LBT 273-275心妤

9580019 詹心妤 第三組

（1） colorless green ideas sleep furiously

（2） furiously sleep ideas green colorless’

Of these two strings （1）sounds more like a sentence than（2）. Thus we may say that（1）is grammatical but（2）is not . Although neither （1）nor（2）says anything about the physical world , it may be conceded that（1）is more likely to have meaning , say to a modern poet , than（2）. There is , then , a certain affinity between meaning and grammatical structure . However , it is also clear from these examples that meaning could not be used as a criterion for grammaticality , for（1）may have no meaning to a prosaic person , but he would nevertheless be perfectly capable of deciding which of the two strings sounds more like a sentence in English than the other .

這兩條件中，(1)比(2)聽起來更像句子。因此，我們可以說(1)是合乎文法的，但(2)沒有。雖然(1)和(2)都沒有說到任何有關自然世界，但對現代詩人來說勉強可以承認(1)比(2)更有意義。此外還有某種特別的密切關係在意義和文法結構中。然而，從這幾個例子也可以非常清楚得知意義不可以被用來當文法的標準，對一般人來說，(1)有可能是不具意義，但他仍有決定英文二條件中哪個更像句子的極佳能力。

How does an adult speaker of English decide on the matter of grammatical structure? Certainly not by estimating the probability of occurrence of the sentence as a whole , because sentence (1) as well as sentence (2) had a zero probability of occurrence before Chomsky made them famous . Thus the difference between (1) and (2), which is fairly obvious to most speakers of English , can not be due to a difference in transitional probabilities of individual words (see also Miller , Galanter and Pribram , 1960) . Jenkins and Palermo (1964)*sought to account for the interpretative difference between the two sentences by postulating transitional probabilities between parts of speech . If this were so , speakers of English would always perceive a chain consisting of adjective , noun , verb , and adverb as grammatical , whereas the reverse order should result in the perception of grammatical chaos . But take the sentences

(3) occasionally call warfare useless

and

(4) useless warfare call occasionally

一個會說英文的成人是如何決定文法結構的問題？無疑地，整體來看並非藉由評估句子發生的機率，因為在Chomsky (喬姆斯基) 讓他們有名前，句子(1)和句子(2) 是零發生率。因此，(1)和(2)之間的差異，對大部分母語是英文的人來說是相當明顯，無法歸結於個別詞上的過度可能性的差異。Jenkins and Palermo 企圖藉由假設詞性中的過度可能性以說明這兩句子的解釋差異。如果是如此，說英語的人會認為形容詞、名詞、動詞、和副詞組成的鏈為文法，而相反的順序會導致文法混亂的理解。但以這些句子(3) occasionally call warfare useless和(4) useless warfare call occasionally為例

String (3) might occur in an instruction booklet on pacifistic rhetoric ; it is judged to be grammatical , and it is susceptible to semantic interpretation . None of this is true of (4) . But the order of parts of speech of (3) is that of the ungrammatical string (2) , whereas the grammatical string (1) has the order of (4) . Clearly , the transitional probability of parts of speech is irrelevant either to our understanding of grammar or to semantic interpretation of a sentence . Nor is it possible to assume that the order of morphemes –less , -s , -ly determines grammaticalness because the sentence

(5) Friendly young dogs seem harmless .

orders these morphemes as in (2) , not (1) .

條件(3)可能會發生在教學冊子裡的和平主義修辭；他被認為是合乎文法且易受語意解釋的影響。(4)中沒有一個是正確的。但是(3)的詞性順序就是不合文法的條件(2)，而合文法的條件(1)有(4)的順序。顯然地，詞性的過渡可能性對於我們的文法理解或句子的語意解釋是毫無關係的。也不可能假設詞素–less , -s , -ly的排列決定了文法，因為句子(5) Friendly young dogs seem harmless . 排列這些詞素是在(2)而非(1)。

One of the many problems that an infant who is learning to speak must face is the fact that the meaning of individual words can not ordinarily give a clue to the meaning of a sentence . This is easily seen by citing such sentences as

(6) The fox chases the dog .

and

(7) The dog chases the fox .

