Epistemology of Ordinary Knowledge [1 ed.] 9781443886277, 9781443880527

Citation preview

Epistemology of Ordinary Knowledge

Epistemology of Ordinary Knowledge Edited by

Mariano L. Bianca and Paolo Piccari

Epistemology of Ordinary Knowledge Edited by Mariano L. Bianca and Paolo Piccari This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Mariano L. Bianca, Paolo Piccari and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-8052-3 ISBN (13): 978-1-4438-8052-7

TABLE OF CONTENTS

Foreword .................................................................................................... ix Part I: Ordinary Knowledge Chapter One ................................................................................................. 3 The Epistemological Structure of Ordinary Knowledge Mariano L. Bianca Chapter Two .............................................................................................. 39 Knowing (Existing) Ordinary Objects Massimo Dell’Utri Chapter Three ............................................................................................ 49 Neoidealism and Neorealism Maurizio Ferraris Chapter Four .............................................................................................. 59 Lebenswelt, The World of Objective Knowledge and Practice: Towards a Phenomenological-Einsteinian Gaze Fabio Minazzi Part II: Common Sense Chapter Five .............................................................................................. 77 From Common Sense Knowledge to Scientific Knowledge (and vice versa) Evandro Agazzi Chapter Six ................................................................................................ 95 On the Differences and Relations between Ordinary Knowledge and Common Sense Paolo Piccari

vi

Table of Contents

Chapter Seven.......................................................................................... 117 The Holistic Theory of Truth and the Epistemic Primacy of Common Sense among All Kinds of Ordinary Knowledge Antonio Livi Chapter Eight ........................................................................................... 127 Common Sense and Truth Conditions Gaetano Piccolo Chapter Nine............................................................................................ 141 A Non-Superficial Conception of Common Sense Lourdes Velázquez Part III: Empirical and Perceptual Concept Formation Chapter Ten ............................................................................................. 147 Perceptual Concepts in Ordinary Knowledge Mariano L. Bianca and Paolo Piccari Chapter Eleven ........................................................................................ 163 Common Sense Concepts between Compositionality and Typicality Effects: A Dual Process Approach Marcello Frixione and Antonio Lieto Chapter Twelve ....................................................................................... 181 An Embodied and Grounded Perspective on Concepts Anna M. Borghi Chapter Thirteen ...................................................................................... 195 On the Social Dimension of Beliefs and Concepts Maria Cristina Amoretti Part IV: Ordinary Reasoning and Historical Knowledge Chapter Fourteen ..................................................................................... 211 Subtleties of Naïve Reasoning: Probability, Confirmation, and Verisimilitude in the Linda Paradox Gustavo Cevolani and Vincenzo Crupi

Epistemology of Ordinary Knowledge

vii

Chapter Fifteen ........................................................................................ 231 The Ordinariness of Reasoning in Arguing Paolo Labinaz Chapter Sixteen ....................................................................................... 247 Regimenting the Ordinary Notions of Knowledge and Justification after Gettier Mario Alai Chapter Seventeen ................................................................................... 263 Memory, History and Ordinary Knowledge Pier Luigi Lecis Chapter Eighteen ..................................................................................... 289 Grasp of Concepts: Common Sense and Expertise in an Inferentialist Framework Pietro Salis Chapter Nineteen ..................................................................................... 299 The Epistemology of Testimony: A Hermeneutical Key Vinicio Busacchi List of Contributors ................................................................................. 307

FOREWORD

Ordinary knowledge has not been a relevant topic of epistemology and only a few essays underline the complexity of such knowledge. The fundamental sensory aspect of ordinary knowledge has led many philosophers to reduce this knowledge to the sensory or, more generally, to perceptual knowledge which refers to entities of the phenomenic world. In fact, ordinary knowledge is not only the result of sensory-perceptual processes, but also of non-perceptual (noetic) contents that are present in any mind. From the epistemological point of view, ordinary knowledge is a form of knowledge that not only allows one to epistemically access the world, but also enables the formulation of models of it, with different degrees of reliability. Usually epistemologists focus their attention on scientific knowledge, believing that ordinary knowledge does not or cannot have an epistemology for it is not in any way rigorous. They claim that ordinary knowledge is biased by subjective perceptions and beliefs, so, it cannot provide a reliable and verifiable account of the phenomenic world. Indeed, human ordinary knowledge, as well as the biological knowledge of other living species, is formulated with specific processes and it is reliable, at least to a certain degree, because life is based on it: each living being needs reliable knowledge of the world in order to survive in its habitat. In humans, ordinary knowledge is far more complex than the biological knowledge of other living beings, because it is not only useful to survive, but also necessary to fulfill the many other needs and desires of any human being; so, ordinary knowledge in humans involves many mental, psychological and social factors and is reliable with reference to the status of the world. The papers collected in this volume analyse different aspects of ordinary knowledge and of its epistemology. Each chapter focuses on a certain significant issue concerning this kind of knowledge. The first chapter focuses on ordinary knowledge from an epistemological point of view, particularly with regard to perceptual/ phenomenic knowledge and self-knowledge, analysing the different mental processes involved and their degrees of epistemological reliability and adequacy. In particular, Bianca examines the complex structure of ordinary knowledge, which he calls the “polignosic structure of ordinary knowledge”, subdivided into different types of knowledge: a) perceptual/

x

Foreword

phenomenic knowledge, b) self-knowledge of our own mental states, contents and processes, c) knowledge of our own selves and the formation of the self-concept, d) knowledge of other minds and other selves, and e) indirect and non-perceptual knowledge of the world. For each kind of ordinary knowledge, Bianca points out how it is formed, its relevancy and its degree of reliability. Another topic in this section is considered by Dell’Utri, who states that acquaintance with ordinary objects is a genuine case of knowledge: being legitimate members of our ontology, ordinary objects can be involved in our practices of knowledge acquisition. If a minimal ontological criterion is adopted – roughly, ‘To be is to be in a causal relation with humans’, provided that the causal relation is not interpreted as a non-Humean relation somehow embedded in the physical structure of reality – then ordinary objects acquire the full right to be part of our ontology. Therefore, according to Dell’Utri, contrary to what some philosophers think, the ‘scientific image’ has no epistemological pre-eminence over the ‘manifest image’: the two images are to be taken on par, adopting a nonreductionist point of view. Dell’Utri also argues in favour of the related thesis according to which knowing ordinary objects can be accounted for on the basis of the classical paradigm for direct and propositional knowledge. Insofar as the latter is concerned, a subject S can be said to know that the table in front of her is yellow if she has a justification for the truth of the belief that the table is yellow. Since in cases like this justification is mainly perceptual, it can be argued that perception is mostly veridical – and sceptical doubts pointless. In his contribution, Ferraris defines Idealism, the philosophical thesis that the external world exists depending on some representation, stating that this argument can be articulated in many ways, which he examines in detail. In fact, these are all variations of transcendentalism and its constitutive misunderstanding, i.e. the so-called ‘transcendental fallacy’: the confusion between ontology (what there is) and epistemology (what we know, or think we know). Ferraris counters Idealism with New Realism, which simply assumes that the subject observes in reality something that is different and independent from him, otherwise what she is doing is not observation, but introspection. According to him, it is banal to observe that knowledge of reality is the result of a constructive process. Instead, it is trivially false to say that reality is the result of a constructive process. Therefore, the new realist thesis sounds like this: ‘knowledge is knowledge of something different and independent of knowledge, otherwise it is not knowledge.’

Epistemology of Ordinary Knowledge

xi

Finally, Minazzi starts his paper with Einstein’s image of scientific knowledge to point out how it can be built by relating two different and antithetical dimensions: that of the world of praxis and that of mathematical abstraction or, in the words of Galileo, the plan of the sensible experiences and that of certain demonstrations. However, Minazzi shows how the plan of sensible experiences constitutes, in turn, a reality that is separated from the plane of the experience of common sense. Using analysis conducted by authors such as Edmund Husserl, Antonio Gramsci, John Dewey and Remo Cantoni, Minazzi shows well how the plan of common sense appears to be a complex plane. In particular, Minazzi argues that this plan of common sense is born and structured in a dynamic relationship with the history of human knowledge and also with the history of criticism of common sense. The phenomenological approach to studying common sense allows Minazzi to show how it is possible to reformulate the notion of objective knowledge - as their common sense as well as the most sophisticated science, resorting to Husserl’s regional ontologies and bearing in mind the role of heuristic Kantian transcendentalism. The second part is devoted to common sense and its relations with ordinary knowledge and scientific knowledge. Agazzi underlines that it is common today to maintain that scientific knowledge and ordinary knowledge are essentially in contrast, though a certain harmony between them was possible in pre-modern times. A historical reconstruction of the growth of modern natural science seems to support this view but, at the same time, cannot adequately explain why such a deep change occurred with contemporary science. According to Agazzi, the reasons can be uncovered by considering the epistemological choices that promoted the birth of modern natural science as early as Galileo’s time, and which consisted of a restriction of investigation to a very limited set of attributes of reality. This determined the specialisation of the general metaphysical principles that were and are included in common sense, from which natural science had to take its start. For a couple of centuries, this contact with common sense was helped by the fact that scientific concepts and models were ‘idealisations’ of ordinary knowledge, but this easy situation could no longer last when contemporary science clearly became a science of the unobservable. In the present situation, common sense necessarily remains the framework in which all sciences find their start. Then they elaborate complex systems of concepts and statements that are very far from the grasp of common sense. Yet all empirical sciences retain the obligation of explicitly linking their theoretical constructions with empirically ascertainable experimental results that must be understandable within the conceptual framework of

xii

Foreword

common sense, which in such a way remains the final appeal even for scientific knowledge. The relationship between ordinary and scientific knowledge is also taken into consideration by Livi, from the point of view of alethic logic. He states that, as the result of a rigorous phenomenology of consciousness, it could be shown that in every thinking subject there are existential certainties whose epistemic justification is founded on the perception of that which necessarily and always presents itself in everyone’s experience as something evident. Such certainties constitute the very first link in the chain of presuppositions; for this reason, they can in no way be subject to doubt. On the contrary, their non-truth is absolutely unthinkable: no one can ever really doubt them, and one must understand that any affirmations to the contrary are merely verbal posturing. Actually, they respond to pragmatic logic, and not expressions of a real certainty, endowed with their own adequate epistemic justification. According to Livi, given that they constitute the nucleus of experience, understood as a body of unmediated knowledge, such certainties are present to consciousness in every moment of the search for truth as the logical presupposition of all knowledge deriving from reflection and inference, both inductive and deductive. For this same reason, such certainties function as an ultimate criterion of truth in verifying any hypothesis successively formulated. They therefore constitute the main alethic presupposition, that is, the presupposition necessary for any ulterior knowledge to be thought of as true. In fact, on the basis of these original truths, every thinking subject verifies, time after time, the admissibility of any hypothesis that presents itself in the search for other truths over the course of his lifetime. As a result, all scientific knowledge should be structured as a system logically compatible with the primary truth of common sense, so as to place the instruments of dialectics (reflection, interpretation, inference) effectively at the service of the search for further truths. The differences and the relationships between common sense and ordinary knowledge are focused on by Piccari. According to him, all organisms are cognitive systems, and life itself owes its preservation to a cognitive process. This process generates biological knowledge, which is formed through the processing of information relating to the external world or our own bodies. In humans, biological knowledge is much more complex than that of other living species due to the presence of the neocortex and the influence of social and cultural life; for him, this kind of knowledge is the ordinary knowledge. From this perspective, common sense as a complex set of beliefs, mental attitudes, views and values not only has an influence on ordinary knowledge, but is also key parts of it.

Epistemology of Ordinary Knowledge

xiii

Ordinary knowledge is not only fit for its purpose, but able to provide articulated semiotic-noetic structures to regulate the behaviour and actions of human beings. Such knowledge also helps to form the noetic conceptual structures that allow the formulation of representations of the human species, societies and any other structures belonging to the human or nonhuman world. From his point of view, Piccolo holds that common sense can be considered adopting a foundationalist approach. Within this foundationalist approach, one can make out lighter forms, in the manner of Wittgenstein in On Certainty, and stricter forms, in the manner of axiomatic systems or methods, as in Aristotle and Descartes respectively. A minimal definition of common sense recalls the Davidsonian idea of positive praesupponendum, understood as a set of few beliefs, which must be shared with other parties so that dialogue can begin. This gaunt structure is placed by Davidson at the base of the cognitive process intended as interpretation. In this sense, Davidson would also permit us to clarify the link between common sense and ordinary knowledge. Finally, in this section, Lourdes Velazquez also states that, according to a wide-spread view, common sense is a domain of shared opinions that people hold regarding almost all aspects of reality, independently of any specialised competence, but as a kind of immediate unreflective acceptance. Precisely because of this, the frontiers of this domain are fuzzy and these opinions are generic rather than general, in the sense that their scope is not circumscribed precisely. In addition, this common sense also incorporates several kinds of reasoning, which are spontaneously applied without being codified in any explicit set of inferential rules. Part 3 deals with some aspects of the formation of ordinary concepts. Bianca and Piccari focus on the nature of perceptual concepts. These concepts are considered a conceptual framework, represented by a fivedimensional vector in which, in addition to the strictly perceptual content (the identitive and specifying perceptual attributes), non-perceptual content (semantic reference and various significances assigned to perceptual concepts by single individuals) is relevant. This model does not consider perceptual concepts the simple results of an empirical generalisation conducted on the basis of different perceptual instances, but highlights the relevance of non-perceptual contents in their formation. Differently from the general model proposed by Bianca and Piccari, Frixione and Lieto consider the notion of typicality. As is known, typicality effects are a well-established phenomenon in the field of cognitive research on common sense concepts. However, typicality effects are hard to reconcile with compositionality, a crucial characteristic of

xiv

Foreword

conceptual systems. They propose to face this problem by adopting a dual process approach, according to which the existence of two different types of cognitive system is assumed. Systems of the first type are phylogenetically older, unconscious, automatic, associative, parallel and fast. Systems of the second type are more recent, conscious, sequential and slow, and are based on explicit rule following. Thus, they advance the hypothesis that conceptual representations should consist of (at least) two different kind of component, each responsible for different processes: second type processes involved in complex inference tasks, and fast and automatic first type processes, which perform categorisation, taking advantage of prototypical information associated with concepts. Ordinary knowledge concepts, furthermore, can be analysed from two different perspectives: an embodied and grounded perspective and a social one. Regarding the first perspective, Borghi underlines that, according to the mainstream view in psychology and neuroscience, concepts are informational units, rather stable, and are represented in a propositional format. In the view that she outlines, concepts instead correspond to patterns of activation of the perception, action and emotional systems which are typically activated when we interact with the entities they refer to. Starting from this embodied and grounded approach to concepts, Borghi focuses on various research lines and presents some experimental evidence concerning the concepts of objects and actions, and abstract concepts. Then she adds that, in order to account for abstract concepts, the embodied and grounded theories should be extended. With regard to the second perspective, Amoretti claims that, according to Donald Davidson, in order to have any concepts or beliefs whatsoever, a creature must ‘triangulate’ (or at least ‘have triangulated’) objects and events in the outside world with other creatures sufficiently similar to itself. More precisely, she argues that there are two prerequisites for thought – that is, fixing the empirical content and having the concept of objectivity – which can only be obtained through ‘triangulation’ and that ‘triangulation’ is the simplest kind of social interaction. Hence, social interaction can be considered fundamental to having concepts and beliefs. However, according to several critics, Davidson’s main argument for the thesis that thought is necessarily social has proven to be seriously flawed. In her paper Amoretti has two aims. On the one hand, she briefly sums up the difficulties with Davidson’s account, showing that the two prerequisites for thought that he recognises are both inadequate and that the very notion of ‘triangulation’ he employs is useless for defending the social dimension of concepts and beliefs. On the other hand, she tries to

Epistemology of Ordinary Knowledge

xv

reconstruct Davidson’s argument by identifying two weaker prerequisites for thought and a different kind of ‘triangulation’. In particular, she claims that the two prerequisites for thought are: fixing common domains of current relevance and a concept of proto-objectivity, conceived as the awareness that there is a difference between agreement and disagreement. Furthermore, she exploits the psychological notion of ‘joint attention’ to sketch a kind of triangulation that may be a genuine example of social interaction, truly necessary for the two prerequisites she identifies, as well as a pre-cognitive and pre-linguistic process. Finally, she explains how this alternative account of the emergence of thought guarantees its social character and is neither circular nor reductive. The last section deals with reasoning, inferential processes and historical knowledge. Cevolani and Crupi discuss the so called ‘Linda paradox’, which is the most well-known instance of a widespread phenomenon – the conjunction fallacy – showing that naïve reasoning systematically violates the rules of probability calculus. As such, in the last few decades, it has been a central issue in the analysis of human reasoning and decision making under uncertainty. In their paper, they present two accounts of the Linda paradox, based on the notions of verisimilitude (or truthlikeness) and confirmation (or inductive support) respectively, as explored within the philosophy of science. Furthermore, they discuss the conceptual relationships between these two accounts and conclude with some brief remarks on the roles of truth and information in human reasoning and cognitive decision theory. For his part, Labinaz, in his paper, criticises what can be called the Cartesian account of reasoning, which is commonly assumed, either implicitly or explicitly, in experimental approaches to human reasoning, and then proposes an alternative account, more closely related to our ordinary reasoning. While the first account, which is based on a strongly individualistic picture of human cognition, considers reasoning a completely self-centred activity aimed at enhancing one’s knowledge and maximising one’s own personal utilities, according to the account Labinaz outlines, reasoning displays a fundamental argumentative, interpersonal dimension. His aim in this paper is to lay the foundations of this theory, starting with some ideas and reflections about human rationality presented (in a non-systematic way) by Paul Grice, which some scholars have recently systematised under the label of argumentative rationality in discussing the rationality of conversational implicature. Alai examines ordinary notions of knowledge and justification after Gettier. Gettier argued that Plato’s definition of knowledge as true justified belief does not do full justice to all the deep features of our common sense

xvi

Foreword

notion of knowledge. But Alai argues that his seminal paper and the ensuing literature might instead be taken as challenges to the ordinary notion of justification. The latter, in fact, implicitly carries two different connotations: one of an objective and external condition, securing truth, and one of a subjective condition, involving internal evidence but failing to entail truth. Thus, for Alai, both conditions must be encompassed by an adequate characterisation of justification. Attempts to explain the objective condition in causal terms encounter various problems. So, in order to avoid them, Alai generalises the concept of causal chain into that of productive chain, and argues that a belief is objectively justified when it is brought about by a normal productive chain, i.e. one where each link is of a type normally produced by facts of the same type as the truthmaker. Moreover, he accounts for the subjective condition of justification by showing that internal evidences like perceptual experiences, collateral beliefs and inferences are typical parts of productive chains. Once justification is thus characterised through both the objective and the subjective component, there is no longer any need to modify Plato’s account of knowledge. But if one prefers to reserve the term ‘justification’ for the subjective component, then the objective component will supply the additional condition required to supplement Plato’s tripartite definition. In any case, truth becomes redundant in the definition of knowledge, since it is already ensured by the normal productive chain. Salis analyses inferential role semantics, which is the idea that conceptual content is determined by its role in reasoning: a concept C is given by the premises we use in order to draw C, and by the conclusions we draw from C. This theory entails that concepts are rich and stratified, and that conceptual competence involves the mastery of a wide range of inferential transitions. He proposes, in this context, a distinction between a common sense version of the grasp of concepts and an expert one. The first idea is that this grasp does not entail mastering all the inferences which are constitutive of C, but just a few. Expert grasp, on the contrary, is a full and qualified mastery of the inferential transitions involving C. Finally, Lecis and Busacchi analyse historical aspects of ordinary knowledge. How do we form the representation of the reality of the past? What roles are played by memory, imagination and the epistemic constructions of historical knowledge? To answer to these questions, Lecis remarks that during the nineties, Paul Ricœur focused his attention on the relation between memory and history, in search of reasons accounting for historians’ naïve realism. The function of the original psychical trace, which a mind receives, is important and irreducible; natural competences and common sense enable us to immediately recognise it as a sign

Epistemology of Ordinary Knowledge

xvii

depending on past circumstances and events. Busacchi, in his paper, develops a comparison between the roles of testimony in ordinary knowledge and in history. Historical knowledge shows a puzzling epistemological structure: firstly because that knowledge is placed between explanation and understanding, according to von Wright and Ricœur, and secondly because historical knowledge works (a) between scientific knowledge and common sense, (b) using materials such as documents, signs and historical remains, mental (subjective) contents such as personal experience and testimony, and subjective reconstruction through a (personal) story of one’s own for that knowledge. Thus Busacchi answers two relevant questions: what is the epistemic status of testimony? And what is the mode of transforming ordinary testimony into a wellcontrolled method of historical knowledge? *** We want to thank the authors of the essays included in this volume for their dedication to the task of formulating an epistemology of ordinary knowledge. We also want to thank the staff at Cambridge Scholars, who made this project both easy and enjoyable. Mariano L. Bianca and Paolo Piccari



PART I ORDINARY KNOWLEDGE



CHAPTER ONE THE EPISTEMOLOGICAL STRUCTURE OF ORDINARY KNOWLEDGE MARIANO L. BIANCA

The epistemological structure of ordinary knowledge is rather complex since such knowledge, on the one hand, is formed by different types of knowledge (perceptual/phenomenic, knowledge of the mind and of the self, knowledge acquired from others, etc.); on the other hand it is always activated with mental and meta-mental processes which involve a great number of factors depending on the conditions in which they occur (previously formed knowledge, motivations and other psychic and existential factors) and from the operational specificity of each mind. For these reasons, even if it is possible to formulate epistemological structures related to the different types of ordinary knowledge of all subjects, any process of ordinary knowledge can be examined in a specific way with reference to different minds and subjects: hence, it is possible to formulate both class epistemologies, referring to the different types of ordinary knowledge, and single epistemologies, referring to single subjects, as clearly happens in clinical analysis and in psychotherapy. This paper analyzes no single epistemologies but class epistemologies, referring to the different types of ordinary knowledge and the related characteristics of reliability, adequacy and control; in the following analysis the most relevant aspects will be analyzed without any deepening, which may be done only by theoretical and experimental research to outline an articulated model of the class epistemologies of ordinary knowledge.

1. Knowledge and its Forms Before analyzing the nature of ordinary knowledge, some specifications about the nature of the mind and knowledge are useful. In this paper, the mind is considered the entirety of the information and processes which are elaborated by the brain with the involvement of one

4

Chapter One

or more cortical areas. The mind is part of the brain but it does not identify with it since not all cerebral processes are mental processes; mental processes are the brain’s processes which involve, besides noncortical areas, different areas of the cerebral cortex and they generate meanings which in many cases, but not always, can be transcribed by means of a system of signs like that of the natural language. We will use the word noesis and its adjective noetic to refer to mind’s activities differentiated as all other cerebral processes that do not largely involve the neocortex; the term nomiosis refers to any kind of mental information and processes and the term nosema refers to each specific result of nomiotic processes. Among mental (or noetic) processes, we distinguish between idetic processes, which are formulated without a direct elaboration of perceptual or empirical data, and perceptual processes, which elaborate perceptual or empirical data. Hereafter, all expressions related to the mind will refer to the configurations and processes which, even if in different ways, are the result of the complex activation of the neocortex, besides other noncortical areas and processes. As far as the notion of knowledge is concerned, it is necessary to point out how it is understood in this context and to specify its different forms. In the history of scientific and philosophical thought, the term knowledge has been assigned different meanings, and it has been used variously to refer to the phenomenic or natural world and to many dimensions or realities, like the theological, the spiritual and the metaphysical. Even if, as is known, there are many differences amongst the gnoseological theories which have been formulated in philosophical, scientific and theological thought, these theories, with the exclusion of the innate conceptions, share the following thesis: General Thesis on the Nature of Knowledge Knowledge is the acquisition, the formation, the possession and the elaboration of information about any kind of something which is considered to exist: acquisition, performed with perceptual and nonperceptual processes and tools; formation, understood as a process that aims to formulate purely idetic something and to assign to them attributes like cognitive, theological, metaphysical, theoretical, non-empirical or non-perceptual something (idetic something); possession, referring to the memorization and therefore to the presence of information in the mind; elaboration, as the collection of processes aimed at the acquisition, formulation and possession and of changes, relations and influences amongst the various pieces of information.

The Epistemological Structure of Ordinary Knowledge

5

The term “knowledge” therefore refers both to the cognitive processes and to the different kinds of information which are elaborated and saved in the memory (the contents of the knowledge). Any form of knowledge is possible only if three entities exist, or are supposed to exist: a) a biological or artificial organism (a subject); b) a biological or artificial apparatus inside such an organism which is able to acquire, formulate and elaborate information, more particularly a brain or a mind, and c) a universe of innumerable something belonging to the phenomenic world or to other realities. Any form of knowledge, moreover, is based on three realistic assumptions: 1) there are entities phenomenically or noetically considered to exist (including the knowing subject); 2) cognitive processes exist only since they are referable to something considered existing, and 3) information on various something can be acquired (empirically) or formed (idetically) and the processes of elaboration of the acquired or formed information can be carried out. Therefore, to claim that it is possible to formulate knowledge of something means to presuppose that it exists or its existence is perceptually or idetically considered (theoretical or metaphysical something). The term phenomenic refers to the so-called natural world and so to anything which is empirically observed with the sensory organs or with non-biological tools like scientific apparatus. The term idetic refers to different types of something: those formulated without any direct empirical or perceptual information, like scientific theoretical entities; those non-empirical (or not perceptual)m formulated in ordinary thought and in memetic thought, including common sense, like aesthetical and ethical values; those peculiar to any mind and those formulated by philosophy, more particularly by metaphysics and theology. It is not possible to speak of knowledge of something which is not considered to exist, no matter what its existence is; in other words, anything which is considered to possess a determined ontic status (including theoretical and mental ones). Indeed, to formulate knowledge of something means first of all to consider it to be phenomenically or idetically existent. The assumptions which have been indicated are realistic, phenomenically and idetically, but not completely, since from them does not derive directly, as a consequence, the fundamental character of standard realism: it is possible to have a strongly reliable and complete knowledge of something. So, to claim that it is possible to formulate knowledge of something, according to what has been said above, does not

 

6

Chapter One

mean to assert that it is completely corresponding, or even isomorph, to the structure of the thing to which it refers. From this perspective, which is proper in the philosophical tradition, the term knowledge is used to refer, on one hand, to the process of acquisition or formation of information, and on the other, to its elaboration and accumulation (possession). This paper adopts the aforementioned gnoseological thesis, which can be applied to anything and any kind of knowledge, such as scientific, ordinary, mental and idetic (theoretical and metaphysical).

2. Ordinary Knowledge and Common Sense Humans formulate ordinary knowledge using tools and methods fitting to their bodily, cerebral and mental structure: the knowledge that human beings assemble in order to acquire or form information on the phenomenic world (including their own body), on their minds and on their selves, on the memetic or socio cultural world, including their fellows, or on any other something considered to be existent, such as idetic something (mental, metaphysical or theological). Therefore ordinary knowledge can be considered to be any knowledge which, for its methods, tools and results, takes place outside scientific and metaphysical knowledge, even if it can share some of the different something to which they refer. This kind of knowledge, although in very different and more or less complex forms, still refers only to the phenomenic world, and is present in all living beings because they are disenergetic or heterotrophic; they do not possesses the energy for their survival, hence they need this knowledge, consisting of information useful for finding the sources of energy necessary to maintain their thermodynamic balance, in other words, to survive in a certain habitat, to reproduce and to compete for energetic resources with the members of their species or of others ( biological knowledge). The ordinary knowledge of Homo is analogous, but only partially, to the biological knowledge which any living being is capable of formulating relative to its habitat and which is necessary to its survival, and for man, in a broader sense to its existence. For these reasons, the ordinary knowledge of Homo has a biological basis and as such it can be considered (at least in many conditions, even if not in all) more or less reliable; in other words, the information that it elaborates corresponds, more or less, to the status of the world, even with the differences deriving from the diversity of sensory apparatuses.

The Epistemological Structure of Ordinary Knowledge

7

Such knowledge, in the strict biological sense, must be reliable, since it is evolution’s expedient which allows the living beings of planet Earth to acquire information about the world which are useful to their survival; if this knowledge were not more or less reliable, it would not accomplish its task and heterotrophic living beings could not survive, or maybe they would not even have been generated by the processes which caused the onset and the developments of life on our planet. However, even if there is this biological root to ordinary knowledge, in Homo, this knowledge is not reducible to this character; it is much more complex, leaving also the task of satisfying biological needs: the biological knowledge which is formed with the sensory organs is only a small, if relevant, part of ordinary knowledge. Epistemology before the modern development of scientific research dealt only with this knowledge and its investigations brought about the formulation of different gnoseological theories, like, for instance, those of Aristotle, Plato, Saint Thomas, Roger Bacon, Locke, Hume, Kant, Husserl and many others. In the history of western philosophy, one of the main investigation fields was gnoseology, which had the purpose of analyzing the methods with which information is acquired, formed and elaborated, not only of the phenomenic entities but also of the theoretical or metaphysical ones considered to exist. This interest was present not only in western civilization, but also in oriental philosophy, like the Vedanta philosophy, which devoted great attention to gnoseology and particularly to the perceptual processes. The development of scientific knowledge in the modern and contemporary era evidenced, as it is known, fundamental differences from ordinary knowledge, since scientific knowledge aimed to overcome the senses, which were considered the main tool of ordinary knowledge; Copernico, in a letter in his De revolutionibus Orbium Coelestium dedicated to Pope Paul III, referred to the fact that his conception surpassed the senses, and in this case that meant overcoming the sensation of the sun moving on the horizon. However, this is not the only difference between ordinary and scientific knowledge. There are many others, which will not be examined here but will be evident in analyzing ordinary knowledge and some of its epistemological aspects. The sensorial character of ordinary knowledge leads many authors, even in the present day, to reduce ordinary knowledge to sensorial knowledge or, more broadly, perceptual, referring to anything of the phenomenic world.

