Part 2. Psycholinguistic theories of speech perception 1. Concepts of speech perception and understanding processes

The perception of speech is a complex and multi-dimensional mental process. In principle, it occurs according to the same general laws as any other kind of perception. In the activity of speech perception one can distinguish two different components. The first of them is the primary formation of the image of perception. The second is to identify the already established image (30, 91, 97, etc.).

The perception of speech includes the reception of audible or visually perceived elements of the language, the establishment of their relationship and the formation of ideas about their meaning. Thus, the perception of speech is carried out at two levels - the actual perception and understanding of the speech utterance.

Understanding is defined in psychology as a deciphering of the general meaning that stands behind the directly perceived speech (sound) stream; it is the process of transforming the actual content of perceived speech into the meaning behind it .

For example, the meaning of the phrase “Cold!” May be different depending on the “non-speech context” and how it is expressed. If this is a mother's address to her child, then he can understand her words as advice to dress warmer. If this is said by someone in the room and is accompanied by a gesture in the direction of the open window, the phrase can be understood as a request to close the window. The same phrase expressed by one of the participants of the children's game “cold - hot” has a completely different meaning.

In the course of understanding, the listener (the recipient) establishes between the words semantic connections, which together constitute the semantic content of this statement. As a result of comprehension, the listener may come to an understanding or a misunderstanding of the semantic content of the utterance. It is important to note that the process of understanding from a psychological point of view is characterized by different depth and accuracy (30, 86).

The process of understanding speech is not a mirror reflection of the process of drawing up a separate utterance or an entire text, but there is always the transformation of this utterance into abbreviated semantic schemes at the level of internal speech, which can then be turned back into utterances (42, 91). The process of understanding speech is the extraction from the flow of information of essential moments or essential meaning. That process, which is usually called speech understanding, pointed out LS. Vygotsky, “is something more and something other than performing a response by an audible signal” (43, p. 117). Understanding of speech includes the active use of speech. L.S. Vygotsky believed that the semantic side of speech, understanding, proceeds in its development from the whole to the part, from sentence to word, and the outer side of speech, its production - from word to sentence.

According to the theoretical concepts of A.A. Brudnogo and L.S. Tsvetkova (30, 244), the initial, most general level of understanding consists in understanding only the main subject of the utterance, i.e., what is being said. At this level, the hearer can only say what was said to him, but cannot reproduce the content of what has been said. The semantic content of what was heard serves as a background against which the recipient can determine the main subject of the utterance.

The second level - the level of understanding of the semantic content - is determined by the understanding of the whole course of presentation of the thoughts of the speaker or writer, the course of its development and argumentation. He is characterized by an understanding not only of what was said, but also of what was said (i.e. remarks of the utterance).

The highest level is determined by the understanding not only of what and what was said, but also for what, why (that is, for what purpose). If necessary, it is also clarified by what language means the speaker expressed his thought. Such an insight into the semantic content of speech allows the listener to understand the speaker's speech motives, to understand everything that implies, the internal logic of his statement. This level of understanding includes an assessment of the linguistic means of expression of thought used by the speaker or writer (91, 111, etc.).

All existing theories (and models) of speech perception can be classified according to two major parameters. The first parameter is the motor or sensory character of the perception process; the second is his active or passive character.

The motor theory in its classical version postulates that in the process of listening to a speech, a person determines the values ​​of the controlling motor signals necessary for the production of a message, similar to the one heard (251). Motor theory of speech perception is described in detail in the works of AA. Leontyeva, L.A. Chistovich, PM Frumkina and others (133, 181, 236).

The motor theory of speech perception in Russian psycholinguistics has received its scientific substantiation in the works of LA Chistovich, her students and followers. The views of the representatives of this school are presented in the well-known collective monograph “Speech. Articulation and Perception ”(181). The process of human recognition of audible speech is characterized in the framework of this theory as follows.

Speech sounds are recorded in memory as a set of characteristics based on their characteristics: vowels are recorded with markers indicating the degree of stress. After perception of the stressed syllable, a conditional word boundary is drawn up, and the person finds the appropriate word. If the decision is made, the boundaries of the segment included in the word are marked, and the dictionary of subsequent elections is shortened. Thus, segments of the message, larger than syllables, acquire a new acoustic parameter — a rhythmic pattern (181, 251).

One of the main tenets of the motor theory is the following: “The articulation signs of speech elements, not the acoustic, are the basis of the phoneme classification” (181, p. 188). The acoustic component of the auditory perception is considered as the second main component of speech perception. At the same time, the motor image of a speech unit and its intended sensory image coincide with each other (133, p. 123).

According to A.A. Leontiev, as the perceptual standard in the processes of semantic speech perception, is the sound image of the whole word. This position is based on the views of the SI. Bernstein, who believed that “to pronounce and distinguish words, all the spoken and heard sounds ... perceived as part of words ... Decomposition of words into separate sounds when using the native language is observed only when there are some hearsay, while listening to slurred speech and the like cases beyond normal speech practice ”(20, p. 25).

On this basis, it follows that the unit of the speech perception process is the whole word (and not a separate sound or syllable). Experimental studies testify in favor of the “word-for-word” character of speech perception (27, 122, etc.). This is consistent with the data of well-known foreign researchers (PO Jacobson, G. Mole, and others.). Thus, according to linguists G. Mol and E. Uhlenbeck, the listener does not perceive phonemes one by one in speech, but identifies and determines meaningful sound unity, that is, words (133, p. 130).

One of the original and promising models of the speech perception process is the concept of “analysis through synthesis” put forward by M. Halle and K. Stevens (1964), which is also shared by N. Chomsky and J. Miller (307). Its basic principle is as follows: in order to adequately perceive a speech statement, a person needs to build his syntactic model, fully or at least partially corresponding to the model that is used by another RF actor (speaker) in the process of generating speech. As E. Lenneberg notes, “the mechanism of understanding basically does not differ with the planning mechanism of a statement when it is produced” (301, p. 106). According to A.A. Leontiev, the concept of analysis through synthesis is generally consistent with the “motor principle” of speech perception (133).

Factors affecting the identification of words have been studied in detail by L. R. Zinder and A. S. Stern (96). In their works, it was found that, when identifying words, various properties of speech signals can be used — their phonetic characteristics (or, when read, graphic), semantic and semantic-grammatical features, as well as their “probabilistic characteristics” (the potential occurrence of a given word in a perceived speech utterance). In various conditions of perception, various signs of the speech signal may be dominant in the identification. So, for example, with relatively good identification conditions (when the signal level is significantly higher than the noise level), phonetic features are used primarily, under “medium” conditions, active reliance on semantic attributes occurs, and under the worst conditions, when the noise level is particularly high, the frequency "(The frequency of use of the word in speech statements, similar to the perceived). This provision is consistent with the concept that these three types of features are the main criteria for finding a word in long-term memory (81, 95, etc.). Based on this, in the process of speech perception, the recipient does not take a particular decision regarding one signal or one separate utterance, but chooses a general strategy for perceiving (identifying) the interlocutor's speech (67, 84, 91).

This interpretation of the speech perception process involves the use of the concept of pre-orientation, during which the recipient receives information about the situation of speech communication, the source of information, and chooses a semantic solution that organizes speech perception. The choice of a specific solution occurs at the next stage of perception and is largely determined by its first step.