The complexities involved in understanding sentences may be further illustrated by the sentence

(8) The dog is chased by the fox

which is understood as having the same meaning as (6) even though the order of subject and object is that of (7) .

在眾多問題中的一個是一位正在學習說話的嬰兒必須面對各個字的意義是無法提供線索給句子意義的真相。在引用的句子裡是易見的，像是(6) The fox chases the dog .和 (7) The dog chases the fox . 這複雜性包含了句子的理解可能進一步地由句子(8) The dog is chased by the fox . 說明。即使句子(8)的主詞和受詞的排序是和 (7) 一樣，句子(8)仍因和 (6)有相同意義而被理解。

Once more it is tempting to explain our understanding of the grammar of these sentences by postulating a simple chain of uni-directional associations between elements .For instance ,take a sentence in which the first noun-phrase is the subject ; here we learn to expect this noun-phrase to be followed by another noun-phrase that serves as object , except if the noun-phrase is followed by an is , the verb by an –ed and the verb-phrase by the word by ; when these morphemes appear , they would signal the reversal of subject and object . Again , the source for our interpretation of the meaning of a sentence must be due to a deeper understanding of grammatical structure than these rules would indicate . The sentence

(9) The fox is interested by virtue of his nature in chasing the dog
may not be elegant and even contrary to fact ; but it is a grammatical string , and its meaning is associated more closely to (6)than to (7) , showing that morphemes do not have the same function in a sentence as , for example , a verbal stimulus in a word-association task . Word association seems to be irrelevant to that process of grammatical analysis which must be performed by ever speaker and language-acquiring infant .

藉由假設要素間的單方向結合的簡單鏈，再一次企圖解釋我們對句子的文法理解。以一句子為例，句中第一個名詞片語是主詞；在這我們可以預期此名詞片語後面會接著另一個名詞片語當作受詞，除了如果第一個名詞片語後接著is，動詞後跟著–ed，動詞片語後跟著by這個字；當這些詞素出現時，他們會發出主詞和受詞逆轉的訊息。此外，我們對句子意義的解釋的來源必須是由於比這些被指出的規則有更身的文法結構的了解。句子(9) The fox is interested by virtue of his nature in chasing the dog 不能是講究精緻的，也不能和事實相反；但他是合於文法的條件，且他的意義和(6)較有關聯，勝於(7)。這表示，詞素在句子中沒有相同的功用，例如，在字詞結合任務中的口頭刺激。字詞的結合似乎與那些每位說話者和語言獲得的嬰兒所執行的文法分析的過程是毫無相關的。

If phonology and syntax pose many unsolved problems , the association between words and things is by no means less difficult to understand . A great number of different words occur in physical and temporal contiguity with one phenomenon (say a dog ) ; some example are : “ wow-wow , ” bark , naughty , dog , big , up on the … , careful . etc . On the other hand , one and the same word , for instance , “ bye-bye ”(or up , or good , or truck ) , may occur in the presence of constantly changing physical stimulus configurations . Why is the child not confused by this ?

如果音韻學和句法提出許多未解決的問題，則詞和事物的結合絕不會較不難理解。大量的不同詞以one現象發生在物理的和時間的一連串事物中(說一隻狗)；一些例子是：“ wow-wow ” 、咆哮、頑皮、狗、大、在…上面，小心等等。換句話說，一個和相同的詞，舉例來說，“ bye-bye ” (或上、好、卡車)，可能發生在經常改變物理刺激一連串事物的面前。為何小孩不會被此搞混？

We cannot help but wonder how an infant at little over one year of age can ever learn to understand and produce this behavior . The number of articles and books that deal with the development of language goes up into the thousands . But only few authors have seen that there is a formidable and totally unsolved problem here . How does the child develop language ? To say vaguely that it must be discrimination learning , secondary reinforcement , or stimulus generalization does not bring us any closer to a solution , because it is not at all clear what has to be discriminated or what is generalized from or to , nor is it clear what is being reinforced , when and how . Oversimplifications , and even representations that are blatantly contrary to observable facts regarding the nature of language , have often led to explanations for language learning that rest on nothing but fiction .