 

8

Chapter One

Even if the ordinary knowledge which is formulated with sensoryperceptual tools is very important and in some way analogous to that of other living species, it is only a part of ordinary knowledge, which interlaces with other types of ordinary knowledge, which may also utilize the sensory/perceptual results, but, as it will be seen, in many cases, leave them out of consideration. Before continuing the analysis of ordinary knowledge, it is useful to pause to analyze the notion of common sense. In the next section it will be asserted that ordinary knowledge is formed by many types of knowledge, which are different not only in their contents but also in the methods and the tools with which they are formulated. In this complexity of different knowledge types, there are also those of so-called common sense, which are learned in a perceptual or non-perceptual way and which are considered a part of the mind of those who belong to a certain culture. Common sense can be understood in two different ways: a) the usual modalities with which ordinary knowledge is formulated, or b) the collection of information of different kinds which is formulated in a certain culture and which is acquired by those who belong to it. In the first, the expression “common sense” is comparable to “ordinary knowledge” as it has been considered here above and as it will later on be analyzed. In the second, the expression instead refers to the memetic world, that is, to that collection of information which is formulated by a certain culture and that differentiates it from any other. In this case we do not refer to the knowledge formulated by single subjects, as happens in ordinary knowledge, but to that formulated by collective or socio-cultural processes which generate visions of the world, values, judgments, prejudices and conceptions regarding different aspects of man, of individual life, of interpersonal relations, of social structure, of the sacred and divine, etc. Any subject belonging to a certain culture can acquire the knowledge of common sense and by absorbing it; even without careful analysis, he uses it (aware or not) as a guide for his behavior and his way of thinking. This does not mean, as it was claimed in the past, that common sense corresponds to the manner of thinking of the human mind, because it is not the result, at least not only, of the mental activity of the individual person, but that of the social-anthropological culture that produces the so called memes. Using the second description, the notion of common sense cannot be confused with that of ordinary knowledge; however, the knowledge of common sense is part of ordinary knowledge since it is considered

The Epistemological Structure of Ordinary Knowledge

9

knowledge which any man can acquire, elaborate and appropriate in the same way in which indirect knowledge is acquired, that is, ordinary knowledge which is not formulated through personal experience, but according to reports, as it frequently happens in daily life, more particularly in the processes of social and scholastic learning (see Section 3.5). Common sense, moreover, according to the second way in which it is understood, plays an important role in the formation of ordinary knowledge; in fact, many but not all the knowledge types which form ordinary knowledge, including the perceptual-sensory type, can be influenced in different ways and with different intensity by the knowledge of common sense acquired and considered reliable or even true, and as such it is a useful tool for interpreting and knowing the world. This does not mean one can assume, like some sociologists and anthropologists claim, that ordinary knowledge is determined by common sense, even if between one and the other there can exist various mutual differences and relations which are not investigated in this paper. For these reasons, it is important to distinguish the two notions. This essay deals with neither the structure of common sense nor the ways in which it is formulated and influences the social tissue and mind of the individual subjects, but it will restrict its attention to ordinary knowledge, its types and the modalities of its formation and control, without forgetting that common sense, as it has been indicated, can influence, but not determine, the formation of ordinary knowledge.

3. The Polignosic Structure of Ordinary Knowledge: The Articulation in Various Types of Epistemologically Different Knowledge Ordinary knowledge, on the whole, is formed by a large number of mental processes and the something to which it refers: ordinary knowledge is formed by very different types of knowledge, which are correlated in various ways according to mental conditions and the tasks they try to accomplish. The types of knowledge which form ordinary knowledge distinguish themselves by the methods and tools which they use, by the something to which they refer and by the purposes they want to achieve. Let us think, for instance, of the difference between ordinary perceptual knowledge and non-perceptual or idetic knowledge; or of the knowledge of one’s own body, on the one hand, and that of one’s own mind on the other; in these types of ordinary knowledge, the methods of acquisition and treatment of

 

10

Chapter One

the information and the tools which are used and their purposes are different. Differently from scientific knowledge, which only refers to the phenomenic world, and from metaphysical knowledge, which deals with entities considered essential beings or the principles of reality as a whole, ordinary knowledge has a complex structure formed of different types of knowledge which not only refer to the phenomenic world, but also to a very large gnoseological spectrum relative to different systems of entities like one’s own body and mind, other minds, and other ethical, metaphysical or memetic entities. This characteristic is one of the most relevant aspects which differentiate ordinary knowledge from scientific and metaphysical knowledge; in other words, the formulation of knowledge that does not refer, or does not directly refer, to the phenomenic world. In ordinary knowledge, it is possible to point out some of the fundamental types of knowledge which form that which can be called the polignosic structure of the ordinary knowledge, or in other words, its gnoseological and epistemological articulation in different types of knowledge and in the related entities which they refer to.

3.1. Pperceptual/Phenomenic Knowledge Pperceptual/phenomenic knowledge is that ordinary knowledge which is formulated through sensation and perception, with the involvement of a large number of extra perceptual factors, cognitive and non-cognitive, such as concepts, categories, feelings or emotions. This knowledge is activated and conceived to achieve purposes related to: 1) the phenomenic world and 2) the individual’s own body. This ordinary knowledge is direct, more or less reliable and adequate and is analogous to that of other living beings. It is direct because it is based on reports in which the information concerns entities to which the activity of the sensory organs refers; this knowledge is determined by the structure of the sensory organs and the brain/mind and, differently from what happens in many living beings, its formulation is not limited to the acquisition of information (the sensory data); the latter is subject to a complex elaboration which involves cognitive structures and noncognitive contents saved in the mind, which influence in different ways, sometime remarkably, perceptions and their positioning within the mind; that is, the way in which the perceptual results are meaningfully correlated with different mental contents. Reliability is a relational notion according to which cognitive information is more or less reliable, not in itself, but in relation to

The Epistemological Structure of Ordinary Knowledge

11

something to which it refers; present in this definition are knowledge and an entity which, in theory, cannot be the same. Cognitive information about something is reliable if it reports its attributes in a more or less complete way; this is also true in cases in which the information is expressed with an assertion. The definition of reliability is therefore applicable not only to the perceptual knowledge of man, but also to the sensory knowledge of any living being. The measure of reliability is the result of a comparison between the information which is formulated about something and such something in its actual condition of being the semantic reference of the information that has been formulated; this information can be linguistically expressed with assertions which speak on such something. For this reason, the notion of reliability is applicable to ordinary perceptual/phenomenic knowledge; such knowledge is more or less reliable since, as it has been stated, in a radical and biological way it is arranged to accomplish tasks connected to survival, even if in Homo it is not limited to these but extends to others, relevant to any subject, such as psychological, relational and existential ones. For this reason, ordinary perceptual/phenomenic knowledge, like biological knowledge, is suitable for supplying a report on something and on its relations with other things; this report refers to the attributes which identify and distinguish it from any other, even if not to all of them. Even if this knowledge is formulated with the activation of different mental non-perceptual contents which can greatly influence it, a perceptual core remains within it which also allows Homo to be able to know in a sufficiently reliable way the environment and its conditions and to recognize the things which are in it in order to be able to act within it to reach certain goals. It is this perceptual core that is more or less reliable in regards to the things to which it refers. This knowledge, at the same time, is adequate since it allows the individual to operate in and on the world and to satisfy specific needs and related goals; its degree of reliability allows this adequacy, but to the adequacy may correspond a low level of reliability. Therefore the reliability and adequacy of perceptual/phenomenic ordinary knowledge are mutually interlaced, but the adequacy does not mean reliability; in fact it can even not be reliable, while adequacy can derive from reliability. The information included in this knowledge is derived from perceptual activities which involve cognitive and non-cognitive contents that can be processed by various elaborations which, on the one hand, correlate them amongst themselves or with other mental contents and, on the other hand, allow the formulation of new contents by means of those tools investigated by logicians and psychologists, like inferences, logical and

 

12

Chapter One

argumentation rules, which can also be useful to formulate descriptions, explanations and predictions. From the epistemological point of view, it has been claimed that this knowledge is direct and more or less reliable, but the critical aspect consists of the fact of its being submitted to control processes in order to evaluate its reliability and adequacy. The fact that such knowledge possesses a biological root, as it has been previously claimed, may induce one to believe that it does need any control, since its reliability is proven by the purposes for which it has been conceived by the evolutionary plan of the living species. This knowledge is not infallible and in strict biological conditions in Homo and in other living beings may not always be adequate for the tasks for which it has been conceived; however, the biological and evolutionary reasons for its presence make it more or less reliable, and a control process can be activated. In fact, this happens very often and automatically ( unawarenessly) in various perceptual conditions; at the same time, control can be also formulated in a more or less intentional way. Of what does this control consist? It is a double control related to reliability and adequacy, which can even lead to the modification of the content formulated or the formulation of new contents relevant to the the same something. Reliability control, which can involve adequacy, is carried out with the following mental and meta-mental processes: 1) the formation of perceptions; 2) the meta-mental process of focusing the perceptual contents; 3) the process of bringing back to the state of awareness the perceptual contents memorized (operation feasible only under some conditions); 4) processes of comparison among various contents related to different modalities of acquisition of information, and 5) focusing on the result of the comparison, evaluation and formulation of a judgment of acceptability (even if not expressly formulated), grounded in the measure of reliability. Control is a meta-mental process, even if it does not refer directly to the individual’s own mental contents in all its modalities. Reliability control can be carried out with the following modalities: The first modality, of a direct type, is the repetition of the perception or its perceptual deepening (first stage): for instance, focusing perceptual attention, like observing something more attentively in order to formulate a comparison between the first perception and the following ones; afterwards, a comparison (second stage) and a subsequent evaluation of reliability (third stage) are formulated, along with the related evaluation and acceptance, even if not expressed explicitly.

The Epistemological Structure of Ordinary Knowledge

13

The second modality consists of the formation of different schemes of inference, like, for instance, in cases in which there is a subsequent carrying out of behaviors or actions so that, if the perceptual contents are adequate, it is inferred that they are sufficiently reliable, at least relative to that condition. The third modality, which can generate inferences, is the comparison between one’s own perception and the perceptions of other subjects and, in cases in which there is sufficient informative overlapping between them, with the exclusion of some idiosyncratic differences, it is inferred that one’s own perception is more or less reliable and adequate. The fourth modality is of a type strictly internal to the mind; in this case, as often happens, the actual perception, with the activation of metamental intentional acts, even if not fully aware, can be compared with previous perceptions of the same something which have been memorized and which can be brought back to the actual status of awareness. It is not taken into consideration that any informative content memorized then brought back to the status of awareness may not correspond to that in the memory and can be simultaneously formulated anew; it is possible to activate a meta-mental act of comparison of the third level (the first relates to the actual perception and the second to perception brought back to the status of awareness) between the current perception and that reported by the memory; with this procedure, the subject may hold that the current perception can be evaluated as more or less reliable and therefore, according to the results of this comparison, he can formulate, explicitly or otherwise, the related judgment of evaluation, including on its adequacy respective of certain conditions. Other modalities of control are indirect and can be of different kinds depending on the something target (the something to which an actual perception refers; subject to meta-mental processes), and at least two can simply be outlined: descriptive linguistic reports formulated by others and surveys carried out (directly or by others) with different kinds of tools, as happens in the case of photos or drawings of the same something target of perception. The process of control is very complex and the indications presented, even if they need theoretical or empirical deepening, are intended to point out that for perceptual/phenomenic ordinary knowledge, the process of control is activated by the subject that formulates it. The crucial expressions in this condition are “sufficiently” and “more or less”, used to refer to the notions of reliability and adequacy. They are expressions which have a subjective character and therefore it is the perceiving subject who measures the reliability and adequacy depending on many factors of

 

14

Chapter One

the perceptual condition (cognitive and non-cognitive factors, including targets, motivations and interests); such subjective measurements by one party may lead to different evaluations of an analogous perception referring to the same object. Therefore, such controls cannot be considered fully neutral and the factors accompanying specific conditions may also significantly influence the measurement of reliability and adequacy. However, these controls are not always or under any conditions unreliable and therefore they allow an evaluation of reliability. Even if they are subjectively activated, they are shared, at least partially, by all subjects and this can be pointed out by any subject and by specific experiments. Subjective influences are comparable in a scientific context to epistemological and epistemic structures, for instance the paradigms, indicated by Kuhn, which operate within the controls of the procedures and the results of scientific research. Furthermore, it is possible to wonder if reliability is merely subjective and as such cannot be controlled and evaluated by a third person; reliability is not merely subjective, thus the measures of reliability formulated by a person can be evaluated by others and this process happens in many perceptual conditions in which more than one subject is presentperceptual. If what is claimed is not accepted, or not completely accepted, then a form of perceptual solipsistic unrealism should be accepted, according to which perceptions can neither be controlled with respect to the world nor compared with those of different subjects in respect to the same something (so-called “perceptual solipsism”) and any subject will formulate on this something perception incomparable with that of any other. Such unrealism, however, cannot be accepted, since, on the grounds of what has previously been affirmed, it is in contrast on one hand with the reliability and adequacy of perceptions which are functional in acting in the world (their biological root) and, on the other, with the perceptual sharing by different subjects of the same something. Finally, it is relevant to note that the tools of the perceptual/sensory knowledge are not useful only to formulate this knowledge but they are active also in other types of ordinary knowledge like the indirect knowledge of the world which avails itself of these tools but it does not formulate a direct knowledge of the world except that related to the sources of the information like videos, books or newspapers.

The Epistemological Structure of Ordinary Knowledge

15

3.2. Knowledge of One’s Own Mental Contents, Processes and States: Self-Knowledge or Meta-Mental Knowledge This type of ordinary knowledge is very different from the perceptual/phenomenic type in four different aspects: 1) it is not knowledge but self-knowledge; 2) it is not about the phenomenic world but about one’s own mind; 3) it does not use sensory/perceptual tools, although it can involve their results, and 4) the notion of reliability is applicable in a very different way from in the case of perceptual/ phenomenic knowledge. Self-knowledge can be considered plausible if three assumptions are accepted: a) the mind can attest to its presence and know its states, processes and contents; b) the mind can activate meta-mental processes which work on mental states, processes and contents already formed and preserved in memory or under formation, and c) the mind can transcribe the results of these meta-mental self-cognitive processes in any system of signs, included linguistic, gestural and behavioral, even if this transcription does not always happen and is not always necessary to the dynamics of thinking; that is, contrary to the propositional conception of thinking, mental activities (included meta-mental processes) very often work in a non-propositional way, so the linguistic cortical areas are not always involved in thinking processes, aware or otherwise. If these three assumptions are accepted, particularly the first two, it can be claimed that it is possible to formulate ordinary knowledge of one’s own mind, whose different something of investigation are its states, processes and contents subject to meta-mental processes. The meta-mental processes can be of three orders: the first order is formed by mental states, processes and contents to which the self-cognitive process pays attention; the second order is that of the processes which deal with the states, processes and contents pointed out in the first order, and the third order, to which many others can be added (a cascade of meta-mental processes of different levels, in the form of a tree with many branches), is that of the processes that act on the mental states, processes and contents of the first and second orders. It is possible to wonder if this self-knowledge (expanded in many meta-levels) is really knowledge, even if it is very different from others; these meta-mental processes fall in the group of those which have been indicated as having the characteristics of any form of knowledge; in other words, the acquisition, formation, possession and elaboration of information about something. In this case, the things in questions are

 

16

Chapter One

different mental something, that is, contents, the processes and states of the individual’s own mind on different meta-levels (see Section 1). The first aspect of this knowledge, as it has been already pointed out, is that of being self-knowledge. This means that the mind, which is the entity which formulates knowledge, is not dealing with something external to it but with its own states, processes and contents and in this way this knowledge is formed by meta-mental processes of different meta-levels. This distinctive aspect is the most significant, not only with respect to any other form of knowledge, like the scientific and the metaphysical, but also the ordinary perceptual/phenomenic knowledge, with the exception, but only partially, of the knowledge of one’s own body, which is knowledge of the perceptual-phenomenic type. The second aspect results from the first: this meta-mental ordinary knowledge does not have the phenomenic world as a semantic reference. Thus a crucial, but easily solvable question, can be raised without having to accept a metaphysical conception of the mind, but maintaining a naturalist one. The mind too is a natural or phenomenic entity, inside the encephalon of which it is a part, even if, as indicated in Section 1, it is formed by a specific class of cerebral processes; at the same time, the mind is not contained only in the brain, but extends beyond it in specific ways and under specific conditions through the signs and expressions of mental states, contents and processes. On the grounds of this claim we can state that self-knowledge too refers to the phenomenic world (cerebral and cerebro-mental processes) and in this way, only in this aspect, it is no different from ordinary perceptual/phenomenic knowledge about the world. It is possible to solve this problem while assuming that the mind does not deal with those phenomenic entities dealt by sensory/perceptual processes; however, it is useful to point it out that these processes are also mental processes, but they refer to different something; meta-mental knowledge does not refer to the phenomenic world if that world includes only that of perceptual/phenomenic knowledge, but if states, processes and mental contents also belong to this world then meta-mental knowledge also refers to the phenomenic world, although it is knowledge of phenomenic entities different from those of perceptual/phenomenic knowledge. These observations let us hold that the mind does not refer only to the phenomenic world, but also to those phenomenic entities that are the states, processes and contents of the mind that generates self-knowledge. On the basis of the characteristics of other forms of knowledge, selfknowledge too is knowledge, for it acquires, formulates and elaborates information on something, and in this case the different something are the

The Epistemological Structure of Ordinary Knowledge

17

mental something (mental states, processes and contents) of the mind that activates this knowledge. One’s own body can also be part of self-knowledge, but only because it is possible to formulate cognitive processes (meta-mental) which deal with the perceptual mental representations of the body which have been saved in the memory, and they are not direct perceptions of the individual’s own body. The third aspect of self-knowledge points out that it is not based on sensory/perceptual processes, since one’s own mental states, processes and contents cannot be perceived, so the so-called “internal senses” are not senses, as such, but a set of meta-mental processes and their results. This statement is acceptable on the basis of the structure of CNS/PNS, in which sensory receptors are predisposed to receive and elaborate information from the phenomenic world, included the body, but do not have any direct neural connection to the cortex so cannot acquire information from it, although, as it is well known, cortical and non-cortical areas are involved in the formation of perceptions. Therefore any knowledge regarding one’s own mind is not perceptual knowledge. However, it can also avail itself of sensory data and of perceptions, but only in cases in which they have been elaborated and are presented as contents of the mind in the form of mental representations. The fourth aspect concerning the reliability of self-knowledge is the crucial one for pointing out the epistemological status of this knowledge, and it will be analyzed after the different types of content, process and mental state that are part of self-knowledge are presented. Self-knowledge of one’s own mind has many targets of semantic reference relating to the following classes of mental states, processes and contents: a) cognitive information; b) pathic or emo-affective information; c) information on status and d) structural and process information. On this occasion, as before, the term “information” and the expression “mental content” will be used with the same semantic reference. The terms “cognitive” and “pathic” or “emo-affective” (noncognitive) refer to the usual way of distinguishing processes and mental contents, cognitively and non-cognitively, as well as relatively to the two cerebral hemispheres and to the cortical and non-cortical areas that are involved. Even if, as it is known, this distinction may be subject to many criticisms, mainly if it is understood in a strong way, it can be accepted either from the analytical point of view or from that of a subject who is able to distinguish, for instance, the difference between solving a logical problem and having feeling within an emotional dynamic.

 

18

Chapter One

Therefore, on one hand, there are the mental activities of the cognitive type, which refer to perception, language, attention, intelligence, concept formation, memory, problem solving, formulation of decisions - topics investigated by cognitive psychology and neuro-psychology. Moreover, evaluations, preferences and motivations are also cognitive activities, even if they involve non-cognitive factors; in the history of philosophy and psychology some of these activities, even if not all, were described with the expression “rational processes”. On the other hand, non-cognitive or pathic (or emo-affective) mental activities are moods, psychic humors, tastes, affections, feelings, passions and emotions involving, besides the cortical areas, various paleoencephalic cerebral structures like reticular formation, roof, hypothalamus, rhinencephalon and limbic system, including the amygdala. The distinction between cognitive and non-cognitive, even if acceptable, since the mind can separately elaborate cognitive and noncognitive information, is that often in mental work these different types of information interlace and are simultaneously involved in different degrees depending on the mental processes activated and on the psychological, existential and mental conditions in which they are activated. These two targets of semantic reference of self-knowledge are very articulated within, so that, for instance, the self-knowledge of mental images is very different from that of non-figurative contents like those related to the individual’s own conceptions about others, society or the world; in the same way, it is different to self-know on the one hand one’s own actual emotion and on the other, a feeling towards a person. This topic is not analyzed in this paper. Status information, meanwhile, is information related to actual or past mental states, that is, the status of the the mental dynamics. A mental state or configuration can be defined as all of the processes which are happening in a certain instant or which have happened in a past instant and which have been memorized; a mental status is not related to the information elaborated nor the results and contents obtained. Let us think of a cognitive status, for instance the activation of an argumentation, or of a non-cognitive status, for instance the condition in which an emotion or a mood is formed. The information elaborated by mental processes is often not a single piece of information, but interlaced informative structures, and in this way the meta-mental dynamics in self-cognitive processes develop by passing from one informative structure to another, depending on the meaningful (or nomiosic) correlations attributed to the global structure of the mind and its connectome.

The Epistemological Structure of Ordinary Knowledge

19

It has to be added that meta-mental self-cognitive processes do not only deal with information and informative structures, but also with the dynamics, processes and modalities of operating unique to each mind; hence, self-knowledge also concerns the modalities with which each mind works. We do not refer to the operating modalities of the human mind, but to the specific dynamics and modalities which are activated by a single mind and that guide its way of operating; for instance, to privilege the role of the imagination, to reason through schemes or even to often use argumentative thinking. Self-knowledge, therefore, is not only related to the informative structures already formed or under formation, but also to the actual ways of using one’s own mind, and in fact each mind has adopted in its history certain modalities and dynamics, which may not be peculiar to that mind, but shared by many human beings. In this type of ordinary knowledge, the information which has been saved in the memory plays an important role. Very often, this information is essential for self-cognitive processes, particularly when this information is brought back to a state of awareness. For this type of ordinary knowledge, there is also the problem of reliability and control. The notion of reliability is applicable, as we saw, to perceptual/phenomenic ordinary knowledge, but it is critical to apply it tout court to knowledge of one’s own mental states, processes and contents. In this condition, there is no subject and something greatly differentiated; on the contrary, there is just the same entity, which formulates statements on itself. In the case of self-knowledge, reliability has to be considered within the mental dynamics in which the mind is capable of making a distinction similar to that between a subject and an object. In this case, the subject is formed by the meta-mental processes activated and the object is formed by the mental something to which such processes refer; obviously, in this condition, an intentional act is present, which, in terms less allusive and confused, consists of the psychological dispositions and motivations which activate the mental processes. Previuosly it has been stated that one of the assumptions for claiming that it is possible to formulate ordinary self-cognitive knowledge is the capability of any mind to know its own states, processes and contents and therefore to activate meta-mental processes to focus on them and make them aware. However, even if it is possible (and it should be) to accept this assumption, it does not say anything on the degree of reliability or the control of self-cognitive knowledge; in other words, what is deemed to be

 

20

Chapter One

known of the own mind could be only the result of mental processes which build contents and do not report others already formulated, as may happen, for instance, in introspective processes. Beyond these conditions, the assumptions formulated show that it is possible to know one’s own mind. What is the reliability of the results of these self-cognitive processes? Do they focus on bringing back to awareness contents, states and processes, or are they instead actual formations of mental contents or even illusions? Even if the illusory aspect is not negligible, any subject has the ability to ascertain that, by activating meta-mental processes, it is possible to know some of the contents of that subject’s own mind; for instance, to find a notion in the mind that is useful for solving a problem, or to understand how to behave in a certain situation based on points of view which have been formulated in the past and brought back to the state of awareness. Therefore, any man is capable of affirming that focusing on his mind can survey, for instance, his own thoughts or his own judgments about something or someone, including himself. There are many ways to ascertain knowledge of one’s own states and mental contents, for instance through the observation of one’s own behavior consequent to certain mental contents. Empirical investigations, too, allow us to support this self-cognitive possibility. More crucial, however, is the control of reliability. It is possible to claim that any subject is capable of controlling the reliability of the results of their self-cognitive meta-mental processes, even if these processes can be greatly invalidated by the psychic, or even, as it has been pointed out, by processes which do not “reflect” mental contents but form them in the moment in which are activated. Reliability, in the case of self-knowledge, refers to the relation between the results of meta-mental processes and the various mental something to which these process refer. Let us consider the following cases. A meta-mental process can aim to bring the memory of a past event back to a state of awareness; this memory is brought back to awareness and its reliability is the relation between the recollection as it is saved in the memory and the information which has been brought back to awareness. Another meta-mental process could aim to focus on an actual emotional state or even a determined judgment or opinion about something, for instance the personal conception of the social conditions in which we live or the behavior of a person. In these cases and many other similar ones, the judgment of a subject of this reliability can be controlled in many ways, the most frequent of which consist in a process of focusing and deepening, which induce one to

The Epistemological Structure of Ordinary Knowledge

21

repeat the meta-mental process and to analyze the results obtained, references to other contents or judgments saved in the memory. In other words, the subject wonders if the result of a meta-mental process corresponds to their own mental state or content. For instance: is my judgment of a determined person just what I have brought back to awareness or is my opinion on the social conditions in which I live really that which I have in my mind now? A process of control is always activated, even if not always in an intentional and aware way, which means it is of the third level (the first is that of focusing on mental information, the second is the result of this focusing and the third is the control of reliability). This process of control may happen either automatically or intentionally and it follows the following dynamics: Motivation and purpose of activating a self-knowledge process ‫ۂ‬ Focusing on a mental something by the meta-mental process ‫ۂ‬ Analysis of the results obtained ‫ۂ‬ Deepening ‫ۂ‬ Control ‫ۂ‬ Modification of the results ‫ۂ‬ Evaluation ‫ۂ‬ Acceptance of the results The process of reliability and its control is therefore quite complex and different from the various types of mental something dealt with by selfcognitive meta-mental processes; for instance, bringing a cognitive content that refers to one’s own life (for instance, which university the individual graduated from) back to awareness is different from bringing back present or past non-cognitive content from one’s own life (for instance, what emotional condition one was in when one was informed of having won an academic competition). Even if, as it is known, there is some difficulty in the processes of selfknowledge, and even more in those of the evaluation/acceptance of their

 

22

Chapter One

results, they can be submitted to self-control, which often means that the subject is capable of repeating them in order to be able to accept the results with a certain degree of reliability with respect to the contents and states of their own mind.

3.3 Knowledge of the one’s own Self and the Concept of Self An analysis similar to that necessary for self-knowledge of one’s own mind can be carried out for the knowledge of one’s own self, which includes the mind but extends to attributes of the subject which do not generally have a mental nature but which can partially if such attributes have a mental representation, such as the attributes of the body (the mental representations of certain parts of the individual’s own body). Knowledge of one’s own self, which is not restricted to that of the mind and the perceptual knowledge of the body, is even more complex and articulated than that of the mind, which, as it is known, leads to the formulation of the concept of self; the self-concept is formed by a complex structure in which various concepts and other judgments and evaluations that are part of the process of self-consideration and of self-esteem are present. Even knowledge of one’s own self has many degrees of reliability and self-control, but, as scientific results point out, cases in which its reliability is very low are many, like in the case, for instance, of many teenagers (but also adults) who assign to their concept of themselves attributes which do not correspond to their real characters; such low reliability may be found even more so in self-deception, which also appears very critical to the process of controlling reliability. However, even with reference to the self, it is possible to trigger, deliberately or otherwise, the verification processes of the results of the meta-mental self-cognitive processes of the self, and consequently also those of control, similar to those indicated for self-knowledge of the subject’s own mind. Finally, it has to be pointed out that the self, with its structures, plays an important role in the formulation not only of self-knowledge, but also of knowledge of one’s own mind.

3.4 Knowledge of Other Minds and Other Selves An important part of ordinary knowledge is the knowledge of other minds and other selves, on which interpersonal and social relations are

The Epistemological Structure of Ordinary Knowledge

23

based: part of establishing interpersonal and social relations is formulating knowledge of others and this knowledge refers to their selves and minds. In this paper, we do not intend to analyze the different modalities of knowledge of other minds and other selves, however it is useful to point out that this knowledge is different from those about the world and the own mind, not only because the something of reference are different, but also because, even using similar tools, it is based on non-perceptual and particularly inferential activities. In fact, to know the minds of other subjects means to formulate inferences, starting from perceptual data about them, their behaviors, their gestural expressions as well as from their linguistic expressions and what they say. Any of this data is the result of perceptual activities and can be evaluated in itself as a manifestation of the other mind and the other self; however, these data, once acquired and elaborated, form knowledge, even if only initial; indeed, very often knowledge of other minds and selves does not limit itself to this data, but they are used to formulate inferences of different kinds, which is not always a rigorous application of inferential processes and their results are not always reliable. In this case, the inferential processes are also not formal, even if, theoretically, they can be formalized and modeled. However, many factors, even if not explicit, which may influence their formulation are involved in them, and in many cases, also in the formation of non-reliable inferences, even if in a certain moment the subject may consider them acceptable. Strictly sensory-perceptual knowledge is fundamental for the formulation of this knowledge, but in a different way, or at least to a different degree, from what happens when the knowledge refers to the phenomenic world. In the case of knowledge of others, perceptual data is always submitted to further complex elaborations, which involve cognitive and non-cognitive factors, evaluations, motivations, judgments and interests; in these cognitive processes, the pathic factors, particularly those which are referred to with the terms “empathy” and “sympathy”, also play an important role. No less essential is the knowledge of common sense, which influences the way in which such knowledge is formed. Let us think, for instance, of many stereotypes related to “the other”, such as those related to gender, ethnicity, skin color or social class, which influence the ordinary processes of knowledge of other minds and other selves. Knowledge of other minds and selves is perceptual and non-perceptual at the same time, and inferences play a significant role. Its contents, results and processes can be considered and analyzed from an epistemological point of view; in this case it is possible to notice not only the different

 

24

Chapter One

modalities of formulation, but also the different manners in which any subject controls reliability and adequacy, which are more articulated than those of the other kinds of ordinary knowledge, because the extra factors that have been indicated are involved in a more significant way. Knowledge of other minds and selves is a form of ordinary knowledge very different from those referring to the subject’s own mind and the phenomenic world, but, due to its importance and mixture with many aspects of daily life, its epistemological statute is more difficult than those of the other forms to analyze and clarify. However, even it is not free of epistemological modalities similar to those of the other forms of ordinary knowledge, and also, meta-mental processes of the third level are carried out by it (these processes refer to its reliability and control, similar to those indicated for perceptual/phenomenic knowledge and selfknowledge, and are not analyzed in this paper).

