Language understanding in psycholinguistics

“Language understanding” is a topic intensively discussed in psycholinguistics. Its research includes: (1) speech perception; (2) vocabulary extraction; (3) sentence processing; (4) discourse understanding. In short, it is the study of how people understand language.

First of all, "speech perception" involves "research methods of speech perception", "conditions of speech perception", "generation of speech signals and acoustic characteristics of speech", "listening and identification of vowels and consonants" ", "Continuous Speech Listening", "Perception of Written Language" and "Speech Perception Model" *** 7 aspects of research. The research method of language perception first used the "vocoder" invented by H. Dudley (1939) to record and analyze language output information. Later, the principle of making a "sound recorder" led to the emergence and development of the "sound spectrogram". This is a product produced during World War II specifically used to analyze speech signals according to the distribution of sound frequencies. In the 1960s and 1970s, influenced by acoustic phonetics and articulatory phonetics, "electromyography" and "electrokymography" appeared, which were used to record muscle contractions respectively. The generated voltage changes and the airflow changes in the mouth and nasal cavity when speaking. Of course, the latest research method is "cineradiography", which is used to record the dynamic changes of pronunciation. In the study of language perception conditions, "context" is the most important condition that affects speech perception. Most of the research on speech signal production is related to "acoustic characteristics" and "listening of vowels and consonants". This is a relatively complex study, a study on "phoneme" listening and discrimination using the English language as the research object. Another area of ??speech perception research is the study of "written language understanding." This involves the exploration of aspects such as "visual induction", "letter recognition", "determination of word meaning", as well as "information memory" and "information organization". Of course, the most important field of speech perception research is the study of "speech perception models". So far, the "Motor Theory" proposed by Liberman (1967, 1970) and his colleagues, and the "Analysis-by-Synthesis Model" proposed by Stevens (1960) ), the “Fuzzy Logical Model” proposed by Marcello (Hassaro, 1987, 1989), the “Cohort Model” proposed by Maslen and Wilson, and the “Cohort Model” proposed by Elman (1984, 1986 The "Trace Model" proposed by ) has become one of the five major models in language perception research and is one of the five important "model" theories that influence psycholinguistics research on language understanding.

Second, in terms of research on "vocabulary extraction", its exploration mainly focuses on "basic elements of words", "research methods of psychological vocabulary", "factors affecting vocabulary extraction and organization" and " Vocabulary extraction model" and other 4 aspects. The study of vocabulary extraction is inseparable from the analysis of the basic elements of the word, because the basic elements of the word are directly related to the meaning of the word. The research method of psychological vocabulary is directly related to the research of language understanding and is an important part of the research of language understanding. There are three main research methods on mental vocabulary: "reaction time experiment", "naming/lexical retrieval" and "speech error analysis". Among them, "speech error analysis" and "naming" are the two most important methods for studying vocabulary extraction. People can figure out how people understand the meaning of words or extract language information through speech errors such as "lip phenomena", "word transposition" and "gaffes" and the analysis of the errors caused. In terms of research on "factors affecting vocabulary retrieval and organization", the research focuses on three aspects: "lexical effect", "semantic effect" and "context effect". The research on vocabulary extraction models is a fruitful research. So far, the "Serial Search Model" represented by the "automatic search model" of vocabulary extraction by Forster (1976) and the "lexical generation model" proposed by Morton (1969, 1979) "Parallel Access Model" represented by "Parallel Access Model" has become the two mainstreams in vocabulary extraction research. In particular, Morton's "lexical generation model" provided important clues for later research on word retrieval in the study of "lexical activation". In addition, another one joining the ranks of “parallel extraction models” is the “connectionist model”. This is an important theory proposed by connectionists in psychology, philosophy and computer science to explain vocabulary extraction. Connectionism believes that the final understanding of a word is completed from the incremental connection of "input" → "feature" → "letter" → "word" of information. In addition to the "connectionist model", there is also an auditory word recognition model called the "Cohort Model".

