) below ) as a marker that denotes some sort of temporal relation . He seems to accept Draganov ’ s ( 1958 ) analysis of it as a past tense marker . I regard it as a modal-aspect marker which denotes no change of situation and which functions independently of a cleft construction .
 
(48 ) a . Zhang San shi 1970-nian biyede1 ( xuesheng ) .
 
張三 是 1970年 畢業 的 （ 學生 ） 。 「 張三兮 是 1970年 畢業 的 （ 學生 ） 。 」 Zhang San is a student who graduated in 1970 .
 
b .
 
Zhang San shi 1970-nian biyede2 張 三 是 1970年 畢業 的 。 「 張三兮 是 1970年 畢業 的 。 」 It is in 1970 that Zhang San was graduated . Although I recognize two de ’ s--one as a modifier formative as the de1 in ( 48 a ) above and the other as a modal aspectual marker as the de2 in ( 48 b ) .
 
13 I see no need of recognizing two shi ’ s , one for equation and one as a focus marker . When the head noun in ( 48 a ) is deleted , ( 48 a ) and ( 48 b ) are homophones . The difference between ( 48 a ) and ( 48 b ) is due to the difference between de1 and de2 , which mark modification and unchanged situation respectively , not due to the alleged difference of two shi ’ s.14209 A similar analysis is sufficient for another pair of homophonous sentences taken from Teng ’ s paper ( translation is my own ) .
 
(49) a .
 
Nide gou shi wo zai gongyuanli zhaodao de1 .
 
你 的 狗 是 我 在 公園 裡 找到 的 。 「 你 的 狗 是 我 tr 公園 內 底 chho0 著 彼 隻 。 」 Your dog is the one that I found in the park .
 
As for your dog , it is correct ( to say ) that it is the one I found in the park .
 
b .
 
Nide gou shi wo zai gongyuanli zhaodao de 2 .
 
你 的 狗 是 我 在 公園 裡 找到 的 。 「 你 的 狗 是 我 tr 公園 內 底 chho0 著 的 。 」 As for your dog , ( it is correct to say ) that I found it in the park . As for your dog , I am the one who found it in the park . I would like to stress that , even though ( 49 b ) is not synonymous with ( 49 a ) but with ( 49 c ) and ( 49 d ) below , there is one thing that ( 49 b ) shares with ( 49 a ) but not with ( 49 c ) or ( 49 d ) . Namely , ( 49 a ) and ( 49 b ) share the same topic as shown in the alternative translations . In spite of the possibility of more than one English translation , ( 49 c ) is not ambiguous according to native intuition . Unless one recognizes that the shi in ( 49 c ) is the same shi in ( 49 d ) , one has no way to account for the judgment that ( 49 c ) and ( 49 d ) are synonymous , as claimed by Teng ( 1979 ) .
 
( 49 ) c . Zai gongyuanli zhaodao nide gou de ( ren ) shi wo .
 
在 公園 裡 找到 你 的 狗 的 （ 人 ） 是 我 。 「 Tr 公園 內 底 chho 0 著 你 的 狗 的 （ 人 ） 是 我 。 」 As for the person who found the dog in the park , ( it is correct to say that ) the per son is me .
 
d .
 
