凱蒂·伯頓 陳偉濟

An experiment that asked different sized groups to invent a new language has revealed that community size plays an important role in determining the type of language that develops. 一項實驗讓人數(shù)不同的幾個小組發(fā)明一種新語言，結(jié)果表明，群體規(guī)模對新語言的類型有重要影響。

The differences between the 6，800 or so languages that currently exist in the world are remarkable. From Cantonese， in which a speaker must perfect six different tones each of which change the meaning of a single word， to Georgian， in which verb endings vary not just according to the tense or plurality （as in English）， but in up to 200 other ways. Grammatical and morphological systems vary hugely. What is the source of this linguistic diversity？ And why might some languages be so much harder to learn than others？

Many researchers have dedicated their time to answering these questions. In particular， extensive research has been carried out into the effect of the physical environment on the development of language. As one theory has it， languages typically spoken in very dry and cold climates （Siberia， for example） tend not to involve tonal distinctions between words （as is the case in Cantonese）—the simple reason being that it is harder to precisely control the vocal cords in dry environments.

Now research is being carried out into the role of social environments in language development. Limor Raviv， a researcher at the Max Planck Institute for Psycholinguistics in Nijmegen is working on this topic. In a recent study， she sought to answer one social question in particular—does the size of a community affect the development of a new language？

To find out the answer， Raviv designed an experiment in which two groups—one with four participants and one with eight—were set the task of creating a new language to describe the same set of 23 simple scenes. Within each group a ‘speaker would see one of four shapes moving in some direction on a screen and type in nonsense words to describe the scene （both its shape and direction）. The ‘listener would then guess which scene the other person was referring to， by selecting one of eight scenes on their own screen. Participants received points for every successful interaction.

Raviv discovered that community size did indeed affect the outcome. In the main， the larger groups created languages with more systematic grammars than the smaller ones. For example， one large group created the term ‘wowo-ik to refer to a specific shape moving up and right， and used ‘wowo-ii to refer to the same shape moving straight up. A regular system like this made it easier to predict the meaning of new labels （so ‘mop-ik meant a different shape going up and right）.

It has already been noted in previous studies that this phenomenon—in which languages spoken in larger communities typically have more systematic grammatical structures—exists. For example， English， now one of the worlds most widely spoken languages， is structurally quite simple when compared to many others. While it might have a large vocabulary， its grammatical structure is relatively easy to follow and even irregularities are still grouped in systematic ways.

Previous hypotheses as to why this might be the case have sought to identify some kind of external influence. One suggestion has been that larger groups have more adult second-language learners who， if the pressure is big enough， result in the simplification of a language. But Raviv now has another theory more directly related to the original community size in which the language developed.

‘I believe that you may get this effect just because the language is spoken by a bigger community and because it passes through more minds， she says. ‘The language evolves in a context where more people need to negotiate about it， to reason about it， and theres more contact with strangers. So we hypothesise that languages that develop in bigger communities evolve to be more systematic to accommodate the difficulty in communicating with more people that you know less about. Highly elaborate structures that really rely on intimate familiarity between every member of the group just cant hold in a big community.

There are clear advantages to the type of systems typically developed by large communities， she argues. ‘It really allows productivity and flexibility. So even two strangers that have never interacted before about something theyve never seen before can immediately use their language and understand each other. They dont need any prior negotiation.

Raviv is now taking this study further and attempting to teach the newly-created languages to other participants in order to verify whether more structured languages really are easier to learn. In doing so she may be able to challenge a widely held assumption—that all languages are similarly easy to acquire. ‘Actually， some studies suggest that this is not the case， she says. ‘Some languages are harder to learn， even for children. For example， Danish is considered to be a really hard language for children to learn and there is evidence that they master some features much later than speakers of other languages， even related ones such as Norwegian or Swedish.

None of this， Raviv makes very clear， makes any one language better than another. ‘All languages are equally good at expressing a message， she says. But experiments such as these might just help us to understand why we speak the way we do and why learning some languages fills us with much more fear than others. ? ? ? ? ? ■

世界上現(xiàn)存6800多種語言，它們之間的差別十分明顯，如粵語和格魯吉亞語。說粵語必須精通六種不同的音調(diào)，對同一個字來說，一種音調(diào)有一個意思;格魯吉亞語中，動詞詞尾不僅有時態(tài)或復(fù)數(shù)（如英語）的變形，還有其他多達200種變化。語言的語法和形態(tài)體系差異很大。語言的這種多樣性源自什么？為什么有些語言比其他語言難學(xué)得多？