Marslen-Wilson (1987) proposed that when a person hears a word, all of its phonetic neighbors will be activated. Of course, no matter which theory is used, it is difficult to explain the entire complex process of language understanding. However, the "activation" theory emphasized by each "model" is still a topic of general significance and worthy of further research and exploration in explaining the complex process of language understanding.

Third, sentence understanding is an important part of language understanding research. Research in this field is divided into: (1) The nature of sentence structure; (2) Syntactic processing; (3) Sentence decomposition and syntactic ambiguity; (4) Sentence decomposition models; (5) Language processing and memory; (6) A processing model of sentence comprehension. Psycholinguists believe that the reason why people can understand language is because the information receiver and the information outputter have the same grammatical rules or "conventions" in their minds. However, the research on sentence understanding is not that simple. It also involves the surface structure and deep structure of the sentence. Just take the sentence "It's too cold here" spoken by A as an example. If B takes the literal meaning of the sentence, then he may not understand the true meaning of A. Therefore, syntactic processing is the second issue explored by psycholinguistics in language understanding. The research on syntactic processing involves four issues: "the relationship between surface structure and deep structure", "the relationship between language ability and pragmatic ability", "decomposition of sentence structure" and "processing of clauses". It mainly explores how people are How to understand compound sentences and their meaning through syntactic processing. In the study of sentence decomposition and syntactic fuzziness, psycholinguists focus on the discussion of two aspects: "local fuzziness" and "permanent fuzziness". Since the "fuzziness" of language is an essential feature of non-artificial language, using the "sentence decomposition model" to study the understanding of fuzzy sentences has become a major means of studying language understanding in psycholinguistics. In terms of research on decomposing fuzzy sentences, J.B.Cleason (1998) divided his models into two types: one is the "garden path model" and the other is the "forced perfection model". The garden path model discusses the process of sentence comprehension based on the principle of "structure first" or "vocabulary first". However, to understand sentences correctly, sentence memory is also an important part. In this regard, psycholinguists focus on research in three aspects: "memory of meaning and surface structure", "semantic reasoning and sentence memory", and "proposition and sentence memory". Of course, to understand how people remember language information, the "processing model of sentence understanding" is the foundation for studying language understanding. So far, the "perception strategy" proposed by Fodor (1974) and others, the "non-switching strategy" proposed by Bever (1970), the "surface strategy" proposed by Kimball (1973) "Seven Principles of Structural Syntax Analysis" and the "enema machine" theory proposed by Frazier (1978) have become the four major main models of sentence understanding and processing (22). In particular, Kimbauer's "7 Principles" (23) basically summarize the general process and research ideas of sentence understanding.

Fourth, the research on "discourse comprehension" can basically be summarized into three aspects: (1) research on discourse coherence and comprehension strategies; (2) research on discourse memory; (3) Research on discourse processing. As far as the study of discourse coherence is concerned, "anaphora" (anaphora) and "cataphora" (cataphora) proposed by Halliday and Hasan (1976) have become a major theme in discourse coherence research. In the study of speech perception, Jakimik & Glenberg (1990) also regarded "up-pointing" and "down-pointing" as important prerequisites for language understanding. But whether it is "pointing up" or "pointing down", it talks about the relationship between the close connection of sentences and the coherence of discourse. In terms of research on discourse comprehension strategies, Clark and Haviland (1977) proposed the "known/unknown strategies", "direct matching strategies", "bridging strategies" and "retrospective known strategies" for discourse information. "Information strategy" has become an important thinking clue for discussing discourse understanding. Regarding the exploration of discourse memory, the "organizational planning framework" of the human cognitive system proposed by Bartlett (1932) still influences current psycholinguistic research on discourse memory. He believes that memory is active, creative and planned. As far as discourse memory is concerned, "proposition" and "inference" have become the center of research. Psycholinguists believe that not only topic identification and sentence combination are closely related to discourse memory, but discourse understanding also contributes to discourse memory. In terms of research on discourse processing, "discourse processing model" has become the central content of research on discourse understanding. Such as "context model", "connectionism model", "construction and combination model", etc., all explain the process of discourse understanding from various angles.