Introduction

The paper under study has been written in line with the immanent-centric theory of language, the variants of which (structuralism, derivational grammar, linguosynergetics) have long been known in linguistics. Keeping the line of immanent-centric research, we further develop the idea that a language is arranged so and not otherwise, in order to maintain the situation of dialogue. A dialogocentric model (theory) of language, from our point of view, is one of the versions of immanent-centric hypotheses about the structure and functioning of the language.

In modern linguistics, the idea of a dialogocentric language is certainly recognized, but it is substantiated from the standpoint of anthropodynamic theories. For example, P. Linell (Linell, 2005) proposes a dialogocentric model of language as a variant of the contextual and interactional theory of language (see also Grice, 1985; Frank, 1986; Dijk van, 1984). The idea of dialogue as a principle of organizing semiotic systems, including language, is discussed, for instance, in (Lotman, 1984).

The interest to a dialogue as to a basic principle of language communication organization in the second half of the 20^th century correlated with the interest of researchers to the structure of a dialogue as a text type. Conversational Analysis became a new research area that had a certain contribution to the dialogue structure research. Particularly, many specialists on dialogue until today use the turn-taking system described in the classical paper of H. Sacks, E. A. Schegloff and G. Jefferson (Sacks, Schegloff and Jefferson, 1974) as a basis for creating their own models of dialogue structure. They in some ways transform, specify or supplement the understanding of turn-construction, turn-allocation components and rules of H. Sacks, E. A. Schegloff and G. Jefferson’s system (see for example the model of the mental processes in (Levinson and Torreira, 2015)).

The hypothesis of dialogic-centricity of a language assumes the actualization of the idea of language algorithmic nature. In a broad scientific context, this idea can be comprehended as a version of the Generative (Transformational) Grammar of N. Chomsky, defending the thesis of innate language ability and recursive rules for deriving statements in any language (Chomsky, 1957). The algorithmic nature of a language means that a language is a recoding algorithm possessing a prognostic orientation. In the field of computer programming artificial languages are viewed as the results of recoding (as algorithmic formations), we consider natural language to have the same algorithmic basis.

Recoding is a prognostic interpretive process that anticipates the perception of a language unit image by the addressee. The code transitions, and, in fact, recoding as an innate mechanism of language / speech (“speech language”) have been described by N. I. Zhinkin, considering code transitions as a process of translating the code of internal speech into an articulated code (Zhinkin, 1964). However, recoding is considered by the researcher as a manifestation of the explicative function, that is, the function of expression of the expressed (N. I. Zhinkin contrasted the communicative function and the explicative function). In the current research, recoding is interpreted as a mechanism of a dialogue, its very existence is conditional upon the adaptation of language to the situation of dialogue, namely the fact that the expressed will be perceived by the addressee, so the perceptual element (perception of the addressee) is genuinely included in the recoding algorithm. Dialogocentric orientation of the recoding inevitably assumes that this process is carried out with allowances made for the image of what will be said. In other words, the language ability of a speaker is arranged in such a way as to be able to identify and distinguish the internal form of what will be said. Otherwise, from our point of view, a speaker would be deprived of the ability to produce and perceive speech, as these language activities by some means involve a premonition of the complete unit image.

In the paper we will try to dwell upon the algorithms of text dynamics in a dialogue, illustrating the connection between the first and the next turn at the propositional level.

The aim of the paper and its methodological grounds

The purpose of the paper is to identify and describe the algorithms for recoding the next turn in a dialogue. The algorithm is considered as an intralinguistic process of prognostic interpretation of a language sign, which determines its speech manifestation in one form or another, under one or another mode. Text recoding algorithms determine the pattern of the next turn.

In order to clarify the research logic, we consider it necessary to outline the methodological grounds of the research.

A. A dialogic utterance, a turn, a dialogic unity

Bearing in mind that a dialogic utterance is always the next turn (Jakubinskij, 1987; Bakhtin, 1997), our further reasoning is based upon the differentiation of a dialogical utterance (text) and a dialogical unity. A dialogical utterance is a language unit represented in speech by a single utterance (a single turn) that implements a dialogical meaning (vide infra). A dialogical unity is a syntagmatic formation that includes two dialogical statements (two turns) (Mihajlov, 1984). In this case, one utterance is considered as an initial turn, and the other – as the next turn.