3.5. Indirect Knowledge of the World Derived from the Perceptions of Others, from Knowledge Learned from Others and from Common Sense A great part of the knowledge that any human being formulates and possesses derives from indirect perceptual knowledge and knowledge acquired from others and from common sense. Indirect perceptual knowledge and that acquired from others and from different sources is second level knowledge (unlike perceptual/ phenomenic and self-knowledge), since the something to which they refer are known by perceiving and elaborating information, which in turn refers directly to these something on one hand, from written sources, like the reading of texts, including literary, such as essays or reviews, or from audiovisual media like films, documentaries, slides etc., and on the other hand, from oral sources like descriptions formulated by others referring to people, animals, or urban or natural events and environments. In this case, the role of the perceptual apparatus is different from its role in perceptual/phenomenic knowledge, whose results directly refer to phenomenic something. In cases of second level knowledge, on the one hand, the perceptual apparatuses operate in the usual way, formulating data referring to the something perceived (for instance, the words written on a page or those uttered by a human or artificial being); on the other hand, the perceptual results are subjected to a process of simulative interpretation for which the data is considered by the subject as if it referred directly to the something described by the texts; indeed this perceptual data refer to texts, and indirectly to something in the world, but

The Epistemological Structure of Ordinary Knowledge

25

the subject considers it as directly referring to the world. In this way, the subject does not take it (the perceptual data of the texts) into consideration like those he formulates directly about the world (except in pathological cases), but he accepts them as if they were so, and consequently he can claim, through simulative interpretations, to possess knowledge of the world derived from these processes involving texts or other sources; for instance, the vision of the subject of a documentary on the Serengeti Plain not only allows him to acquire knowledge about it, but he may consider it as if it were knowledge that he has formulated in a direct way, even if he is always capable (unlike in pathological cases) to distinguish direct perceptions from indirect ones. Simulative interpretation consists of the acquisition of indirect knowledge and considering it as if it has been perceived directly. In this case, the indirect perceptual data on something, similar to direct data, is subject to processes of control, which consist of comparisons with other data which are present in the mind and also of the acquisition of further data. Amongst this data must be included the theoretical, imaginary and fantastic, for which different processes of knowledge and control are performed, which are not investigated in this paper. Common sense is also inherent to the above indicated process so, as already mentioned, the contents of common sense include indirect knowledge which is acquired and owned. The processes examined, once again, show that ordinary knowledge is not only, as stated by folk psychology but also by some researchers and epistemologists, a simple, ingenuous or even unreliable knowledge in which the people are only passive subjects, acquirin information without activating any critical or reliability processes.

4. Descriptions, Explanations, Predictions and Cognitive Macrostructures Except in some cases, knowledge, no matter of which type, is elaborated with the purpose of formulating descriptions, explanations and predictions useful for being and acting in the world. Ordinary knowledge, too, is activated to formulate descriptions, explanations and predictions regarding the phenomenic world, other minds and selves, the social, one’s own mind and one’s own self. If it was not like this, it could not accomplish its biological, existential, interpersonal and cultural purposes. Descriptions, explanations and predictions refer, on the one hand, to one’s own mind and self, and on the other hand, to the world, including other minds and other selves. In the first case, the formulation of self-

 

26

Chapter One

knowledge is grounded on motivations, purposes and targets, and involves the activation of meta-mental processes. The latter are refer to the states, processes and contents of the subject’s own mind in order, first of all, to attest to their cognitive or non-cognitive type; from this attestation, the processes which allow the formulation of a description of the ascertained state/content can be activated, which consists of the formulation of a sequence of the information present in it. For instance, information related to a judgment of some event which has been previously formulated, or related to a feeling in the past for a person. In the case of descriptions, the processes of reliability and control are also activated, which allow the subject to potentially ascertain the result of the processes of description and therefore the correspondence between the descriptions and states, processes and contents of the subject’s own mind (this correspondence is deeply analyzed in my essay: M. Bianca, La mente immaginale, 2.3). In the second case, descriptions related to the world (including other minds and selves) are formulated in such a way that the available data is useful for knowing it better or for activating certain behaviors. In ordinary knowledge, similar to scientific knowledge, explanations are formulated for both one’s own mind and the world; for instance, the explanation for the presence of an emotional state, a choice, oa determined way of thinking or acting and the cause of a phenomenic event. In ordinary knowledge, such explanations are usually not as accurate as those of scientific knowledge. They are not formulated by adopting certain methods and their goal is not usually precisely that of formulating a reliable explanation, but that of supplying an adequate one, that is, a useful one for the subject understanding him or herself and the world in order to formulate their own vision of it or to activate a determined behavior; in this paper, the tools with which they are formulated, which are often inferential ones like induction and abduction etc. or various kinds of argumentation, are not analyzed. Explanation is an essential process of ordinary knowledge and for this reason, not only is a large amount of information, which is the result of processes of explanation referring to the world, present in it, but explicative activity is also an important part of the ordinary knowledge of any subject; in fact, such knowledge not only describes the world, the subject’s own mind and the self, but it is also engaged with complex processes to formulate explanations, and in this way it accomplishes its task: to be useful and adequate (even if more or less reliable) for being and acting in the world. As happens in scientific knowledge, descriptions and explanations formulated by ordinary cognitive processes are also useful for formulating

The Epistemological Structure of Ordinary Knowledge

27

predictions about the subject’s own self and about the world (including other selves), and therefore for making choices and activating behaviors: predictions are the results of inductive processes and inferences from perceptual data, and of other reasoning grounded on one’s own conceptions, ideals and judgments on the self and on the world. A further aspect of ordinary cognitive processes refers to those which we call cognitive macrostructures (or superstructures). This expression refers to the fact that ordinary knowledge, similar to what happens in scientific knowledge, not only formulates knowledge on something; the knowledge formulated is not only classified and arranged in order to be able to be used in different mental circumstances, but are also intercorrelated so that they form cognitive macrostructures, which operate: a) to acquire new knowledge; b) to classify new knowledge; c) to correlate different types of knowledge and d) to guide the cognitive processes. Such macrostructures may be considered ways to formulate knowledge, and at the same time, modalities of considering the different something to which cognitive attention is turned.

5. Aware and Unaware Knowledge In the analysis of the various forms of ordinary knowledge, a fundamental aspect has not been analyzed: that of awareness. In general it has to be pointed out that when someone claims to know some attribute of something, it means that they are capable of expressing such attributes with explicit awareness in any system of signs whatsoever: if, for instance, it is stated that a person knows something of their own mind, then it means that person is able to formulate propositions, to carry out behaviors, to make a drawing or a sketch which represents this something of their own mind. Explicit awareness is awareness in which the self (and the subject) are aware. Definition: A mental configuration (or mental state) and its information is awarenessly explicit if the subject is capable to formulate propositions in any language, even if simple and iconic-symbolic, or in other expressive gestural or behavioral signs which point out and notice that he is in that determined neuronal configuration and to make explicit the informative content or to activate the consequent behaviors. It is possible to prove explicit awareness in the following ways: A) To pronounce in an effective and/or explicit way propositions which:

 

28

Chapter One

1) confirm the subject is in a such configuration 2) describe in a more or less complete way the information which is present in this configuration B) To codify with propositional signs or signs of any other nature the condition of being in such a configuration C) To activate the consequent behaviors or actions which confirm the subject is in such a configuration D) To ascertain and/or to notice the subject is in a determined configuration without the same configuration having been codified according to ways A or B E) To preserve such ascertainment for a short or long period of time (preservation of information) The criterion of explicit awareness arises from these conditions. Definition of the criterion of explicit awareness: The neuronal configuration CONF in the mind of a subject S is explicitely aware for S if both the conditions D and E are met and if at least one of the conditions A, B, or C can be effectively satisfiable. From this criterion derives the fact that the possibility of conditions A, B and C being effectively satisfied do not exclude one another. The notion of effective satisfiability states that a certain mental configuration, once the conditions under D and E have been satisfied, is aware only if it can change itself with one of the conditions A, B or C; this distinguishes an aware neuronal configuration from any other neuronal configuration. If none of conditions A, B or C can be satisfied, the configuration is not actually aware, even if conditions D and E have been satisfied. Theoretically, this criterion of explicit awareness should be applied to any form of ordinary knowledge; however. it cannot be applied to all ordinary knowledge in both directions (i.e. with reference to the processes and with reference to the contents). Indeed, is it possible to state that ordinary knowledge as a whole involves awareness? Is it plausible to state that any knowledge formulated by the mind in any of the different ways which have been indicated, can be explicitly expressed and therefore be aware? In other words, can it be affirmed that it is possible to be aware of all the cognitive information that one’s own mind has elaborated, elaborates or will elaborate? Can it be stated that any contents of ordinary knowledge and any gnoseological mental process has, as a result, aware contents? These are questions which are obviously not put in similar ways for other types of knowledge, like the scientific, but

The Epistemological Structure of Ordinary Knowledge

29

which are essential for ordinary knowledge and particularly for perceptual knowledge and that of one’s own mental states, processes and contents. If these questions were answered in the negative, then the same notions of ordinary knowledge and self-knowledge would be meaningless. If an unreal skepticism is not accepted, then it is not possible to answer these questions completely negatively; however, neither is it possible to answer them in a fully affirmative way. This way of operating safeguards any analysis of ordinary knowledge, including the one formulated in this paper. The topic is very difficult and complicated to tackle, but on this occasion, it is necessary to formulate at least some observations. First of all, obviously, not all the knowledge we possess in our minds is actually aware, but all of it might have been in the past and could be in the future; maybe not all, but certainly a large amount of it. Therefore, within ordinary knowledge, some knowledge is present in the state of awareness, currently or in another past or future instant: a thesis fully acceptable and subjectively ascertainable, but one which is obvious. It does not solve the problem, but it facilitates the search for its solution. Different bits of ordinary knowledge are unaware in a certain instant, and in any instant there is only aware knowledge, so, except for this, none of the cognitive information formulated and present in the mind is aware. Is it possible to conclude that, among this actually unaware information, there is some which is never aware; in other words, the notion of explicit awareness cannot be applied to them? That there is knowledge in the mind which never reaches the state of awareness, and as such, is not aware? To state that they are not aware does not mean that they are implicit; in fact implicit knowledge is so because it is not expressed in a certain instant, but it could be in a subsequent moment and as such is not definitively unaware. The subjective and experimental evidence underline that the mind does not need to consciouskly and explicitly express all the knowledge that it formulates, even those about the subject’s own self, so there is knowledge formulated by the mind that the subject cannot claim to possess. Knowledge which, even if it is not aware, can influence mental dynamics. In this paper, we do not analyze the difficulties of this thesis; however, another important aspect has to be pointed out: what is aware does not always correspond to what is present in the mind, so it could be stated that the knowledge formulated in the subject’s own mind is much more than that which is awarenessly ascertained; ordinary knowledge of the various indicated kinds is wider than that which can be expressed. Hence, once again, the criterion of explicit knowledge cannot be considered the only

 

30

Chapter One

criterion for measuring the quantity of different ordinary knowledge formulated by a mind. The question of awareness in reference to ordinary knowledge is even more critical if, instead of the mental contents, the cognitive mental processes are considered. This is a topic which is not analyzed, but it is possible to state that, in most cognitive activities, we are not aware of the processes that we are activating. Obviously in this case we do not consider the unawareness of the underlying physical and chemical processes. Finally, it is relevant to note that many, but not all, unaware contents may be brought back to awareness in a determined instant through metamental processes aimed to do just that; an activity of the mind to which we have previously referred and one which is fundamental for the selfknowledge of one’s own mind.

6. Formation, Interrelation Dynamics and Operations on Knowledge and Self-Knowledge In the preceding sections, the different types of ordinary knowledge and the modalities of their specific formation have been described. However, the analysis that has been presented is very idealized and simplified, since the formation of any ordinary knowledge follows, on one hand, a specific modality (for instance a perceptual one), while on the other hand, it can be influenced by various mental factors, including those which are not pertinent to its content (for instance, non-cognitive information related to cognitive content). In forming ordinary knowledge it is necessary to distinguish the formation of specific knowledge from the global formation of ordinary knowledge. For the first, besides the already indicated modalities, it is necessary to point out that the different types of knowledge do not form in a neutral and isolated way, but are subject to the mental influences of already acquired knowledge, the psychic existential conditions in which they are formulated and the goal which it is intended to reach with their formulation. Any ordinary knowledge, including perceptual, is formed with the intervention of various information present in the mind, which, even if to a different degree, may guide its formation. Let us think, for instance, of the influence of conceptual structures, of conceptions of the world, of macrostructures and of various (cognitive and non-cognitive) prejudices. This does not mean that the intervening factors in the formation of any ordinary knowledge are such that they invalidate the cognitive processes and so allow denial of their reliability.

The Epistemological Structure of Ordinary Knowledge

31

The second important factor is constituted by the conditions in which the knowledge is formulated. These conditions are not only mental but also psychical and existential; the mental ones are those previously indicated, which include the information which intervenes in the formation of knowledge. The psychical and existential refer to the global state of the self in the moment in which a gnoseological process is activated; these include, on one hand, moods, self-considerations and judgments on one’s own existential condition and, on the other hand, relations with others and the world. The third factor is related to the purposes that the formation of the knowledge is intended to fulfill. For instance, it is very different to perceptually know an object just enough to attest its presence than it is to be able to use it to perform an action; in a similar way, the learning process is different to acquiring the knowledge of others in order to take an examination (for instance, that learned in class or read in a book), or even to satisfy one’s own cognitive curiosity or for the understanding of a problem which has been formulated. Although in everyday life knowledge can often be acquired automatically and in a disinterested way (e.g. by hearing other people’s words, watching an advertisement or perceiving objects while walking), often, the various types of ordinary knowledge are guided by motivations and purposes for which those which it is intended to formulate are not only chosen, but they are also formulated in a certain way rather than in another; for instance, knowledge of a person may be very different if one intends to establish with that person an affective rather than a working relation. In both cases, different mental and psychical factors can be activated, which influence the formation of the knowledge. These three factors, then, point out that ordinary knowledge is the result of a great amount of information which cannot be reduced to that strictly related to the modalities of acquisition and formation (perceptual and non-perceptual, see Section 1). As far the formulation of global ordinary knowledge is concerned, it is also necessary to point out its complexity, which in this case is due to its topicity/atopicity and therefore to the interrelations amongst various knowledge (the meanings of the words “topicity” and “atopicity” derive from the ancient Greek word topos, which means place, like the word “topology” in mathematics). The topicity/atopicity of ordinary knowledge shows that any ordinary knowledge is topic, since it is a carrier of specific thematic information, like, for instance, that related to the perception of an object or referring to a determined argument, like a political conception.

 

32

Chapter One

At the same time, it is also atopic, and this means that the thematic specific information is interrelated with other information which can be tied together in the large diversity of meaningful connections typical of any subject, for instance, those related to a political conception can be correlated to others related to anthropological or religious conceptions, or even to moral or psychological preferences. Topicity and atopicity derive from the fact that the formulation of global ordinary knowledge is at the same time sectorial and non-sectorial; on one hand, it is formed from various knowledge referring to specific ambits, and on the other, different knowledge is interrelated amongst itself, with many significant bonds, and so all knowledge considerably influences all other knowledge. For instance, the knowledge of a wide geographical area, like a town, is at the same time sectorial, referring to how the space is formed, and non-sectorial, related to non-geographical judgments and considerations, or to experiences relating to places in that area. In a similar way, for instance, psychological knowledge of others can be sectorial if it refers to their behaviors, but at the same it can be nonsectorial since they involve, for example, philosophical, political or religious ideas. For this reason, it has to be pointed out that there is often a continuous interrelation amongst different bits of ordinary knowledge, not only in mental processes like inferences but also in the formulation of different knowledge. From this perspective, global ordinary knowledge is at the same time formed by sectorial knowledge and by non-sectorial knowledge, and by continuous interrelations amongst them. The book of one’s own ordinary knowledge is a sui generis book, very different from the handbooks of history, geography, physics or chemistry; such a book could be considered a hypertext in which thematic sets of ordinary knowledge and a continuous and complex and significant process of feedback among them are contained. Ordinary knowledge is classified and organized in a “register” so that the mind can easily approach not only the knowledge, but also its previously formed correlations with other knowledge. For instance, in a subject involved in politics, all his political knowledge are placed on a register, and at the same time they have specific connections with others, like personal and social knowledge; in a different way, in another subject, similar knowledge of a political nature can be strictly correlated to the subject’s own religious convictions. At the moment of their formation or acquisition (see the distinction we have previously made, Section 1) the knowledge is classified and recorded in registers in which there is other knowledge of the same thematic nature. The various knowledge, therefore,

The Epistemological Structure of Ordinary Knowledge

33

except for in some specific cases, is never an isolate contents of the mind, but is part of groups or classes which contain knowledge of the same nature and also knowledge which may belong to different classes but which has been formed with some meaning to other classes. In this sense, ordinary knowledge of the different types which have been indicated is always an open structure. which can change its mutual interrelations and can therefore be modified under different conditions or in different periods of the history of each mind. These considerations lead directly to the claim that the book of ordinary knowledge is dynamic, not only with reference to the relations amongst the various thematic classes and the interrelations amongst knowledge of different types, but also with reference to their change over the history of the mind, and therefore of the self and of any subject. Ordinary knowledge as a whole and various specific knowledge are subject to a continuous dynamic, which may cause the strengthening of some knowledge and the modification of other knowledge, and also a web of significant ramifications between the various knowledge types. This topic is one of the most significant aspects of the structure of the ordinary knowledge of a subject, and in this paper is not analyzed, not only due to the intrinsic difficulty, but also because there are no results of specific research; in fact, each mind formulates its own system of ordinary knowledge for which, in theory, knowing the structure of the system and its way of operating is possible, starting from certain knowledge and noting that which is related to it. For instance, starting from the religious ideas of a subject it is possible to pinpont his ethic or aesthetic conceptions. In this paper, it is only pointed out that all global knowledge of a subject is formed by a structure constituted by a single bit of knowledge and by classes, different kinds of knowledge and influences and their interrelations, and this structure is dynamic, so it can change in certain conditions or over the history of each mind. Ordinary knowledge is not only dynamic, but it is possible to operate on it with awareness and intentionality. If on one hand ordinary knowledge changes itself autonomously, on the other, it can be intentionally modified. In general, it is possible to state that it is possible not only to know the ordinary knowledge contained in one’s own mind, but also to ascertain its dynamic and to operate on it. Operations on one’s own mind are many, and here we will only mention that they can be carried out according to three modalities. The first relates to single-topic knowledge, on which it is possible to operate intentionally and examine globally or according to its particular contents, as well as to modify them; the second modality refers to the relations

 

34

Chapter One

amongst different bits of sectorial/thematic knowledge or among knowledge of different types or sectors, for instance, knowledge relating to others can be correlated with moral knowledge on the grounds of modifications to ethical macro-structures. The third modality is related to the relevance assigned to the knowledge. In fact, the mental organization of ordinary knowledge is not only based on the thematic information or relations among the knowledge, but also on the relevance assigned to each of them: in other words, some knowledge is considered more relevant than other knowledge in any condition, or becomes so only in some specific conditions; usually, the knowledge considered the most relevant is that related to one’s own existence. From this perspective, any mind, and more widely any self, organizes its own knowledge in a gnosic structure on the grounds of three factors: thematicity, correlations and relevance. The knowledge is grouped by its thematic likeness (for instance, information referring to a person or to a place), on the correlations which have been previously formulated (for instance the relation between the knowledge of man and anthropological or religious conceptions) and on the grounds of its relevance: for instance, ethical, religious or professional knowledge will be considered more relevant by some people and less relevant by others. Every mind constructs its own gnosic structure, that is, the organization of all its knowledge (by acquisition or formation); it can be analogous for different minds, but at the same time is very different as far as relevance is concerned, which is specific to each mind. This aspect implies that ordinary knowledge cannot simply be considered a heap of differing knowledge, but is a very complex structure, made up of many sub-structures, which not only builds itself but can change during the course of its existence according to the dynamics and history of each mind. The formation of the gnosic structure is a fundamental topic to the analysis of ordinary knowledge, for which deep research is necessary to allow us to know how the human mind works, and particularly the minds of individual people, in organizing its own knowledge in order to have easy access to it and modify it and to be able to perform mental activities, to make choices and activate adequate behaviors related to one’s own existence; the organization of the gnosic structure is the result not only of autonomous mental processes but also of intentional and meta-mental activities.

The Epistemological Structure of Ordinary Knowledge

35

7. Conclusions This paper has presented a general model of ordinary knowledge. First of all, it has been stated that ordinary knowledge is formed by different types of knowledge; these types of knowledge are epistemologically different in the tools and methods they use, in their formulation and in the processes relating to their reliability, adequacy and control, and therefore their acceptance. More particularly, the epistemological difference between perceptual knowledge and self-knowledge of one’s own mind and self has been investigated; as part of this, some difficulties have been noticed relating to reliability and control. Moreover, it has been pointed out that, even if the mind operates in a topic and sectorial way to elaborate specific information, like perceptual information or that relevant to the space in which someone lives, at the same time, even if in a different measure, in any of the topic gnoseological processes, a great number of factors are involved, even if they are different from those typical of a topic elaboration. For instance, the elaboration of visual perceptions is not limited to the elaboration of visual information, but may involve other cognitive activities, like judgments or considerations on the subject of reference. Ordinary knowledge is at the same time topic and atopic. Furthermore, the ordinary gnoseological processes, similar to what happens for scientific activities, are always guided by mental paradigms, that is by concepts, visions of the world or other information which can influence in different ways the formation of specific knowledge, its consideration and its classification within the global gnosic structure of each mind. Ordinary knowledge not only formulates knowledge, but generates cognitive macrostructures formed by different kinds of information or by correlations among them. Ordinary knowledge, in its articulation of the different types of knowledge, is not a mere accumulation of knowledge, but is subject to a global organizational process which generates that which is called the gnosic structure of each mind. This is a dynamic structure, subject to continuous autonomous (or intentional) changes, so modifications can be carried out over the history of any mind, not only of single bits of information and their correlations, but also of its global organization. The ordinary knowledge of Homo, moreover, has a peculiar aspect which consists of the fact that any mind with meta-mental processes can operate intentionally and with awareness on itself in order to modify

 

36

Chapter One

single bits of information, correlations between information and aspects of the global organization of the gnosic structure. The analysis which has been presented and the model which has been formulated have pointed out how complex ordinary knowledge is. It has epistemological statutes and in many aspects is more complex than scientific and metaphysical knowledge, for example, the great number of factors which intervene in its formation, control, and even more in the use which is made of it in the daily life: ordinary knowledge, even if it has its own cognition, is formulated to satisfy needs, intentions and wishes and is guided by determined motivations. In this work, many arguments have been simply presented and not investigated in depth, not only because it was decided to formulate a general and schematic model, but also because there are no suitable neuroscientific results related to ordinary knowledge, its formulation or its mental organization. Therefore, only further neuro-scientific and psychological research will allow us to understand and explain the structure of ordinary knowledge, which, obviously, is part of a wider knowledge of the human mind.

References Bartsch, R., Dynamic Conceptual Semantics, CSLI, Stanford, 1998. Bianca, M., La costruzione della scienza, Franco Angeli, Milano, 1984. —. Capire la scienza, Le Monnier, Firenze, 1990. —. Rappresentazioni mentali e conoscenza, Franco Angeli, Milano, 2005. —. La mente immaginale, Franco Angeli, Milano, 2009 Douglas Fields, R., The Other Brain, Simon & Schuster, New York, 2009. Dretske, F., Knowledge and the flow of information, MIT Press, Cambridge, MA, 1981. Gallagher, S., Zahavi, D., The Phenomenological Mind, Routledge, London, 2012. Gertler, B., Self-Knowledge, The Stanford Encyclopedia of Philosophy (Spring 2011 Edition), Edward N. Zalta (ed.). Hardin, R., How do you Know? The Economics of Ordinary Knowledge, Princeton University Press, Princeton, 2009. Lars-Gunnar, L. The mind considered as a system of meaning structures: Elementary meaning structures, Scandinavian Journal of Psychology, 22, 1981, 145-160. Seung, S., Connectome: how the brain’s wiring makes us who we are, Harcourt, New York, 2012.

The Epistemological Structure of Ordinary Knowledge

37

Rogers, T.T., J.L.McClelland, Semantic Cognition, MIT Press, Cambridge MA, 2004. Rookes, P., J.Willson, La percezione, Il Mulino, Bologna. Sheldrake, R., Mind, Memory and Archetype. Morphic Resonance and the Collective Un conscious, Psychological Perspective, 18(1), 1987, 9-25. Tononi, G., Consciousness and the Integration of Information in the Brain, Consciousness. At the Frontiers of neurosciences, 77, 1998, 245-280. Wolfe J.M., (ed.), Sensazione e percezione, Zanichelli, Bologna, 2007.

 



CHAPTER TWO KNOWING (EXISTING) ORDINARY OBJECTS MASSIMO DELL’UTRI

1. What the Reader Will Find Despite having a title that sounds epistemological, this paper is first and foremost an attempt to shed light on the coordinates along which a plausible form of ontology may be illustrated. In what follows the reader will find a way to delineate a view of the kind of objects that possess full citizenship in our world, paying particular attention to ordinary objects. However, this is not a “philosophically neutral” enterprise, free of any assumption whatsoever. Every effort at delineating our ontology avails itself of a criterion, even if implicitly. In this paper a criterion based on the notion of cause will be put forward. But this notion is susceptible to more than one interpretation, and the scope of our ontology may change according to the interpretation one makes. Since the idea according to which ordinary objects exist and are therefore fully knowable is a core element of common sense, and common sense is in turn a core element of pragmatism, this paper finds its place in a broad pragmatist framework. Because of this framework, one of the paper’s assumptions is that metaphysics and ontology, on the one hand, and epistemology, on the other, are strictly intertwined.

2. A Criterion for Ontological Membership It goes without saying that you can know an object only if it exists. Therefore, before tackling the epistemology of ordinary knowledge, one’s first task should be to find an answer to the question “Do ordinary objects exist?”—equivalently, “Do they belong to our ontology?” In order to accomplish this task we need a criterion for ontological membership, and my suggestion is to exploit what has been called Alexander’s Dictum: “To

40

Chapter Two

be real is to have causal powers.”1 The criterion is named after the British philosopher Samuel Alexander, born in Sydney, Australia, in 1859, philosophically trained in Oxford, and Professor at the University of Manchester from 1893. He never actually gave the wording to this criterion, though. It was Jaegwon Kim who, reasoning about the reality of mental properties, 2 distilled the dictum from the following passage, in which Alexander curtly dismisses epiphenomenalism. The latter, he claims, supposes something to exist in nature which has nothing to do, no purpose to serve, a species of noblesse which depends on the work of its inferiors, but is kept for show and might as well, and undoubtedly would in time be abolished (Alexander 1927, 8).

It has however been noticed 3 that the wording of the Dictum is inadequate. In fact, the equivalence it establishes between existing and causing (having causal powers) is a partial description of what can be taken as being. It is as if only causes existed, but not their effects. Nevertheless, the remedy is ready at hand. It is provided, for example, by Plato in the following passage: I suggest that anything has real being that is so constituted as to possess any sort of power either to affect anything else or to be affected, in however small a degree, by the most insignificant agent, though it be only once. I am proposing as a mark to distinguish real things that they are nothing but power (Plato, Sophist, 247e).

The things that populate our ontology are thus the ones that possess the power to affect and/or to be affected. Consequently, the tenet we may now distil from the latter quotation is the following: “To be real is either to cause or be caused.” Following Eugene Mills, I will dub it the Stranger’s Dictum. In fact, in the dialogue it is the Eleatic Stranger that utters these words, and—given the difficulty in attributing to Plato the idea the words express—it is advisable to do so. Hence, the Stranger’s Dictum allows us to answer in the affirmative the question we raised at the beginning of this section: since ordinary objects cause countless effects and are caused in a lot of ways, they exist. As the reader will see below, I believe that many other things can be taken to be existent on the basis of the Stranger’s Dictum, and that this in turn 1

I would like to thank Achille Varzi for bringing the Dictum to my attention. See Kim 1993, 348. 3 Mills 2003, 105. 2

Knowing (Existing) Ordinary Objects 

41

allows us to describe a multilayered reality. However, in order to address the topic of this volume let us stick to the level of ordinary objects and conclude that this level exists. It obviously exists, we may add, as the belief in its existence is a central feature of common sense. Moreover— sceptics apart—who feels seriously inclined to deny it?

3. The Reductionist Objection Someone is. She or he might claim that common sense is not sufficiently refined, and therefore does not manage to capture the causal relations actually present in the world. What matters—she or he would add—is what science tells us, what can be proved to exist thanks to the technological equipment our best scientific theories (the best available theory in physics for example) guide us to devise. This is a “reductionist” objection in that it reduces the phenomenological complexity of the world to the level analysed by the natural sciences. To use a distinction Wilfrid Sellars made famous, it is the scientific image that we should appeal to— not the manifest image—if we want to gather a plausible picture of the furniture of the universe. Sellars, by the way, was not a reductionist: he argued for a fusion of the two images in a sort of stereoscopic view which would “transcend the dualism”4 between them. To grasp the idea behind the objection we may refer to the well-known “two tables” the physicist Arthur Eddington envisaged in his 1927 Gifford Lectures at the University of Edinburgh. He started by claiming that there are duplicates of every object around us, so he was actually sitting at two tables: the former is quite familiar, being the type of commonplace object we are accustomed to from birth; the latter is a more recent acquaintance, something introduced to us by physics. And what physics tells us in introducing table number two is that it is nearly all empty space in which numerous electric charges are rushing about with great speed. 5 The rationale behind the objection, then, is that if we want to find the entities which possess real causal power we have to look at the myriad of microparticles swarming around within the microlevel studied by natural science—the macrolevel being only an appearance, devoid of any scientific interest.6 Therefore, if these are the real causes existing in the

4

Sellars 1962, 40. See Eddington 1928, Introduction. 6 I do not want to attribute this contention to Eddington. Even if he showed a clear inclination toward it, he did not take anything for granted and engaged in a deep critical discussion about the relationship between the two pictures of the world. 5

 

42

Chapter Two

world, then the Stranger’s Dictum cannot but prompt us to say that ordinary objects do not exist. But—we would like to ask—why should one be inclined to uphold such a reductionist stance?

4. Non-Humean Causation There may be different answers to this question depending on the aims and interests implicit in those who ask the question itself. If our interest is metaphysical and our aim is to discover the broad metaphysical framework that this reductionist view can be placed within, I think we may say that behind this view there is a strong realist inclination toward causation. This “strength” is due to the fact that the kind of realism involved here is radically non-epistemic, i.e. characterised by a conception of the world as completely independent of our cognitive abilities—a world so cognitively detached from the human mind that a description of the world and its relation to the mind could be given only from a God’s eye view. Only a superhuman creature could be in the right position from which to survey how things—in the broadest sense of the term—stand. Despite the dichotomy the supporter of this view sees between common sense and science, and despite her or his inclination to favour the latter, I think he is moved by a commonsensical (and wrong) intuition, according to which causation is something completely independent of the human cognitive sphere, something humans are not but subjected to. If David Hume saw an unavoidable human trace in causation, taking it as eminently psychological, what invigorates the reductionist stance we are talking about is non-human non-Humean causation, to rephrase a well-known Quinean pun. It is a relation embedded in the metaphysical structure of the world considered eminently non-human, the only world which is of interest to the reductionist philosopher: a world which justifies talk about dichotomies such as objective-subjective, mind-world, reality-appearance and factvalue. Is this a plausible conception of causation? The answer I am going to give is twofold. For one thing, I believe this conception has as little bite as the realist attitude which lies behind it. What I am alluding to is closely connected with what has been called “the Kantian problem”, a problem telescoped into the following question: “How can we obtain cognitive access to mindindependent objects?” This is an intrinsically unsolvable problem, given its dependence on the very little perspicuity of posing the concept of something which by definition is beyond the reach of the human mind,

Knowing (Existing) Ordinary Objects 

43

like the concept of the “thing in itself”. Thinking that there is a way the world is in itself, a way based on a purported essential and “fixed in advance” structure of the world, and a “fixed in advance” way to correspond to the human mind, is altogether pointless, given the impossibility of gaining a clear picture of the entire situation—even in principle. Only a superhuman creature, as I mentioned above, could grasp it. It seems that entertaining such a metaphysical belief amounts to a refusal to accept ourselves as we really are, hence an impossible effort to deny the intrinsic limits which constitute our nature—to deny the human condition itself. This has been brilliantly expressed by two leading neo-pragmatists, Richard Rorty and Hilary Putnam. The former has made it clear that we will always be held captive by some picture or other, for this is merely to say we shall never escape from language or from metaphor—never see either God or the Intrinsic Nature of Reality face to face (Rorty 1994, 80),

whereas the latter has claimed that elements of what we call “language” or “mind” penetrate so deeply into what we call “reality” that the very project of representing ourselves as being “mappers” of something “language-independent” is fatally compromised from the very start. […] Realism [with a capital “r”] is an impossible attempt to view the world from Nowhere (Putnam 1990, 28).

