INFERENCE AS A “LIVE” PEG OF COMPREHENSION (PSYCHOLINGUISTIC APPROACH)

Голубева О.В.

Кандидат филологических наук, докторант, Тверской государственный университет

ВЫВОДНОЕ ЗНАНИЕ КАК «ЖИВАЯ» ОПОРА ПОНИМАНИЯ (ПСИХОЛИНГВИСТИЧЕСКИЙ ПОДХОД)

Аннотация

Статья посвящена проблеме выводного знания, признаваемого «живой» причинной имплицитной опорой естественно протекающего процесса понимания значения слова. Психолингвистическая структура значения предполагает учёт продуктов перцептивно-когнитивно-эмоционально-оценочного опыта индивида, вербального и невербального, представленных как свидетельства объяснения «для себя» смысла поступающей по различным каналам информации

Ключевые слова: выводное знание, эвиденциально-смысловое переживание значения, «живое» знание.

Golubeva O.V.

PhD in Philology, doctoral candidate, Tver State University

INFERENCE AS A “LIVE” PEG OF COMPREHENSION (PSYCHOLINGUISTIC APPROACH)

Abstract

The paper covers research of the inference phenomenon, which is defined as a “live” casual peg in natural processes of meaning comprehension. Psycholinguistic structure of a word meaning includes a number of productions of perceptive-cognitive-evaluative-emotional individual experience, verbal / nonverbal, defined as evidences of meaningful explanation “for oneself” of the input information getting through various channels.

Keywords: inference, evidential sense-formation, “live” knowledge.

The important cross-cutting issues of modern science are the studies of comprehension process, as well as its outputs, verbalized or remaining inner knowledge. The object of research is the inference defined as a cognitive process based on prior knowledge / evidences (state, event, action, or behavior). As a result, human reasoning capabilities are considered to peg on causality as a universal phenomenon leading to causation networks construction and allowing machines “to mimic complex human reasoning mechanisms” [4: 50–51].

Some approaches in cognitive computing and AI, psychology, neuroscience, linguistics work out models of inferential human information processing: mental models (P. Johnson-Laird [3]); situation models (T. van Dijk, W. Kintsch); neuron mental models based on if-then relations (P. Thagard); experiential simulations of the situation (R. Zwaan); perception-based generalization (ground cognition theories by L. Barsalou [1]), etc. Different approaches to modelling highlight the problem of meaning formation, how results of human experience are incorporated into the structure of a word meaning. They are considered to be implicit semantic or pragmatic component (presupposition or implicature), which causes a default meaning (S. Levinson’s defaultism theory, etc.).

Meaning, including such productions, is a dynamic structure or a semantic-associative network, functionally asymmetric due to a number of cognitive profiles for lexical concepts (the theory of lexical concepts and cognitive models by V. Evans, a dynamic model of meaning by I. Kecskes, etc). The idea of perceptive roots of category formation through simulations of a referent functioning allow to suggest a prototype (the theory of prototypes by E. Rosch, G. Lakoff). It is implicitly included in the content a word means as an amodal idea of the best referential sample. But due to psychological findings what a word means can be regarded as a multimodal representation of an object based on mental imagery (A. Paivio DCT conceptual peg hypothesis; mental imagery theory by S. Kosslyn; ground cognition theories by L. Barsalou, etc).

Various theories generated in different branches of science require integration, and Russian psycholinguistics can be one of the ways of a complex “vision” of meaning processing phenomenon. Due to L. Vygotsky, N.I. Zhinkin, A.A. Leontyev, A.A. Zalevskaya meaning is a production of social and individual experience gained in a dynamic process of mutual activities. As a dynamic structure (do not mix up with an entry in a dictionary!) it is characterized by dual nature: it is shared by all members of society due to generalized core (common knowledge) and individual “visions” of what a word means pegged on person’s evidences. A verbal input has a “live” meaning which “emerges only in case when a person finds some grounds in his / her social and personal, nonverbal and verbal experience” [5: 153] (the interfacial theory of word meaning by A.A. Zalevskaya).

