survive the remaining days of the week.

Tue. : Could have been a day off because the seminars had not started, but had to go to PolyU to attend a prep meeting for the validation of Msc. therapy. Still hadn’t recovered from the Monday exhaustion but had to write notes for CBS500. Felt sleepy during the meeting and had nothing to do there. Left earlier.

Wed. : a day off, started with taking the heavy trumbo all the way to F’s school; preparing the DALS lecture notes for Thurs. by reading more of Neil Smith’s Chomsky: Ideas and Ideals. I do respect the generative paradigm a lot, but are sometimes put off by the attitudes of some of the staunch followers who acted as if only they had the whole truth. One repeated said during a lecture "Chomsky hasn’t heard about it. Chomsky doesn’t know about it. I am going to tell Chomsky about it." When each of the above sentences got uttered more than a dozen times, they became pause fillers. Pathetic!

Thurs. :saw the acquisition of Betty Birner’s new book Intro. Pragmatics by the PolyU Library in e-form. Experimented with the on-line reading. Then to PolyU. Discussed Y-3 undergraduates on final year project. Had a great night with research students teaching DALS 6050 Linguistic Theories. Wished I had prepared more on the Generative Paradigm. Walked with some nice participants on the campus. Two accidents. 1. Forgot to take the right U-disk, but got it during the break. 2. Forgot to take my mobile phone after the class, but got it back at the securities. Lucky. Got home but could not go to sleep. In fact, woke up only after 2 hours’ sleep. Thought it was because of the strong milk tea for supper. Thinking over the good moments of exchanges that evening in class. It was relaxing and fun, while everyone was serious and inquisitive.

Fri.: took the trumbo again for F’s to school. Then made a dash for PolyU to attend a confirmation viva. Got excited by the topic and talked a lot with the participants. Again it was relaxing and enjoyable, with serious contributions.

Administration wise, if things could go wrong, they would. Had to deal with appeals and potential appeals from …  And the afternoon was spent on Speech Therapy validation. Had to eat during the meeting to catch up with time so that I could go to my even lectures without a supper. Took the shuttle to HHB lecture theatre. Impressed by the classroom, which seemed to worth the travel. Gave a 3-hour lecture. Hope students could catch up with reading. May use the lecture notes as the main reading rather than the textbook. Did not finish the notes but got to the middle.

Lots of papers are approaching the deadlines. 6 of them. Need to work over them during the weekend.

Read Full Post »

Betty J. Birner  9781405175838 cover

 

本科目第二階段討論語用學,教材是新近出版的 Introduction to Pragmatics. 我已經請理工大學圖書館購買了電子版。

如何閱讀電子版:

1. 進入學校圖書館系統,可以直接點擊本鏈接 http://library.polyu.edu.hk/search~S6?/aBirner%2C+Betty/abirner+betty/1%2C1%2C6%2CB/frameset&FF=abirner+betty+j&4%2C%2C6

2. 在網上直接閱讀(有時限,且好像兩人無法同時閱讀),或是在書頁上方提供的eBray瀏覽器(先要註冊)里把各章生成pdf文件下載到自己電腦閱讀或打印。

3. 注意:生成的pdf文件只供個人使用,請同學 不要轉發,特別是不要在公共網站上轉發,因為生成的文件上每頁都有使用者信息,出版社有可能會起訴轉發電子版的用戶。請大家一定要謹慎使用。

Read Full Post »

第二講 基礎邏輯工具 (續)【互參:教材第二~三章】

星星 謂詞邏輯

命題邏輯雖然刻畫了複合命題的內部層次關係,但對各原子命題內部的結構不做分析,那是謂詞邏輯(predicate logic)的任務。謂詞邏輯把一個命題拆分成謂詞和論元兩大部分。以自然語言的句子打比方,一個主-謂-賓分明的句子的謂語動詞就是典型的謂詞,而主語和賓語就是兩個論元。當然,謂詞邏輯與自然語言的句子並不完全等同,但就這個句型而言,兩者是重合的。換一個角度看,謂詞就像是代數式中的運算符即算子,而論元仿佛是式中的數字或字母。謂詞需要與論元相結合,才能構成完整的式子。而且,謂詞對論元的數目也是有規定的,不能多,也不能少。多了會剩下用不了的論元無法進入句子,少了相應的邏輯式就不完整,不能成為合格的式子。這些是邏輯上的嚴格規定。自然語言中的論元常有缺失,因為可以從上下文補出。另外,自然語言中能擔任論元的成分較複雜,不限於名詞組,有時候子句和介詞詞組也可以充當論元。不過這更多地涉及句法的問題。