B. A dialogical utterance as the dialogical meaning actualizer

Considering the dialogical statement as a turn of the dialogical unity (dialogue), following M. M. Bakhtin, we believe that the dialogical statement realizes not only the referential meaning, but also the dialogical one. We assume that the dialogical meaning has a phatic referent, that is the Other as an image of a communication participant. According to the concept which we are developing, the dialogical meaning of a statement is related to modal predicates that correlate with receptive dialogical modalities – an agreement, a disagreement, and a neutral modal predicate. Based on the versatile nature of a dialogical meaning, we have identified some genesis models of a dialogical utterance: agreement model, contrast model and adjunction model.

C. The addressee in the dialogue

Usually, the addressee in the dialogue is the person to whom a dialogical turn is faced directly. This detail can probably be explained by the fact that for a long time linguists paid their attention to oral speech in a situation involving the visual presence of the addressee. The consequence of this comprehension was that specialists in pragmalinguistics started to distinguish between the addressee and the listener, who is present in the act of communication, but who is not the addressee of the message. The paper develops the idea according to which the addressee of the message can be any speaker who can potentially respond to the message. The way a question is put, in our opinion, is quite justified, because in written communication, especially in the Internet dialogue (network communication), the addressee of the message is not only the author of the turn, but also any other speaker who could potentially leave a comment. It is mainly because all comments are saved and available to anyone who decides to participate in the discussion under study.

When identifying algorithms of textual dynamics, we implement a so called semasiological approach, that is, we go in the direction from the statement to those procedures that determine its external pattern. However, the very description of the algorithms for creating the next turn in a dialogue is carried out from the standpoint of the onomasiological approach, which involves modeling of the addressant’s (speaker’s) speech activity. Combination of onomasiological and semasiological research logic corresponds to the general tendency of linguistic research dealing with the form of a language unit, but describing it as a result of some certain language-speech procedures.

The next turn in a dialogue: the problem of genesis

The issue of the next turn emergence in a dialogue, most probably, is actualized for the first time in the work of L. P. Jakubinskij “On Dialogical Speech” subsequent to the decision on the specifics of replication. According to L. P. Jakubinskij, the appearance of the next turn in a dialogue is a manifestation of the speaker’s dialogical ability, determining the manifestation of replication mechanism in the reply situation. As a consequence, replication can be viewed as a mechanism of language ability that determines the emergence of utterances in a dialogue (dialogical utterances).

Later, the idea of replication was developed in scientific papers that postulate the idea of turn-taking process in a dialogue and the conditions specifying such a process. In other words, the replication problem is comprehended as a problem of turn-taking. In this regard, linguists were faced with the task of explaining the appearance of the next turn in a dialogue. The explanation was provided as follows: in the part “given” there are two turns, in the part “it needs being explained” there is an answer to the question of how these two turns are connected.

The most widespread is the pragmatic theory of replication (the 80-90s of the 20th century), which explains the exchange of turns by means of communication participants’ intention to cooperate (Grice, 1985). According to this theory, the emergence of the next turn is a pragmatic process of illocutionary coordination in the stimulus-response model (Baranov and Krejdlin, 1992; Gerasimova, 1986; Paducheva, 1982), and it is a process anticipated by the addressant. The notion of illocutionary coercion becomes central in this approach and it generates the internal structure of the minimum dialogical unit and “connects speech acts in a real dialogue, and the latter therefore receive new content characteristics that reflect their functioning in natural language communication” (Baranov and Krejdlin, 1992: 85).

For example, it is assumed that a question can acquire the following versions of the text: an echo-question, a clarification-seeking question, an answer (a negation or an assertion); a request can obtain a consent, a refusal, a clarification-seeking question, and an echo-question etc. Pragmatic interpretation of a dialogue is discussed not only in the Speech Act Theory, but also in such theories as J. Verschueren’s Linguistic Adaptation Theory and H. Spencer-Oatey’s Rapport Management Theory (Verschueren, 1999; Spencer-Oatey, 2008), as well as in the framework of some other modern theories (Bragina and Sharonov, 2019; Wang and Yang, 2020; Kolmogorova, Kalinin and Malikova, 2021). It can be noticed that the pragmatic approach does not take into account the real activity of the addressee, which can hardly be described in the stimulus-response model.

