Intentional Primitives for Communication

A solution to this problem can be found in a work by Airenti, Bara & Colombetti (1993). They acknowledge the crucial role played in communication by common knowledge, but view it as neither an unattainable ideal, nor as an objective fact to be indefatigably assessed. Rather, they define common knowledge as a primitive mental state type called shared belief: an agent shares the belief thatp with a partner if she believes both thatp and that the partner shares the belief thatp with her.

Shared belief, when conceived of as a subjective mental state type, turns out to be ontologi-cally on a par with (and therefore irreducible to) private beliefs. The relationship between the two mental state types is inferential. For instance, if I am aware that you and I have been copre-sent to a certain conspicuous event (call itp) like, say, Ann uttering "I think it's about to rain" I can legitimately infer that I share with you thatp. Reciprocally, if I share with you that q, then I can legitimately infer that you believe thatq, that you share thatq with me, that you believe that I believe thatq, and so on. On this account, the drawback in the classic formulation pointed at by Clark & Marshall (1981) is avoided, while the circular nature of mutuality is preserved.

The status of shared beliefs is analogous to that of private ones also from the epistemological point of view. Since sharedness is a matter of attribution, an agent has neither the necessity nor the possibility to know what is "objectively" shared with a partner. Sharedness does not require fancy abilities like telepathy or an endless circularity of confirmations; nor does it require any more reference to objective facts in the external world than ordinary beliefs do. The possibility that I takep to be shared with you, whereas you do not believe p or do not takep to be shared with me may give rise to various kinds of failures, misunderstandings, or deceptions, but cre­ates no more epistemological difficulties than standard private beliefs do. Therefore, Sperber & Wilson's (1986) objections against the concept of mutuality do not apply to this account.

The intention to communicate is subject to the same kind of analysis: for the reasons we have just viewed with respect to shared belief, it cannot be defined as a conjunction of private inten­tions. Airenti, Bara & Colombetti (1993) instead define communicative intention as another primitive mental state type: in particular, as an agent's intention to overtly make some of her mental states (including the communicative intention itself) shared with a partner.

Shared beliefs, private beliefs, and communicative intentions partake in a specific cognitive dynamic, dubbed the conversation game by the authors, that accounts for the process whereby an agent understands a partner's communicative act and generates an appropriate response. The conversation game consists in a sequence of defeasible inferences and metainferences leading from the agent's comprehension of the literal meaning of the utterance to his comprehension of its pragmatic meaning, to a consequent modification of his mental states about the domain of discourse, and finally to the planning and the generation of a response. The whole process is given a detailed formal account in a descriptive logic of mental states.

Technicalities aside, the key points made by Airenti, Bara & Colombetti (1993), from the viewpoint of our current discussion, are their arguments that communication requires a dedi­cated set of mental states (namely, shared belief and communicative intention) and that the mental events underlying it consist in a dedicated cognitive dynamic (that is, a specific, well-identifiable sequence of mental states).

Given the framework I have described in "Intentional Cognitive Architectures" the ability to communicate thus meets the requirements for being considered as competence, that is, as a specific cognitive subsystem, or faculty.3

To say that human communication is competence does not mean that it has to result from the

activity of a module, at least in Fodor's (1983) acceptance of what a module is. Modules are blind, rigid, mandatory, "informationally encapsulated" input systems in charge of providing central cognition with the raw material needed to form representations; the typical representa­tives are language and vision as they are conceived of in classic Chomskyan linguistics or in Marr's (1982) theory. As explicitly theorized by Fodor himself (1983), a modularview of input systems can peacefully coexist with a domain-general conception of "central" cognition.

Communication, however, is a flexible, inferential, deliberative activity that cannot be yielded by the functioning of a module: it has therefore be conceived of as the province of Intentional cognition. At the same time, it requires a dedicated competence, made up of special­ized Intentional primitives that find their exclusive role in it. (Actually, it would make no sense for an intelligent but nonsocial species to entertain these architectural characteristics.) Thus, the domain-specific view of the architecture of the mind/brain that I am proposing questions the existence of a general-purpose, equipotential central system, not of the modules.

Задание 1. Цель: определить уровень знания характерных особенностей предста­вителей мира животных (домашних и диких).

Оборудование: дидактическая игра для детей младшей группы «Угадай, кто где живет» (Приложение 2).

Детям предлагалось поиграть в игру «Угадай, кто где живет». Перед ребенком раскладывали две картинки: одна с изображением деревенского дома, вторая – с изображением леса и картинки с изображением разных домашних и диких животных. Ребенку нужно было «поселить» каждое животное в среду его обитания.

Оценка результатов деятельности.

Высокий уровень (3 балла). Дошкольник различает и называет при рассмотрении картинок домашних и диких животных. Выделяет характерные особенности строения животных, их внешнего облика. Имеет представление о диких и домашних животных (живут рядом с человеком, приносят пользу, живут в лесу, сами добывают пищу). Узнает основные признаки типичных представителей разных групп животных (у птиц – крылья, хвост, клюв, тело покрыто перьями; у животных – лапы, хвост, уши, тело покрыто шерстью; у рыб – туловище, хвост, плавники, тело покрыто чешуей).