__UNDEF__ Shi wo zai g ong y uanl i zh aod ao nide gou de .
 
是 我 在 公園 裡 找到 你 的 狗 的 。 「 是 我 tr 公園 內 底 chho0 著 你 的 狗 的 。 」 ( It is correct to say that ) I found your dog in the park . It was I who found your dog in the park .2105 . In Teng ’ s view , and in the view of many others who have discussed shi-predicates , the element that immediately follows shi is always the focus . I have given many examples to support a different claim . The focus function of shi is much more limited than most researchers think . Shi , in fact , is best translated as ‘ it is correct ( to say ) that ’ or ‘ it is the case that ’ , as I have done for sentences ( 49 a-d ) , and bushi as ‘ it is not correct ( to say ) that ’ or ‘ it is not the case that ’. The focus function of shi is limited to selection and inclusion of focus within the predicate which may include unfocused elements . As for factors which help with identification of focus within a shi-predicate , we have mentioned four : ( 1 ) the presence of a split constituent , ( 2 ) an element being optional or obligatory , ( 3 ) an element being a subject or outer adverb , and ( 4 ) an element being at the left-most or right-most position of the predicated constituent . It should be noted that , in giving English translations of other shi-predicated sentences , I try to make them sound natural to English speakers and , therefore , they do not faithfully reflect Chinese structures . The basic meaning of shi or bushi predication is that of the truth-value of a truth table in modern logic . Focus plays no function in the logical negation nor in the basic meaning of bushi.15 Within a shi predicate , identification of focus is the function of the above-mentioned four factors concerning each element .6. Teng recognizes that an object cannot be focused on by a cleft construction . We pointed out that the verb complement or the main verb preceded by an internal adverb also has this restriction . Such constraints are , in my opinion , evidence of the verb status of shi in a “ cleft ” sentence . Shi-predication requires that a verbal topic be nominalized if the shi-predicated element is a noun and if a general head noun is to be inserted . Such a general head noun should denote a general category of which the predicated element is a member . These requirements are not explainable if shi is a focus marker but are explainable if shi is a verb that denotes confirmation and equation/category - subsumption .7. Finally , Teng has presented and evaluated two alternatives for the derivation of cleft sentences : the complement sentence analysis and the simplex analysis . Recognizing problems of both , he left the question of derivation open . I have not addressed the problem of postulating deep structures , rather , I have described shi-predication on the basis of surface structures which are outputs of subjectivization and inputs to topicalization , repetition , or shi-predication .4. CONCLUSION In conclusion , I would like to make the following generalizations about focus devices in Mandarin Chinese .2111. There is no device that functions only to mark focus . Each focus device has some basic semantic or syntactic function of its own besides the subsidiary function of marking focus . Topicalization , repetition , and deletion of unfocused elements are mainly to make a topic out of the unfocused elements or to shorten a sentence . Selecting a different part of the sentence for inclusion in the shi-predicate or a predicate led by another modal involves use of a marker that denotes confirmation , denial , or another modal . Question forms are for interrogation . Parallel constructions are for showing contrast in a similar situation .2. Such being the nature of focus devices , each device has its own syntactic constraints which may result in an ineffective or imprecise indication of intended focus . It is very likely , therefore , that a speaker may use a focus device to express focus , and yet the addressee may not identify the intended focus .3. Though the question of the span ( or scope ) of focus is still pending , we may acknowledge the possibility of more than one focus in the same sentence . However , while multiple focuses are expressible by some focus devices ( such as questions and parallel constructions ) , they cannot be handled by others ( shi-predication in particular ) , and we find the principle of unified focalization working when more than one focus device is used in the same sentence .4. Most focus devices adhere to the general principle that the predicate marked by shi and other modals is the slot for focus , and the topic is the slot for unfocused elements . Deletion and topicalization of unfocused elements deduces these elements that are in the shi-predicate for easier identification of focus . When an element is repeated in the topic and predicate in the same sentence , it is discounted as a focus element in spite of its appearance in the predicate .212 In his article entitled “ Givenness , Contrastiveness , Definiteness , Subjects , Topics and Point of View ” , Chafe ( 1976 ) discusses six different “ package ” statuses of a noun which concern the speaker ’ s assessment of the addressee ’ s state of mind about the noun ’ s referent at the speech time . There is no doubt that focus is a “ package ” status of an element ( which can be a noun or some other type of expression ) because it is the speaker ’ s assessment of whether or not a certain element of the sentence he is uttering will receive special attention from the addressee . Among his six package statuses , the closest equivalent to focus is contrastiveness , which he also calls ‘ focus of contrast .’ In our view , contrast and focus seem to be two different notions and need to be distinguished . We may mention two types of sentences in order to show that wherever there is a contrast , there is a focus , but there can be focus without contrast . By focus of contrast , Chafe would include topicalized elements that are in a parallel construction and which are marked by wa in Japanese .
 
(50 ) a . Kore wa oisii ga are wa oisikunai .
 
This is delicious , but that is not .
 
In Japanese , wa marks contrast as well as a topic .
 
While we recognize that the contrasted elements marked by wa above are also focused , we consider the main focuses to be in the predicate .
 
To shift the main focus from the predicate to are ‘ that ’ , wa should be changed into ga as in : ( 5O ) b . Iya . Are ga oisii .
 
Kore wa titto mo oisikunai . no - that-delicious this - at all - delicious not .
 
No , it is that that is delicious .
 
This is not delicious at all .
 
We have shown by Chinese examples that a topic is never focused .
 
An exception may be found in topics that are in contrast in a parallel construction , as the double-lined elements below .
 
( 51 ) ( * ) Zheitiao yu ( * ) wo ( + ) yao chi yutou ; ( * ) neitiao yu ( * ) wo ( + ) yao chi yuwei .
 