In the 90s of the 20th century, there was developed a receptive theory of replication, presented in the works of N. D. Arutyunova and her disciples. N. D. Arutyunova, dwelling on the phenomenon of the next turn in a dialogue, concludes that its genuine emergence is the result of the addressee’s receptive activity, and therefore the next turn is related to dialogical modalities – either agreement or disagreement (Arutyunova, 1990). According to N. D. Arutyunova, the next turns create a favorable ground for the use of modal words (doubtfully, of course, etc.); <…> they indicate the area of secondary predicates of truthful assessment (an assertion or a negation) (Arutyunova, 1990: 177). However, even with this approach, a certain amount of turns, non-actualizing either agreement or disagreement, remains unaccounted for.

Based on the ideas of N. D. Arutyunova and M. M. Bakhtin, we believe that the emergence of the next turn in a dialogue is associated with receptive modalities of agreement, disagreement and neutral dialogical modality, which determine the phatic patterns of dialogical text actualization. At the same time, receptive modalities illustrate the peculiarities of dialogical comprehension (according to M. M. Bakhtin), acting as a model marker of creating the next turn, but their presence does not fully explain the variability of the textual form of the next turn in a dialogue.

The current research in some ways corresponds with the general features of turn-taking process in conversation, such as, for example, Local Management System and Interactionally Managed System in a classical paper of (Sacks, Schegloff and Jefferson, 1974). In the dialogue type “Internet news article + Comments” we can observe many same facts as to the oral conversation, for example, about not fixed turn order and turn size, not specified length of conversation etc. (see Sacks, Schegloff and Jefferson, 1974: 700-701). That confirms the fact about central position of conversation among the speech-exchange systems.

In comparison with some other modern researches on the topic of turn-taking process, which often deal with oral dialogues (conversations), our research takes an Internet news article and comments to it as a specific dialogical system. That is why such facts as pauses, gaps, overlaps etc., which are often analyzed in the papers about turn-taking in a dialogue, are not in focus in the current research. While other researches concentrate mainly on phonetic and syntactic components of turn-taking (see, for example, (Levinson and Torreira, 2015)), we analyze the propositional component, concentrating on semantic corresponding between turns.

This paper discusses the issue of variability of the textual form of the next turn in a dialogue. The form of the next turn is perceived as a way of its structuring (internal programming). It is important to stress that we do not set a task of describing the variation of the textual form, our efforts are specifically aimed at finding the reasons for explaining its variations. The paper actualizes the idea according to which the emergence of the response in a dialogue is the implementation of text encoding algorithms that determine the image of the next turn in a dialogue (taking into account the perception of the addressee). We have identified two algorithms for textual dynamics – an analytical and a synthetic one. The analytical algorithm determines a segmented image of the next turn, its syntagmatic distancing from the initial utterance, and the synthetic algorithm determines a segmented image of the next turn, the element of which is the initial utterance. In both cases, there is an inclusion of the image of the addressee (we include a semantic component “a perceptual image of the addressee” in the meaning of the term “the image of the addressee”) in the recoding algorithm. In the case of analytical algorithm, the image of the addressee is situational, i.e. it is implied that the addressee is included into the communicative situation. In the case of a synthetic algorithm the image of the addressee is extra-situational, i.e. it is assumed that the addressee is not inscribed in the communicative situation. We consider text algorithms to be mechanisms for the internal binding of utterances in a dialogue (Glekner, 1980). Text recoding algorithms, in our opinion, perform a programming function at the replication stage, where replication is a component of a speech generation model in a dialogue.

Research material

To identify the algorithms of text dynamics in a dialogue, we turned to the analysis of a social network dialogue presented by the news article and comments to it in the amount of 258 units[1]. There are various reasons for choosing a virtual dialogue as a research material, including the fact that it is a variant of [oral / written] conversational speech, in which a dialogue is presented as a natural form of speech (Shcherba, 1915; Jakubinskij, 1987). P. Linell, dwelling on modern linguistic theories, characterizes linguistics as a manifestation of written-linguistic bias (Linell, 2005), assuming that most theories are based on written (monologue) speech and not on oral (we would add “written” as well) interactive language, i.e. a dialogue. In this context, the Internet dialogue lets us identify natural algorithms of textual dynamics in a dialogue, which, in all probability, can be considered as a manifestation of innate language ability.