 

在教材第一章裡作者已經提到,不及物動詞指謂的是具有相關動詞語義特性的所有個體的集合,那是帶一個論元的謂詞的指謂,即只有主語,沒有賓語的情況。本章既然討論了論元的數目,就引出了相關的問題:帶兩個論元的謂詞指謂什麼?作者在本章沒有討論,我們替她補上。只帶主語論元的謂詞,指謂的是一個特性,所以指謂具有該特性的個體所構成的集合,而相關論元所指謂的個體就是該集合中的一個成員(這裡暫時假設主語論元由專名充任,更複雜的情況要等到介紹了量化之後才能說得清楚)。帶兩個論元的謂詞,指謂的不是特性,而是兩個個體之間的一種關係。所以這種謂詞指謂的是這種關係的集合,也就是所有構成這種關係的兩兩配對的個體的集合。這種兩兩配對是有序的,否則無法分清哪個是主語,哪個是賓語。這種配對叫做有序偶(ordered pair)。那麼相關的集合就是有序偶的集合(set of ordered pairs)。讀者可以進一步推算出帶三個論元的謂詞(比如三價動詞)的指謂內容的集合論定義。另一些問題是:自然語言裡有沒有帶四個論元的謂詞?為什麼?論元的數目是如何確定的?除了論元和謂詞外還有什麼範疇?這些涉及到論元的定義和語法語義的介面關係,在教材的第十、十一章會繼續討論。

 

星星 邏輯量詞

教材第三章討論謂詞邏輯裡的量化問題。前面的討論對充當論元的子句和介詞詞組的表達僅採取了簡化的處理辦法。另一個簡化的處理辦法是讓名詞性論元都指謂特定的單一個體,不涉及變項(variables),這些指謂單一個體的名詞在一階謂詞邏輯中都表達為個體常項(individual constants)。其實,謂詞邏輯的論元還可以由變項充任。變項就像代數式中的x, y, z,其指謂對象是可以變化的。變項需要約束,在一階邏輯裡,約束變項的是量化詞(quantifiers),簡稱量詞。在一階邏輯裏,量詞只有兩個:全稱量詞(universal quantifier ")和存在量詞(existential quantifier $)。沒有量詞約束的變項是自由變項(free variable),含帶自由變項的式子不是完整的命題,而是開放性命題(open proposition),又叫做開語句,或是命題函項(propositional function)。量詞規定了變項的賦值次數。"xF規定在F中出現的一個或多個x,其取值需要窮盡論域中的所有可能個體,只有每次取值都使F為真,"xF才為真。$yY規定在在Y中出現的一個或多個y,只要對其一次賦值能使Y為真,$yY就為真。在謂詞邏輯中,出現在某個量詞右面的式子包括其中的受該量詞約束的變項都處在該量詞的轄域(scope)內。如果一個式子中有兩個或更多的量詞,那麼出現在左邊的量詞的轄域就比出現在其右邊的量詞的轄域要大,這時,不同量詞約束的變項要用不同的字母來表示,以免引起混亂,參見(1).

(1)                "x$yF

上式中F是元符號,代表原子命題或是複合命題,內含謂詞、論元,並可以包含作為論元的變項xy. 說“可以包含”,是因為如果x,y不出現在F中,那麼它們只是作為引導變項出現在式中,而不是論元,那樣量詞就會做空約束(vacuous quantification)。多數邏輯系統允許空約束的存在,但在形式句法理論裡多不允許。

                除了變數的約束問題外,還有量化域的限制問題(restriction on quantifier domain)。現在我們改用具體的謂詞來寫式子。假如我們有"xF’x,且F’=放假【這裡上標的撇號表示相關詞語的指謂意義,而不是它們的詞彙形式】,那麼整個式子的意思是“每個個體都放假了”,也就是說世界上的所有個體都放假了,包括人、狗、桌子、樹,等等。這顯然不是我們經常需要表達的意思。假如要表達的是“每個人都放假了”,甚至是“xx中學的初中部教師每個人都放假了”或“xx中學的初中部有個教師放假了”,那就要收窄變項的取值範圍,收窄的方法是增加限定性謂詞,例如(2)(3)中的B’或是更多的謂詞。