So the weakness of the radical non-epistemic conception of realism reverberates within the conception of causation we detected behind the reductionist stance. But—and this is the other part of my answer to the question about the purported plausibility of the non-Humean conception of causation—nothing beyond the radical non-epistemic conception of realism and an unthinking product of common sense could justify nonHumean causation. To see this, let us turn to an alternative current of thought about causality.

5. Relevance and Causation One of the leading authors in this school of thought is Ludwig Wittgenstein. In fact, he never held that causation is a relation we could find “out there”, so to speak, ready to be grasped via experience or the like. “There is no causal nexus” and “A necessity for one thing to happen because another has happened does not exist. There is only logical necessity” are typical expressions of his stance toward causation, to be

 

44

Chapter Two

found in the Tractatus Logico-Philosophicus (Sections 5.136 and 6.37, respectively). The kind of logical necessity referred to in the second quotation was to be dropped by the end of the 1930s, when he began to think that the “necessity” of the causal nexus depends on our practice—not on a nonexistent logic of the world. There is no cause-effect relation, there are no causal facts, yet we cannot help thinking of the world as causally structured. Strikingly anticipating a theme of the last period of his philosophical reflection, Wittgenstein started to argue for the idea according to which, in actual fact, in our verbal and non-verbal practice we simply do not doubt that the world possesses such a structure, even though it does not actually exist. That the world is causally structured constitutes part of that certainty which lies at the bottom of our conceptual scheme and makes our lives possible.7 While we are on the topic, I would like to draw attention to a passage which is related to the difference we mentioned above—that between the micro- and macrolevels of reality, where according to the reductionist line only the former has real causal powers. Wittgenstein stresses that the mere fact that there is this distinction cannot individuate relations between objects at the microlevel as actual causal ones and the relations present between objects at the macrolevel as depending on the microlevel for the bringing about of phenomena. His point is that any proof to the effect that, on the one hand, there is an upward direction of causation and, on the other hand, the link between the levels is not accidental, is still lacking. Imagine we had seeds which came from two plants. We find that the two plants have exactly the same seed. But if it comes from a poppy, the seed produces a poppy again; if it comes from a rose it produces a rose. We are then inclined to say there must be some difference in the seed itself. But if we did discover a difference, we should not know whether it was a relevant difference (whether it affected the outcome) (Wittgenstein 1937-38, 433).

Its sketchiness notwithstanding, Wittgenstein’s argument helps to highlight the idea that the macrolevel is after all autonomous, and does not ontologically and epistemologically depend on a more basic level. The other thinker I would like to mention as representative of the same point of view about causation is Hilary Putnam. Putnam relates to a concept criticised by John Stuart Mill, i.e. the concept of total cause. Far 7

In the Introduction to the Italian edition of Wittgenstein 1937-38 Alberto Voltolini convincingly maintains that this is a Kantian trait on Wittgenstein’s part, not a Humean one.

Knowing (Existing) Ordinary Objects 

45

from having a cognate in ordinary language, the concept of total cause reveals itself to be nothing but an idealisation, given that it presupposes, for one thing, that “cause” is physically definable and, for another, that a cause is a necessary and sufficient condition for its effect (in a deterministic universe). However, according to Putnam, nothing could be further from the truth. What usually happens is that we see a (total) cause as being divided into parts and pick out only the part of interest that we deem to be “the” cause, pushing the rest into the background. What obviously follows on from this is that “one man’s […] ‘background condition’ can easily be another man’s ‘cause’” (Putnam 1983, 214). With further consideration, we may realise that “causes” […] is often paraphrasable as “explains”. It rarely or never means “is the total cause of”. […] “Causal powers” are properties that explain something, given background conditions and given standards of salience and relevance (Putnam 1983, 215).

This last quotation leads us directly to the point. Standards of salience and pertinence are inseparable from the process through which we come to detect the cause of an event, and this in turn means that nothing can count as a cause if it does not speak to us, as it were. Causes have an unavoidable human trace, since “salience and relevance are attributes of thought and reasoning, not of nature” (Putnam 1983, 215), not of a purported noumenal structure of the world we would never know. This kind of human cause is causation enough—we need look no further. Causation […] certainly exists in our “life world”. […] The world of “ordinary language” (the world in which we actually live) is full of causes and effects. It is only when we insists that the world of ordinary language (or the Lebenswelt) is defective (an ontological “jungle”, vague, gappy, and so on) and look for a “true world” (free of vagueness, of gaps, of any element that can be regarded as a “human projection”) that we end up feeling forced to choose between the picture of “a physical world with a built-in structure” and “a physical universe with a structure imposed by the mind” (Putnam 1984, 89).

That causation amounts to a deeply human relation is something Sellars seems to agree with when urging us to harmonise the scientific image with the manifest image in a stereoscopic view. Here is what he claims at the very end of his “Philosophy and the Scientific Image of Man”: to complete the scientific image we need to enrich it not with more ways of saying what is the case, but with the language of community and individual intentions, so that by construing the actions we intend to do and the

 

46

Chapter Two circumstances in which we intend to do them in scientific terms, we directly relate the world as conceived by scientific theory to our purposes, and make it our world and no longer an alien appendage to the world in which we do our living (Sellars 1962, 40).

It seems to me that his insistence on making the world become our world is no less than an acknowledgment that the only world that exists is the Lebenswelt Putnam refers to—in the wake of Edmund Husserl, of course. Furthermore, it seems to me that from all this we may draw the following schematic moral—a moral endowed with an inherently pragmatic flavour. All in all, what the thinkers we have mentioned thus far urge us to do is to abolish three dichotomies: the one between scientific and manifest images (Sellars), the one between genuine reality and human projection (Putnam), and the one between ordinary and deep language (Wittgenstein).

6. What to Make of All This It is the latter threefold abolishment that brings to the fore the pragmatist network we have mentioned—the network the present essay is placed within. In fact, a network of this kind urges us to give philosophical priority to the verbal and non-verbal practices our concrete lives are immersed in. And this, coupled with the claim that in our lives we cannot afford an objective unassailable certainty, leads us to subscribe to the epistemological doctrine of fallibilism. Here we have the real basis which allows us to conclude that reductionism is implausible, that there is no layer of reality to be taken as metaphysically privileged, and that any kind of “strong” realism is untenable. Thanks to the Stranger’s Dictum, therefore, this basis justifies the multi-faceted ontology we have been alluding to, an ontology which comprises, for instance, things such as ideals and values—ethical, aesthetical, or religious—provided these things cause people to display specific patterns of behaviour and are revised by people in particular ways. Above all, our ontology features ordinary objects, given the central role common sense plays in the ontological enterprise. Thus, having been cleared on the ontological level, common sense can stake its claim to the epistemological level as well and state that ordinary objects are susceptible to genuine ordinary knowledge, however wobbly it may be. This gives the present essay a kind of circular progression, since in a sense we have arrived where we started—but it is a virtuous progression, not a vicious one, allowing us to “know the place for the first time”, as T.S. Eliot’s Little Gidding puts it.

Knowing (Existing) Ordinary Objects 

47

Again, nobody would deny this—barring a radical sceptic or a foundationalist in epistemology. But it is here that fallibilism shows all its stature, giving suitable support to maintain that empirical knowledge, like its sophisticated extension, science, is rational, not because it has a foundation but because it is a self-correcting enterprise which can put any claim in jeopardy, though not all at once (Sellars 1956, 170).

References Alexander, S. 1927, Space, Time, and Deity, London: Macmillan, vol. 2 Eddington, A.S. 1928, The Nature of the Physical World, Cambridge: Cambridge University Press. Kim, J. 1993, Supervenience and Mind. Selected Philosophical Essays, Cambridge: Cambridge University Press. Mills, E. 2003, “Introduction”, Topoi 22 (2), Causal Criteria of Reality, E. Mills (ed.), 105-109. Putnam, H. 1983, “Why There Isn’t a Ready-Made World”, in H. Putnam, Realism and Reason, Philosophical Papers, Vol. 3, Cambridge: Cambridge University Press, 205-228. —. 1984, “Is the Causal Structure of the Physical Itself Something Physical?” in Putnam 1990, 80-95. —. 1990, Realism with a Human Face, Cambridge (MA): Harvard University Press. Rorty, R. 1994, “John Searle on Realism and Relativism”, in R. Rorty, Truth and Progress, Philosophical Papers, Vol. 3, Cambridge: Cambridge University Press, 1998, 63-83. Sellars, W. 1956, “Empiricism and the Philosophy of Mind”, in W. Sellars, Science, Perception and Reality, Atascadero (CA): Ridgeview Publishing Company, 1963, 127-196. —. 1962, “Philosophy and the Scientific Image of Man”, in W. Sellars, Science, Perception and Reality, Atascadero (CA): Ridgeview Publishing Company, 1963, 1-40. Wittgenstein, L. 1922, Tractatus Logico-Philosophicus, London: Kegan Paul, Trench, Trubner & Co. —. 1937-38, “Cause and Effect: Intuitive Awareness”, Philosophia 6 (34), 409-425; “Appendix A: Immediately Aware of the Cause”, 429434; “Appendix B: Can We Know Anything but Data?” 435-445.

 



CHAPTER THREE NEOIDEALISM AND NEOREALISM MAURIZIO FERRARIS

I define “idealism” as the philosophical thesis that: (1) The external world exists as depending on some representation Obviously, this argument can be articulated in many ways, which we will examine in detail. In fact these are all variations of transcendentalism and its constitutive misunderstanding, i.e. the "transcendental fallacy": the confusion between ontology (what there is) and epistemology (what we know, or think we know). It is a very natural confusion, something very similar to the “stimulus error”, in which, after closing their eyes and being asked what they see, someone says “nothing” (while the truth is that they are seeing phosphenes, consecutive images and so on). The subject is not giving a description, s/he is proposing a naive theory of vision: the eye is like a camera, so when the lens is closed there is nothing or, at most, perfect darkness. From this point of view, the boutade that Ramses II did not die of TB because Koch isolated the tuberculosis bacillus only in 1882 is but a magnified variation of the stimulus error. Let us examine the ways in which this fallacy is articulated.

1. Correlationism The first and most generic articulation is correlationism, 8 according to which: (2) The external world only exists in correlation with a knowing subject 8

Meillassoux 2006.

Chapter Three

50

This thesis obviously raises problems. Meillassoux, who coined the term, posits – through the argument of the ''arche-fossil" – that the main problem of correlationism is to explain the existence of objects like fossils, proving the existence of a past prior to the appearance of any I think. However, I do not think that is a conclusive argument, since it is not inconceivable that God created us two minutes ago with all our memories. The real problem to which correlationism cannot respond is solipsism. If correlationism were true, we would have at least as many mental worlds as there are subjects, and the passage from one mental world to another would be inexplicable. In fact, the world of correlationism is made up by monads: subjects representing objects within themselves. This, ultimately, is the world represented by Kant in his refutation of idealism. So, rather than in pre-existence, correlationism finds its fundamental obstacles in interobservation 9 and interrelation: 10 different subjects can observe the same object; different objects, humans and other living things can interact in the same space. To explain these circumstances with correlationism, one would have to invoke a pre-established harmony. It seems much more likely to believe that interaction is guaranteed by the properties of objects and other human and animal agents, which are independent of the correlation. Since the idealist cannot think that interaction is ensured by the objects, she has to argue that it is made possible by the subjects – indeed, by the single subject. In this sense, correlationism evolves into constructivism.

2. Constructivism The fundamental thesis of constructivism11 is that: (3) The external world is constructed by the subject Saying that the external world is constructed by the subject means claiming that it is dependent on the subject. Constructivism therefore has to clarify what this dependence is.

3. Causal dependence Causal dependence is an ontological dependence in which: 9

Bozzi 1990. Ferraris 2001. 11 Boghossian 2006. 10

Neoidealism and Norealism

51

(a) The external world is causally dependent on the subject That is, the subject is the cause (the creator) of the world. Strange as it may seem, this thesis was supported in full seriousness by the Italian neoidealist Giovanni Gentile: “Berkeley in the beginning of the Eighteenth century expressed very clearly the following concept. Reality is conceivable only in so far as the reality conceived is in relation to the activity which conceives it, and in that relation it is not only a possible object of knowledge, it is a present and actual one. To conceive reality is to conceive, at the same time and as one with it, the mind in which that reality is represented; and therefore the concept of a material reality is absurd.”12 Let us analyse this sentence. It claims that “Reality is conceivable only in so far as the reality conceived is in relation to the activity which conceives it”, which is tautological: in order to think, I have to think. Then Gentile goes on to say that only in relation to the thinking subject can reality be “not only a possible object of knowledge” but “a present and actual one”. Hence, with a perfect non sequitur, he gets to the conclusion that “To conceive reality is to conceive, at the same time and as one with it, the mind in which that reality is represented; and therefore the concept of a material reality is absurd.” This is absurd, because it assumes that only what is actually present in my thought is real – which classifies as unreal anything that is not actually present in my thought. If this were true, there would be openly absurd consequences. There would be no difference between introspection and knowledge of the outside world. All things past, from dinosaurs to the Sumerians, would be present exactly like the thoughts that think of them. All things future would be no less present than things past (and therefore there would be no difference between possible and actual). Everything Giovanni Gentile ignored (and there is good reason to believe that there were few things he did not know) would be non-existent; on the other hand, all that Giovanni Gentile thought of would exist, including Pegasus. I say “Giovanni Gentile” without including other individuals because the only subject of which, according to his own starting assumption, Giovanni Gentile had knowledge (and therefore existed) was Giovanni Gentile. Since Giovanni Gentile ceased to exist on April 15th, 1944, there is no longer anything – in fact, to be exact, nothing ever existed, including Giovanni Gentile.

12

Gentile 1916, my emphasis.

 

Chapter Three

52

One would be happy to leave such theses to philosophical folklore, as they seem to confirm the widespread view that antirealism only exists in the mind of realists (thus, according to Gentile, but only according to Gentile, it exists). However, the idea that being depends on thought was revived in perfect seriousness by Foster,13 as part of a rehabilitation of Berkeley's idealism, and by Severino, 14 as part of a rehabilitation of Gentile's idealism. As for the latter, here is what he writes: "To refute idealism, Ferraris calls for the existence of the infinite things that existed before man, the obstacles encountered by mankind, the unpredictability of events. The idealist answers, rightly, that one could not speak of all these situations unless they were thought of and that therefore they are not beyond thinking, they are not independent from it. Thought includes in its content the very human individuals who are born and die." And then: "If 'realism' means that certain entities may exist even if there were no mankind, then realism is a form of nihilism (i.e., a self-contradictory thesis) - like idealism." Given that "man" is a sum of entities many of which have existed since before the concept of "man" was drawn up (and since Severino existed even before coming across the concept of "man" at some point in his childhood), it seems really difficult to see the existence of the universe as dependent on the concept of "man".

4. Conceptual Dependence To avoid the misinterpretations related to causal dependence, idealists sometimes speak of "conceptual dependence": (b) The external world is conceptually dependent on the subject This is one of the possible outcomes of Kant’s famous statement: “Intuitions without concepts are blind.” Which may, however, be interpreted in two ways: (1) the weak form, in which, without the concept of “dinosaur”, we would not recognise a dinosaur if we saw one, and (2) the strong form, in which, without the concept of “dinosaur”, we would not see a dinosaur if we saw one. When it comes to defending Kant, it is usually said that he meant conceptuality is reconstructive of experience in general. But had he meant this, he would not have written the Critique of 13 14

Foster 2008. Severino 2012.

Neoidealism and Norealism

53

Pure Reason, only the Metaphysical Foundations Of Natural Science, namely, a theory of science. If he wrote the Critique of Pure Reason, it is because he meant conceptuality is constitutive of experience in general. 15 Strong conceptual dependence (ontological dependence) can be traced back to causal dependence and is subject to the same criticism. Weak conceptual dependence (epistemological dependence) is not a real dependence. In fact, it is trivially false to state that Tyrannosaurus Rex depends on our conceptual schemes, just as it is trivially true to state that the name “Tyrannosaurus Rex” depends on our conceptual schemes. It is also trivially true to argue that the name "Tyrannosaurus Rex" is very useful for our knowledge of the Tyrannosaurus Rex. But it is no less trivially true to argue that if there is one thing that Tyrannosaurus Rex has never known, it is that its name was "Tyrannosaurus rex" – which did not prevent it from having exactly all the features of Tyrannosaurus Rex.

5. Representational Dependence This issue emerges very clearly in the weak conceptual dependence that Rorty calls "representational dependence”: “None of us antirepresentationalists have ever doubted that most things in the universe are causally independent of us. What we question is whether they are representationally independent of us.”16 That is to say: (c) The external world representationally depends on the subject However, either "representational dependence" means that the name "dinosaur" depends on us, and then it is not a serious dependence, or it means that a dinosaur's being depends on us. But then, given that when there were dinosaurs we were not there, when there were dinosaurs (paradoxically) there were no dinosaurs. If we try to give concrete shape to representational dependence, we will realise that the technical term hides a conceptual confusion between ontology (what there is, which is independent of our representations) and epistemology (what we think we know, which may depend on our representations - still it is not representations that make statements true, but that to which the representations relate). At this point, perhaps an antirealist might argue: how can you prove the existence of a dinosaur independent of thought? But the answer is 15 16

Ferraris 2004. Rorty 1998.

 

54

Chapter Three

simple: "The burden of proof is on you. You should be the one who proves the dinosaur’s dependence on thought, which is what, so far, you have not done.”

6. Culturalism Contemporary neoidealism rests on precisely this fake dependence. De facto, representational dependence is merely related to nomenclature: it asserts that the names of the known objects depend on the knowing subjects. However, neoidealism also asserts that this dependence is an authentic dependence, because it is a cultural dependence. In this sense, neoidealism is a form of culturalism, whose thesis is that: (4) The external world culturally depends on the subject All this shows that neoidealism starts from a completely inconsistent assumption. In representational dependence, epistemology is nomenclative with respect to ontology. It does not seem very interesting to know that the word "dog" refers to a dog, and most of all it does not seem like a philosophical thesis. And yet, from this very obviousness started the most widespread form of twentieth-century idealism, i.e. culturalism. Typically, culturalism is manifested in theses like "being that can be understood is language" (where "being", "understanding" and "language" are used in ways so vague that they can mean anything) or "there is nothing outside the text" (which Derrida later tried to present as equivalent to a sentence of a very different kind: "there is nothing outside the context"). In fact, culturalist theses seem very challenging from an ontological standpoint, because they appear to configure the world's causal dependence on language, texts and the world of culture. However, based on what we have just seen, this is not a causal dependence, but a representational dependence. One cannot speak, then, of dependence in any serious sense of the term: it is a pseudo-dependence presented as dependence. However, that is apparently enough to give a status to culture without going manifestly against common sense.

Neoidealism and Norealism

55

7. Antirealism The thesis of causal dependence and conceptual dependence is supported by classical idealism. But, as we have seen, it is poorly practiced, even though it still has some followers. Surely, neither Gadamer, nor Rorty, nor Derrida, nor Vattimo would ever argue that dinosaurs depend on us (and in fact they criticise Gentile and Severino). But, then, what do proponents of culturalism argue? What is their ontological commitment? Culturalism does not clearly commit ontologically to anything, but acts as if this (antirealist) ontological commitment were obvious. As Harman noted,17 this strategy usually manifests itself in asserting that the issue of realism or antirealism is surpassed, or that one should look for a third position beyond realism and antirealism, which is in fact the most common strategy in debates on new realism.18 However, while the explicit ontological commitment is weak, the implicit ontological commitment is very strong, and consists of the theses: (5) There are not a subject and an object; there is the relation between subject and object Those supporting the thesis that "There are not a subject and an object, there is only the relation between subject and object" should be asked to show clearly where the subject ends and the object begins. If they answered with a punctual statement (for instance, that the subject ends at the level of the looking eye, while the object is what is being looked at by the eye) this would prove that the previous statement is false. It is not true that "There are not a subject and an object, there is only the relation between subject and object", because the answer (be it right or wrong) shows that in principle it is possible to distinguish the subject from the object. If instead – as is much more likely – they replied with something indistinct, restating the indistinguishability of object and subject, then they should explain why they keep talking about objects and subjects. In fact, it follows from this thesis that they can only know a subject, their own, in a radical form of solipsism. Of course, this happens when you consider the thesis in the ontological form, that is, precisely in the formulation "There are not a subject and an object, there is only the relation between subject and object". On the 17 18

Harman 2014. Scarpa 2013.

 

56

Chapter Three

contrary, if one simply claimed that knowledge is the relation between a knowing subject and a known object, in which the main interest is directed to the latter, one would affirm something totally obvious and understandable. What appeared to be a bold ontological theory proves to be an epistemological platitude, a principle of common sense that nobody would ever deny, but that, because of this, is really not very interesting. Things do not change if you reformulate the antirealist thesis in a more sober fashion: (6) There are not ontology and epistemology, there is only the relation between ontology and epistemology Far from being ontologically modest, this thesis is extremely binding (and ultimately no less binding than that advocated by causal dependence). This statement describes the cognitive process, the one by which a person knows an object (epistemology) as what is known (ontology), which, in hindsight, collapses the distinction between knowledge of the outside world and introspection. It does not take into account that between being and knowing there is an essential asymmetry, which is constitutive of being as much as of knowledge. In fact, without knowledge there would still be whole regions of being (all natural objects and all social objects), while without being, there would be no form of knowledge, which is always knowledge of something

8. New Realism If this is how things are, new realism simply assumes that the observer observes in the observed something different and independent of the observer, otherwise what s/he is doing is not observation, but introspection. For new realism, snow is white if and only if snow is white regardless of the sentence "Snow is white if and only if snow is white". It is banal to observe that knowledge of reality is the result of a constructive process. Instead, it is trivially false to say that reality is the result of a constructive process. The new realist thesis therefore sounds like: (7) Knowledge is knowledge of something different and independent of knowledge, otherwise it is not knowledge

Neoidealism and Norealism

57

Obviously, at this point it all comes down to answering the question: "How do we know what is there?" But that is a different question from the ontological question, which is, and remains: "What is there?"

References Bozzi, P., Fisica ingenua, Garzanti, Milano 1990. Ferraris, M., Il modo esterno, Bompiani, Milano 2001. Ferraris, M., Goodbye Kant! What Still Stands of the Critique of Pure Reason, SUNY Press, New York 2014. Foster, J., A World for Us: The Case for Phenomenalistic Idealism, Oxford, Oxford University Press, 2008. Gentile, G., Teoria generale dello spirito come atto puro, 1916 (ed. Le Lettere 2003). English translation: Theory of Mind as Pure Act, London, 1922. Harman, G., (2014), “Foreword” to Ferraris, M., Manifesto of New Realism, SUNY Press, New York 2014. Marconi, D., “Realismo minimale”, in M. De Caro – M. Ferraris, eds., Bentornata realtà, Einaudi, Torino 2012113-137. Meillassoux, Après la finitude, Seuil, Paris 2006. Rorty, R., “Truth and Some Philosophers”, in Id., Truth and Progress: Philosophical Papers, Cambridge UP, Cambridge 1998. Scarpa, R., Caso nuovo realismo. La lingua del dibattito filosofico contemporaneo, Milano-Udine, Mimesis 2013. Severino, E., “È il crepuscolo delle tradizioni”, Il Corriere della sera, 15 November 2012.

 



CHAPTER FOUR LEBENSWELT, THE WORLD OF OBJECTIVE KNOWLEDGE AND PRACTICE: TOWARDS A PHENOMENOLOGICALEINSTEINIAN GAZE FABIO MINAZZI

«Denn was ist der g e s u n d e V e r s t a n d ? Es ist der g e m e i n e V e r s t a n d, so ferm er richtig, urteilt. Und was ist nun der gemeine Verstand? Er ist das Vermögen der Erkenntnis der Regeln in abstracto ist. So wird der gemeine Verstand die Regel: dass alles, was geschieht, vermittelst seiner Ursache bestimmt sei, kaum verstehen, niemals aber so im allgemeinen einsehen können. […] Gemeiner Verstand hat also weiter keinen Gebrauch, als so fern er seine Regeln (obgleich dieselben ihm wirklich a priori, und unabhängig von der Erfahrung einzusehen, gehört vor den spekulativen Verstand, und liegt ganz ausser dem Gesichtskreise des gemeinen Verstandes.» Immanuel Kant, Prolegomena zu einer jeden künftigen Metaphysik, die als Wissenschaft wird auftreten können [1783, Auflösung der allgemeinen Frage der Prolegomena 369-370]

1. Einstein's Conception of Knowledge In a letter dated May 7th, 1952 from Albert Einstein (1879-1955) to his lifelong friend Maurice Solovine (1875-1958), 19 the physicist from Ulm

19 See A. Einstein, Lettres à Maurice Solovine, Gauthier-Villars, Paris 1956. It should be remembered that in 1903 Einstein, together with Solovine and Konrad Habicht, founded the “Akademie Olympia”, a small and restricted circle in which, in a more convivial atmosphere, the three friends were accustomed to reading and discussing such classics as Hume, Mach, Spinoza, Riemann and Poincaré.

60

Chapter Four

traced the following drawing to illustrate his epistemological image of human knowledge: Fig. 1.

Line E indicates the plane of the Lebenswelt, or, as Einstein wrote, the plane of “immediate experiences”, which forms the common sense world in which all men are immersed – including Nobel Prize winners! – in the course of their lives. But this plane, in addition to indicating the set of experiences peculiar and specific to “common sense”, also indicates the set of experiences of a man in the course of his life. Therefore, every man, at least with respect to the plane of the Lebenswelt, not only finds himself inserted in it inasmuch as he has been literally “catapulted” into it with his birth (being thus in a certain historical moment, in a given environment, in a given historical/social/linguistic/civil, etc. society), but he also constitutes a highly specific and autonomous element in it, since every man, always with respect to this plane of common sense, constitutes, in his own actual concrete individuality, a history and a particular process. Each of us coincides with his own history, because, to quote the words of Hegel, was der Mensch tut, das ist er: man is the sum of his actions, his practice of what he has done; in short, of his history. Above this horizon constituting the Lebenswelt, Einstein then outlines the arc of a parabola which “floats” freely above the level of common sense, enabling us to reach some of the axioms (A), from which we draw conclusions (S, S ', S ... n). “Psychologically,” explains Einstein, “the As are connected to the Es. But there is no logical route leading from the Es to the As, but only an intuitive connection (psychological), which is always ‘re-turning’.” From the axioms A are derived, by a strictly deductive path, the consequences (S, S’… Sn) “that can claim to be true”.

Lebenswelt, The World of Objective Knowledge and Practice

61

Last but not least, the Es are correlated with the Lebenswelt through experimental experience, mediated, of course, by the technological dimension. “Closer examination,” specifies Einstein again, “shows that this procedure belongs also to the extralogical (intuitive) sphere, for the relation between the notions show up in S and the immediate experiences are not logical in nature. But the relation between Ss and Es is (pragmatically) much less certain than the relation between the As and the Es. (Take the notion ‘dog’ and the corresponding immediate experiences.) If such a relation could not be set up with a high degree of certainty (though it may be beyond the reach of logic), logical machinery would have no value in the ‘comprehension of reality’ (example: theology). What this all boils down to is the eternally problematic connection between the world of ideas and that which can be experienced (immediate experiences of the senses).”

In my opinion, this powerful Einsteinian image of scientific knowledge is particularly interesting for a number of reasons, in the first place because it posits a bond, direct and fruitful, between the plane of the Lebenswelt and that of the more abstract theoretical construct (represented by the axioms A and the consequences that can be derived from them by a strictly deductive method). Moreover, if, secondly, one looks at Einstein’s scheme as a whole, one can clearly “see” how human knowledge is constructed precisely to the degree that one is capable of establishing a quite specific and very special bond between the plane of Lebenswelt and the plane of abstract thought, or the Logos. Nor is this all, because, thirdly, the very presence of that segment of the parabola that, floating above the line of the Lebenswelt, takes us towards the axioms, reminds us that within the very construction of the most abstract thought and the highly rigorous and formal delineation of different scientific theories, the human imagination still plays a very special part that has always to be borne in mind. On the other hand, Einstein does not omit, fourthly, to emphasize that this whole intellectual construct, which can take its cue from our own fantastic creativity, then merges with a logical-deductive mathematical rigour, would end up by being a mere “fantasy”, without any real verification and cognitive scope, if it failed its final and decisive objective of being finally able to connect the plane of the consequences of S, S'... Sn to the plane of the Lebenswelt. But precisely this final “connection”, which seems to close the virtuous circle of human knowledge, is in fact in

 

62

Chapter Four

its turn a highly problematic and programmatically open element, because it in turn is based on the decisive mediation of technology, through which the cognitive scope of the different abstract theories that enable us to delineate certain predictions is “tested” – be “verified” or “falsified” – within the world of everyday human experience. By this Einstein seems to then remind us, fifthly, that the use of multiple technological innovations in the common world of praxis, typical of all human beings, constitutes a kind of decisive and highly structured “social” laboratory, within which is daily verified the cognitive capacity peculiar and specific to the different theories. It is a capacity that is “tested”, however, without letting us ever forget its intrinsically “problematic” character, because every “connection” and every “relation” that can possibly occur between the world of abstract thought and the concrete dimension of the world of praxis prevents us from developing an exhaustive, real and effective picture of the world, since the virtue most proper to each technological mediation is that of seizing only one aspect, always very partial and sharply limited, of the world in which humanity lives. Whence the inherent “historicity” of each single technological discovery, which, at the very moment that it incorporates into the technical devices the positive resolution of one particular problem, yet never makes the claim to be able to develop absolutely an untranscendable solution to the specific problem for which the technological device itself has been devised, predisposed and constructed. Hence the continuous and programmatic openness of the technological dimension to the intrinsic, critical superseding of itself, which makes it possible to rapidly bring out the peculiar historicity that has always characterized not only the dimension of technology, but also of the very nature of scientific knowledge. This was clearly understood also by a buccinator philosopher of modern science, Francis Bacon, for whom, as is well known, natura non nisi parendo vincitur. The inherent strength of technology is rooted precisely in its ability to “know how to command” nature “by obeying it”, i.e. by being capable of incorporating, within a specific and particular technical device, the natural laws that are respected and followed at the same time as they are being used to achieve a purpose that in nature by itself does not exist. The strength and inherent fecundity of this Einsteinian conception of knowledge is rooted in the multiple factors by which Einstein succeeds in conveying to us the whole of the polymorphism intrinsic to human knowledge without reducing it to any prejudicial model and without falling into any facile epistemological dogmatism. In fact, if one looks at Einstein’s drawing, keeping in mind both the so-called “Cartesian syndrome” (according to which, as is well known, science is reduced

Lebenswelt, The World of Objective Knowledge and Practice

63

solely to its supposed and unique “scientific method”) and the epistemological debate of the last three centuries, one immediately and very easily comprehends how the German physicist programmatically avoided reducing the scientific process to a single method. Whether this is deductive, inductive, conventional, formalistic, idoneistic, verificationist, falsificationist, or anything else. If at all, in the light of Einstein’s drawing, all these different and even conflicting scientific “methods” can be comprised – and replaced within the actual practice of scientific knowledge – as the different methodological phases of a much more complex and unified cognitive process, so that scientific knowledge is divided into a plurality of ploys and modes of epistemological awareness. But the inherent plasticity of Einstein’s scheme also enables us to focus the overall articulation of the complex scientific process, since the latter not only uses multiple and even different methodological strategies, but is also always rooted in a precise and historically determined common sense, within which it is elevated critically and to which it then returns, making it fertile and even modifying it profoundly, thereby initiating a highly articulated and complex social process, through which technical-scientific heritage is diversified and articulated, setting also in critical motion common knowledge itself.