Considering the text of the news article as an initial utterance, we believe that its content can be presented in the form of a proposition: Импортозамещениеевропейскихсыровпривелоктому, чтоценазакилограммроссийскогосырасталанамногодороже.Между тем, чтобы просто научиться делать приличные аналоги европейских сыров в условиях эмбарго, и производством конкурентного на мировом рынке продукта, лежит огромная дистанция. Вряд ли сыр, сделанный в России, можно будет продать по цене 40 евро за килограмм где-то кроме России[2]. (The result of import substitution for European cheese is a higher price for one kilogram of Russian cheese. There is a great difference between simply learning how to make presentable analogues of European cheese under the conditions of embargo and producing a globally competitive product. It is doubtful that Russian cheese can be sold at 40 euro per kilogram anywhere except Russia.)

Research results and discussion

On the basis of empirical material, two types of algorithms of textual dynamics in dialogue are distinguished: analytical and synthetic. The analytical algorithm of text dynamics is manifested in the actualization of the next turn, which is implemented in a segmented way, i.e. outside the initial utterance. As well as the synthetic algorithm of text dynamics is manifested in actualization of the next turn that is not segmented with an initial utterance.

A. Analytical algorithms of text dynamics in a dialogue

Analytical algorithms of text dynamics fractionalize the dialogical meaning of dialogical unity into the meanings realized in the initial and next turns. They are differentiated depending on the form of the initial utterance, which can be either monological or polylogical. The monological form of the initial utterance is observed if the response is addressed to only one initial utterance. The polylogical form of the initial utterance is observed if the response is addressed to several initial utterances at the same time. The algorithm of response to the monological utterance is divergence (see Figure 1), the algorithm of response to the polylogical utterance is neutralization (see Figure 2).

Figure 1. Realization of divergence algorithm

Рисунок 1. Реализация алгоритма дивергенции

Figure 2. Realization of neutralization algorithm

Рисунок2. Реализация алгоритма нейтрализации

It should be noted that the analytical algorithms of text dynamics include a situational image of the addressee, i.e. the addressee understands the context.

Divergence as an algorithm of text dynamics in a dialogue

Divergence is the algorithm of text dynamics, in which the dialogical meaning of the initial utterance is split into individual statements that have their own dialogical meaning.

The divergence of the analyzed initial utterance is carried out in accordance with different models: contrast model, agreement model and adjunction model. There are some examples below.

Contrast model as an implementation of the divergence algorithm

Actualization of the contrast model means the realization of the dialogical meaning of disagreement / disapproval with the subject (agent) of the initial utterance.

По-моему автор переборщил со своим гурманными пристрастиями
Уже давно кушаю местные сыры не дороже 800-900 руб/кг и они меня вполне устраивают! (I think the author is really an avid foodie. I have been eating local cheeses at 800-900 rubles per kilo and I am quite satisfied!)[3]

The example shows that the speaker does not agree with the author’s opinion about Russian cheeses (the initial utterance), he believes that Russian cheeses are not behind foreign cheeses in flavor.

Agreement model as an implementation of the divergence algorithm

Actualization of the agreement model means the realization of the dialogical meaning of agreement with the subject (agent) of the initial utterance.

Ну в общем понятно импортозаместилось, только вышло не дешевле, а дороже в полтора два раза т.к. нет конкуренции и производств, всё развалено а создавать с нуля о-о-0-чень дорого! (Okey, we have the example of import substitution, but as a result the product is not cheaper, it is half or even twice as expensive as the previous variants. It is all because there is no competition in this sphere, everything has broken up, and to create something new is incredibly expensive.)

As we can see, the speaker approves the author’s position: Russian cheeses production under the conditions of breakup of course negatively affects the quality of the product.

Adjunction model as an implementation of the divergence algorithm

Actualization of the adjunction model means the realization of a neutral dialogical meaning, i.e. a speaker neither agrees nor disagrees with the initial utterance, his/her opinion may be ambiguous (arguable).

