The ultimate purpose of this study is to initiate a working computer system of having stories created and evaluated as literature. In order to have the system work well, first the classification has to be arranged of the essential elements of literature into the logical and the non-logical. Secondly, the language and its birth, and the relations between words, stories, and literature have to be examined again. So, in this paper, as the first stage of the reconsideration of them, the research is done as to the possibility of transferring literature (which is commonly considered to be made by writers) to readers, and the possibility of distinguishing clearly the elements of stories from those of literature. If the logical and the non-logical were combined very perfectly and mysteriously and they were unable to be classified, I could not refrain from expressing my warm admiration for the subtle combination of mathematics (the logical) and literature (the non-logical) that supports philosophy.
Represented by story grammar, the conventional analysis with a cognitive science approach regarded narratives as a process in which the character solved problems. This idea was invoked by folklore narratives and is insufficient for a display of narrative structures. With the idea that three conditions compose narratives in general—that are, 1) actual temporality, 2) flow of time, and 3) detachment from the narrator, this paper aimed to show that narratives are comprised of four layers: speech, foreground and background, and comment. As for the background, there are two types that exist: the microscopic type which magnifies the process part of motion, and the macroscopic type which compresses multiple events and presents them as a whole. Furthermore, to explain that these four layers are in figure and ground relationships with each other, and reflect these layers linearly onto discourse, recognition of presence of specific border expressions for each borderline is described.
Digital technologies have revolutionized textual practices. How does language on computer screen work differently from language on the page? We analyze the potential of new forms of text, new forms of rhetoric, and new approaches to literary theory.
A hackneyed analogy between a living thing and a narrative provided us with authentically scientific approach by Propp in the field of narratology. The aim of this article is to make a rough sketch of the present and future situation of numerical analysis in folktale studies which we have started from the medieval French fabliau. By using Tesnière's structural syntax model, we can describe a motif, the minimal unit for the analysis, as a tree diagram that represents the dependency relations between a predicate and the other terms. We decompose further this tree diagram into terms and sub-decompose the latter into semes in order to obtain data that represents simultaneously the syntagmatic and paradigmatic structure of a narrative at a deep semantic level. This data can be handled by various computational tools based on a few models of genome informatics. Methodologically speaking, post-genome informatics which is now trying, from the end of reduction to the constituents, a synthesis of them to elucidate the vital phenomena as a complex system will be the fundamental model for the study of narratology also including the automatic narrative generation system.
In this paper, we describe how the ordinary subjects in literatures such as understanding, emotion and common sense are formalized and represented in logic from the viewpoint of communication. First, we consider about performance of communication between 2 agents and make a criterion of success/failure of communication clear and give the meaning of “understanding” in literature. Second, we introduce a model of communication with multi-layered protocols analogous to the protocol stack of computer communication. Using the model, we analyze the performance of communication and illustrate each layer by failure cases of communication caused by some obstacle at the layer. Third, we analyze emotion in literature from logical viewpoints, then explain emergence and dissolve of emotion in a fable. Furthermore, we analyze logical function and role of emotion in understanding and accepting a short detective story. Finally, we show characteristics of common sense from both semantic and expressive viewpoints in literature. Based on the characteristics, we describe an experiment of common sense extraction from a sample short detective story.
This paper presents cognitive/computational models of rhetoric I have proposed: an affective-similarity-based model of metaphor comprehension and an implicit-display-theoretic model of irony interpretation. This paper also mentions poetic/aesthetic effects of metaphor and irony evoked through the interpretation process.
This study deals with just a emotion of emotions. When we see good movies or wonderful landscapes, read heart-warming stories, our hearts are frequently moved. This phenomenon evokes fairly certain emotions. However, it has hardly been studied in psychology. Many researches by the author were reviewed in order to investigate and elucidate mechanisms of evoking emotional responses of “Kandoh (the state of being emotionally moved)”. On the basis of these researches, first, the various types of “Kandoh” are categorized, and secondly a comprehensive process model is proposed. The outlines of this model are as follows. It is suggested that event related knowledge and information are very important factors for an audience (that is spectators, listeners, viewers, readers, and so on) to become highly involved, and then high involvement states elicit psychological and physiological stress. Therefore, desirable developments of a story produce stress reduction or relaxation, and finally people evoke “Kandoh”. Based on these results, some characteristic aspects of our cognitive system and some significance of emotional responses of “Kandoh” are discussed.
The present paper describes a cognitive and computational model of literature appreciation. Thinking and emotion processes involved in literary text processing are analyzed and a layered model of comprehension-appreciation is proposed. Based on psychological experiments, a computational algorithm for a story appreciation mechanism was constructed on a goal/plan based knowledge architecture. Our analysis of aesthetic emotions related to rhetoric aspects of literary texts suggests a cognitive (appraisal and action readiness) structure for aesthetic experiences. Partial implementations of aesthetic processing evoked by a passage from a poetry are described. Finally, the paper discusses an artificial brain approach aimed to unify the mind level theories and the brain level findings.