2. The dynamic and Contradictory Polymorphism of Common Sense An intellectual of the highest importance, Antonio Gramsci (18911937) reflected acutely on the role, importance and polymorphic complexity of common sense, which is not, however, devoid of certain constituent ambiguities. In his Prison Notebooks he devoted numerous reflections in various notes to “common sense” and its relation to “good sense” (grasped in their critical-dialectical tension, according to the wellknown epigram of the poet Giuseppe Giusti, 1809-1850).20 Very briefly, Gramsci sees that: Every social stratum has its own 'common sense' and its own 'good sense', which are basically the most widespread conception of life and of man. Every philosophical current leaves behind a sedimentation of ‘common sense’: this is the document of its historical effectiveness. Common sense 20

“Il buon senso, che un dì fu caposcuola/ or nelle nostre scuole è morto affatto. / La scienza sua figliola, / l’uccise per veder com’era fatto.” (“Good sense that once ruled far and wide, / Now in our schools to rest is laid. / Science its once beloved child, / Killed it to see how it was made.”)

 

64

Chapter Four is not something rigid and immobile, but is continually transforming itself, enriching itself with scientific ideas and with philosophical opinions that have entered ordinary life. ‘Common sense’ is the folklore of philosophy, and is always half-way between folklore properly speaking (i.e. as it is commonly understood) and the philosophy, science and economics of the specialists. Common sense creates the folklore future, that is as a relatively rigid phase of popular knowledge at a given place and time (Prison Notebooks, notebook 24 [XXVII], §4).21

Viewed in this perspective, common sense is perceived by Gramsci as a mobile and dynamic dimension that grows, develops and changes in close contact with the deepening of the cultural and scientific-technical heritage in a specific historical society and also in direct relation to the different social strata (i.e. classes) that configure a specific socio-historical situation. The historical dynamism inherent in common sense is therefore perceived as twofold: first, common sense can in fact be read as a sedimentation of specific research programs (and then, at least from this point of view, the objective historical actuality of any intellectual movement will be recorded in its capacity to turn into common sense the vision of life and man of broad strata of society). Secondly, common sense also exercises a form of resistance against the genesis of new common senses that arise out of new research programs and enter into critical tension with the old common sense. In this second perspective, there is then generated a specific tension and also a veritable struggle between the old common sense, or “good sense”, and a new common sense that opens up a path by seeking to rejuvenate, heal and even mock the old common sense, in order to replace it with a new perception of life and the world. Within this precise and complex historical dialectic the specific value, at once historical and cultural, of common sense can be clearly understood. In fact, Gramsci observes in Notebook 10: In what exactly does the merit of what is normally termed ‘common sense’ or ‘good sense’ consist? Not just in the fact that, if only implicitly, common sense applies the principle of causality, but in the much more limited fact that in a whole range of judgments common sense identifies the exact cause, simple and to hand, and does not let itself be led astray by fancy quibbles and pseudo-profound, pseudo-scientific metaphysical mumbo-jumbo etc. It was natural that ‘common sense’ should have been 21

A. Gramsci, Quaderni del carcere, critical edition of the Istituto Gramsci, edited by Valentino Gerrratana, Einaudi, Turin 1975, 4 Vols., Vol. III, 2271; the quotations which follow immediately in the text are taken from Vol. II, 1334-1335 and 958-959 respectively.

Lebenswelt, The World of Objective Knowledge and Practice

65

exalted in the seventeenth and eighteenth centuries, when there was a principle of reaction to the authority of Aristotle and the Bible: it was discovered indeed that in ‘common sense’ there was a certain measure of ‘experimentalism’ and direct observation of reality, though empirical and limited. Even today, when a similar state of affairs exists, we find the same favourable judgment on common sense, although the situation has in fact changed and ‘common sense’ today has a much more limited intrinsic merit (Book 10 [xxxiii], §1).

In this specific interpretation, common sense thus seems to possess a critical function of its own, highly specific and autonomous, precisely because it is not only formed by a particular historical sedimentation produced by the various specific forms of culture (at least of those that have actually succeeded in exerting a historical influence), but it also seems to constitute in itself a form of resistance, perhaps of empiricalexperimental derivation, to anything that might appear as mere sophistry or metaphysical mystification, in short, the quibbling of pettifoggers. However, it does not escape Gramsci that this same component of empiricist resistance to the “theoretical” in common sense is, however, interwoven, and at the same time has a specific underlying ambiguity that denotes a certain closure and dogmatic rigidity in it: In common parlance ‘theory’ is used in a negative sense, as ‘doctrinaire’ or even as ‘abstractionist’. It has met the same fate as the term ‘idealist’, which from its technical philosophical meaning has come to mean ‘dreamer of hazy ideas’, etc. That certain terms have acquired this pejorative significance has not happened by chance. It is a common-sense reaction against certain kinds of cultural degeneration, etc., but ‘common sense’ in turn became the Philistiniser, the mummifier of a justified reaction into a permanent state of mind, into an intellectual laziness that is equally degenerative and repulsive as the phenomenon that it wishes to combat. ‘Good sense’ has reacted, ‘common sense’ has embalmed the reaction and made it a ‘theoretical’, ‘doctrinaire’, ‘idealistic’ canon (Quaderno 8 [XXVIII], § 28).

Common sense is, therefore, a sort of Janus, with two faces: on the one hand it is capable of reacting critically to all the metaphysical obscurities of the various theories and new abstract and improbable cultures, but on the other hand it has the far from slight defect of systematically rigidifying its resistances in a form of dogmatic “intellectual laziness”, which ends up assuming a curvature that is “equally degenerative and repulsive as the phenomenon that it sought to oppose”. Hence the twofold intellectual (and civil!) commitment which must always distinguish any serious and conscious movement of thought, which must always be capable of coming

 

66

Chapter Four

into direct and fruitful contact with the common sense of its time, in order to criticize it and supersede it by undertaking the delineation of a new common sense, while, on the other hand, this same movement of thought must not allow itself to be ensnared and fossilized dogmatically by that same new common sense that it helped to forge and create historically. Common sense progresses and develops, in this way, along a “regional” dividing line, which is very delicate, avoiding the two opposite extremes of metaphysical abstraction and the dogmatism of inertia typical of tradition; it must at the same time be capable of being innovative, “theoretical” and ideally constructive in the context of thought, without ever losing sight of the concrete experience peculiar to the historical world of praxis common to all people within a given horizon of life. We are thus faced with a highly complex movement, precisely because, as noted by the refined philosopher of anthropology Remo Cantoni (1914-1978) in Tragico e senso comune, “common sense, which is a characteristic moment of the ‘collective consciousness’ and of the ‘objective spirit’, can be defined as a cultural habitat in which we grow up and live”.22 Not surprisingly for a Romantic and idealist philosopher such as Georg Wilhelm Friedrich Hegel (1770-1831), science – at least in the dialectical, rational and global interpretation outlined in Die Phänomenologie des Geistes (1807) – always requires “us to abandon the life of the object [sich dem Leben des Gegenstandes zu übergeben]” and that we “must be aware of it and express its inner necessity [die innere Notwendigkeit desselben vor sich zu haben und auszusprechen]”. 23 Certainly, in the Vorrette of the Phenomenology of Spirit, Hegel launches a direct attack against common sense or good sense [Gesunder Menschneverstand], but this critical stance concerns, at the same time, both the dogmatism of common sense and the schematizing abstractness of the intellect that is not capable of grasping true knowledge. In this Hegelian perspective, the abstraction of the intellect and the dogmatism of common sense are in fact allies against both scientific knowledge and philosophy itself. And the common thread running through their uncritical alliance is rooted in their shared inability to understand and properly perform the arduous and difficult “work of conception” [der Arbeit des Begriffs], which constitutes the heart of the progress of scientific and 22

R. Cantoni, Tragico e senso comune, Mangiarotti Editore, Cremona 1963, 14, volume later reissued in Id., Il senso del tragico e il piacere, Editoriale Nuova, Milano 1978, 7-8. 23 See. G. W. F. Hegel, Die Phänomenologie des Geistes present in the Sämtliche Werke. Kritische Ausgabe, edited by J. Hoffmeister, Hamburg 1952, 6 ed., 45, while the quotation that immediately follows in the text is taken from pp. 28-29.

Lebenswelt, The World of Objective Knowledge and Practice

67

philosophical knowledge. This last is characterized precisely by the epistemological effort of being always able to adhere critically to the object of its study. This can only lead Hegel once again to emphasize a dogmatic limit intrinsic to common sense, since “if in knowing something is presupposed as known and it is assumed as such, they end up deceiving themselves in the most ordinary of ways, and others are also deceived; such knowledge, without understanding how it happens, does not take a step forward, despite all the great talk that is done”. For this reason, as wisely noted by Cantoni, who bears in mind this Hegelian analysis (linking it, moreover, to other, similar, classic examples presented by Plato and Kant), “the appeal to common sense amounts, in conclusion, to dwelling on the known, without realizing that the hardest things to know are precisely things known. Common sense is a knowledge that does not proceed, that insists on repetition, a wordy superficial knowledge that does not see the movement, life and relations between things”.24 On the other hand, as we saw emerging from Gramsci’s reflections, in common sense there is present an intrinsic positivity concerning the world of practice peculiar to everyone, which philosophical-scientific thought itself cannot by any means renounce. In short, to put it again in Cantoni’s words, “when philosophy defines itself as a critique of common sense, philosophy is right because its task is precisely to criticize the dogmatic aspects of common sense, but the definition becomes incomplete and erroneous if the philosopher fails to add that the ultimate purpose of philosophy is the reformation of common sense, a return to common sense regenerated”. The considerations invoked above thus make it possible, in the final analysis, to understand that there is a complex continuity and an equally complex specific discontinuity between the dimension of common sense and scientific investigation. But within the ever-existing and variously articulated dialectic established, on the historical level, between common sense and scientific knowledge, what can the precise heuristic function of the former in relation to the latter ever be? Does there exist, then, a decisive role that common sense can perform in relation to scientific knowledge? And, if this is positive, what exactly is this role? Taking his cue from these questions, the scholar and epistemologist Evandro Agazzi (1934-) has pointed out that, in his view, the specific function of common sense derives

24

R. Cantoni, Tragico e senso comune, op. cit., 39 (in Il senso del tragico e il piacere, op. cit., the passage quoted on p. 29); the quotation that follows in the text is taken from p. 23 (15-16 of Il senso del tragico etc).

 

68

Chapter Four just from being common, understanding this characteristic no longer in the usual sense of being (ideally) ‘shared’ by all men, but in that of containing a global vision of reality and all its dimensions. With this, it is an antidote to the reductionist tendency that relentlessly assails the various sciences (and perhaps science as a whole). For common sense no aspect of reality, no “attribute”, is meaningless, since it contains the repository of all cognitive and non-cognitive experiences of which humanity is capable and therefore it is certainly not able to ‘fill in’ but undoubtedly to keep open that horizon of the totality that, conversely, every science programmatically ignores in order to delimit its field.25

In short, to the extent that every scientific discipline must inevitably always be able to circumscribe and define, with the utmost rigor, the scope of its investigation, thus attaining the construction and delineation of its own autonomous and specific “regional ontology”, then against the tendency of every single science to close itself programmatically within this ambit, so promoting a reductionist approach to life, common sense exerts its healthy and positive function, because it is the dimension with which it is possible to gain (critical!) awareness of the dogmatic limits of any scientific reductionism. This is precisely what happened historically when the practical and technological aspects of the sciences, in their extraordinary practical and effective impact with the world of practice, did not automatically generate a mythical improvement in human life, but instead ended up creating new problems, new anxieties and also “fear” and “suspicion” towards the development of the techno-sciences themselves. But how was it possible for such an embarrassing development to displace, more or less critically, all the finest mythical dreams of every positivist (more or less conscious) about “the magnificent and progressive” outcomes of the techno-sciences? This failure resulted precisely from the very “narrowness” of the scientific view of the world, since a wide range of existential factors, in the sense that they emerge directly from the world of life, has forced our culture to realize the reduced character which is inherent in the scientific image, in the types of conceptualization, in the methods of investigation which it adopts in gazing at reality, which are not sufficient to tackle the ‘non-scientific’ and ‘non-technological’ problems that the techno-scientific developments

25

E. Agazzi, Continuità e discontinuità fra scienza e senso comune in Valori e limiti del senso comune, edited by E. Agazzi, Franco Angeli, Milano 2004, 355, italics in the text, while the following quotation is taken from the essay by Agazzi, Il senso comune e l’unità dell’esperienza, 37-38, italics in the text.

Lebenswelt, The World of Objective Knowledge and Practice

69

themselves are raising and vice versa that are obvious to everyone just because they fall within the experience of every human being, they are a conspicuous part of the present unity of experience. On the basis of this awareness it is possible, in fact, to enhance science and technology for what they effectively are and give to humans and, without, moreover, absolutizing them, which, after all corresponds to the balanced view that we usually attribute to ‘common sense’ understood not so much as an image of unsophisticated reality or the world, but rather as an attitude of wisdom.

3. Common Sense and Phenomenological Attitude But even in this, its valuable heuristic function, common sense is always presented as a perpetually dynamic dimension and in intrinsic tension. So much so that even a thinker like John Dewey (1859-1952) rightly pointed out that the relation of continuous osmosis is established between the world of common sense and that of the use of technical instruments and of technology itself. In his Logic: The Theory of Inquiry, Dewey wrote: Common sense in respect to both its content of ideas and beliefs, and its methods of procedure, is anything but a constant. Both its content and its methods alter from time to time not merely in detail but in general pattern. Every invention of a new tool and utensil, every improvement in technique, makes some difference in what is used and enjoyed and in the inquiries that arise with reference to use and enjoyment, with respect to both significance and meaning.26

Dewey's insistence on technique is valuable for several reasons, not only because it is precisely on the plane of the social use of technology that the inherent limitations of the positivist view of the growth of scientific knowledge emerge, as the mythical and uncritical automatic driving force for the improvement of human society as a whole, but also because reflection on the dimension of technique helps us better critically understand both the always “limited” character of technical-scientific action and its full cultural value. In fact, every technical object, by its intrinsic nature, has, in its own actual configuration, the positive resolution of a given problem (namely that particular problem for which the particular technical device was invented and built). From this specific and fundamental point of view, then, the technical object has to be interpreted 26

J. Dewey, Logica: The Theory of Inquiry, Henry Holt and Company, New York 1938, 64.

 

70

Chapter Four

as a sort of “materialization” of human thought itself, which, in the technical device, takes concrete form in an apparatus by which – in the words of Francis Bacon (1561-1626) – natura non nisi parendo vincitur. Thanks to the fruitful mediation of the technical object, humanity is thus able to “control nature by obeying it”, namely by being able to relate the world of its free thought to the world of nature, which follows “universal and necessary”, “deaf and inexorable” laws, in the words of Galileo Galilei (1564-1642). In this way technology is a fundamental link, à la Simondon (1924-1989), between the natural and the artificial, a dimension within which – albeit by virtue of a historical sedimentation formed by “an immense deposit of hard work”, to quote the nineteenthcentury thinker Carlo Cattaneo (1801-1869) – humanity has gradually constructed its own “artificial world”, within which it has not only drawn itself gradually up from the horizon of barbarism, but it has also achieved, not without contradictions, its historical adventure, differentiating itself from animal life.27 In short, seen in this perspective, common sense always lives in close, direct, plastic and fruitful pragmatic contact, in the world of practice, with all of these – although very different and even conflicting – technological practices, through which humanity has variously constructed different societies, always historically determined. But it is precisely within this very special world of human practice, always closely intertwined with the dimension of technology and the use of the various and extraordinary technical instruments invented historically, that humanity also risks getting lost in a “naturalistic” attitude. A “naturalistic” attitude that, in reality, is always the result of a particular and specific world of technological practices, historically determined, which, however, are not perceived in this, their intrinsically historical and conceptual dimension, but are instead experienced (and practised!) in a way that is absolutely – and uncritically – naive. If within this particular ambit of problems concerning experience we also wished to refer to the phenomenological reflection inaugurated by Edmund Husserl (1859-1938), we might recall that the author of the Logische Untersuchungen rightly insisted on the desirability of avoiding a traditional fundamental equivocation, by virtue of which the experience is systematically confused with the object, manifested in the same object that manifests itself as such. Husserl writes:

27

For a discussion of these issues the reader is referred to the contribution by the present writer: F. Minazzi, “Salire sulle proprie spalle?” Simondon e la trasduttività dell’ordine del reale, «aut aut», n. 361, January-March 2014, 110129.

Lebenswelt, The World of Objective Knowledge and Practice

71

Die Äquivokation, welche es gestattet, als Erscheinung nicht nur das Erlebnis, in dem das Erscheinen des Objektes besteht (z. B. das konkrete Wahrnehmungserlebnis, in dem uns das Objekt vermeintlich selbs gegenwärtig ist), sondern aus das erscheinende Objekt als solches zu bezeichnen, kann nicht scharf genug betont werden. Der Trug dieser Äquivokation verschwindet sofort, sowie man sich phänomenologische Rechenschaft darüber gibt, was denn vom erscheinenden Objekt als solchem im Erlebnis der Erscheinung reel vorfindlich sei. Die Dingerscheinung (das Erlebnis) ist nicht das erscheinende Ding (das uns vermeintlich in leibhaftinger Selbstheit ‘Gegenüberstehend’). Als dem Bewusstseinszusammenhang zugehörig, erlebe wir die Erscheinungen, als der phänomenalen Welt zugehörig, erscheinen uns die Dinge. Die Erscheinungen selbst erscheinen nicht, sie werden erlebt.28

Hence, for Husserl, the manifestation of the thing (namely experience) does not coincide with the thing that is manifested, because his concept of experience does not coincide with the common concept of lived experience. For Husserl, in fact, one has always to take into consideration the intentional content of the experience that opens up the phenomenological dimension of the human experience peculiar to the Lebenswelt. For Husserl, certain external events signify the possession of certain acts of perception or knowledge directed towards these external events: Es besagt nicht mehr, als dass gewisse Inhalte Bestandstücke in einer Bewusstseinseinheit, im phänomenologich einhaitlichen Bewusstseinsstrom eines empirischen Ich sind. Dieser selbst ist ein reeles Ganzes, das sich aus mannigfachen Teilen reel zusammensetzt, und jeder solche Teil heisst „erlebt“In diesem Sinne ist das, was das Ich oder das Bewusstsein erlebt, eben sein Erlebnis,Zwischen den erlebten oder bewussten Inhalt und dem Erlebnis selbst ist kein Unterschied.

The phenomenological concept of experience therefore refers programmatically to intentionality: Sind die sogenannten immanenten Inhalte vielmehr bloss intentionale (intendierte), so sind andererseits die wahrhaft immanenten Inhalte, die zum reellen Bestande der intentionale Erlebnisse gehörigen, nicht intentional: sie bauen den Akt auf, ermöglichen als die notwendigen

28

E. Husserl, Logiche Untersuchungen, Max Niemeyer, Halle 1928, 2 Vols. in 3 Tomes, Vol. II, Tome 1,. 349-350, (fifth inquiry, § 2, the spacing is indicated by italics). The quotations that follow in the text are taken from the following pages of the same tome: 352 (§ 3) and 374 (§ 11) respectively.

 

72

Chapter Four Anhaltspunkte die Intention, aber sie sind nicht selbst intendiert, sie sind nicht die Gegenstände, die im Akt vorgestellt sind. Ich sehe nicht Farbenempfindungen, sondern gefärbte Dinge, ich höre nicht Tonempfindungen, sondern gefärbte Dinge, ich höre nicht Tonempfindungen, sondern das Lied der Sängerin usw.

In this respect, an eminent scholar and equally valid interpreter of Husserl's thought, Vanni Sofia Rovighi, in her personal copy of the Logische Untesuchungen, noted as her personal marginalia to this crucial passage in Husserl: This does not mean to say that the object of direct knowledge is the thing and not my modification, but it does mean that for there to be knowledge it is necessary for there to be more than sensations. Sensation is not a cognitive act, it is not intentional.29

In this is rooted, to be precise, not just Husserl’s critique of the prospect of empiricism derived from Hume, but also his basic concord with the critical lesson of Kantian transcendentalism, by virtue of which sensation, in order to become a fully cognitive act, always requires the intervention of intentionality, by which experience is structured and oriented according to a specific plan of intelligibility of the world. This observation by Husserl, whose precise meaning is reinforced by the annotation of the Italian scholar, enables us, then, to grasp the critical distance that lies between natural and phenomenological reflection. If, in the former, the act of cognition is obscured by the role of the cognitive subject – according to a metaphysical tradition which is hard to eradicate and which was inaugurated by Descartes (which does not, however, as we often read, find its supposed crowning achievement in the Kantian lection) – so giving rise to a relation between the subject of the awareness and the object known, in contrast to Husserl one has always to look at the lived act, since the self then refers to the object precisely through the act and within this act. But in every act, for Husserl, the self always intentionally refers to a given object. As Vanni Rovighi rightly points out, again in her marginalia, “in the act of knowledge an intentional identity of subject and object” is attained. This then enables us to understand how, for Husserl, 29

Vanni Rovighi’s personal copy of the Logische Untersuchungen was acquired by me in a bibliographic study in Bologna in 1991; the annotation cited appears in the margin on p. 374 of the first tome of the second volume; the second note by Vanni Rovighi is taken from p. 376 of the same volume, as a comment on a passage in § 12.

Lebenswelt, The World of Objective Knowledge and Practice

73

the “perception” itself, the “vision” itself, etc., of any reality always contains something beyond what is actually perceived and seen. Why? Because every act of perception, like every act of seeing, constitutes a “having-in-person” and, at the same time, a “design” [Vorhaben], a presuming [Vormeinen]. For this reason, Husserl sees every experience increasingly as an endlessly open induction and prediction. This, however, helps us to remember that the experience of the Lebenswelt constitutes an indispensable dimension, in which arises, always dialectically, the technical and scientific heritage which is actually available to us within a given society, historically configured.

 

PART II COMMON SENSE

CHAPTER FIVE FROM COMMON SENSE KNOWLEDGE TO SCIENTIFIC KNOWLEDGE (AND VICE VERSA) EVANDRO AGAZZI

1. Science and Common Sense Any science (in the modern meaning of the term) is an investigation of a limited number of attributes (that is, properties and relations) that common sense assigns to reality or, more precisely, to a specific part of it. So mechanics (the first ‘modern science’ historically speaking) consists of studying the ‘local motion’ of bodies, that is, it deals with those ‘things’ to which common sense assigns the properties of matter and motion, and it investigates only such attributes. To this purpose it introduces some specific predicates: ‘mass’ (which technically represents the intuitive notion of ‘quantity of matter’), ‘position’ (which ‘places’ a body with respect to a spatial reference), ‘displacement’ (which means the change of place of a body in space), ‘motion’ (which indicates the process of displacement), ‘duration’ (which means the interval of time in which the motion happens) and ‘speed’ (defined as the relation between the space covered by a moving object and the time employed, which characterizes the movement itself). These concepts are clearly taken from common sense, but in a certain way they are already idealized. For instance, even the geometric properties of a body are ignored: not only is its ‘shape’ not interesting, but also its spatial ‘extension’, since in mechanics, bodies can be reduced to material points and all their mass is deemed be concentrated in only one point, that is, the barycentre. Mass then assumes one of the fundamental characteristics of substantiality, which is permanence or conservation (as indicated in the fundamental principle of mechanics, that is, the conservation of mass).

78

Chapter Five

Even common sense can accept without difficulty such idealizations, but things change when motion is taken into consideration. According to common sense, it is an ‘event’ which occurs whenever a body abandons its natural state of rest and, therefore, this requires a causal explanation (as was considered in Aristotelian and, in general, pre-modern physics). According to modern mechanics, rest and motion are instead two different forms in which motion can be described with respect to two different systems of reference, and motion is an entity which possesses the same characteristic of ‘substantiality’ as matter, in the sense that it may change, however it remains (law of the conservation of the quantity of motion). Therefore, motion does not require a causal explanation; only its variation requires one, as indicated by the principle of inertia: a body not influenced by any force remains in its state of rest or uniform rectilinear motion. Common sense is therefore ‘caught’ but not fully ‘disproved’; in fact, common sense’s conviction that a cause is necessary to change the state of rest of a body and move it is respected, since if such an event occurs (in relation to a specific system of reference) we have a variation of motion for which mechanics also requires a cause called ‘force’. Another surprise may occur: the concept of ‘displacement’ receives more precise determination in mechanics, since it is connected with the geometric notion of ‘distance’, which is the length of the rectilinear segment connecting two points. Therefore the ‘fundamental’ motion which does not require any cause to be explained is uniform rectilinear motion, in which there is no change of speed. Circular motion, according to common sense, is very similar, but in mechanics it is not: in fact, while the ‘module’ of the speed’s vector does not change, its direction changes and this change requires a cause (that is, a force) which explains it. The change of speed is called acceleration: once again we ascertain a continuity and a discontinuity with respect to common sense, since according to it, acceleration means an increase of the absolute value of the speed at which it opposes slowing down or deceleration. In mechanics, acceleration is instead the mere variation of speed, and this includes indifferently an increase or decrease of the absolute value, as well as its change of direction. Therefore, the only thing to be explained in mechanics is the acceleration of bodies which are in motion, and the concept of force was introduced as the cause of it. A further step in the ‘overcoming’ of common sense depends on the fact that modern science started with a project of quantification and mathematization of the attributes taken into consideration, which are expressed in magnitudes. This perspective is not alien to common sense, which does not ignore the comparison of quantities, but in general it

From Common Sense Knowledge to Scientific Knowledge

79

accepts intuitive ‘estimations’, more or less coarse and subjective, while science worries about identifying the measurement’s criteria, which allows for the determination of the objective and numerical values of the various quantities, and this then allows the use of mathematics for its investigations. This consideration is reflected also in the specification of concepts. We have already said, for instance, that the concept of mass can be intuitively defined as ‘quantity of matter’ (Newton adopts this expression in intuitively describing this concept), however, its use could be ambiguous in comparing masses. For example, looking at a heap of straw and at a small cube of iron, one is easily induced to consider the mass of the first to be bigger than that of the second. For this reason, the well-known riddle: ‘Which weighs more, a kilo of straw or a kilo of iron?’ can induce a reckless interlocutor to answer that a kilo of straw weighs more, realizing only afterwards that they have the same weight. However, it is not guaranteed that he is fully convinced that they have the same mass, since it is rather spontaneous to believe that the bigger the mass of a body the bigger the volume and, in fact, things are ‘more or less’ like this. However, common sense also accepts another criterion for comparing masses, which is tied both to motion and to force: having two bodies lying in a state of rest on a flat surface, if we try to move them by pushing and we realize that one ‘opposes resistance’ to motion more than the other, since it requires a bigger ‘effort’ to be moved, then we will say that this happens because one has a mass bigger than the other. In this way, mass is defined as opposition to motion and its evaluation consists of the ‘force’ necessary to produce a certain acceleration. And this is just the common sense ground which underlies the mechanical definition for measuring mass expressed by the fundamental law of Newton’s dynamics: f = ma. It explains that the acceleration impressed upon a body by a certain force is inversely proportional to its mass (or that it is directly proportional to the force applied) and the proportionality coefficient is its mass. Even in common sense, one finds the tendency to say that the bigger the mass of a body, the bigger is its weight: once again, the intuitive reference is to the ‘effort’ necessary to lift or to hold each of the two bodies, and therefore, again, there is the reference to a force. This way is also used in mechanics, in which masses are usually measured in ‘units of weight’ (like the kilogram), even knowing that weight is a force (the force of gravity) while mass is not. This conventional assimilation is permitted thanks to the practically constant value of gravitational acceleration on the earth’s surface. It is not, however, guaranteed beforehand that the ‘inertial’ characterization of mass as resistance to motion and the ‘gravitational’ one

80

Chapter Five

are equivalent or that, when measured, they will give the same values. The existence of such equivalence has been theoretically proved only with the advent of relativistic mechanics. The few elementary analyses presented till now are sufficient to show how science starts from common sense and then gradually departs from it. It is, however, important to point out that science cannot ‘lose contact’ with it, since its non-eliminable task remains that of supplying explanations for empirical facts accessible to common sense, that is, falling within what we can call ordinary knowledge (which is not limited to the simple level of sensations and perceptions but also includes a series of organizations and conceptualizations both ‘ordinary’ and ‘common’). So, for instance, Copernican theory is justified since it is capable of formulating an explanation of ‘celestial’ phenomena empirically observed more satisfactory than that of Ptolemaic theory. Kepler’s laws, in turn, represent a further improvement of Copernican theory by introducing, besides the ellipticity of the planetary orbits, some precise statements referring to the motion of the planets (which allowed it to explain some other phenomena which were previously misunderstood). As to the Newtonian theory of gravitation, it allows us to explain these same laws of Kepler. It is not important that this happens at the cost of certain idealizations which overcome common sense (like the ‘fact’ of considering the sun and the planets material points, ignoring their volumes, which is justified by their limited volumes compared with the distances between them); what counts is the fact that in such a way a series of actually observable astronomic phenomena are more completely explainable and understandable by common sense, and in such a way fall within ordinary knowledge. This attitude of trust and ‘continuity’ was reinforced when it was ascertained that this ‘celestial mechanics’ had been able to predict the appearance of the Halley comet and, later, that of two planets (Neptune and Pluto) previously unobserved. More of this reconciliation with common sense happened in terms of the reality of the everyday physical world, since the original mechanics of a material point can be extended to include solid bodies endowed with a volume, as well as liquids, and to take into consideration additional forces different from gravitation (like the elastic forces). The initial support for this extension from the mechanics of the point to the ‘mechanics of systems’ still comes from the convictions of common sense and, more particularly, from the intuitive notion of the additivity of masses: it is obvious to common sense that if many objects are heaped up, the mass of the heap is the sum of the masses of the single objects, and symmetrically,

From Common Sense Knowledge to Scientific Knowledge

81

if we divide a body into parts, its mass is equal to the sum of the masses of the parts, no matter how big their number. A second common sense intuition which concerns the divisibility of the bodies is grafted here. Common geometrical intuition spontaneously admits that a solid body can be indefinitely divided into always smaller parts, and at the same time, it admits that such a division must ideally come to an end with indivisible entities that ‘have no parts’, that is geometrical points (so described by Euclid). In applying this intuition to a material body, we end up admitting that the indefinite division of a body ends with material points, that is, with material entities not further divisible: these are the ‘atoms’, already introduced by an ancient school of thought and returned to favor in the seventeenth century: they were proposed as the conceptual background of Newtonian mechanics. Points (both geometrical and material) are conceived as ‘infinitely small’; therefore in order to form a body, their number has to be ‘infinitely big’ and - since we are also accustomed to thinking that a body is the ‘sum’ of all its points - we also think that its extension and its mass are the result of the sum of infinite addends. All this, apart from a certain haziness, is acceptable to common sense, but things change when the generic idea of sum is translated into what is the mathematically defined result of an ‘addition’. In an addition, finite (and not infinitely small) addends are added and the number of the addends must also be finite, since otherwise the ‘sum’, that is the result of the addition, is not attained. It was thanks to the results of the development of infinitesimal calculus (to which Newton himself substantially contributed) that it was possible to define mathematically the concepts of infinitely small (‘infinitesimal’) and of infinitely big (‘infinite’) in order to justify the use of ‘sums of infinite finite terms’ (‘infinite series’) and ‘sums of infinite infinitesimal terms’ ( ‘definite integrals’), establishing the correct conditions for the existence and the calculation of the respective results. By using this abstract tool, it was possible to obtain the calculation of the mass of a material system (for instance, a solid body) as well as of other mechanical properties, as ‘sum’ of the corresponding properties of its parts, through ‘integration’. The infinitesimal inverse operation (that is, ‘derivation’), in turn, allowed us to state precisely and make computable certain seemingly obvious but rather undetermined common sense notions, like that of instantaneous speed and acceleration, or of the direction of a curve at any arbitrary point. Therefore, Newtonian mechanics was not actually restricted to the study of a few attributes of physical reality, like matter and motion, but rested on a real cosmological vision of this reality, like, in effect, the atomistic conception. The proof of this fact is that Newtonian optics is also

82

Chapter Five

arranged within this view, developing a corpuscular theory of light, whose rays are conceived as swarms of particles moving in rectilinear trajectories in empty space and which, clashing against bodies, are reflected, refracted and absorbed. This theory explains the empirical laws of geometric optics, and also, for instance, the existence of different colors.