Завышенные цены в отсутствии конкуренции. Рашка же. Впрочем, ничего нового. (High prices and lack of competition. That’s Russia. In fact there is nothing new about this.)

The paragraph above is the example of the speaker’s neutral attitude to the author’s opinion. The speaker looks at the situation in the context of the events that are typical for Russia, as a result we can see a neutral reaction in the speech: Впрочем, ничегонового (In fact there is nothing new about this.).

Neutralization as an algorithm of text dynamics

Neutralization is the algorithm of text dynamics, the result of which is the combination of dialogical meanings of several initial utterances in the next turn, so the dialogical meaning of the next turn is addressed to all initial utterances. In some previous works we described the effect of neutralization as a derivational mechanism that determines the convergence of utterances in a dialogue. Rethinking of the neutralization nature lets us illustrate its manifestation as an algorithm of textual dynamics in a dialogue.

There are different implementation models of neutralization algorithm, but in the framework of this paper we are going to consider as an example the model where initial utterances, created by the agreement model, are neutralized in the next turn, created by the contrast model.

Откуда знать хуже или лучше? Или у людей такая память на вкусы?.... Дайте весь импортный ассортимент в продажу, сравним. Атакэтоблаблаблаполучается) (How do they know if Russian cheeses are better or worse? Is their memory about flavors so good?.. Give us all assortment of imported cheeses for sale, we will compare first. What I see now is just blah-blah-blah))

This turn is a reply to a series of some previous turns that evaluate the import substitution in Russia in a negative way. The author of this turn disagree with the comments of some speakers, because there are no enough reasons in them. According to the author of this turn, it is possible to make a conclusion about quality of Russian cheeses in comparison with imported ones only when we are familiar with “all assortment of imported cheeses”.

Another example:

Столько разговоров, столько претензий. Делайте сами тогда сыр, в чем проблема? Такие все знатоки сыра прям. Не ну если брать сыр за 500 р кг, то конечно наверное в сравнении сыр за 1000 р будет отличаться. Так не берите дешевый сыр. И как всегда у них там и сыры вкуснее, и трава зеленее. Я к сыру равнодушный человек, но бывала в Европе и пробовала там сыр, плюс привозили в качестве презента друзья. Ну сыр, да сыр. Че все так пищат с этого сыра. Вот хамон понравился, да. Хотя наши мясные деликатесы тоже есть очень даже. (So many talks, so many claims. Make cheese by yourself then, what’s the problem? I doubt that all of you are cheese experts. Okey, if you buy cheese at 500 rubles per kilo, of course you will feel the difference in comparison with the cheese at 1000 rubles per kilo. So just do not buy cheap cheese. And as usual “their cheese is tastier, their grass is greener”. I am indifferent to cheese. But I went to Europe and ate their cheese, and my friends also brought me cheese from there as a present. So it’s just cheese. Why is everybody so mad about this cheese? As for me, I like their jamon more. But at the same time some Russian meat delicacies are also quite good.)

This turn is a reply to those participants of discussion who criticize Russian cheeses. The author of the reply does not take this kind of criticism seriously: Такиевсезнатокисырапрям. (I doubt that all of you are cheese experts.).

A complete algorithm for the implementation of various models of neutralization is schematically represented in Figure 3:

Figure 3. A complete algorithm scheme of various neutralization models implementation

Рисунок3. Схематичное представление полного алгоритма реализации различных моделей нейтрализации

B. Synthetic algorithms of text dynamics in a dialogue

The implementation of a dialogical statement can be associated with the actualization of synthetic algorithms of textual dynamics, which are manifested in cases when the next turn in a dialogue includes the original utterance (see Fig. 4). In other words, the dialogical meaning of the next turn includes the dialogical meaning of the initial utterance. Citation and retelling can be distinguished here as algorithms for the interaction between two turns.

Under implementing these algorithms, a speaker takes into account that the addressee is out of the situation, i.e. may be unfamiliar with the topic of the discussion.

Figure 4. Synthetic algorithm realization of text dynamics in a dialogue

Рисунок 4. Синтетический алгоритм реализации динамики текста в диалоге

Citation as an algorithm of text dynamics in a dialogue

Citation is an algorithm of text dynamics, the implementation of which assumes that the genesis of the next turn in a dialogue is carried out by actualization of its dialogical meaning based on the inclusion of the initial utterance into the next turn. This algorithm is closely related to the phenomenon of dialogical citation, which was described in detail by N. D. Arutyunova (Arutyunova, 1986).

