汉语句子理解中句法与语义相互作用的研究

【作者】 刘宇；

【导师】 冯建新；

【作者基本信息】 陕西师范大学 ， 基础心理学， 2009， 硕士

【摘要】 由于句子一直以来被认为字词组合以表意义的理想单位,在对语言理解的多层次研究中有关句子理解的研究对于了解人的语言心理机制具有很重要的意义。六十年代以来,心理学关于句子理解的研究主要集中在句法分析过程。但是进一步的实验证据更倾向于将语义作为一个重要因素加以考虑,所有的研究者都承认句法分析的最终结果要受到语义因素的影响,同时语义的作用表现在对每个短语甚至每个词的分析上,而不是在确定整个句子的结构之后才进行。但对于二者之间的相互作用的具体机制,相互之间的独立性以及加工过程中的时间进程,尽管近年来已进行了大量的研究,但是相关理论模型在跨语言,语种时总会产生矛盾,由于汉语的独特特点,对于汉语句子理解的句法与语义加工的探讨就尤为重要,鉴于动态助词在汉语句子中的重要作用,对其的加工的观察和分析可以从一个独特的视角观察汉语句子加工中语义与句法的相互作用机制。本研究通过三个实验,采用句子验证任务,比较系统地考察了在不同的动态助词合理性条件下,以及不同呈现时间中,汉语句子理解中句法和语义加工的相互作用和其时间进程的特点。研究一:利用移动窗口技术,要求被试对呈现在屏幕上的汉字进行识别和加工,并在加工完成后手动切换到下一个字,直到完成对整个句子的阅读加工,在对呈现的所有单字作出反应后,完成一个句子再认任务,记录汉语单字加工时间(包括实意单字词,例如:打,坐,红,我;和动态助词了,着,过)和阅读完成句子所需要的时间,以便考察两者之间在阅读识别时反应时间的差异。研究二:该实验在研究一基础上,运用自控速度移动窗口阅读方式,对四种不同情况的动态助词条件下的汉语简单句的阅读理解速度和动态助词加工的时间进行了考察,通过在句子呈现结束后的句法与语义违背句的快速判断任务探测句法与语义的作用模式。研究三:在本实验中,继续使用了在研究二中使用的四种条件的含有动态助词的实验材料,但改变了呈现方式,不再由被试自己控制逐字呈现而是将整句句子快速呈现,要求被试完成与研究二相同的判断任务,在呈现过程中句中的动态助词的呈现时间被设置为早晚两种情况,以进一步观察句子理解的时间进程。并验证了研究二的试验结果。结果和主要结论:(1)汉语读者对句子的表征是随着句子的展开而逐渐增加的,并会及时加工联结频率较高的非句法信息,与交互作用理论的描述相符;(2)在句子加工的每一个时间进程上,均会产生暂时的句法结构,并对其后的续接成分做出预期,这也说明交互作用理论更符合汉语句子理解的情况;(3)通过不同呈现方式的相互印证,发现移动窗口技术可以得到比较一致的反应结果。这与以前一些研究对这种实验技术获得的实验结果提出的质疑不同,因此可以利用该技术所提供的数据结果推导句法与语义加工的行为模型。更多还原

【Abstract】 In opsite to the simple and experience of reading and understanding sentence in actual life , the complxcity and unsure of mechanism on which the people relie to acomplish the transformation from the deep structure to the surface level result in the argument on how semantic and synatic function in the sentence paser. The effort of researches in the past decades has yet convinced us with experimantial evidence that sentence processing really depend on the the intergration of both elements . However , further discussion of the working style of above two systems is required to define the interaction of them in Chinese.When people read and process sentence in Chinese , inevitably there are two definitive factors: sequence of the lexion and the assistant of the auxiliary word that affect the sentence comprehension . Within several decades As an alternative to this proposal, we consider the possiblity that through contraling the variable of the auxiliary word expose the inner mechanism of such framework. There were three experiments designed for this object.The first experiment utilize the method of moving-window technique(by means of keyboard-controled sequential display techinque ,the reading time of each individual character was mearsured as on-line indicator of comprehension process )to compare the recations under both the conditions of with or without auxiliary word after the verb .In the second experiment ,the auxiliary word is assigned to different time sequence, in another word ,the subjects will read the Chinese sentence in the screen ,in which the auxiliary word appear in various time point compared with other elements of the whole item .Experiment3 used a flexible format that detect how the range between verb and AW influence the paser of semantic and syntax .The result showed a significant difference in diverse point according to the relation to the verb .The finding are discussed in relation to existing computational models in sentence reading .The main results and conclusions:(1) Native speaker of Chinese symbols the sentence according to the extent of the sentence reading .In addition ,the readers will process the high-frequency non- synatic infromation as soon as it is detected ,this is consist with the theory of constraint-satisfaction model.(2) In each stage of sentence process, there is a temporary synatic structure which is the antipithy for the hereafter components. It indecates that the theory of constraint-satisfaction model provide a better interpre of the sentence comprehension in Chinese language. (3) Through the analysis of data from different experiments, the research assured that the technology of moving-windows can reach a credibility conclusion.更多还原