2. Representations of the World The atomistic conception was certainly widespread, but was not the only possibility and, in fact, in the same century it was opposed by a ‘continuistic’ conception of the physical world which was defended especially by Descartes, according to which the space is a ‘plenum’ of matter in which motions are transmitted by contiguity. In optics, this conception gave rise to the wave theory of light (developed by the Cartesian physicist Huygens), according to which light is formed of waves which propagate in a ‘medium’, later called ‘luminipherous ether’. Clashing against bodies, these waves are reflected, refracted, absorbed, etc. according to modalities that make it possible to explain optics’ empirical laws and the existence of colors. The ‘continuous’ vision of the cosmos (according to which all space is filled up with an invisible, intangible and rather imaginary ‘cosmic ether’ of a material nature) seems less plausible than the ‘discrete’ atomistic conception, but this too had its conceptual difficulties, starting from the very problem of ‘the existence of the vacuum’, to which is added that of the nature of the gravitational force which acts amongst bodies as an attracting ‘action at a distance’, at odds with the intuitive conception of the contiguity of causal actions. In the end, the preference in favor of one or the other vision could only be determined by their success in satisfying the common sense requirements tied up with empirical confirmations. Historical developments did not give rise to the prevalence of either of the two. The Newtonian atomistic vision prevailed in the wide field of proper mechanics, undermining in the same country, France, during the eighteen century, Cartesian physics, which had not been able to significantly overcome the merely intuitive level of the cosmological representations of its founder or to be expressed in adequate mathematic forms. However, in the field of optics, notwithstanding the initial predominance of corpuscular theory, largely due to Newton’s authoritativeness, it had to give way to wave theory when - in the early decades of the nineteenth century - this (unlike corpuscular theory) proved capable both of explaining the phenomena of interference and diffraction

From Common Sense Knowledge to Scientific Knowledge

83

of light and of correctly predicting how light’s speed had to change when passing from a less dense to a more dense medium. ‘Luminipherous ether’ acquired the right to be accepted in the ontology of physics, and this right increased when Maxwell proposed a geometriccontinuistic theory of electromagnetic phenomena (initially presented simply as a ‘different’ mathematical representation with respect to the interpretation of the most famous physicists of the European continent, who drew their inspiration from the conceptual corpuscular model). When, based on pure theoretical considerations, he discovered that the speed of propagation of electromagnetic vibrations in the continuous ‘medium’, as he himself conceived it, was equal to the propagation of light in the ‘vacuum’, he concluded that light also consists of electromagnetic vibrations (electromagnetic theory of light) and the ‘electromagnetic field’ (as he called it) was identified as that ‘ether’ which was already accepted in optics. It has to be pointed out, however, that the two models were intended to be strictly ‘mechanical’, since the undulatory one also used mathematical concepts and methods belonging to other branches of mechanics (more particularly taken from the theory of elasticity). He considered it to be the task of future generations to devise an adequate mechanical representation of the structure of this ether, of which he simply postulated certain general properties. It is therefore possible to say that, in the second half of the nineteenth century, the two models shared the field of physics: the corpuscular supported mechanics, acoustics, and the theory of heat, while the undulatory dominated in optics and electromagnetism; Maxwell himself used the first in his research on gasses and the second in that on electromagnetism. Both models have been developed thanks to progress in the experimental techniques of observation (for instance in the measurement of the speed of light) and in the mathematical tools (for instance, the theory of the field makes a great use of partial derivative equations, harmonic functions, etc., which were discovered and studied only recently). Therefore, the two models could only collapse for ‘internal’ reasons, that is, if serious difficulties were discovered in developing them ‘in detail’. This actually happened. Various failures occurred in the attempts made by eminent scientists to elaborate a consistent mechanical model of the cosmic ether, suitable to reflect the often divergent properties of electromagnetic waves, which marked the decline of this concept even before it was swept away by the theory of relativity. On the other hand, the kinetic theory of matter (more particularly of gasses) did not succeed in adequately explaining the phenomena and principles of thermodynamics,

84

Chapter Five

notwithstanding the mathematical skill of the illustrious physicists who undertook this task, adopting probabilistic-statistic methods. It is possible, therefore, to conclude that, as early as the end of the nineteenth century, physics was in a situation of great discontinuity with reference to the interpretative requirements of its results claimed by common sense.

3. The Specialization of Concepts and Principles There is a deeper side of our problem that we want to briefly examine. We have seen that the Newtonian concept of force plays the role of the concept of cause in mechanics. Someone might be pleased to notice that the principle of causality (considered with diffidence by many, since it is not only one of the principles more clearly accepted by common sense but also a pillar of the abhorred metaphysics) must be accepted, even if in a masked form, by science itself. It is necessary to realize, however, that force not only represents a specialization or ‘restriction’ to the physical phenomena of the concept of cause, but is a real ‘impoverishment’ of this concept. In fact, the forces taken into consideration by physics are considered to act on bodies ‘from the outside’, and as such they possess the characteristic features of what in the tradition was called ‘efficient cause’. This, however, was only one of the causes admitted, and in a certain sense, the least important (its privileged field of application was that of the artificial production of objects and processes). In the case of natural events, more importance was given to the causes which, in Aristotelian terminology, are named formal and final, and which offer an explanation for the behavior of entities as a consequence of their essence (the Aristotelian form) or nature. In this essence are, in fact, inscribed the courses which will be followed by an entity in its development, which consists in the passage from potentiality to actuality and tends, as its end or télos, toward the realization of the act for which the essence plays, from this point of view, the role of final cause. Galileo had already characterized natural science as a waiver to investigate the essence of natural objects, to restrict the study to that of some of their attributes (named by him ‘affections’) and Newton had strongly reaffirmed this point of view. Since then, the ostracism of final causes has become a real dogma in modern science in general, also extending to those sciences (like biology) in which a ‘finality’ seems to be unavoidable in characterizing organs and mainly ‘functions’ (indeed, the

From Common Sense Knowledge to Scientific Knowledge

85

great biologist Monod did not overcome this point of view, but instead of using the suspect term of ‘teleology’ he used that of ‘teleonomy’). From this point of view, ordinary knowledge is richer than scientific, since it avails itself of concepts given up by the latter. This scientific conception of causality, moreover, has accompanied the outlined impoverishment with gratuitous inflation. While the principle of causality only affirms that each variation has a cause, physics adds that each cause necessarily generates a well-determined effect. Here we have a transition from mere causality to determinism, which, however, has been identified with causality and, when quantum physics showed the limits of determinism, many felt authorized to affirm that with this the venerable principle of causality was disproved. It is, however, useful to understand why this identification was made. In actual fact, physical causality expresses itself in the physical laws according to which a precise ‘effect’ must follow from certain ‘initial conditions’ according to the casual relation expressed by the law in question. Now, if it is not expected that such an effect always, necessarily follows, not only the explanatory function of the laws gets lost, but also the possibility of making binding predictions, which is the necessary condition of any experimental control; in such a way, one of the main pillars of modern science (that has adopted the experimental method as its main tool of research) would collapse. It is possible to wonder, however, whether determinism has only this pragmatic justification, and it is possible to realize that it is not like this. In fact, the notion of physical laws rests, as an ontological postulate, on the concept of the uniformity of nature, which is perhaps also a common sense conviction, but maybe not completely. However, at this point this topic is not particularly important, it being sufficient for us to have given an example of the complexity of the transformations, often unobserved, to which the concepts and principles, also very general, are subjected in passing from common sense to science; the consequence then is that when these are ‘sent back’ from science to common sense, they may generate unjustified difficulties, and certainly many ambiguities. We have claimed that the aim of the deterministic conception was to guarantee the exactness of scientific predictions and that this conception was thrown into crisis by quantum indeterminism. In fact, it was already in crisis within classical mechanics, because of the merely statistic treatments which had to be used in order to bring thermodynamics back, not into particle mechanics, but into the ivory tower of celestial mechanics. In fact, if we consider a system formed by bodies subject only to gravitational

86

Chapter Five

forces, whose number is bigger than two (the so-called ‘problem of three bodies’) and if we aim, once its initial state has been determined, to ‘determine’ its state after a certain interval of time, we find that (due to the mutual interplay of the gravitational forces amongst these bodies), such determination is not possible in the following sense. Not even classical physicists thought that magnitude could be measured with infinite precision. However, (besides being confident that more and more perfect instruments could allow the attainment of always narrower approximation intervals) they counted on the fact that by applying the laws of classical physics, ‘small variations’ in the determination of the initial conditions would also be reflected in ‘small variations’ in the determination of the final result (‘condition of linearity’), so the exactness of ‘deterministic’ predictions is the same as that of the specification of the initial conditions. In the case of the problem under examination, i.e. vice versa, small variations in the initial conditions are enough to cause very large differences in the final state, that is, differences which largely exceed the degree of precision that we would like to achieve. In this sense, the ideal of classical determinism already seemed to be utopian for classical science too, and the ground for the studies on non-linearity and complexity which characterized the last decades of science was laid.

4. Discontinuity Between Science and Common Sense The discontinuities between science and common sense already appeared significant at the end of the nineteenth century and it is easy to see that they depended on the more and more accentuated separation which was taking place between the contents of science and the possibilities of representation offered by the daily experience, that is, by what we may call ordinary knowledge, even if ‘idealized’. The corpuscle of the old atomistic theory could be conceived as an idealization of a grain of sand, and the waves of the old cosmic ether as an idealization of the waves which are formed and propagate, interfere, strengthen and neutralize in a pond into which some stones have been thrown. But when these intuitive models must be filled with complications in order to satisfy the explanatory function assigned to them, and moreover, when they collapse, common sense has no more support to represent the scientific knowledge. Indeed, the same scientists were in a bad situation: let us think of the moment in which, once the possibility of proposing mechanical models of the ‘ether’ (even if highly theoretical and sophisticated) disappeared, the

From Common Sense Knowledge to Scientific Knowledge

87

affirmation that the electro-magnetic field is a mere field of force or energy was accepted. Not only was common sense incapable of understanding the ontology of this entity, but in the beginning, many scientists ended up assigning to it a simple mathematical existence (let us recall the declaration of Hertz: “the electro-magnetic field is the equations of Maxwell”). Little by little, scientists became familiar with this expression and today they have no difficulty in developing a ‘physics of the fields’, even ‘absorbing’ into it the notion of the elementary particle (which becomes a ‘singularity’ of the field in a mathematical sense), and by now almost all fundamental physics is dealt with in terms of fields. However, this means that the sciences have created their own ‘regional ontology’, abstract and very far from that of common sense. This is not strange, since ‘contemporary’ physics can be differentiated from ‘modern’ physics particularly because it has largely become a ‘science of the unobservable’ (at least in the immediate and intuitive meaning of this term, as it appears in ‘ordinary knowledge’), and it has lost the requirement of visualizability, which is the strongest element in the relationship with common sense. Now we will briefly mention some other factors (also well-known) which have contributed to widening the gap between science and common sense, between scientific and ordinary knowledge. For common sense and for classical physics, the concept of the simultaneity of two events has an absolute value (they are such if they happen in the same temporal instant). Einstein, however, made it clear that this concept has a real ‘physical meaning’ only if we are able to specify how to determine such simultaneity, and his analysis induced him to admit that it depends on the reference system, and whether it obtains results or not depends on the relative motion of the reference systems. This intrinsic simultaneity of two events can be conceived by common sense only with explicit reference to ‘absolute time’, but this has been excluded by the theory of relativity, which has also made a more radical change, showing that space and time are interdependent. We will now examine other properties that common sense considers ‘intrinsic’ and invariable for a certain body, like the mass and length of a rigid body. Relativity theory shows that they depend on the speed of translation of the body, which in turn is relative to the system of reference. Moreover, according to the well-known Einsteinian formula, mass can change into energy and vice versa, so the old law of the conservation of mass must be reformulated in a more complicated formula of conservation of energy (or of mass-energy). Note that, in this way of thinking, the old conviction of the uniformity of nature, expressed by physical laws, has not

88

Chapter Five

been abandoned, since the theory of relativity represents a great effort to find a formulation of the laws of nature which is invariant with respect to the relativity of the possible measures in the different systems of reference. Considering quantum physics now, it is enough to remember that it places a theoretic (and not only practical) limit on the exact and simultaneous determination of the conjugated magnitudes of a microphysical system (Heisenberg’s indetermination relations) and this is also reflected in the possibility of exact prediction allowed by physical laws (which has only a statistical character). All this is also a consequence of the fact that during measurements it is necessary to realize a ‘preparation’, which makes impossible a clear separation between the observer (or measurement instrument) and the object observed, eliminating one of the more spontaneous characteristics which common sense (and traditional physics) implies for an ‘objective’ knowledge of the world. Finally, even the two rival visions (corpuscular and undulatory) are bound to live together, not actually in separated fields of physics, but in the description of the behavior of the same entity (the principle of complementarity). We can stop here and inquire how scientists accept theories so ‘strange’ to common sense. The answer is that they have not only received many incontrovertible experimental confirmations and made feasible important technological applications, but also that relativity and quantum physics have never been experimentally disproved and therefore must be considered ‘confirmed’ even more was classical physics in its domains. These very experimental confirmations and these successful applications represent the ‘coming back’ to common sense that we have already mentioned, even if they have missed nearly all the possibilities of representation of common sense widely adopted by the old physics which, of course, can be attempted by new physics with new and more complicated tools.

5. The Regional Ontology of Science We can now wonder why this separation has occurred and the way toward the answer is offered by a careful consideration of what we observed at the beginning, that is, that any science is established by directing its cognitive attention toward a limited number of attributes of reality, which thereby form its regional ontology. These attributes are initially perceived at the level of ordinary knowledge, and are translated into concepts or predicates which, rather

From Common Sense Knowledge to Scientific Knowledge

89

than stating them ‘precisely’, specify them so that they can be treated by means of the intellectual and empirical tools that science intends to use and has concretely at its disposal. In this way, such predicates assume a technical meaning deriving from their theoretical context, which only partially overlaps with that of common sense and ordinary knowledge. This happens even more for concepts and predicates introduced starting from the first ones on the grounds of conventional definitions (think e.g. of the concept of entropy). Moreover, sciences with empirical content (like the natural sciences) must also associate at least some of these predicates with operational standardized procedures of referentiality, so that they can express immediately true statements about things (even if limited to the consideration of the selected attributes corresponding to the ‘point of view’ from which they investigate the world). The regional ontology which therefore develops is influenced by these conditions and for this reason it may become difficult to grasp for common sense, even when the discourse of science ‘returns’ to it in the way already examined. With the progress of research, new sciences or scientific branches are instituted from a starting point which is no longer the general ontology of common sense, but some of the regional ontologies already available; these new disciplines choose their objects on the ground of new criteria and new theoretical and empirical tools available, and in this way they move even further away from common sense. This process is natural and legitimate but its very nature imposes necessary precautions, which consist of not forgetting the regionality or the ‘specific delimitation’ of scientific discourses, and therefore of avoiding giving them universal meaning and range. More particularly this happens when are ‘re-launched’ in ordinary speech concepts and propositions of a specific science which, although it uses terms of ordinary knowledge, gives them a different meaning. This at least generates misunderstandings, and it may also bring false reasoning which perhaps pretends to ‘refute’ the convictions of common sense. This possibility of refutation, however, is not at all excluded (and we have already seen some examples), since ordinary knowledge is not something sacred and intangible, and sometimes incorporates mistakes that can be unmasked through a common sense analysis and reasoning, taking into consideration the cognitive results of the sciences. With this, we recognize the historical nature even of the knowledge called ‘ordinary’, that is, the fact that its contents improve, change and are enriched with the historical development of knowledge; this is tantamount to acknowledging that, little by little, many items of scientific knowledge

90

Chapter Five

are ‘absorbed’ by common sense and become part of ordinary knowledge. Indeed, this process is greatly desirable, since it means that the more reliable common sense is that which has better received the cognitive conquests of the sciences (and not only of them). All this poses again the question of the relativization of common sense: instead of a ‘cultural relativization’ we have here an ‘historical relativization’. However, this does not imply relativism; human reason remains a tool to tell what is ‘historically outdated’ from what can be considered valid, not outside the historical conditions but in any historical condition (without pretending that this operation is always fully and irrefutably successful). As we can recognize, for instance, that the duty of revenge can be a common sense moral imperative in certain communities and, nevertheless, we can clearly show, with common sense evidence, that it is not morally valid, so we can try to demonstrate, with common sense evidence, that certain historically shared convictions are false, and this is the case whether they concern, for instance, the moral and the political sphere, or the domains in which the sciences can offer a cognitive light.

6. The Dimension of Totality The non-eliminable role that ordinary knowledge retains in comparison with scientific knowledge remains to be examined, which does not allow the consideration of the first only as the source of the second, and as the more or less passive ‘user’ of its results. This role comes to ordinary knowledge (that often we also refer to as knowledge of common sense) from the fact of being common, taking this characteristic no longer in the sense of being ‘shared’ (ideally) by all men, but in the sense of containing a global vision of reality in all of its dimensions. In this sense, it is an antidote against the reductionist trend, which implacably tempts the various sciences (and maybe science on the whole). For common sense, no aspect or ‘attribute’ of reality is meaningless, since it contains the deposit of all the cognitive and noncognitive experiences of man and is therefore capable certainly not of ‘filling’ but of keeping open that ‘horizon of the whole’ which, vice versa, any science undoubtedly ignores in order to delimit its domain. In other words, it is common sense which advises us that a certain scientific answer ‘ignores certain aspects’ of things or of a certain entity or of a certain problem which in fact are important; such importance cannot be perceived from within a single scientific perspective, while common sense is able to discover and explain it. In other words, common sense can be seen as direct access to that world of life that Husserl reproached

From Common Sense Knowledge to Scientific Knowledge

91

science for having distorted. Unfair charge if we consider the world of life that we observe in our days. Actually, science has literally invaded this world. This charge, however, appears more justified if it is meant to indicate the increasing “distance” between the sciences and the life world. Indeed, science has made more difficult to confer a sense to this world: this conferring comes from other non scientific sources and its value can be increased by common sense, not as a denial, but as a completion of the ‘scientific vision of the world’.

7. Scientism as Opposition between Science and Common Sense We can define scientism as that conception which assigns to science an absolute value, maintaining it to be the only authentic form of human knowledge, which therefore constitutes the more adequate tool to solve all human problems: the strictly cognitive and the practical (insofar as the only ‘rational’ way to solve a practical problem is to adequately analyze it from a cognitive point of view in order to be able to choose those ‘solutions’ which are dictated by the results of such an analysis and by the knowledge of the ‘adequate tools’ which are suggested or even imposed by this analysis). It is therefore clear that, if science is considered the most suitable tool for cognitively analyzing a problem, its solutions will also be ones which science can offer as ‘applied knowledge’, that is, as a direct matrix of technology. Therefore we can consider scientism a techno-scientific ideology. Like many ideologies, it overburdens itself with a utopian dimension: the passage towards utopia takes place when this conception gives rise to a ‘practical commitment’, that is, the commitment to promote and make triumph this vision of the world and of life. This aspect is not trivial, since it shows that even in science it is possible to find something that can give sense to life. From a historical point of view, we can find in the positivism of the nineteenth century the philosophical current which explicitly initiated scientism, without forgetting that its premises are met throughout the course of ‘modern’ philosophy. This philosophical current developed in parallel with the developments of modern science, and from this it drew inspiration not only about the way to conceive adequate forms of knowledge (from Descartes to Kant, even regarding different issues), but also about the practical benefits that science can offer. The philosophical fortune of positivism actually ended at the end of the nineteenth century, and in the twentieth century it was revived only within an important but limited school (new positivism). However, scientism has

92

Chapter Five

remained deeply rooted in the common mentality of western society. It is enough to think that, more or less, we have all been convinced to live in a situation of progress, since we know ‘much more and much better’ than our ancestors, and we have available an enormous quantity of ‘tools’ to guarantee our welfare and our health which were not available to them; it takes little to realize that this increased knowledge and these tools are what has been offered by the developments of the sciences and technologies. Therefore, we all share a bit of scientism and this may even induce us to think that the scientific perspective is by now part of common sense (especially if we intend it in the meaning we have previously outlined). However, things appear vastly different if we consider the difficulties present in this perspective, which are generated just by the relationship that science maintains with common sense. A not insignificant difficulty in the actual realization of scientism’s program consists in the circumstances already analyzed, i.e. that man has a vision of the world and life rooted in common sense which constitutes the frame of reference and meaning in which he places himself, and ‘recognizes’ and ‘understands’ himself in everyday life (we would therefore say in ‘ordinary life’). The various sciences (both the natural and the ‘human’) have induced many modifications to this vision that we might call ‘spontaneous’. Initially they could appear to be idealizations or in-depth studies of certain elements of the frame offered by ordinary knowledge, more or less easily integrable in it; however, with the passing of time, they appeared more and more to be ‘falsifications’ of this frame. It is sufficient to think of the notions and anti-intuitive theories which characterize the contemporary natural sciences. In this way, a real discrepancy occurred between what the American philosopher Wilfrid Sellars called the manifest image and the scientific image of the world, adding that the first is substantially wrong while the second is true. Therefore, if we sincerely want to promote the progress of humankind (simply meaning by ‘progress’ the passage from mistakes to truth), we must do our best to make sure the scientific image replaces the manifest image. This is clearly a utopian aspect, one that Sellars recognizes as an implication (even if not expressly declared) of his scientism. Utopia does not consist, in this case, of a simple practical difficulty, but of a more radical difficulty ‘of principle’. In fact, the scientific image cannot constitute itself unless starting from common sense and from its image of the world - not in a merely intuitive and almost ‘visual’ sense, but in a much deeper one, since the common sense vision which constitutes ordinary knowledge also includes a range of general concepts, categorizations, classifications, linguistic codifications and more or less

From Common Sense Knowledge to Scientific Knowledge

93

implicit ontological and metaphysical conceptions, which have developed over the millennia of human history and which are necessarily employed in the elaboration of the scientific image, which, therefore, cannot at the same time consider the vision of common sense wrong and use it as a reliable tool for its own constitution. Apart from this methodological difficulty, a further difficulty is met when one tries to determine of what that scientific image which should replace the wrong common sense image consists. In fact, the very idea of scientific image is highly problematic, since the various sciences offer many partial ‘images’, tailored according to their specific criteria and methods of investigation, and it is not possible to see how this plurality of images can generate a single, unitary and global ‘scientific image’, in which man can recognize and place himself. Finally, even if this could happen, we would not know on which scientific image we should rely, since it is well known that all the sciences are in a perpetual flow, continually surpassing and correcting their theoretical conceptions; once again we should rely on the utopian achievement of a final scientific image, wholly undetermined and unforeseeable, which should then replace the common sense manifest image. But in the meantime, what should we do? These critical reflections, which, for brevity we have referred to as Sellars’ position, actually strike scientism as such and show that its weakness does not reside in prizing science but in carrying it to the extreme, that is, in ascribing to science the exclusive right of attaining true knowledge, not acknowledging that its range is only partial compared to the broadness and complexity of that world of life which is accepted in the ‘manifest image’ of common sense. By this, we neither want to affirm that the image of common sense is fully valid, nor deny that it develops proportionally to the reception of aspects of the scientific image that it is able to metabolize inside itself. We only want to say that the scientific image is a partial vision of reality which, in order not to become a reduced vision, must carefully consider the vision of the whole of reality as it is accepted in the fullness of human life, with all the variety of accesses, problematizations, and attributions of sense that it puts into practice. What we want to maintain is that the vision of common sense does not coincide with a ‘trivial’ image of the world (which, in the end, is the ‘manifest image’ of Sellars) but is just a frame of reference in which all the factors of the ‘fullness’ of human life appear (in an unanalyzed way); factors which are generally not transformed into an ‘image’, so the existence of a ‘common sense image’ of the world is no less problematic than that of a really identifiable scientific image.

94

Chapter Five

In this sense, it is even possible to realize that, if we want to speak of an opposition between science and common sense, in the end it is the second that wins, as it results from the crisis that has pervaded scientism in recent decades. Scientism’s utopia underwent a strong crisis around the middle of the twentieth century, first on the practical plane and afterwards also on the cognitive one. From the first point of view, doubts have been raised about the fact that the applications of science (technological conquests) are destined to improve human conditions. Instead it has been more and more underlined that they can be dangerous, possibly even for the survival of humankind, and in any case they can worsen, in many ways, our ‘quality of life’. From the second point of view (the cognitive), it is sufficient to note how, in the second half of the twentieth century, some anti-realistic interpretations increased within the philosophy of science, which have overemphasized the fallibility of science, its dependence on social contexts and linguistic-cultural influences, even denying its characteristics, not only of ‘true’ knowledge but even of objective, rigorous, reliable knowledge, superior to pre-scientific and extra-scientific forms of knowledge. This aspect, although very relevant as a factor in the disintegration of scientism, interests us less in this context; in fact, it has been mainly the ‘practical’ aspect (that is, that which discredited the myth of a technoscience which by itself contributes to the improvement of quality of life) that has given rise to a diffuse anti-scientific and anti-technological reaction in many layers of western culture. What does all this mean? It clearly means that a large range of existential factors, in the sense that they directly emerge from the world of life, has obliged our culture to realize the limited character inherent to the scientific image, to the types of conceptualization and the methods of investigation adopted to face reality: they are not even sufficient to solve the non-scientific and non-technological problems which the developments of techno-science are creating and which, vice versa, are evident to everybody and - just because they are part of the life of any human being - form a large part of the actual unity of the experience; that is, they are part of ordinary knowledge in the most authentic meaning of the notion. Therefore, on the grounds of this awareness it is possible to prize the value of science and technology for what they really are and for what they give to man, without, however, making them absolute. This, after all, corresponds to the balanced vision that we are used to assigning to common sense, understood not as an image of reality or as a nonsophisticated way of reasoning, but as an attitude of wisdom.

CHAPTER SIX ON THE DIFFERENCES AND RELATIONS BETWEEN ORDINARY KNOWLEDGE AND COMMON SENSE PAOLO PICCARI

1. Preliminary Remarks on Common Sense According to its main acceptation in cognitive science, common sense is what is common to the understanding and cognitive background of all human beings, to their cognitive processes, unsupported by specific competences, and to the systems of belief that stem from them and that refer to a world of inanimate objects and living beings. As observed by Luigi Pareyson, the history of philosophy is marked by the problem of the relations between philosophic thought and common sense and the solutions that have been proposed.1 Especially in modern philosophy, the term common sense seems decidedly polysemic. Indeed, no philosopher has ever specified which of the various ways he uses it and this has led to problems of interpretation. In the acceptation “common manner of thinking of most individuals”, based on a complex of theoretical and practical notions shared by the entire or most of the human race, the expression common sense has been widespread in the classical lexicon, probably since Cicero, who, following the logic and ethic of the Stoics, considered sensus communis as the object of consensus of the human race (consensus omnium populorum et gentium), a set of innate truths and principles not in opposition to philosophical truths but in continuity with them.2

1

L. Pareyson, Truth and Interpretation, transl. by R.T. Valgenti, State University of New York Press, Albany 2013. 2 See P. Piccari, Conoscenza ordinaria e senso comune, Franco Angeli, Milano 2011.