Below is an example of the implementation of citation as a text algorithm actualized in the dialogue.

Initialutterance (thecommentofaspeaker):Основной итог импортозамещения - это ощутимое повышение цен. Самый простой сыр Российский подорожал за 8 лет вдвое. Тот же Российский брендованный - ещё больше. Обычный сыр по цене стал деликатесом. Деликатесныежесырысталидлямногихпоцененедоступны (The main result of import substitution is a perceptible price increase. The price of the most common Russian cheese has increased twofold in the last 8 years. The price of Russian branded cheese has increased even more. An ordinary cheese has become a delicacy while specialty cheeses’ price has become so high so it is impossible for many people to buy them.).

Response utterances (reactions to the comment):

1. Не надо указывать 8 лет, нужно говорить за последнее почти десятилетие. (To say “in the last 8 years” is not enough, it should be said “almost in the past decade”.)
2. Какое вдвое? В 2014г российский сыр стоил около 120р сейчас 650!!! (Is it really “twofold”? In 2014 the price of Russian cheese was about 120 rubles, now it is 650!!!)

As we can see, the speakers use some fragments of the initial utterance in their responses: 8 лет (in the last 8 years), вдвое (twofold). That is to say that the dialogical meaning realized by the response utterances is based on the evaluation of the initial utterance’s dictum.

Retelling as an algorithm of text dynamics in a dialogue

Retelling is a synthetic algorithm of text dynamics, the implementation of which is associated with the fact that under creating a next turn in a dialogue one or another dialogical meaning is actualized. Realization of this algorithm means that the initial utterance is not included into the response utterance in the original form, but “dissolves” in the next turn as a manifestation of a generalized or personalized “voice” (a semantic position).

The following turn is the example of the response utterance constructed by using the algorithm of retelling:

Нгс подает статьи с такой подоплекой мол жили мы хорошо, но наше бесталковое правительсво сделало все, чтобы мы жили плохо.
Поколение неумех блогеров, журналистов, менеджеров, экономистов, филологов, историков, юристов ..... ни чего не умея делать руками только и могут критиковать (Ngspublishesarticleswiththemainideathatourlifewasgoodinthepast, butourstupidgovernmenthasdoneeverythingtomakeitbadnow. The current generation is the generation of bloggers, journalists, managers, economists, philologists, historians, lawyers … They have no manual skills, but are masters of criticism.).

This turn has been created using the algorithm of retelling. The speaker seems to retell the content of a series of articles that are criticizing the government: Нгсподаетстатьистакойподоплекоймолжилимыхорошо, нонашебесталковоеправительсвосделаловсе, чтобымыжилиплохо (Ngs publishes articles with the main idea that our life was good in the past, but our stupid government has done everything to make it bad now.).

Conclusion

The algorithmic nature of a language is manifested in the acts of recoding as a process of prognostic interpretation and is adapted to serve the dialogue. Recoding at the level of dialogical unity appears in the recoding of the initial utterance and its dialogical meaning into the form of a response, which is carried out in accordance with different models: agreement, contrast or adjunction models.

The paper highlights two algorithms of text dynamics, manifested in the dialogue: analytical and synthetic. The analytical algorithm determines the segmented image of the next turn, its syntagmatic distancing from the initial utterance. The synthetic algorithm, in contrast, determines the non-segmented image of the next turn, the element of which is the initial utterance. In the first case, the image of the addressee is constitutive, i.e. it is assumed that the addressee is inscribed in a communicative situation. In the second case the image of the addressee is extra-situational, i.e. it is assumed that the addressee is not inscribed in the communicative situation (he/she is not so familiar with the topic under discussion).

We can see that key algorithms of text dynamics correlate with two typological ways of expressing grammatical meaning – analytical and synthetic. The language has analytical and synthetic algorithms, the manifestations of which can be observed on different language levels – not only on the grammatical (syntactic) level but also on the levels of text and discourse.

The attained results clarify the ideas about actualization patterns of a dialogical utterance and the specifics of its genesis.