我們禁不住想知道嬰兒是如何在經過不到一歲數的時間裡就學會理解和產生這行為。處理語言發展的書和文章，已經是數以萬計了。但只有少數的作者看見這裡有巨大的且完全未解決的問題。孩子是如何發展語言？含糊地說，這必須是區別學習、次級增強，或刺激概念化並不會使我們離解答更近，因為並不是要清楚明嘹地區別什麼，或歸納成什麼，或從什麼歸納而成，亦不是要增強什麼、何時和如何。過渡單純化的事物以及與有關語言本質的顯見事實相反的相同表述會導致對基於想像的語言習得的說明解釋。

The problems involved in language development cannot be understood in the absence of an analysis of the structure of language ; and it is quite possible that the proper understanding of language structure is dependent upon empirical investigations into the acquisition process .

這類包含語言發展的問題，在語言結構分析的缺乏是無法理解的；而且語言結構的事當理解是有可能的，取決於獲得過程中的糾驗方面的研究調查。

LBT310-315雅雯

LBT310-315雅雯

An interesting question concerns the role of intelligence in the acquisition of language. Is mastery of this, in a sense, highly abstract behavior dependent upon measurable intelligence? The problem is complicated (1) by the definition of intelligence and (2) by the changing intelligence quotients with chronological age among the feeble-minded. An individual whose cognitive status remains constant on a level comparable to that of the normal three-year-old appears to have a steadily falling IQ throughout childhood due to the peculiar way in which this figure is computed. The situation is well-illustrated in the scattergram of Fig7-10. (Compare also Zeaman and house, 1962.) The study of the mongoloid population, as well as that of additional case of mental retardation, indicates that there is a certain “IQ threshold value” that varies with age and that must be attained for language to be acquired. Individuals below this threshold have varying degrees of language primitivity, as illustrated in Fig. 4.3(Chapter Four). It is noteworthy that this threshold is relatively low. If we take a population whose IQ is at or just above threshold, which is the case of mongoloids intelligence figures correlate quite poorly with language development. Only if we confine our observations to the low grades of feeblemindedness can a relationship between intelligence and language learning be established.

一個有趣的問題是智力在語言習得上扮演著什麼樣的角色。智力是否支配著語言?在某些意義而言，高度抽象化行為是否取決於智力？這個問題隨著下列兩個原因變的複雜，(1)智力的定義(2)智力商數隨著年齡的變化與否。經由這個圖計算的特殊方式，比較一個認知狀態維持恆定的個體，和在幼年時期有穩定下滑IQ的正常三歲個體個體。這個情形在7-10的散佈圖中是很好說明(也比較於1962年Zeaman和House)。關於唐氏症人口的研究和心智遲緩的案例一樣，指出特定的「智力閾值」，其隨著年齡的變動，而且智力對於語言獲得是需要的。個人在智力閾值方面低下會有不同程度的語言結構，圖4-3解釋(第四章)顯著的低於閾值。如果我們對唐氏症的案例中作抽樣，個體智力剛好在閾值或在閾值之上，智力型態和語言發展的關係是相當的薄弱地。除非我們限定觀察嚴重智力缺陷的小孩，那麼智力與語言發展是有相關性的。

Among the mongoloids, whose prognosis for mental development is not the worst, chronological age is a much better predictor for language development than computed IQ’s.

在唐氏症患者中，認知發展預後表現不差的個體，心智年齡是跟IQ比起來對於語言發展比較好的預測值。

The relationship between physical maturation and language development has been treated in Chapter Four. Relevant to the same topic is Table7.4.( The criteria for “language developing “was the predominance of words and phrases in all utterances and absence or at most a modicum of random babbling.) The relationship between development of gait and of language appears to be roughly similar to that of normal children; there is greater likelihood for language acquisition after gait is established than before. The development of hand preference is particularly interesting. Right-handedness emerges at the time that language unfolds, even though this occurs at a considerably later time than in normal children.