這 條 魚 （ * 是 ） 我 （ 是 ） 要 吃 魚頭 （ * 是 ） 那 條 魚 （ * 是 ） 我 （ 是 ） 要 吃 魚尾 . So far as this fish is concerned , I want to eat the head ; ( but ) so far as that fish is concerned , I want to eat the tail It seems to me , however , that although the topics in the above example are focused , they are not the main focus . The major focuses are on the elements that are in the prominent position of the predicate . This analysis is supported by the fact that the topic cannot be included in a shi-predicate , but the predicate can . If the above view is correct , we can at least say that , while contrast may always entail focus , it is not necessarily the main focus .213 Chafe , on the other hand , may not necessarily include in his focus of contrast a question form or new information that , in the speaker ’ s judgment , would attract the addressee ’ s special attention . I view any question form to be the focus because it is sure to attract the addressee ’ s attention . Syntactically , it shares certain constraints with other focus devices in Chinese . In Japanese , wa can never be added to a question word . In English , there is also a constraint between a cleft sentence and a question form . All these can be naturally accounted for by the principle of unified focalization . Unless we recognize question forms as a type of focus device , we have to resort to some ad hoc postulate to account for the constraints commonly shared by three unrelated languages . There is , therefore , at least one type of focus that is not normally considered contrastive , namely the question form . Focus , therefore , does not entail contrast , but contrast does entail focus . NOTES * A part of this paper was presented at a faculty-student seminar of the East Asian Languages Department , University of Hawaii and at the 1979 MLA annual meeting in San Francisco . I would like to express my thanks to those who attended the seminar for their valuable comments and criticisms . I would like to also express my gratitude to Dr . John Haig and Mr . James Chow for reading the first draft and giving me many valuable comments .1. See Teng ( 1979 ) for an example of such an attempt .2. See Chafe ( 1976 ) for a description of statuses of a sentence element on this sort of criteria which he calls package statuses .3. Here I disagree with Chafe ( 1976 ) , who views that there can be only one possible candidate in “ contrastiveness ”. 4. See 4.6 for the view that these constraints can be explained naturally by regarding shi as a copula verb rather than a focus marker .5. Sentence ( 10 a ) may be said when the speaker has heard a sentence , but perhaps did not hear clearly some part of the sentence . In such a sentence , the focus is shifted to the part that is questioned.The rest of the sentence becomes assumed in spite of the shi-predication . It should be kept in mind that none of the constraints discussed in this paper are applicable to those unusual cases of clarifying a sentence that has been said .6. See Li and Thompson ( 1979 ) for a discussion of ma-questions in Mandarin .2147. The possibilities of using shi are more complicated than what these examples may indicate . See Chao ( 1968 ) for a description of choice-type questions in Mandarin .8. See R . Cheng ( 1987 c ) for the meanings of sentence-final le .9. Cf . Zhu ( 1961 ) and footnote 10.10 . There is some reason for regarding both de ’ s as a modifier formative and nominalizer . We may postulate a head noun after de , qingxing ‘ unchanged situation ’, which is obligatorily deleted . Such a head noun would be similar to a modal noun yangzi ‘ it has the outlook that .
 
.
 
.
 
’ , as in Ta zuotian maile shu de yangzi ‘ It looks like he bought a book yesterday .’ If we do so ( i.e . postulating an obligatorily deleted head noun ) , we can say that there is only one de which is a modifier formative and nominalizer . See Paris ( 1979 ) , A . Hashimoto (1969 ) , and Zhu ( 1961 ) for different justifications of the nominalizer status of de .11 . Zhu ( 1978 ) has analyzed the patterns of de-sentences and copular sentences very extensively . However , focus is not the central theme of his paper , nor does he discuss the notion of the cleft sentence .12. See Chao ’ s ( 1968 , 8.1.8 ) description of various meanings of shi other than these two .13. I agree with Zhu in his 1978 article where he does not recognize the focus marker shi in addition to the copular shi . I disagree with him with respect to de1 and de2. He regards both as the same de . The discrepancy seems to result from his failure to recognize a semantic difference between ( 48 a ) and ( 49 a ) , and from ( 48 b ) and ( 49 b ) respectively .14. See footnote 10 of this paper . Teng , during the MLA panel , commented that there are two types of shi : one stressed and the other unstressed . The stressed one is a verb and the unstressed one is a focus marker . I recognize the different manifestations of shi as discussed by Chao ( 1968 ) , who regards the difference as the signaling of contradiction ( shi X as opposed to bushi X ) and contrast ( shi X as opposed to shi Y ) respectively . In our view , there is only one copular shi that can be stressed or unstressed , as most other lexical items in Mandarin . Both cases of shi can be considered focus devices , though the location of focus is different . In shi X ( as in shi wo ‘ it is me ’ ) , shi is focused as opposed to bushi X ( bushi wo ‘ it is not me ’ ) . In shi X ( as in shi wo ‘ it is me ’ ) , X is focused as opposed to shi Y ( shi ta ‘ it is he ’ ) .
 