96

Chapter Six

We owe the first authentic defence of common sense to Thomas Reid, founder of the Scottish School of Common Sense,3 which inspired the mature phase of G.E. Moore. According to Reid, the belief in common sense must necessarily be accepted as ultimate and on a par, not because we have any proof of its truth, but because attempts to refute or derive it lead to paradoxical and contradictory conclusions, for two main reasons: a) common sense even governs the language and behaviour of those who refute it, and b) the judgement of philosophers has no greater authority in these matters than that of the common people, because, “regarding a matter of common sense, everyone is as competent a judge as is a mathematician for a mathematical demonstration”.4 It is therefore evident that since Reid, in British philosophy, common sense has been considered a nucleus of specific theses taken as principles or self-evident truths. In the Scottish School of Common Sense, the notion of common sense takes a prominent position as an instrument of “self-criticism” and renewal of philosophy. It has reappeared in this form in analytical philosophy since its constitution, reputedly offering a bulwark against any type of sceptical phenomenism or constructionism, its principles being epistemic or logical foundations, that is, intuitively evident truths. If the expression common sense, in a wide current acceptation, indicates the pre-philosophical and pre-scientific thought that guides and disciplines linguistic acts, or otherwise the jumble of cognitions and presuppositions that enable individuals to understand language and to formulate inferences, for Reid, instinctive belief has specific contents: it is indeed “common” belief, shared by all humans, in the existence of the external world and of the perceptual subject, the difference between the subject and the external world. This is why common sense is generally used in analytical philosophy to claim that phenomenism is senseless and that the external world exists independently of our representations of it and our existence. The different acceptations that the expression common sense has acquired over the centuries are perhaps sufficient to understand the ultimate difficulty not only of defining its meaning, but also of establishing whether it is epistemologically more appropriate to propose distinct levels of analysis that take its semantic stratification and heterogeneity into account. First of all, before attempting to draw a line around the meaning of common sense beyond its “positive” or “negative” connotations, it is worthwhile examining the adjective “common”, which 3

See S.A. Grove, The Scottish Philosophy of Common Sense, Clarendon Press, Oxford 1960. 4 Th. Reid, Essays on the Intellectual Power of Man [1785], VI, 4, ed. by B.A. Brody, The MIT Press, Cambridge (MA) 1969, 611.

On the Differences and Relations between Ordinary Knowledge

97

by virtue of its polysemic nature lends itself to different interpretations: it may mean “shared by various individuals” or “belonging to all”, as well as “ordinary, everyday, usual”. Besides these acceptations, we must also consider the negative connotations of the term “common”, namely secondrate, mundane, vulgar, trivial and so forth. In any case, the prevailing concept of common sense seems to be “common perception”, “common thought”, “common reason”, based on knowledge, ideas, opinions, principles, prejudices and reasoning about which all humans should agree, since their simplicity does not call for specific competence. According to this concept, common sense is a sort of large store of knowledge, ideas, opinions, prejudices and reasoning that everyone should accept because it can be inferred from ordinary experience without any specific competence; a network of certainties that tradition and experience have gathered in a theoretically unauthenticated form, albeit founded on general consensus; a series of beliefs considered “common” to individuals of a given historical and cultural period, not based on specialist competence but widely accepted as “correct” and “non problematical”, even in the absence of a reasoned basis.5 Being a set of beliefs shared by most individuals, common sense becomes a source from which to draw propositions and general principles for elaborating plausible forms of reasoning. From this perspective, common sense is like an articulated system of widely shared beliefs, considered obvious and therefore ranking as common convictions: thus, for example, common sense beliefs include the ideas that the world is older than we are, that it existed before we were born and that it will continue to exist after our death. This system sustains a constellation of inferential or argumentative procedures that are also used in a spontaneous and unconscious way. In the field of cognitive psychology and science, expressions such as “common sense belief”, “common sense reasoning”, and “world of common sense” are often used in contraposition to beliefs, knowledge, reasoning and worlds with different characteristics. In this context, common sense seems an uncertain cognitive field due to the imprecision (or falsity) of its assumptions and the incorrectness of its inferential procedures, despite the fact that its function is undoubtedly useful for the human species, if not indispensable. This notion of common sense is certainly problematical, first of all because it assumes that all humans share at least part of their “ordinary

5 See E. Agazzi, Normatività logica e ragionamento di senso comune, in F. Castellani, L. Montecucco (eds), Normatività logica e ragionamento di senso comune, il Mulino, Bologna 1991, 45-46.

98

Chapter Six

experience”, understood as a sequence of different empirical instances. In actual fact, similar empirical instances can be interpreted in very different ways by different individuals, giving rise to opinions and judgements which are often so much at variance as to discredit the idea that common sense can be understood as a body of ideas and judgements largely shared on the basis of facts. How, then, is it possible to delimit a nucleus of ideas, opinions, judgements, prejudices, beliefs and inferential procedures that all humans should share? In fact, here it is clear that common sense loses the universal character, attributed by some, and acquires the characteristics of a “cultural system”, referring to opinions found in most members of a specific social group in a given historical moment. If we want to define common sense as a complex system of beliefs about the nature of the physical, human and social world, largely accepted by the individuals of a society, by means of which beliefs it is possible to explain events, analyse and interpret one’s own and others’ behaviour and therefore act in the world, we cannot ignore the fact that each tiny part of such a system is connected through an intricate network of cognitive processes. In considering the notion of common sense in its two distinct semantic units, we cannot fail to observe that as a whole, what it refers to is anything but “common” or “universally agreed”, but rather denotes variability and uncertainty in different historical, cultural and geographic contexts, observable just as much in an ethnic group as in a small community of individuals. In any case, it is empirically difficult to establish the degree of diffusion and sharing (commonality) of a system of beliefs, that is, how common it is to societies, classes, communities or small groups. So if we exclude the possibility of “common sense” as a system of beliefs shared by all human beings, since no beliefs are innate, it seems reasonable to define common sense as a set of beliefs (supporting rules of conduct, social conventions, mental attitudes and values) shared by a human group in a given historical moment, hence the legitimacy of the connotative character “common” attributed to this set. Thus defined, common sense does not coincide, as it might seem, with the notion of “ideology” formulated in Marx’s German Ideology to refer to the ideas of the dominant classes, which thus become prevalent. Indeed, common sense in the above acceptation does not form the set of beliefs of a dominant class or classes, but is rather a collection of subsets of beliefs of

On the Differences and Relations between Ordinary Knowledge

99

different subgroups, opinion movements, political factions, youth movements or lobbies, present in complex post-industrial societies.6 From this point of view, common sense can be considered a superset of its various parts or subsets, each of which is a set of beliefs shared by some of the members of a group or social stratum, who use it to elaborate their Weltanschauung: that view of the world proper to each individual, which is the basis for their daily actions, as well as many of their choices and the sense attributed to their existence. If common sense CS is a superset of widely held beliefs in a human group, we can write the collection of subsets of beliefs of CS as S1, S2, S3..., Sn, obtaining the expression CS = S1, S2, S3..., Sn.. Defined in this way, common sense fulfils various functions in a human group, the most important of which are: a) interpreting the human, social and family world; b) planning and carrying out actions; c) evaluating and judging actions and events, and d) predicting and preventing potentially harmful events.

2. Knowledge and Survival What is the role of knowledge in human life and more generally in the life of living organisms? According to the solid Aristotelian tradition, knowledge is an end in itself and has no practical utility.7 It is, however, doubtful that knowledge is an end in itself. Knowledge is indispensable for the survival of the Earth’s organisms: being heterotrophs,8 they survive by exploiting energy sources and avoiding dangers (Fig. 1). 6 I use the term “set” (and its hyponym “subset”) as an intuitive and primitive concept for a broad collection of elements (in this case beliefs), following George Cantor’s set theory. 7 See Aristotle, Metaphysica, A 2, 982 b 20-24, ed. by D.W. Ross, Clarendon, Oxford 1924. 8 Heterotrophy (from the Greek ‫ۇ‬IJİȡȠȢ (different) and IJȡȠijȒ (nourishment)) is the nutritional condition of an organism unable to synthesise its own nourishment autonomously from inorganic substances. To survive it requires organic compounds pre-synthesised by other organisms, which are known as autotrophic, for example plants with chlorophyll. Autotrophic organisms are those able to live solely on simple inorganic substances, such as Plantae, which only require carbon dioxide from the air, water and mineral salts from the soil. Plants are photoautotrophic because they use the sun as energy source. In much rarer cases, such as those of certain bacteria, the organism obtains the energy it requires from the oxidation of inorganic substances (chemoautotroph). Heterotrophic organisms feed on organic substances produced by autotrophic organisms: Animalia feed directly (herbivores) or indirectly (carnivores) on plants. An important case of

Chapter Six

100

Living Organisms

Autotrophic Nutrition

Photoautotrophy (Plants)

Chemoautotrophy (some Bacteria)

Heterotrophic Nutrition

by Absorption (some Bacteria and Fungi)

by Ingestion (Animals)

Fig. 1 Autotrophic and heterotrophic organisms

To achieve this aim, all beings need to know the environment in which they live and acquire information that enables them to behave in a way that enables them to survive.9 This type of knowledge can be defined as “biological”, since it is necessary for survival; it is present in all organisms, even to some extent in those without a central nervous system (CNS). Thus, all organisms are cognitive systems and life itself owes its existence and conservation to a cognitive process. Even elementary organisms, such as prokaryotes, which appeared on the planet about four billion years ago, are cognitive systems. Life in its various forms depends on a cognitive process, because every new item of knowledge acquired increases the possibility of obtaining energy. All forms of life are cognitive systems that conserve themselves through cognitive processes. In the case of prokaryotes, the quantity of information is clearly modest; it is elementary knowledge to enable these organisms to respond adequately to stimuli from the environment, which they receive with rudimentary sensors. More evolved organisms need more information, especially if they live in a complex environment. In order to respond to environmental stimuli and have a concrete possibility of survival, they require more articulated knowledge, which they acquire through much more sophisticated sensors than those of prokaryotes. In this sense, knowledge fulfils a biological function, because the survival of organisms depends on their identifying

heterotrophism is that of decomposers, which feed on the organic detritus of plants and animals in soil. All animals, protozoa and fungi and almost all bacteria are heterotrophs. 9 On the role of knowledge in nature and its biological function, see M.L. Bianca, Rappresentazioni mentali e conoscenza. Un modello teorico-formale delle rappresentazioni mentali, Franco Angeli, Milano 2005, 15-43.

On the Differences and Relations between Ordinary Knowledge

101

particularly favourable objects and places, such as food and shelter, as well as on avoiding poisonous substances and dangerous situations. Biological knowledge therefore consists of a set of information regarding the external world or one’s own body, acquired through sensory organs, which in the case of more evolved animals, are then processed by specific CNS structures. It is knowledge of the phenomenal world, the primary function of which is to enable identification and use of energy resources to ensure individual survival and the survival of the species, or more precisely, thermodynamic equilibrium of living organisms. Clearly this form of knowledge has different levels of complexity in different animal species. For example, Thompson gazelles recognise lions when they see them, and for an alerted gazelle, a lion is a lion irrespective of where it is viewed from: close or distant, stationary or in movement. The mere appearance of the lion, however, is not sufficient to trigger flight behaviour in the gazelle, unless the lion seems disposed to attack. Herds of gazelles often approach lions (as close as 100-200 metres) to check their movements. If the lion moves, the gazelles or a buck follow it, aware of the danger and ready to flee and alert the herd at the first sign of attack.10 Thus the mental representation “lion” in the gazelle’s CNS, which takes the place of a given lion, is a brain state that can trigger another brain state, that of flight, hence flight behaviour and flight itself. Biological knowledge is therefore based on perceptual activity, both in humans and in other animals. According to the empiricist tradition, perception is the foundation of cognitive experience, because individuals access the external world epistemically through perceptual processes; in other words, perceptions are relations with the world. If, for example, we limit our consideration to visual perceptual activity, it is clear that there is a passive condition with respect to the perceptual stimulus. I can inhibit my visual perception by closing my eyes, but when I reopen them I cannot prevent the retinal photoreceptors from acquiring and transducing physical information. In other words, I cannot avoid receiving information under conditions in which I am exposed to receiving it. Individuals are evidently able to perceive the world visually without any type of thought, though they cannot say that they “see something” but rather that they “experience something visually”. This expression cannot be reduced or likened to simply “seeing something” because it refers to the

10

See D.R. Griffin, Animal Minds, The University of Chicago Press, Chicago 1992.

102

Chapter Six

experiential condition of the subject during the visual perceptual process and his relation to the world.11 This is why subjective conditions are important when visual perceptual stimuli are received: not only do they regulate the processing of visual information, but they may also orientate visual perceptual experiences, selectively anticipating some element of the information. This aspect is clearly interesting for contemporary epistemology and hermeneutics, since from this perspective, what humans experience at least partly depends on being disposed to certain interpretative choices, which determine on one hand which of the many stimuli will be selected, and on the other, the way in which they will be processed. It follows that visual perceptions are always collocated within subjective global experience, where interest and need factors often intervene and may orientate actions and the formation of certain kinds of knowledge. In any case, the relation between perception and cognition is clearly a controversial philosophical problem, involving many hypotheses of the cognitive sciences, which seem to indicate the inflexible otherness of the two, even if in some cases we glimpse the widespread idea that perception and cognition are sequential functions. For example, we collect and codify sensory information that is then ordered, classified and conceptualised by cognitive operations.

3. Perception and Ordinary Knowledge 3.1 The Perceptual Experience Perception is undoubtedly a fundamental process for the conservation of life on our planet, and in one form or another it concerns all living organisms. As we saw, it is the primary source of biological knowledge. The innate perceptual and cognitive instruments that formed in the course of phylogenesis are our first tools for “reading” and “interpreting” the world. These tools enable us to establish causes, formulate choices and make predictions, which, however, are always limited to the world in which we evolved. Clearly, each species lives “immersed” in a particular perceptual and cognitive world that is different from that of any other species: what Uexküll called the Umwelt of every species, namely the life dimension of that particular animal. From an epistemological viewpoint, this is the notion of dominion, of a space correlated with the cognitive instances 11

For reconnaissance on the topic, see L. Foglia, Percezione visiva. Prospettive filosofiche ed empiriche, Franco Angeli, Milano 2010.

On the Differences and Relations between Ordinary Knowledge

103

of a given species.12 For example, in the environment where our species was selected, every change in the structure of a scenario involved a preferential interpretation: change is interpreted attributing movement to one or more subjects with respect to an immobile background. This can certainly lead to epistemological errors, such as thinking that the Sun orbits the Earth, but it is still important from an evolutionary point of view to have a perceptual system that allows us to distinguish a moving subject from its immobile background. It is therefore possible to define perception as a set of biological processes that allow information about the world to be acquired for the purpose of knowing its state and acting appropriately towards specific objectives. Why do living beings have a perceptual system that collects information on their environment? Is it possible to find a reason that enables us to answer this question in a satisfactory way? The capacity to perceive, which is a fundamental characteristic of life on earth, comes from the fact that evolution determined the appearance of organisms that are not self-sufficient; as already underlined, living beings have to seek energy in their environment in order to keep themselves in thermodynamic equilibrium. In this context, life is inextricably linked to its environment. Living organisms, generated in the terrestrial environment, could only evolve for life here, acquiring tools to enable it; hence the appearance of perceptual systems for energy sources and locomotor systems to reach them. Each species lives “immersed” in a particular perceptual and cognitive world, different from that of other species. Within this specific domain, the perceptual system can fulfil the functions of orientation for which it was selected in the course of phylogenesis. All living organisms need to acquire energy from outside, which is why they must have sensors to allow them to recognise and distinguish energy sources. Hence the correlated need to have mechanisms to enable physical movement in space. If this were not the model of life on the planet, in other words, if organisms were not dysenergetic, then sensors would no longer be useful and it would be legitimate to wonder why an organism had eyes. Under earth’s conditions, however, the answer to the question can only be based on objective biological-evolutionary evidence: eyes are for knowing this world, because they “are made” for this world. Perceptual systems therefore evolved in the course of phylogenesis to “read” and “interpret” this and only this world (even if they could fulfil their function in another world similar to our own). Since they were “made” for this world, we may well ask if evolution failed by favouring the appearance 12

See J. Von Uexküll, Umwelt und Innenwelt der Tiere, Springer, Berlin 1909.

104

Chapter Six

of useless perceptual systems or whether evolution was at least partly successful in favouring the appearance of perceptual systems “adapted” or at least suitable for acting in this world. If the answer were that biological evolution failed, then we would have to conclude, for example, that our eyes have no function and so we could do without them. However, biological evidence seems to contradict this thesis, if only for the fact that, pathology aside, eyes are useful for detecting the position of objects in space, which is fundamental for moving about in the world. If the answer were that biological evolution was at least partly successful, then the question would be naïve and the answer evident. Accepting such an answer would mean recognising that perceptions provide variably reliable information on how the world is; on how it “appears” and not on its structural constitution or on the “laws” that guide and regulate its processes. In the opposite case, we would have to maintain that biological evolution gave humans false and useless systems of perception, which are unable to provide reliable and adequate information for knowing the environment and for acting within it. In other words, our systems of perception would be deceptive. In philosophy, in fact, there have been recurrent attempts to cast doubt on the reliability of perceptions on the basis that sensory organs are deceptive. In the first of his Meditationes, Descartes stated that our knowledge comes from the senses, which are sometimes deceptive and therefore unreliable; in this way he favoured the development of a sceptical attitude. Although in certain cases the information provided by sensory organs proves to be wrong, in many others it is entirely reliable. He perfectly exemplified the condition of extreme scepticism with his famous deceptive god/evil demon argument: Verumtamen infixa quaedem est meae menti vetus opinio, Deum esse qui potest omnia, & a quo talis, qualis existo, sum creatus. Unde autem scio illum fecisse ut plane sit terra, nullum coelum, nulla res extensa, nulla figura, nulla magnitudo, nullus locus, & tamen haec omnia non aliter quam nunc mihi videantur existere? […]Supponam igitur non optimum Deum, fontem veritatis, sed genium aliquem malignum, eundemque summe potentem & callidum, omnem sua industriam in eo posuisse, ut me falleret: putabo caeulum, aërem, terram, colores, figuras, sonos, cunctaque externa nihil aliud esse quam ludificationes somniorum, quibus insidias credulitati meae tetendit.13

13

R. Descartes, Meditationes de prima philosophia [1641], I, 23-29, introduction and notes by G. Rodis-Lewis, Vrin, Paris 1978, 23.

On the Differences and Relations between Ordinary Knowledge

105

How can we exclude the possibility of being in a similar condition? This is the sceptic’s challenge in a nutshell: it casts doubt on the real existence of the external world and the possibility of knowing it. A radically nonepistemic concept of truth underpins this attitude. It is a concept that substantiates all forms of scepticism regarding the existence of the external world and that justifies arguments like the one according to which humans do not live the “normal” cognitive lives they believe they are living because they are under the influence of a powerful evil demon, who makes them believe they are perceiving the world exactly as they have always perceived it, whereas they have been asleep for years in their beds. Yet perception has characteristics that make it unmistakeable, so it is difficult to confuse dream with reality. According to Locke, if you look at the sun in a dream, the eyes do not hurt.14 The Cartesian hypothesis has a well-known contemporary version: the so-called “brains in a vat” argument of Putnam,15 which holds that we can have systematically false knowledge of the external world because everything we perceive is the result of manipulation of our phenomenology of experience by an evil scientist who keeps our brains alive in a vat of nutrients, connected to a computer which deceives us that everything is completely normal. We perceive people, objects, sky and so forth, but in actual fact our experience is the result of electronic impulses sent by the computer to our nerve endings, simulating our mind’s needs. The sceptic sustains that we cannot determine whether or not we are brains in a vat: if everything we believe and relate to is only an illusion, it means that we have no secure tools to uphold the legitimacy of our perceptual representations of the world. The brains-in-a-vat argument evidently does not dispute the existence of the world, but rather that the world we know is actually the existing world.16 Putnam’s reply is based on the causal theory of reference, as well as the old argument of the self-refuting character of scepticism. According to Putnam, if we were brains in a vat, we could not realise this fact while living an apparently “normal” life and believing our experiences and perceptions were ordinary. Indeed, we could not even raise the question, since as brains in a vat, unaware of our condition, our cognitive horizon would be precluded: as 14 J. Locke, An Essay Concerning Human Understanding [1690], IV, 2-14, ed. by P.H. Nidditch, Clarendon, Oxford 1975. 15 See H. Putnam, Reason, Truth, and History, Cambridge (MA), Cambridge University Press, 1981. 16 See M. Devitt, Realism and Truth, Princeton University Press, Princeton 1991.2 60-82; M. Dell’Utri, Choosing Conceptions of Realism: The Case of the Brains in a Vat, in «Mind» 99, 1990, 79-90.

106

Chapter Six

brains in a vat we cannot refer to any real object but only to imaginary objects. Thus, when our brain thinks “we are brains in a vat” it refers to an imaginary vat, an imaginary liquid, and so forth. It follows that were we effectively brains in a vat, the assumption “we are brains in a vat”, which is supposed to refer to the real world, would be false, since brains in a vat can only refer to an imaginary world. Terms such as “chair”, “bed”, “tree”, etc. would not be determined by their respective physical objects (chairs, beds, trees and so on) but simply by stimulation from the computer guided by the evil scientist. Thus scepticism is refuted and with it all forms of metaphysical realism. The sceptic and the metaphysical realist agree on the existence of a completely preconstituted world, independent of our representations, although the former insists on its unknowability while the latter aspires to a correspondence theory of truth. On their side, psychologists can cite many examples to show a certain unreliability of perceptions, but at the same time they are forced to cite many more examples of reliable perceptions, as in most situations of everyday life. Considerations about the biological objective of perception so far formulated suggest that it is a simple process, whereas it is actually particularly complex due to the variety of systems involved and their wide diversity in living organisms. In humans, processes of perception are clearly much more articulated than in other species, although they share similar objectives: this depends on the greater complexity of the human CNS, which not only receives and processes information from the world, but subjects it to innumerable neuro-mental processes, in turn influenced by signals from the body, relations the body establishes with the environment, information about the self’s personality and cultural superstructure, which have a great effect not only on simple perceptual data but also on how it is processed in the CNS. These aspects make the processing and codification of information much more complex in humans than in other species. Considering the complexity of perception in humans, it is legitimate to ask whether the final elaboration of the information received from the world always reliably and adequately reflects the structure of the world. This question is not worthwhile for the perception of a gazelle or a warthog, since it seems that in other animals the processing of sensory information does not involve errors that would lead, for example, to zebras being mistaken for lions. Indeed, beyond reception difficulties, their perceptions match states of the world. In this regard, it is interesting to consider the meaning we assign to the terms “reliable” and “adequate” with reference to knowledge of the world. Knowledge is defined as reliable if it somehow corresponds to the structure of things at the time it is formulated. The term “adequate” refers to

On the Differences and Relations between Ordinary Knowledge

107

the relationship between knowledge and its intended purpose: knowledge is adequate if it is useful for reaching a certain aim.17 What can we observe in humans? If, for example, we examine a simple visual perception, such as a book on a table, we can observe that the perceptual content does not consist solely of the visual information that arrives from the object to the eyes and the visual cortex, but also of other information existing in the CNS. This is why we not only visually perceive an object with certain physical attributes, but a “book”, which is such because it has been identified cognitively, that is, we assigned it attributes not directly derived from photonic reception, such as “an object useful for learning”, “an object consisting of printed pages”, “the book on my desk”, etc. Non phenomenal factors, such as the concepts “book”, “page”, etc., are involved in the formation of this perception. What, then, is human perception? It cannot be reduced to mere reception of information from the external world, because perception is a more complex phenomenon involving more than merely phenomenal information, i.e. information derived from sensory reception. What role does perception play in the mind’s economy besides the main role of recognising world entities? Is perception a merely cognitive process or can it trigger various other processes, such as emotion or affect, or even behaviours and actions? Beyond the specific contributions of philosophical investigation, the study of perception has acquired more fertile articulation and methodology through empirical research. For example, experimental research has revealed that perception is a specific experience, above all a cognitive experience, in which many aspects of the perceiver are involved: mental, bodily and emotional conditions, conceptual schemes, and often the objectives and motivations that precede (causally or otherwise) a perceptual process.

3.2 The Visual Perceptual Experience If we return to considering visual perception, we observe that seeing, looking and watching, i.e. the reception of visual information, is a particular perceptual experience. Thus we do not see something but we visually experience it: the expression “visually experience something” refers to the processing of visual perceptual information, image formation and memorisation.

17

I am obviously not asserting a “necessary correspondence” between knowledge and the structure of the things it refers to.

108

Chapter Six

Visual perception, like other types of perception, is not an independent mental process, but occurs within a broad process that we indicate with the expression “visual experience”. The theatre of visual experience, like other types of perceptual experience, is not limited to the brain areas assigned to processing visual information, but involves other mental content and bodily states, as well as worldly objects that stimulate retinal photoreceptors (rods and cones). This is why we need to consider a subject’s state and relationship to the world. Every visual experience is a relation between the subject and the world: “perceive visually” means establishing a relationship with the world, characterised by the reception and processing of visual information. Visual information is therefore the relation between subject S who has a system of sensors for receiving photons (or electromagnetic waves) and a source that emits them; in a visual process, S receives photons (or EM waves) that may trigger a neurophysiological process. A fundamental element of the relationship between the perceiving subject who receives the information and the emitting object is a modification of subject S caused by the reception of stimuli emitted by object O. Subject S and object O must also be in the same space and time and S must be exposed and disposed to the reception of a stimulus at the moment it is emitted; S is exposed if temporally and spatially present at the moment of emission of the stimulus, and is disposed if able to receive the stimulus. Spatial-temporal coincidence and the dynamics of exposure and disposition are necessary conditions for any type of perceptual experience. These conditions only permit the experience when they are met, otherwise there can be no experience, as when a television is turned on but S is not present or able to receive the stimulus. It is worth noting, however, that the disposition to receive a stimulus is not a sufficient condition for reception, because it is also necessary to be able to receive it; for example, one can be disposed to receive a stimulus consisting of sounds of a few decibels but they will not be received if the subject has impaired hearing. To be able to receive means having the necessary instruments to receive certain information or stimuli. The perceptual experience does not consist solely of the reception and processing of perceptual information, but is a complex process that establishes an articulated relationship between S and O. It follows that to experience, as we have noted, not only involves the mind but the entire subjectivity. The results are derived from perceptual processes that generate different modifications in the mind; for example, subject S changes when it receives photons/waves from O, but O does not change (except in special cases) due to the presence of S or the processes occurring in the mind of S.

On the Differences and Relations between Ordinary Knowledge

109

Seeing my desk is a visual experience that generates modifications in my mind, but the desk remains unchanged whether or not I observe it. The fact of observing it does not trigger any causal modification in its matter or function. Object O is therefore the cause of my neuro-mental processes, whereas my visual perception of O does not cause any modification in O. In the visual experience, it is the subject that sees, so not only the organs of sight, from the retina to the primary and associative visual cortex, are involved. Received visual information is not only processed, but triggers a complex response in the subject, involving cortical, subcortical and noncortical, as well as motor areas of the brain. Thus the perceptual experience consists of a wide, articulated, radiating neurophysiological process, the information of which circulates in a branchover process, involving different areas and structures of the brain. Sensorimotor areas are involved too, so the body is not only disposed and active during perception but changes as well, by virtue of the results of the branch-over processing of perceptual information. The biological function of vision is to enable humans to act adequately, because there is a structural relationship between images and the world. The images are not “photos”, although they “report” the figured attributes of objects in the physical world through specific neuronal structures. The same is true, albeit in a different way, for the other forms of perception, such as hearing, touch, etc. It is legitimate to ask whether the results of visual perceptual processes provide a reliable representation of the world or a representation that matches the state of things. This question can be overcome if we consider the biological and mental role of vision. Now for an ingenuous question: why do we have eyes? Simple reflection on individual experience leads to the realisation that eyes are the biological structure enabling visual knowledge of the world, indispensable for moving about, avoiding obstacles, seeking energy sources, establishing relations with the environment and the objects or other beings it contains. This question is often asked by psychologists and philosophers and is only apparently ingenuous, being fundamental in substance. According to Dretske, it is necessary to distinguish between seeing objects, which is simple seeing, i.e. a type of seeing independent of recognition, and seeing facts, which is epistemic seeing through recognition based on the use of concepts.18 Simple seeing is preconceptual, whereas

18

See F. Dretske, Simple Seeing [1979], in Id., Perception, Knowledge and Belief, Cambridge University Press, Cambridge 2000, 97-112; Id., Seeing, Believing and Knowing, in D.N. Osherson, S.M. Kosslyn, J.M. Hollerbach (eds.), Visual

110

Chapter Six

epistemic seeing is conceptually mediated and requires at least a vague notion of the perceived object. According to this theory, it is possible to see an object without necessarily recognising it cognitively, i.e. it is possible to see without having any belief about the object, considering recognition equivalent to a belief. Philosophers who maintain this position, while admitting that in many cases what we perceive visually determines our beliefs and classifications, observe that there are cases in which visual perception is completely independent of our beliefs, consisting solely in the detection of information from the environment. For example, I can walk along a corridor that contains an ottoman. I may not have noticed that it has damask upholstery with silver stitching: when I perceived the ottoman visually I did not also “see” its particular upholstery. This means that simple seeing consists in distinguishing objects as discrete units with respect to their background. In other words, if subject S sees an object (e.g. a book) simply, it means that he sees certain attributes of the book, i.e. he sees the book distinct from the other objects in his visual field without necessarily believing that it has the appearance of a book or that it is a book. Thus simple seeing does not involve perceptual beliefs; in other words, seeing a book does not imply believing that it looks like or is a book. However, if I see a book simply, from a certain point of view it looks like a book: can I visually perceive an object that looks like a book without believing that what I perceive visually has the appearance of a book? According to Dretske’s, I could be faced with a book in full sight on the table without my eyes avoiding perception of it; if I am distracted, my visual system perceives the book by “simply” distinguishing it as a discrete unit with respect to the background. On the other hand, when we recognise and classify an object, that is, when we “see” a book and believe it has the appearance of a book or is a book, then simple seeing becomes another type of experience, namely epistemic seeing. Let us consider the different roles of simple and epistemic seeing with regard to ordinary knowledge. In simple seeing, the subject obtains visual information on the phenomenal world, much of which, however, is not immediately used by the perceiver. When we obtain a certain type of information by centring our attention on an object, that is epistemic seeing, or seeing in which we obtain knowledge of certain characteristics of the

Cognition and Action, MIT Press, Cambridge (MA) 1990, 129-148; Id., Naturalizing the Mind, MIT Press, Cambridge (MA) 1995.