A consideration on the extreme north literary text has been done in this paper. Based on the mathematical proof and other theoretical deduction, here the incompleteness of the autonomous generation of the extreme north literary text has been shown. The extreme north literary text has much similarity to the production of it within a cognitive process, however, it is not in today's computation from the axiomatic definition of it. Moreover, general text generation within computation is far more difficult than the extreme north literary text. Thus, it is obvious that general text generation is impossible by the today's computation. As an alternation, the other computation based on circular structure and function has been proposed. Moreover, the alt-computation might be a breakthrough to this incompleteness issue. For constructing a complete language performance system is the final and the most integrative problem for A. I., this proposition might be not only the milestone of the language system study, but also of the fundamental A. I. study.
Differences in discourse type or genre produce various effects upon readers. Recent empirical research on the cognitive processes of discourse comprehension shows that mature human readers utilize cognitive processes called “comprehension control systems”. These develop in a specialized way for each type or genre of discourse. Comprehension control systems for natural language discourse have to be built-in if they are to be as flexible as those of adult human readers, so they must be designed and constructed as such. This is expected to enable the discourse comprehension system to utilize efficiently the fictional information in literary texts without confusing fictional and actual situations.
The rhetoric of the film means combination of film techniques based on a film director's purpose. Classification of the rhetoric of the film based on cognitive and computational approach is done from perspective of the director, the film itself, and the viewer. The classification can make clear the possibility of the film rhetoric. Then, film generation by computer technology based on the classification is proposed. We generate different types of the rhetoric from the same story according to a strategy for intensity of cognitive effects.
We research Geino organization as a concrete example of complex social phenomena, based on the model of multiple narrative structures. Geino organization consists of not only a Geino production organization which produces and develops Geino events, but also groups of receiver which receive and response to Geino events. At first, this paper shows the characteristics and a macro framework of Geino information system as a model of multiple narrative structures, secondary introduces the relations between Geino information system and social science research (marketing, business, etc.), and finally explains the possible social applications of Geino information system research.
In this paper, a novel interdisciplinary approach to literary phenomena is proposed. I call this approach “expanded literary theory”. The expanded literary theory is a framework for research and practice that intends to develop an integrated model of literary phenomena with cognition, text, and institution by the fusion of literary theories, narratology, cognitive/computational science, and social sciences such as management and marketing, and aims at the dynamical analysis of literature, the experimental creation and practice of new literary genres. Here, after I describe the concept and objectives of the expanded literary theory, I consider its methodology and some trials.
In this paper we addressed the problem how social interactions, especially P2P (Peer to Peer) communication, set trends. We made a multiagent-based model to investigate the issue. The agents are computer programs that act autonomously and behave individually. In this artificial society, the agents demand goods that they want to consume as well as they can produce and consume goods. They try to barter their goods with one of the other agents that is selected randomly. They make their own evaluation of each good based not on global information in the market but on local information shared with their trading partners: They assume the goods, which their trading partners demand, to be popular. They can also demand the goods that they think very popular by themselves and exchange their evaluation of the popularity when they trade with other agents. We pointed out that, in this situation, agents' evaluation of the popularity can be concentrated on some good one after another, which can explain a mechanism of the concentration of popularity.
This study consists of a simulation and an experiment. The purpose of the simulation was to confirm that the individual differences in performance found in spaced learning using the Low-First method were mainly due to individual differences in working-memory capacity. The Low-First method is a spaced learning method which was derived from a reactivation theory of spacing effects in order to produce the largest possible spacing effects. The aims of the experiment were to improve the Low-First method by making it adaptive to the individual differences in working-memory capacity, and to further prove the effectiveness of the modified Low-First method. In the simulation, the learning processes of the participants with the three highest and the three lowest probabilities of recall were simulated using the reactivation model, changing only the activity decay speed. It was found that the estimates of probability of recall closely approximated the test results, indicating that individual differences in working-memory capacity were the primary cause of the disparities in probability of recall. The Low-First method was then modified by incorporating a new principle defining the error limit of a learning session to make it adaptive to the individual differences. In the experiment, two groups of participants with large and small working-memory capacities studied using either the original Low-First method or the modified Low-First method. The results showed that the modified Low-First method enhanced the effectiveness, time-efficiency, and smoothness of learning, especially for the participants with smaller working-memory capacities.
In regard to the causal induction based on contingency information, the probabilistic contrast, or the difference between the probability of the effect (e) in the presence of the cause (c), and the probability of the effect in the absence of the cause have been regarded as the appropriate measure of perceived causality in many studies. This article proposes a new model of causal induction which is called the dual-factor heuristics (DH) model. The DH model is based on the two important factors of causal inference: P(e|c), the predictability of the effect, and P(c|e), the compatibility of the cause. In an experiment, participants made causal judgments from sequential presentations of information about occurrences and non-occurrences of an effect in the presence and absence of possible causes. Participants' judgments were better predicted by the DH model than models based on the probabilistic contrast. Effects of the probabilistic contrast, however, barely detected, and it was considered to be caused by mixed strategies of the participants. The results were discussed from the viewpoint of adaptive rationality.
Social networks are defined as the entity of various relations among people or groups. Social network analysis employs the view of relational structuralism. In this view, the behavior, thoughts, recognition, or mental health of people who are surrounded by social networks receives effects from the configurations of the networks or their positions in the networks. Social network analysis takes advantage of mathematical models to represent the theories of social positions or roles. Social network analysis and other theories are compared both theoretically and methodologically to clarify the bases of social network analysis.