身體成熟和語言發展的關係已經在第四章討論。有關的議題主要是表7.4(對語言發展的標準判斷是每個言語的優勢辭彙和片語以及缺少或只是少量的隨意亂語) 。步伐的發展和語言的關係大致上和正常小孩相似。對於語言習得相較於在學步之前，有一個可能性是建立在學步之後。另外對於慣用手的發展也是很有趣的。呈現慣用右手的現象同時語言出現，儘管語言發展明顯晚於正常小孩。

表格7.4語言發展階段和各式動作技能關係

語言發展
	總是喃喃聲		語言 發展		總計
動作技能
走路.跑步	14		39		53
東倒西歪走或更差	7		1		8
總計	21		40		61
		X2=8.94	P<0.01
著裝	4		27		31
需要幫助穿 衣	17		13		30
總計	21		40		61
		X2=12.93	P<0.001
自己餵食良好	11		35		46
餵食能力弱	10		5		15
總計	21		40		61
		X2=7.36	P<0.01
慣用右手	6		23		29
慣用左手	3		3		6
兩手同用	3		3		6
沒有特定左右手習慣	9		11		20
總計	21		40		61

表格7.5

	念名字 長度	字彙大小	理解能力	發音	語言發展階段
念名字 長度		0.94	0.92.	0.60	0.95
字彙大小			0.94	0.23	0.95
理解能力				0.30	0.95
發音					0.41
語言發展階段

Mongoloid children are known for their simple but affectionate and pleasant personalities. They are eager to please those surrounding them and, of course, are more dependent on parental care than the ordinary child. They have a tendency toward clowning, and they love to imitate. (Such generalizations are possible because they are remarkably alike in their disposition!) Because of their retardation both their state of dependency and their babbling phase are protracted often years beyond the normal duration of these periods. Considering these data, we may well wonder how the development of language differs in these children from that of the rest of the population. If dependence upon adults, extensive babbling, and propensity for imitation were sufficient factors for language development, these children should develop better language than others. Naturally, their mentation must be deemed insufficient for rapid language progress. On the other hand, some do eventually develop all the essentials of language, and, in these cases, we can hardly suppose that they eventually improve their mentation. Once more we are brought to believe that there is an immanent schedule of evolvement in which apparently one set of events sets the stage for a subsequent set, and so on. However, in the case of mongoloids, where worried parents make often desperate efforts to teach their child to speak, where bodily imitation is frequently specifically rewarded by those tending to the child’s needs, we might expect that the children differ among themselves and from other children in “their strategies” for language acquisition. Would it not be possible, theoretically, that one child first tried to perfect his articulation before trying to increase his vocabulary; another might always try to make sentences out of the ten words he knows; or still another might have all his needs taken care of by his family and therefore content himself with learning to understand language without making an effort to speak himself.

唐氏症小孩最顯著特徵是其簡單卻又柔情愉悅的心情的特質，他們渴望和週遭的人和平相處，也比一般大眾依賴著雙親的照顧。他們有開玩笑、幽默的傾向也喜歡模仿。(上述普遍性是可能的，因為他們在各種性情顯著相似!) 因為遲緩的情況，不管是依賴的階段或是喃喃聲時期都長時間擴展的比正常發展時期晚。就這些資料數據而言，我們會懷疑這些孩子語言的發展是否和其他人不同。若是依賴成人有過度的喃喃學語以及模仿的傾向是影響語言發展的重要因素，這些孩子理論上比其他孩童有好的語言發展。自然地，他們的認知不被認為是對快速語言進步的足夠因素。換句話說，一些人確實發展基本的語言而且我們很難假設最後(徹底)改善他們的智力。再一次地，我們被促使的相信有一個天生固有的發展計劃表，也就是一個情況(A)產生在另一個情況(B)之前，而另一個情況(B)以先前情況(A)為前提建立，以此類推。然而在唐氏症患者裡，擔心的家長越是極度渴望且努力使孩子說話，身體上的(有形)的模仿是受到照顧者的鼓舞。我們可能期待著這些孩子不同於自己且和其他孩子在「他們的策略」對語言的發展有不同。或許這是不可能的，理論上一個小孩在試著增加詞彙之前，第一次試圖完美化他的發音;另一個小孩可能總是用他認識的十個字造句子;再另一個小孩，有出自家庭關心的所有需求，因此對於自己學習理解語言感到滿足而沒有意願發展言語。