(2)                "x(B’x ® F’x)

(3)                $y(B’y Ù F’y)

還有一種限定性量化表達法(restricted quantification),更方便自然語言的語義研究,它把量化式分成三個部分,分別是量詞、限定部分(the restrictor)以及核心部分(the nuclear scope)。這種表達法可以把上述兩個式子轉換成以下兩個形式:

(4)                "x(B’x) F’x

(5)                $y (B’y) F’y

qq
bb
cc
dd
 

對限定性謂詞的討論可以延伸到對定語形容詞(attributive adjective)的討論方面,這在教材第三章3.3節扼要地涉及。自然語言中的形容詞有時可以用做謂語,在漢語中是形容詞直接做謂語,在英語中要用連接動詞過渡。這種用法都可以在一階謂詞邏輯中翻譯成謂詞。定語形容詞在自然語言的句子中不是謂語,在相應的一階謂詞邏輯中卻可以像自然語言中的謂語一樣譯成謂詞,只是謂語譯成的謂詞是“主謂詞”,而定語形容詞譯成的謂詞是起限定作用的謂詞,出現在量化邏輯式的限定部分:

(6)                一輛黑轎車在等瑪麗。

(7)                $x((Black’(x) Ù Limousine’(x)) Ù Wait-for’(x, m) ) [m = 瑪麗]

這裡的“黑轎車”中的形容詞“黑”是個交集性形容詞(intersective adjective),這種形容詞A所指謂的集合與被修飾名詞N所指謂的集合相交,交集中的個體既是集合A的成員,又是N的成員。但是,也有一些定語形容詞是非交集性形容詞(non-intersective adjective),比如下列結構中的劃線部分:fake banknote, alleged affair, false teeth, imitation jewelry, stone roses, 漢語中的鈔等等。假的鈔票不是鈔票,所以鈔票沒有交集。這種形容詞又被稱作負性質修飾語(privative modifier)。教材中的簡單化處理是把負性質修飾語同其後的名詞處理為一個謂詞,而不是交集性的兩個謂詞的合取式。其實處理這個問題還有很多困難,比如如何分析fake rhinestone jewelry, 人造真皮,偽金幣?如果說“萊茵石首飾”本身已是有色玻璃仿寶石製品,那麼“假萊茵石首飾”是否又可以通過邏輯上的雙重否定律得到“真首飾”的解讀?為什麼不能?如果把“人造真皮”處理成負性質修飾語“人造”與“真皮”的組合,那又如何處理“真皮”?“偽金幣”也是同樣的問題。另外,六七十年代在中國流行的“假領子”,其實是真領子,只不過它是假的襯衫領子,因為領子以下並沒有襯衫的其餘部分;“假酒”多半還是酒,只不過是仿冒名牌的真酒;“偽軍”並非不是軍隊,而是不被正史認可的軍隊。這些問題說明對負性質修飾語的分析需要考慮很多因素,也需要更複雜的邏輯和語義理論,其前沿性和複雜性遠非基礎教科書所能負荷。

教材第三章3.4節介紹一階謂詞邏輯裡的兩個量詞之間的互釋性,這需要通過否定來實現。3.5節介紹量詞轄域的描寫和轄域的歧義現象。前面說過,謂詞邏輯中的轄域確定是線性一維的,左邊的量詞轄域大過右邊的,但自然語言的句子結構有內部的層次,不能單憑語序來確定。所以作者介紹了簡單的樹形結構表達法來說明自然語言量詞轄域的判定方法。這個題目又是個很大的課題,涵蓋了句法上對句子的邏輯式(LF)及其表徵的大量研究。究其精髓,是對自然語言的句子進行分析變換,將其整飭表徵成准邏輯的式子,以利語義解釋,包括對量詞轄域的正確表徵和量化依存關係的準確解釋。與此相關的是,未經分析表徵的自然語言的語句具有一定的結構歧義,所以含帶多個量詞的句子會出現轄域歧義的現象,對這種句子可以做兩種或更多的邏輯翻譯表徵,不同表徵中的量詞居於不同的位置。這樣量詞轄域的不同就能導致不同的解釋。

Read Full Post »

Older Posts »