21515 . Cf .
 
Allwood et al .
 
( 1977 ) , who have stated , “ Ordinary language seems to allow negation below the sentence level , but this is not possible in propositional logic .
 
” ( p .
 
31 ) . REFERENCES Allwood , J .
 
, Anderson , L .
 
, and Dahl , T.M . 1977 .
 
Logic in Linguistics . Cambridge University Press . Chafe , Wallace L .
 
1976 .
 
“ Givenness , Contrastiveness , Definiteness , Subjects , Topics , and Point of View ” . In Charles N . Li ’ s Subject and Topic . Academic Press , Inc . New York . Chao , Y.R .1968 . A Grammar of Spoken Chinese . University of California Press . Cheng , Robert L . 1966 Some Aspects of Mandarin Syntax . Indiana University . Doctoral dissertation .
 
.
 
1967 .
 
“ Universe-Scope Relation and Mandarin Noun Phrase ” . POLA Project of Linguistics Analysis Reports , II , No .
 
3:1 - 182 .
 
.
 
1971 .
 
“ Universe and Scope Relation in Chinese and Japanese .
 
” .
 
1976 .
 
“ Time Relations in Chinese ” . Manuscript . Oragunov , A . A.1995 . Xiandai Hanyu Yufa Yanjiu ( Study of Mosdern Chinese syntax .
 
) , Vol.1. trans by Zu-ging Zheng.Peking ; kexue Publishing Company . Hashimoto , Anne .1969 .
 
“ The Verb ‘ To Be ’ in Modern Chinese ”. Foundations of Language Series 9.4:72-111. Householder , F.W . and Cheng , Robert L .1971 .
 
“ Universe and Scope Relation in Chinese and Japanese ” . Journal of Hawaiian Association of Language Teachers , Vol .13 , Nos .1 and 2.216 Li , Charles N . and Thompson , Sandra .1979 .
 
“ The Pragmatics of Two Types of Yes-No Questions in Mandarin and Its Universal Implications ” . To appear in Papers from the 15th Regional Meeting of Chicago Linguistics Society . Paris , Marie-Claude .1979. Nominalization in Mandarin Chinese . Department de Recherches Linguistiques , Universite Paris VII . Teng , Shou-hsin .1979 .
 
“ Remarks on Cleft Sentences in Chinese ” . JCL , Vol .7:101-115 . Zhu , Dexi . 1961 .
 
“ Shuo de ’ ( On ‘ de ’ ) ” . Zhongguo Yuwen 12:1-15 . Zhu , Dexi . 1978 .
 
“ De ’ Zi Jiegou he Panduan-ju ( De-Construction and Copular Sentences ) ” . Zhongguo Yuwen 1:23 - 27 and 2:104 - 109 .
 
217 華語 焦點 的 表示法 摘要 這裡 華語 有關 的 焦點 的 表示法 ， 是 指 可以 使 語意 重點 更 容易 認定 的 各 種 句型 結構 。 語意 重點 ， 是 指 句子 成份 中 有 可能 被 別的 成份 代替 ， 而 成為 說話 時 最 受 重視 與 關心 的 成份 。 例如 在 “ 我 在 花園 裡 看到 了 一 朵 花 ” 這 句 話 裡 的 “ 一 朵 花” ， 如果 被 認為 是 一 個 可能 被 “ 一 隻 鳥” ﹑ “ 一 隻 貓” ﹑ “ 一 個 池” 等 可 代替 的 成份 ， 那 末 “ 一 朵 花” 就 是 這 句 話 的 語意 重點 或 焦點 。 至於 句子 裡 的 非 重點 成份 ， 是 指 說話 者 認為 言談 訊息 裡 不 重要 的 部份 ， 不 重要 成份 都 是 舊 的 或 對方 已 知 的 訊息 。 華語 有關 語意 重點 的 表示法 包括 ﹕ （1） 以 省略 非 焦點 成份 來 簡縮 句子 ； （2） 把 全部 或 部份 的 非 焦點 成份 主題化 ﹔ （3） 以 “ 是 ” （ 「 是 謂語化 」 ） 或 其他 助動詞 或 副詞 來 標示 含有 語意 重點 的 謂語 的 開端 ﹔ （4） 使用 平行 結構 與 （5） 疑問 結構 等 。 疑問 結構 是 一 種 表現 焦點 的 方法 ， 因為 說話 者 要求 對話 者 答覆 疑問 結構 裡 的 重點 成分 ， 而且 與 其他 的 焦點 表示 法 （ 尤其 是 「 是 謂語化 」 ）