	語言發展
	總是 喃喃聲		語言發展		總計
說出名字
說出每個人名字	1		36		37
說出一些名字	20		4		24
總計	21		40		61
		X2=38.43		P<0.001
字彙
大於50 字	2		40		42
50或更少	18		0		18
總計	20		40		60
		X2=47.23		P<0.001
理解
好	5		39		44
有限	16		1		17
總計	21		40		61
		X2=33.62		P<0.001

Our investigation have shown that this is not so. In all the patients studied, the sequence of learning phases and the synchrony of emergence of different language aspects remained undisturbed by the disease. The correlation matrix of Table 7.5 shows that progress in one field of language learning is well correlated with progress in all other fields, except for articulation.

我們的研究調查發現並非如此。在所有病人報告裡，一連串學習階段和同步出現的不一樣語言方面仍然不因這疾病受干擾。表格7.5的相關矩陣顯示除了發音以外，語言學習的某一範疇的發展是和所有其他範疇發展相關的。

The variables of the matrix are all dichotomized and qualitative; the data for the correlations are shown in the six contingencies shown in Table 7.6 and 7.7.

這多變的矩陣是交叉且性質上的。這些相關數據是以這六個情況顯示在表格7.6和7.7。

表格7.7.發音和理解的關係。字彙和語言發展的階段。

	發音
	理解50﹪或更少		理解大於50﹪		總計
理解
好	33		10		43
有限	15		2		17
總計	48		12		60
		X2=0.42		不顯著的
字彙
多於五十字	31		10		41
五十字或更少	16		2		18
總計	47		12		59
		X2=0.67		不顯著的
語言發展
語言發展	29		10		39
大部分喃喃聲	19		2		21
總計	48		12		60
		X2=1.32		不顯著的

The simultaneous unfolding of language,“across the board.” is of great importance for language theories. There is no a priori reason why a child who has stopped babbling and whose utterances are always at saying words or phases, should also have a vocabulary of 50 words or more (he might be content to say the same five words again and again; or he might still be babbling randomly at times while already in possession of some hundred different words); nor is it immediately obvious why the postbabbler should have adequate understanding of spoken commands (he might have learned to imitate like a parrot); or that he should have the same facility in naming people as in naming classes of objects. The one exception to the rule of simultaneous unfolding of language skills is the children’s articulation. The lag is not due to structural abnormalities of fauces or tongue (see complete report, Lenneberg et al..1964)

這些同時呈現的語言，「全面性」是對語言理論有很大重要性的。那裏沒有一個猜測性的理由，為什麼一個小孩誰會喃喃自語和誰的語音總是在說字和片語也理應有50或更多的字彙。(他可能滿足於說一樣的五個字一遍又一遍，或者在他早已經具備數百個不一樣字詞時，可能有時候仍在隨意地牙牙學語)。為什麼postbabbler對於說話能力應該有足夠的認識，這也不是直接顯著的(他可能學著模仿像隻鸚鵡)，或者他可能在說出人們的名字時和說出各式各類的物體目標時有一樣的能力。這些同時呈現的語言技巧規則，惟一的例外是小孩子們的發音。遲滯並不是由於咽喉和舌頭構造上異常。(見完整報告Lenneberg et al..1964)