許多研究人員已投入時間來解開這些疑問，特別是在物理環(huán)境對語言發(fā)展的影響方面，開展了大量研究。有一種理論認為，通常在非常干燥寒冷的氣候下（如西伯利亞）使用的語言，字詞間往往沒有像粵語那樣的音調(diào)差異，原因很簡單，干燥環(huán)境下，更不容易準(zhǔn)確控制聲帶。

有關(guān)社會環(huán)境對語言發(fā)展的影響，目前正在開展研究。馬克斯·普朗克心理語言學(xué)研究所（位于荷蘭奈梅亨）研究員利莫爾·拉維夫正在研究這個課題。在最近的一項研究中，她試圖回答一個特別的社會問題——群體規(guī)模會影響新語言的發(fā)展嗎？

為了找到答案，拉維夫設(shè)計了一個實驗。實驗分兩組，一組有四名參與者，另一組有八名;要求這兩組成員創(chuàng)造一種新語言來描述相同的23個簡單場景。每一組的“說者”會看到屏幕上四個形狀中的一個朝某個方向移動，然后輸入無意義的單詞來描述看到的場景（包括形狀和方向）?！奥犝摺眲t要猜測“說者”所描述的場景，在自己屏幕上的八個場景中選擇一個作為答案。參與者的每一次成功互動都會獲得積分。

拉維夫發(fā)現(xiàn)，群體規(guī)模的的確確會影響結(jié)果。大體上，比起人數(shù)少的小組，人數(shù)多的小組創(chuàng)造的語言語法更為系統(tǒng)。例如，一個人數(shù)多的小組創(chuàng)造了術(shù)語wowo-ik來指代向右上方移動的特定形狀，使用wowo-ii來指代向正上方移動的同一形狀。像這樣有規(guī)則的系統(tǒng)更容易預(yù)測新詞語的含義（mop-ik便意味著向右上移動的另一種形狀）。

先前的研究便已注意到這種現(xiàn)象的存在：規(guī)模大的群體所說的語言通常具有更系統(tǒng)的語法結(jié)構(gòu)。英語就是一例，作為目前世界上使用最廣泛的一大語言，與許多其他語言相比，其結(jié)構(gòu)非常簡單。盡管英語可能詞匯量很大，但其語法結(jié)構(gòu)相對容易理解，即使是不規(guī)則的部分，仍然可以系統(tǒng)地分門別類。

至于為什么可能是這種情況，以往的假設(shè)一直試圖找出某種外部影響。其中一種說法是，大的群體有更多第二語言的成人學(xué)習(xí)者，一旦影響夠大，語言就會被簡化?，F(xiàn)在，拉維夫提出了另一種理論，該理論直接涉及語言發(fā)展之初的群體規(guī)模。

“我認為之所以可能有這種效應(yīng)，只是因為使用語言的群體更龐大，因為語言歷經(jīng)了更多人的思考。”她說，“在更多人需要用語言進行協(xié)商、推理，并且和陌生人有更多往來的環(huán)境下，語言才會逐步發(fā)展。所以我們提出這樣的假說：為解決與更多不甚了解的人進行交流的困難，越龐大的群體造就的語言越系統(tǒng)。高度復(fù)雜的結(jié)構(gòu)十分依賴群體每個成員間的高度熟悉，完全不適用于龐大的群體。”

她認為，通常由龐大群體發(fā)展起來的語言體系具有明顯優(yōu)勢。“這類語言利于產(chǎn)出、富有彈性。即使是對未知事物未曾有過交流的兩個陌生人，也能立即使用語言溝通并相互理解。他們不需要任何事先的協(xié)商?！?/p>

目前，拉維夫正在進一步研究，嘗試向其他參與者教授新造的語言，以驗證結(jié)構(gòu)化更強的語言是否真的更容易學(xué)習(xí)。此舉也許會質(zhì)疑人們廣泛認可的觀點——所有語言的習(xí)得都一樣容易。“實際上，有些研究表明情況并非如此?！彼f，“即使是對于孩子，有些語言也比較難學(xué)。例如，丹麥語被認為是兒童很難學(xué)會的語言。有證據(jù)表明，他們掌握某些語言特征的時間比說其他語言的兒童晚得多，即使與挪威語或瑞典語等相似的語言比，也是如此?！?/p>

拉維夫明確指出，這些都不能說明某種語言比另一種語言更好。她說：“在傳遞信息上，所有語言不相上下?！钡@樣的實驗或能幫助理解我們自身說話方式形成的原因，以及為什么學(xué)習(xí)某些語言會比學(xué)習(xí)其他語言讓我們畏懼得多?！? ? ? □

（譯者為“《英語世界》杯”翻譯大賽獲獎?wù)?單位：廣東科學(xué)院）