So psycholinguistic approach to meaning as a “live”, dynamic structure based on individual’s experience allows to map it as a number of evidences or implicit pegs activated by a word, relevant ad hoc and revealing causal connections between the components of the structure. Some of them are in greater use and become more relevant and stable (default inferences). These evidences, marked in the psycholinguistic structure of meaning as associates, must be truthful, for an individual can explain to him / herself why they are relevant on-line to judge what a word means.

Experimental findings assert that individuals highlight not only generic attributes which are necessary to classify a real / virtual object but depict it by means of perceptive and emotional-evaluative features revealing a certain facet of person’s “vision” of it. For example, during free associative experiments 1183 respondents were asked to write down the words that help them explain the meaning the following terms (the stimuli) investment (инвестиция), mutual fund (взаимный фонд), crisis (кризис), loan (кредит), security (ценная бумага), profit (прибыль). 1241 reactions were analyzed [2] and the outputs provide some conclusions:

1) while explaining to “oneself” individuals refer to their object-oriented experience (e.g. loan – машина, пылесос, иду в банк, кредитная карта (car, vacuum cleaner, go to bank, credit card), etc.; generalized knowledge (investment – капитал, финансы, бизнес (capital, finance, business), etc.; emotional-evaluative attitude (mutual fund – помощь, доверие, любовь, что-то хорошее (assistance, trust, love, something good), etc. Most of respondents (72%) assign evaluative attributes (positive for profit, mutual fund, investment and negative for loan, crisis) to a potential referent forming a certain facet of inner “vision” of the word meaning;

2) verbal responses (associates) refer to certain inferential contexts or contextual conditions of a personal “vision” of a relevant situation or event. For example, respondents were asked to complete the sentence Если компания хочет получить прибыль … (If a company wants to get profit it …). The majority of answers (64%) show that respondents have a causal scheme to suggest a process of context assignment: “cause (premises) – ways to achieve effect”. They interpret the company is only going to get profit and were ready to offer their “recipes” for achieving the result;

3) the facet of inner imagery “vision” of an object’s main features and a potential inner context of self-explanation (a scheme) are justified by a “history” of getting experience of processing various knowledge of an object as a fragment of the world (physical and social) keeping in mind the ways of verbal outputs of this process.

The above-mentioned allows to hypothesize that evidences (inferential productions) as “live” pegs in comprehension process can be defined as inner sense-formation (hereinafter – ISF). It is a procedure of activating person’s internal potential, including a facet of an object “vision” implying contextual samples of satisfaction. The procedure of integration of external (verbalized, situational) and internal premises is based on establishing inferential / causal links, or models of feature assignment in accord with subjective prior experience and assessment of its relevance. The findings on inference-influenced meaning suppose the idea of a functional “cloud service” implied by a word which stores and provides subjectively true explanations justified by person’s experience. The ISF hypothesis adopts a number of insights from earlier theories of comprehension, specifically developed in psychology, linguistics, psycholinguistics, logics, AI providing an overlap with well-known approaches to inference. The goal of it is to contribute to currently existing ground theory of word / text comprehension providing a framework that coherently accounts for verbal and inner premises integration.

References

1. Barsalou L.W. Simulation, situated conceptualization, and prediction // Philosophical Transactions of the Royal Society B: Biological Sciences. – 2009. – Vol. 364. – P. 1281–1289.
2. Golubeva O.V. Basic provisions of individual evidential sense-formation theory // Vestnik TvGU. – 2015. – Vol. 4. – Р. 7–21.
3. Johnson-Laird P.N., Lotstein М., Byrne R.M.J. The consistency of disjunctive assertions // Memory & cognition. – 2012. – Vol. 40. – № 5. – Р. 769–778.
4. Wang Y. On cognitive Models of causal inferences and causation networks // International Journal of Software Science and Computational Intelligence. – 2011. – Vol. 3(1). – P. 50–60.
5. Zalevskaya A.A. Interfacial theory of word meaning: a psycholinguistic approach. – London: IASHE, 2014. – 179 p.