Savage-Rumbaugh E S, Murphy J, Sevcik R A, Brakke K E, Williams S L, Rumbaugh D M
Monogr Soc Res Child Dev. 1993;58(3-4):1-222.
Previous investigations of the linguistic capacities of apes have focused on the ape's ability to produce words, and there has been little concern for comprehension. By contrast, it is increasingly recognized that comprehension precedes production in the language development of normal human children, and it may indeed guide production. It has been demonstrated that some species can process speech sounds categorically in a manner similar to that observed in humans. Consequently, it should be possible for such species to comprehend language if they have the cognitive capacity to understand word-referent relations and syntactic structure. Popular theories of human language acquisition suggest that the ability to process syntactic information is unique to humans and reflects a novel biological adaptation not seen in other animals. The current report addresses this issue through systematic experimental comparisons of the language comprehension skills of a 2-year-old child and an 8 year-old bonobo (Pan paniscus) who was raised in a language environment similar to that in which children are raised but specifically modified to be appropriate for an ape. Both subjects (child and bonobo) were exposed to spoken English and lexigrams from infancy, and neither was trained to comprehend speech. A common caretaker participated in the rearing of both subjects. All language acquisition was through observational learning. Without prior training, subjects were asked to respond to the same 660 novel sentences. All responses were videotaped and scored for accuracy of comprehension of the English language. The results indicated that both subjects comprehended novel requests and simple syntactic devices. The bonobo decoded the syntactic device of word recursion with higher accuracy than the child; however, the child tended to do better than the bonobo on the conjunctive, a structure that places a greater burden on short-term memory. Both subjects performed as well on sentences that required the ability to reverse work order as they did on sentences that did not require this capacity. These results are discussed in light of a model of the evolution of language that suggests that the potential for language comprehension preceded the appearance of speech by several million years at minimum. The onset of speech is linked to the appearance of fully adapted bipedalism, which necessitated reorientation of the laryngeal tract and made closure of the soft palate possible. For the first time, such closure permitted mammals to easily produce sounds that could be interpreted by the mammalian auditory system in a categorical manner. When these sounds were paired with the previously extant capacity to produce vowels, it became possible to form "bounded vowels" or sound units that could readily be discriminated as units by the auditory system. It is suggested that this physical adaptation allowed the extant cognitive capacity of the hominids to embark on a speech-like mode of communication.
以往对猿类语言能力的研究主要集中在猿类产生词语的能力上，而对其理解能力关注甚少。相比之下，人们越来越认识到，在正常人类儿童的语言发展过程中，理解先于表达，而且理解确实可能引导表达。已经证明，一些物种能够以类似于人类的方式对语音进行分类处理。因此，如果这些物种具有理解词语指代关系和句法结构的认知能力，那么它们就应该有可能理解语言。关于人类语言习得的流行理论认为，处理句法信息的能力是人类独有的，反映了一种在其他动物身上未见的新的生物适应性。本报告通过对一名2岁儿童和一只8岁倭黑猩猩（倭黑猩猩属）的语言理解能力进行系统的实验比较来探讨这个问题。这只倭黑猩猩生长在一个与儿童成长环境相似但经过专门调整以适合猿类的语言环境中。两个受试对象（儿童和倭黑猩猩）从婴儿期就接触英语口语和符号语言，且都未接受过理解言语的训练。由同一位照料者参与抚养这两个受试对象。所有语言习得都是通过观察学习。在没有事先训练的情况下，要求受试对象对同样的660个新句子做出反应。所有反应都进行了录像，并根据对英语的理解准确性进行评分。结果表明，两个受试对象都理解新的要求和简单的句法结构。倭黑猩猩对词语递归的句法结构解码的准确性高于儿童；然而，在连词结构上儿童往往比倭黑猩猩表现得更好，因为连词结构对短期记忆的要求更高。在需要颠倒语序能力的句子和不需要这种能力的句子上，两个受试对象的表现一样好。根据一种语言进化模型对这些结果进行了讨论，该模型表明，语言理解的潜力至少在言语出现前几百万年就已存在。言语的出现与完全适应的双足行走的出现有关，双足行走需要重新调整喉道方向，并使软腭的闭合成为可能。第一次，这种闭合使哺乳动物能够轻松发出能够被哺乳动物听觉系统以分类方式解读的声音。当这些声音与先前已有的发出元音的能力相结合时，就有可能形成“有界元音”或声音单元，这些声音单元能够很容易地被听觉系统作为单元区分开来。有人认为，这种身体适应性使原始人类现有的认知能力能够开启一种类似言语的交流方式。