Средний уровень (2 балла). Ребенок допускает незначительные ошибки при различении и назывании домашних и диких животных, их детенышей, птиц, рыб. Не всегда выделяет характерные особенности строения животных, их внешнего облика.

Называет не все признаки типичных представителей разных групп животных или называет их с помощью наводящих вопросов взрослого. Допускает неточности.

Низкий уровень (0-1 балл). Часто допускает ошибки при различении и назывании домашних и диких животных, затрудняется в выделении особенностей строения животных, их внешнего облика.

 

 

Mind/Brain Dynamics

An agent's mind at any slice of time can now be described as a set of mental states with their contents, like Ann is bored with staying at home or Bob believes that Ann believes that it is about to rain. Since an adaptive mind/brain has to keep tightly coupled to a world that is dy­namic, however, it is better described as whole sequences of mental states across time than as isolated, instantaneous states. Let us call such sequences cognitive dynamics (Tirassa 1997).

Once unfolded, the umbrella episode of "Mentalist Theories of Communication" reports two such dynamics that, taken together, provide the set of events that a theory of communication is expected to describe and explain. One dynamics is Ann's: she looks out of the window, notices that there are clouds in the sky, forms the expectations that it will rain and that Bob will get wet, decides to suggest to him to take an umbrella, and finally plans and executes a suitable communicative act. As for Bob, he is dressing and probably minding his own business, but immedi­ately shifts his attention toward Ann, reads her behavior as communicative, comprehends it, re­flects upon its meaning, makes up his mind in accordance, and finally decides to take an um­brella and maybe to thank Ann for her kindness. Clearly, these events are intimately tied to each other: each is caused by the one that immediately precedes and causes in its turn the one that will immediately follow. Where does this causality come from?

Unless one is willing to be a dualist and to accept the problems that then follow conscious­ness and Intentionality have to be conceived as material properties of an agent's functioning brain — hence talk of mind/brain, rather than mind (or brain) alone, throughout this article. Like all the material properties of a physical object, the state of an agent's mind/brain at any slice of time will therefore play a causal role in the state of her mind/brain at the following slice of time, together with cooccurring factors that may affect this functioning, like the activity of sensory receptors or of a drug.

An agent's cognitive dynamics across time thus results from the interaction of her mind/brain with the surrounding (mental, bodily, physical, and social) environment. The specific pattern of this interaction is rooted in turn in the phylogenetic and ontogenetic events that have shaped the agent's mind/brain.

Cognitive Architecture

We can now define the architecture of an agent's mind as the set of cognitive dynamics she may entertain; communication, as we will see in the next section, is one such dynamic.

It can clearly not be the goal of cognitive science to describe each and every particular mental state or cognitive dynamic an agent may entertain, that is, to list the whole set of her possible thoughts; nor can it be the goal of a theory in pragmatics to describe each and every possible instance of communicative interaction. That would correspond to conceiving of the goal of, say, linguistics as the list of all the possible sentences in all the possible languages, an idea that has been repeatedly proved absurd.

Something more general is needed. The solution, in pragmatics as in linguistics or in cogni­tive science at large, is to describe the system under study as the engine that is capable of gen­erating all and only the relevant mental phenomena, be they instances of communicative interac­tion, sentences, or whatever. As far as mentalist theories of pragmatics are concerned, this cor­responds to describing what I have called the Intentional primitives of a communicating mind. In this interpretation, it is the aim of such theories to provide an abstract (that is, not exten-sional, or conceptually dependent on specific examples) definition of communication as the type of cognitive dynamic that a certain agent (or species of agents) will entertain in a certain type of situation.

Each mentalist theory of pragmatics can therefore be distinguished according to what cogni­tive dynamic it defines as communication; that is, to precisely what Intentional primitives it takes to underlie communication and in what type of situation it takes those primitives to partici­pate. This will, by necessity, build on some specific view of human cognition: each theory can thus be located within the various debates on the nature and the architecture of the mind/brain. In this perspective, the issue that is most crucial to the relationships between pragmatics and cognitive science at large (and therefore the one that will be discussed here) is the view of com­munication as competence or performance.

Задание 3. Цель: определить уровень знания характерных особенностей

времен года.

Оборудование: Дидактическая игра «Времена года».

Проведение: Ребенку предлагались картинки с изображением времен года и предлагалось подобрать к ним соответствующие картинки (картинки, подходящие по содержанию к каждому времени года, на которых изображены явления живой природы, деятельность людей в определенное время года) (Приложение 3). Воспитатель по ходу выполнения задавал вопросы по каждому времени года.

Оценка результатов деятельности.

Высокий уровень (3 балла). Ребенок знает времена года, правильно называет их. Знает характерные признаки каждого времени года. Замечает и называет состояние погоды, явления природы (тепло, холодно, идет дождь, снег, светит солнце, дует ветер), их взаимосвязь (солнце – светло, тепло; нет солнца – пасмурно, и т.д.

Средний уровень (2 балла). Ребенок правильно называет времена года. В назывании явлений природы допускает незначительные ошибки. В основном знает характерные признаки каждого времени года, но иногда допускает незначительные ошибки.

Низкий уровень (1 балл). Ребенок не всегда правильно называет времена года. Затрудняется в определении явлений природы, состояния погоды. Не знает характерных признаков разных времен года.