On the Differences and Relations between Ordinary Knowledge

111

object through our visual system. If simple seeing is a source of knowledge, epistemic seeing is a type of knowledge. Thus there are two types of seeing: simple seeing, which is a source of knowledge, and epistemic seeing, which is a type of knowledge and presupposes simple seeing. Of course it is possible to move from simple to epistemic seeing by exercising our classification capacity, when we extract salient information from the information forming the content of the first type of seeing. Making a distinction between simple vision and epistemic vision is not easy, mainly because it involves the dichotomy between perception and cognition. One of the arguments sustaining this thesis is that it is intuitively evident that an individual can visually perceive an object without saying what it is. In other words, it is possible to see objects simply, having a direct relationship with them. Indeed, an individual who has no notion of physics does not see a synchrotron as an accelerator of particles but simply sees something, without drawing on cognitive expertise or conceptualisation. Beyond Dreske’s theory, the schematic nature of which raises many doubts, it is nevertheless plausible to distinguish between perceptual states, the content of which is not conceptual, and cognitive states, the content of which is conceptual. The former can happen without the perceiver having the concepts necessary to specify their content. It follows that, while it is necessary to have the concepts “eagle” and “sky” to believe that “an eagle is hovering in the sky”, it is not necessary to have those concepts to perceive “an eagle in the sky”. A non conceptual state S is therefore a state with a content that may be characterised by specific contents, possession of which, however, is not necessary for a subject to be in state S. This means that state S has a non-conceptual content P if the subject does not have specific concepts to characterise P. Unlike Dretske, McDowell holds that experience not only presupposes knowledge but is already a conceptual content per se; conceptual capacity is the only way to regiment and organise sensory stimuli.19 Perceptual states therefore have a conceptual content; for every perceptual state there exists a concept that can specifically characterise the content of that state. Actually, when a subject perceives a state of things, that state is represented in some specific way, whatever that may be, and according to the point of view from which it is perceived. If this is true, we have to admit that perceptual states have a conceptual content, because otherwise how could the content of a perception “report” a state of things in a certain way 19

See J. McDowell, Mind and World, Harvard University Press, Cambridge (MA) 1994.

112

Chapter Six

without the subject having concepts by which to characterise this specific way of “reporting” a state of things? We could object that if, for example, an individual without any notion of physics and a physicist were faced with a cyclotron, the contents of their respective perceptions would be different; they are both at the same observation point but only the scientist “sees” the cyclotron as a cyclotron. The other does not “see” the cyclotron as such. Thus, while the non-conceptualist perspective has some aspects that seem able to solve some of the difficulties inherent to the conceptualist approach, it still has to be considered in the light of certain recent studies on perceptual processes, which seem to confirm the argument that the process by which visual stimuli are processed involves both information from the phenomenal world and signals from different areas of the brain cortex. This means that to see we not only use information already acquired, processed and organised, also by concept formulation, but also a series of information of non-retinal provenance (beliefs, expectations, associations, etc.) useful for discriminating and recognising objects quickly. In other terms, visual perception of an object never previously experienced, or not “conceptualisable” or “recognisable”, is a perceptual instance and the potential start of a process of perceptual instantiation, which may not but can conclude with a generalisation conducted by the perceiving subject if there are further subsequent perceptual instances of similar objects. It is therefore possible to consider the content of a visual perceptual state as a minimum conceptual unit, since it is often but not always a prelude to a process of perceptual conceptualisation.20

3.3 Ordinary Knowledge Ordinary knowledge, which also forms on the basis of perceptual activity, especially visual perception, as discussed above, is a more evolved and sophisticated form of knowledge than biological knowledge (regarding survival), not only by virtue of the different conditions that make it possible, but also the breadth of its articulation, mainly determined by the cultural modalities typical of a complex cultural community. One of the functional 20 On the nature and the formation of empirical concepts, see M. Bianca - P. Piccari, Conceptual Framework: A Frequency Model, in Id. (eds.), Concept Formation, Special Issue of «Anthropology & Philosophy», 10, 2009/2010, 9-29. See also J.J. Prinz, Furnishing the Mind. Concepts and Their Perceptual Basis, The MIT Press, Cambridge (MA) 2002; G.L. Murphy, The Big Books of Concepts, The MIT Press, Cambridge (MA) 2002; E. Machery, Doing without Concepts, Oxford University Press, New York 2009.

On the Differences and Relations between Ordinary Knowledge

113

modalities of the organisation of ordinary knowledge in the world, through the use of sensory organs, is mental representation, a neuronal operation triggered by sensory stimuli that activate a number of interconnected neural networks. The life of individuals is also made possible by the fact that in most cases, each has adequate knowledge of self and environment. It is evidently an anomalous form of knowledge in the sense that it regards self-knowledge of a biological kind, concerned with needs and capacities, as well as knowledge of the environment, if not why things happen. These forms of knowledge (self-knowledge and knowledge of the environment) are not only important because they ensure individual survival, but also because they enable life to unfold according to the various expectations of humans and other higher animals. In fact, humans and all other organisms need to know themselves, their fellows and their habitat in order to satisfy their natural needs, including eating and drinking and defending themselves from events like earthquakes and floods. Ordinary knowledge does not consist solely of information derived from experience, but also of knowledge of other conspecifics assimilated by learning. This process developed in an articulated way from the first knowledge of small groups or tribes to current education systems. From this perspective, it is clear that ordinary knowledge is also permeated by the results of scientific research, which are often learned from traditional education and training agencies, and which also favour thought and perceptual modes useful for collecting information derived from experience. This consideration seems to introduce an element of incoherence into the notion of ordinary knowledge so far discussed. Although there is undoubtedly constitutive otherness between ordinary and scientific knowledge from an epistemological point of view, there can actually be as many beliefs based exclusively on perceptual information as knowledge or world-views derived from scientific theories.

4. Conclusion We have analysed ordinary knowledge and common sense to define their structures and clarify their functions in human life. In particular, we underlined the importance of biological knowledge, which enables beings to acquire the information necessary for their survival. This form of knowledge, which is formed through the processing of information about the external world or the body, acquired through the sensory organs, works in a similar way in humans as in other animal species. However, in humans, biological knowledge is especially complex due to the human neocortex and

114

Chapter Six

the influence of supra-individual information structures generated by social and cultural life. Exceptions aside, this means that biological knowledge in humans cannot be reduced to the outcomes of sensory-perceptual processes. Indeed, biological knowledge in humans is formed through information collected by receptor organs and information from different areas of the CNS. Cognitive processes became increasingly sophisticated in Homo with the advent of the extraordinarily complex neocortex, and in almost all cases cannot be described in purely biological terms. In our species the neocortex constitutes about 90% of the brain’s surface and represents the latest phase of a process that began with the evolution of the CNS in the first mammals. Thus, biological knowledge in Homo becomes ordinary knowledge, which, as we have seen, is much more complex than the mere reception of information and its subsequent processing. In this sense, ordinary knowledge exceeds and transcends biological knowledge, although it encloses information derived from sensory-perceptual activity that can also be modified by other information from various parts of the CNS. Unlike biological knowledge, this knowledge does not map isomorphically onto the external world but differs from it epistemically. In other words, isomorphism with respect to world structures that characterises perceptions, making them reliable and adequate, may be altered or impaired to different degrees by subsequent processing of perceptual information.21 Indeed, in ordinary knowledge, perceptions are processed and can be “contaminated” by non-perceptual information; the contents of the primary processing of stimuli from the external world or from the body meet and join with information present in the neural configurations of each CNS. Of course, perceptions are not without noetic content, which may be derived indirectly from sensory and cortical processing. From this perspective, ordinary knowledge can be considered the result of informative elements derived from sensory-perceptual activity (biological knowledge in the strictest sense) and from the noetic content of the 21 The term “isomorphism” is generally used in mathematics when two formal structures can be mapped onto each other, so that for each part of one there is only one corresponding part of the other, where corresponding means that the two parts play similar roles in their structures. In other words, it is a structure-conserving biunivocal correspondence between two sets with structures (ordered, algebraic or topological). Application of this notion to mental representations means that the main characteristics of the objects they refer to are codified in them, thus establishing a semantic relationship between the physical object and the mental representation. This relationship makes the mental representation a structure “in the place of” a phenomenal object.

On the Differences and Relations between Ordinary Knowledge

115

individual mind. It therefore consists of structured perceptual and nonperceptual information that fulfils different functions and often overlaps and interweaves in the many ways that determine the appearance of noetic and semiotic structures, which may have different degrees of reliability and adequacy with respect to the objects of the phenomenal world. In this sense, ordinary knowledge, being the result of further cortical and non-cortical processing, is a collection of multiple informative contents, which on one hand are derived from sensory-perceptual activity, and on the other, from the noetic tissue of the individual mind and from the information acquired through the superstructures of different human cultures. Only in some cases does ordinary knowledge endeavour to formulate explanations, and even then, it is in an approximate manner and without a well-grounded and accurate check of their “truth”; indeed, they are generally idiosyncratic or shared only by specific social groups. From the perspective of ordinary knowledge, knowing why things happen is not of interest; the constatation that they occur under certain conditions is sufficient. As we have seen, ordinary knowledge has precise limits, which nevertheless do not prevent it from being a fundamental tool, not only for knowing the world, but also for knowing the self, the body and the mind. Although quite different from scientific knowledge in methods and aims, it shares its gnoseological reference to the world. The world that ordinary knowledge and common sense (with its many subsets of beliefs) are concerned with is the world of objects and events in which humans move and act: it is the Lebenswelt, the world of life, with its beliefs, habits and judgements that regulate our habitual way of evaluating and interpreting the objects we experience and the events interwoven with our existence. Despite the limits of the epistemological range and reliability of ordinary knowledge, it is not only adequate for its purposes but also capable of providing broad and articulated semiotic-noetic structures that make it possible for individuals to regulate their behaviour and actions. It also contributes to the conceptual noetic structures that enable representations of the world, the human species, societies and all the other structures of the human and non-human worlds. It is largely by virtue of ordinary knowledge that individuals have at their disposal interpretative, normative, gnoseological and ontological models for their noetic activity. At this point it is legitimate to ask about the relationship between ordinary knowledge and common sense, as well as how they affect each other. Before answering this question, it is essential to say that common sense, understood as the set of beliefs variably grounded and widespread in a human group in a given historical moment, is the cognitive substrate

116

Chapter Six

of ordinary knowledge; a particularly useful substrate for processing perceptual information from the world, needed for action, reasoning, speaking and building an individual Weltanschauung. Within ordinary knowledge there is also knowledge of common sense, which is not processed by single individuals belonging to a given culture, but produced by sociocultural processes capable of generating representations of the world or conceptions of life, interpersonal relations and social structure. All individuals belonging to a given culture can acquire the knowledge of common sense (often without checking it) and use it as an “instruction manual” for decision making and actions. It follows that, while common sense does not determine the nature of ordinary knowledge, it is nevertheless able to influence it, “participating” considerably in its formation and revision through the noetic contribution of all individuals and by virtue of the ordination and regulatory function exercised by the superstructures that characterise many human cultures. Common sense, as a nucleus of the beliefs, mental attitudes, dispositions and values of a culture, renders ordinary knowledge a constellation of informative elements, coming both from sensoryperceptual activity and the noetic domain of individual minds, which originates from the mental activity and the information acquired through the superstructures of human cultures.

CHAPTER SEVEN THE HOLISTIC THEORY OF TRUTH AND THE EPISTEMIC PRIMACY OF COMMON SENSE AMONG ALL KINDS OF ORDINARY KNOWLEDGE ANTONIO LIVI

What I am trying to defend with my “philosophy of common sense” (see Livi 2013) is precisely the epistemic primacy of common sense among all kinds of ordinary knowledge, in order to save a holistic theory of truth (see Livi 2014). But my purpose is understandable only by taking into account my own notion of “common sense” – which is very different from the sociological or the psychological one. Actually, many contemporary essays on this topic stand in the tradition of past and current common sense philosophers like Thomas Reid, George Berkeley, Henry Sidgwick, George E. Moore, James B. Conant, Radu J. Bogdan and Noha Lemos, who defend common sense; yet they go beyond their accounts by not only defending common sense but also considering what common sense means (cf., among the many authors I could mention, Davidson 1984; Agazzi 2007; Boulter 2007; Piccari 2011). Unfortunately, none of these scholars realized the possibility or the necessity of overcoming the limits of phenomenological research (social psychology, cultural anthropology, sociology of culture, and so on). A clear example of this is the research performed by Marion Ledwig, from Sweden. Besides giving a historical exegesis of common sense in Thomas Reid and showing parallels in Austin, Searle, Moore, and Wittgenstein, he also discovered common sense in Hume’s An Enquiry Concerning the Principles of Morals and in Kant’s Critique of Pure Reason. But the final interpretation of common sense made by this author reaches not at all the level of alethic logic. With his essay he aimed only to make clear how far common sense generalizes, whether proverbs are a form of common sense, and whether

118

Chapter Seven

common sense can be found in the common knowledge assumption in game theory. He also holds that folk psychology should be considered as a common sense psychology (see Ledwig 2007). On the contrary, my own notion of common sense pertains to epistemology, which is the main issue of philosophical logic (see Agazzi 1981). Actually, my philosophy of common sense should be understood as something similar to what Roderick Chilshom called “the foundations of knowing” (cf. Chilshom 1972). In other words, common sense, as I conceive it, is the first step of a theoretical process which leads to surpassing simply semantic holism, i.e. the holism of meaning (see Tarskj 1944), in order to take into account alethic holism, i.e. the holism of truth. This is made possible by identifying a set of logical connections between judgments based on the truth as the basic value of judgments (see Davidson 1990a, 1990b; Frankfurt 2006). The result is an axiomatic system of epistemic logic based on the acknowledgement of the real dependence of every judgment on the truth of its necessary presuppositions, or the logical conditions of the possibility of them being true. This is the meaning of what I maintain is the basic law of thinking, according to the most rigorous phenomenology of the mind’s processes – which are all directed, in any case, to the consciousness of truth, i.e. to the certainty that the contents of my judgment, here and now, are really true, and I cannot absolutely suppose the contrary to be true. This can happen only when my judgment is strongly founded in its presuppositions, so that I realize that it is just the necessary result of all true knowledge I have already obtained and assured with my former certain judgments. Then, this is the general framework of what I conceive as the holism of truth. According to this logical system, any thought of truth – and any assertion which can express it – is linked with all the other thoughts in its very epistemic justification, through the need of finding its own premise and presuppositions. In such a holistic system, my notion of common sense retains a very narrow extension, since it refers only to a few determinate primary certainties which are the common presupposition of both ordinary and scientific knowledge in all their forms and degrees. In other words, “common sense” is just the hard core of the holistic system of truth. I reached such a conclusion taking into account the basic data of cognitive science (see Smith 1995a), the most advanced studies on the philosophy of the mind (see Searle 2004), and the best results of the phenomenology of consciousness – which makes use both of subjective introspection and the analysis of intersubjective communication. I realized that in the consciousness of every thinking subject there are certainties about the

The Holistic Theory of Truth

119

“real world” – certainties whose epistemic justification is founded on the immediate evidence of existing beings which are necessarily and always present in everyone’s experience. In Searle’s philosophy of the mind, this permanent presence of existing beings is called “original” or “intrinsic intentionality”: Where the mind is concerned we also need a distinction between original or intrinsic intentionality on the one hand and derived intentionality on the other. For example I have in my head information about how to get to San Jose. I have a set of true beliefs about the way to San Jose. This information and these beliefs in me are examples of original or intrinsic intentionality. The map in front of me also contains information about how to get to San Jose, and it contains symbols and expressions that refer to or are about or represent cities, highways, and the like. But the sense in which the map contains intentionality in the form of information, reference, aboutness, and representations is derived from the original intentionality of the map makers and users. Intrinsically the map is just a sheet of cellulose fibers with ink stains on it. Any intentionality it has is imposed on it by the original intentionality of humans. So there are two distinctions to keep in mind, first between observer-independent and observer-dependent phenomena, and second between original and derived intentionality. They are systematically related: derived intentionality is always observerdependent (SEARLE 2004, p. 7).

But I maintain much more. In my system, such certainties constitute the very first link in the chain of presuppositions, so they can in no way be subject to doubt. This means that their non-truth is absolutely unthinkable; actually, no one can ever really doubt them, and one must understand that any affirmations to the contrary are merely verbal posturing. They respond to some pragmatic logic, and not the expressions of a real certainty, endowed with its own adequate epistemic justification. Given that they constitute the nucleus of experience, understood as a body of unmediated knowledge, such certainties are present to consciousness in every moment of the search for truth as the logical presupposition of all knowledge deriving from reflection and inference, both inductive and deductive. For this same reason, such certainties function as an ultimate criterion of truth to verify any hypothesis successively formulated. They therefore constitute the main alethic presupposition, that is, the presupposition necessary for any ulterior knowledge to be thought of as true. In fact, on the basis of these original truths, every thinking subject verifies, time after time, the admissibility of any hypothesis – formulated by himself or proposed by other subjects through one of the ways of communicating thought – that presents itself in the search for other truths over the course

120

Chapter Seven

of his lifetime. As a result, all scientific knowledge should be structured as a system logically compatible with the primary truth of common sense, so as to place the instruments of dialectics (reflection, interpretation, inference) effectively at the service of the search for further truths. In order to makes it easier for my thesis about primary truth to be critically evaluated from different historical and theoretical perspectives, it seems to me appropriate to expound here my proposal in its essential terms. In this way it will become clearer which aspects of my analysis of the holistic structure of thought are obvious and given, and which are instead debatable, as they are not universally accepted (being strongly polemic towards philosophical rationalism and theological fideism) and which, finally, are universally ignored. It is these last two aspects that I propose to highlight and re-evaluate within the limits allowed by this study. I hold that the modern term “common sense”, used by several important European philosophers of the eighteenth and the nineteenth centuries (see Mesolella 2011), can still be used in philosophy today, but only meaning by it just what I mean, i.e. the primary evidence that constitutes the “basic truth” of each thinking subject due to: i) its absolute immediacy ʊ being knowable without any former presupposition, and ii) its necessary function as an alethic foundation of thought in all its forms ʊ being, in the structure of epistemic holism, the very presupposition of any other assertion with a claim to truth. Over the last twenty years I have tried to deepen this theory of alethic logic as a system by returning repeatedly to the same theme and by stressing its very nature, that of a logical system (cf. Livi 2005, 2010, 2013a, 2013b, 2013c). This is just what some scholars have underlined when identifying the structure of alethic logic according to my thought (see Rego 2011; Renzi 2012). For a clearer explanation I now want to present some synthetic theses which summarize my own notion of “common sense” as the beginning of knowledge and the unique basic criterion for all possible statements being considered true. A first thesis is that there exists in human knowledge a realm of certainties that express what every thinking subject understands directly from experience, even before attempting to interpret – through reasoning – what he has experienced. Inasmuch as they are derived directly from experience, these certainties require no other epistemic justification than the absolute evidence that their object exists. Consequently, even when their object is reached through spontaneous inference (this the case of the certainty about the existence of God), they manifest a decisively intuitive character; since the inference, in this case, is not a formal scientific

The Holistic Theory of Truth

121

reasoning, nor does it regard particular anthropological or cultural conditions. A second thesis is that those certainties constitute our primary knowledge of concrete reality, that is, the awareness of the reality that de facto presents itself to every thinking subject (the “world in which one is” and one’s own “being in the world” as a personal being). At the same time, in an indissoluble cognitive unity, they represent our primary knowledge of the universal, that is, our intuition of first principles, both those of a metaphysical-logical nature (which render essentially intelligible, though always problematic, the real from which and in which one lives) and those of a metaphysical-ethical nature (which reveal the meaning and significance of one’s own individual and social life, in both temporal and eternal perspectives). A third thesis is that such certainties, being connatural to human intelligence – that is, being a basic characteristic of human nature as such (at least as long as human nature has at its disposal the necessary material conditions to express itself in ordinary empirical life) – are the patrimony of all men and women, they are universal in time and in space, they are a constant in the midst of all the variables of culture and social conditions, just as they are a constant throughout all the variations (advances and regressions) of an individual’s intellectual developments. A fourth thesis is that this common basis of certain and indubitable knowledge – indubitable in itself, even if someone like Descartes can say (though not think) he has called them into question through meditative thought – is precisely what allows for the communication of knowledge among individuals. This communication begins with the acquisition of language and is further developed through the educational process and the relationship between parents and children at the developmental age. It also allows for communication of knowledge among different cultures, overcoming (at least potentially) the barriers of incommunicability resulting from the psychological, anthropological (both ethnic and political), philosophical-religious and technical-scientific differences, which, along with the historical component, characterize the various linguistic structures (both structures of meaning and of significance). A fifth thesis is that those certainties, despite the fact that they obviously have a pre-philosophical nature, are in direct relation to philosophy, especially to philosophy as metaphysics. Actually, certainties of common sense are in all respects coextensive with metaphysical discourse, and are different from it only because of the very scientific dimension of metaphysics, which consists of performing a systematic reflection, in creating new hypotheses of interpretation, and in elaborating

122

Chapter Seven

conceptual definitions and logical derivations. The certainties of common sense are derived so directly from experience that they furnish philosophical reflection with its necessary material and the first condition for its statements being true. Actually, they furnish metaphysics with the perennial substance of its truest scientific conclusions, those which over the centuries – considering philosophy in its historic dimension – have allowed some historians to speak of a “philosophia perennis”. Incidentally, I maintain that metaphysics, when it assumes common sense certainties as its alethic background, is necessarily realistic; moreover, it is the only kind of true realism (see Gilson 1990; Reffes 2008; Livi 2010b; Livi 2014b). A sixth thesis is that those same certainties are the conditions of possibility (ex parte obiecti, since they furnish the universal objective horizon from which the specific formal object is extracted) of all particular sciences; sciences of nature, mathematical sciences, human or social sciences, and so on. But, in addition to furnishing the sciences with the objective premises of scientific research, that is, the field of objective reality (presupposed as real and rational) to be delineated and analyzed, those certainties of which we speak are also the “human” premises of scientific work. They are the necessary conditions for exercising such sciences, and therefore the conditions of possibility ex parte subiecti: because they provide the basic reasons, they make it possible for man to be a scientist, and also identify the basic logical instruments, such as the principles of identity and non-contradiction, the principles of efficient and final causality, and the roles of inference through induction and deduction (see Tarskj 1959; Agazzi 1981; Nagel 1961). A seventh thesis deals with the different qualities of common sense and science. I maintain that the organic body of certainties of which I speak is in itself qualitatively superior to science, including within the notion of “science”, both the metaphysical as well as the particular sciences, that is, all reflexive and systematic knowledge, mediated by reasoning and culture and equipped with its own technical methodology. The superiority of these certainties lies precisely in the quality of the certainty itself. While the certainty of those direct and universal statements is unconditional and absolute, scientific certainty always has limiting characteristics; it is a subjective certainty linked to privileged conditions of experience or intellectual capacity, or it is a certainty available only to members of a specific cultural community equipped with its own research instruments, or it is a certainty that can be reached by all humankind only in relation to particular historical events, or a certain historical level of technological development, or a particular historical perspective on human

The Holistic Theory of Truth

123

events, or it is a provisional (hypothetical) certainty, susceptible to falsification or at least revision and adjustment, when it is not a bold working hypothesis of an instrumental nature (that is, as an efficacious tool for research, for science itself, or a mere instrument at the service of technical work or other practical ends). To sum up, while the certainties of direct experience are (per se) incontrovertible, the certainties of science are (per se) debatable, or at least relative, perfectible, revisable; the former belong to all people at all times, the latter belong to some people and only in specific moments of personal or collective history. Secondly, the superiority of these certainties over science is also a primacy in truth: they come first, both logically and chronologically (because every reflection presupposes some direct knowledge to which to return, dealing as we are with secundae intentiones), and are also elevated above other certainties, in the sense that they cannot be contradicted (by the truth; that is, proven false), to the point that every scientific thesis that contradicts those certainties is for that very reason vitiated by error (even if that error can only be demonstrated at the level of science). An eighth thesis deals with the relationship between particular sciences and metaphysics. I think that a dialogue between particular sciences and metaphysics is certainly necessary, and de facto always beginning anew. However, the threats of reciprocal interference or methodological commingling can be avoided, as can attempts to annex one to the other, deriving from epistemological errors (as in the age of classical philosophical cosmology or the age of modern positivism) if a basis of conceptual agreement is found. This basis can be found precisely by returning to the common epistemological derivation of the certainties and contents of direct experience, thus circumventing any attempt at an impossible, unmediated translation (without the mediation of those basic certainties) of the technical language of metaphysics into the technical language of the other sciences, or vice-versa.

References Agazzi E.,1981, Modern Logic. A Survey, edited by Evandro Agazzi. Dordrecht: Reidel, 1981. —. 2007,Valore e limiti del senso comune, edited by Evandro Agazzi. Milano: Franco Angeli, 2007. Boulter, S., 2007, The Rediscovery of Common Sense Philosophy. New York: Palgrave Macmillan, 2007. Chilshom, R., 1972, The Foundations of Knowing. Minneapolis: University of Minnesota Press, 1972.

124

Chapter Seven

Davidson, D., 1984, Inquires into Truth and Interpretation. Oxford: Clarendon Press, 1984. —. 1990a, “The Structure and Content of Truth”, The Journal of Philosophy, 1990, 279-328. —. 1990b, Truth and Predication. Cambridge, MA: Harvard University Press, 1990. Frankfurt, H.G., 2006, On Truth. New York: Alfred A. Knof Publisher, 2006. Gilson, E., 1990, Methodical Realism, Front Royal: Christendom Press, 1990. Kleist, E.E., 2000, Judging Appearances: a Phenomenological Study of the Kantian Sensus commun. Dordrecht: Kluwer, 2000. Livi, A., 2005, Senso comune e logica aletica, Roma: Casa Editrice Leonardo da Vinci, 2010. —. 2010, Metafisica e senso comune. Sullo statuto epistemologico della “filosofia prima”, Roma: Casa Editrice Leonardo da Vinci, 2010. —. 2013a, A Philosophy of Common Sense. The Modern Discovery of the Epistemic Foundations of Science and Belief. Aurora (Colorado): The Davies Group Publisher, 2013. —. 2013b, “Why Common Sense, When Assumed As the Primary Truth, Furnishes Alethic Logic System With its Very Epistemic Justification”, in La certezza della verità. Il sistema della logica aletica e il procedimento della giustificazione epistemica, edited by Antonio Livi, Roma: Casa Editrice Leonardo da Vinci, 2013, 19-30. —. 2013c, “Perché ogni ricerca di un’adeguata giustificazione epistemica presuppone una coerente teoria sistemica della verità”, in La certezza della verità. Il sistema della logica aletica e il procedimento della giustificazione epistemica, edited by Antonio Livi, Roma: Casa Editrice Leonardo da Vinci, 2013, 217-230. Ledwig, M., 2007, Common Sense: Its History, Method, and Applicability. New York & Oxford: Peter Lang International Academic Publishers, 2007. Mesolella, M., 2012, I filosofi moderni del senso comune, edited by Mario Mesolella, Roma: Casa Editrice Leonardo da Vinci, 2013. Nagel, E., 1961, Science and Common Sense, in The Structure of the Science: problems in the Logic of Scientific Explanation. Indianapolis, Indiana: Hackett, 1961, 547-605. Newell, D.J., 1980, Philosophy and Common Sense. Washington: University Press of America, 1980. Piccari, P., 2011, Co noscenza ordinaria e senso comune, with an Introduction by Mariano L. Bianca. Milano: Angeli, 2011.

The Holistic Theory of Truth

125

Reffes, A.J., 2008, The Philosophy of Realism, Common Sense, and Ordinary Experience. Berlin: Progress Publishers, 2008. Rego, T., 2011, La filosofia del sentido común en Aristóteles. La doctrina aristotélica de la ‘koinai doxai’, los ‘endoxa’ y los primeros principios de la demonstración en comparación con la teoria de Antonio Livi acerca de la sprimeras verades existenciales. Roma: Casa Editrice Leonardo da Vinci, 2011. Renzi, F., 2012, La logica aletica e la sua funzione critica, Roma: Casa Editrice Leonardo da Vinci, 2012. Searle, J.R., 2004, Mind: A Brief Introduction. New York: Oxford University Press, 2004. Smith, B., 1995a, «Formal Ontology, Common Sense, and Cognitive Science», in International Journal of Human-Computer Studies, 43 (1995), 641-667. —. 1995b, «The Structure of Common Sense World», in Acta Philosophica Fennica, 58 (1995), 290-317. Tarskj, A., 1944, «The Semantic Conception of Truth», in Philosophy and Phenomenological Research, 4 (1944), 341-376. —. 1959, Introduction to Logic and Methodology of Deductive Sciences. Oxford: Blackwell, 1959.

CHAPTER EIGHT COMMON SENSE AND TRUTH CONDITIONS GAETANO PICCOLO

1. Definition of Common Sense The path proposed in this article is inspired by the discussion that took place at a conference in Arezzo (7-9 October 2013),1 in particular in the session dedicated to the possibility of a foundation for common sense and the relationship between common sense and ordinary knowledge. Ordinary knowledge recalls the various hypotheses about the formation of concepts, the relationship between perception and cognition, and the relationship between knowledge (epistemology) and objects (ontology). According to Maurizio Ferraris, for example, while epistemology is linguistic, conscious and teleological, ontology is neither linguistic, nor conscious, nor teleological. In this context, one might say that "to exist is to resist".2 Gestalt psychology, for its part, has shown how our perception is taken as the privileged point of view. One can think of optical illusions, in which the perception persists even when we have understood that it is an illusion. Recent contributions from neuroscience have also highlighted the fact that the mind reformulates beliefs on the basis of what has been archived. From an evolutionary point of view, ordinary knowledge prefers to trust in its own point of view, considering it correct; otherwise one would need to constantly question the knowledge that followed. The tendency to consider correct or to give preference to one’s own point of view also emerges in situations in which we assess the probability 1

See my report on the conference in http://www.philosophicalnews.com/ (November 2013), 7. 2 This is a thesis developed in M. Ferraris, Exist is to resist, in M. De Caro, M. Ferraris (ed.), Bentornata realtà: il nuovo realismo in discussione, Einaudi, Torino 2012, 139 -165.

128

Chapter Eight

of a certain event happening. Ordinary knowledge is statistically more prone to taking account of descriptive data rather than logical calculation. The classic reference here is to what has become known as "Linda’s case". Linda is 31, unmarried, extroverted, bright, a graduate in philosophy, a politically active student and supports anti-nuclear ideology. With this description, three alternatives are proposed: 1) Linda is a shop assistant; 2) Linda is a feminist; 3) Linda is a shop assistant and a feminist. In 90% of cases, option 1 is considered less likely than option 3. We know that the truth of 3 depends on 1 and 2, so option 3 is actually less than or equally as probable as option 1. Investigators have deduced that 3 is perceived as more representative of reality compared to 1.3 I begin my reflection with the definition proposed by E. Agazzi during the Arezzo Conference: “Common sense is something about which we believe that everyone should agree. It is not obvious; in fact philosophy is the criticism of the obvious, and seeks rather that which is evident, i.e. what imposes itself as inherently true. Common sense then consists in the basic element to begin a discussion: concepts, judgments, principles, representations... we could say the Lebenswelt.”4 The discussion was therefore able to concentrate at this stage on the nature of these basic elements. Usually, the different positions at play are all forms of foundationalism, i.e. tending towards the idea of common sense as that which is needed to lay the groundwork for the construction of knowledge. These different positions may be considered in two large groups: first, a lightweight, non-rigid foundationalism, according to Wittgenstein in his On Certainty (at least according to the interpretation that considers shared cultural structures as those beyond which you cannot go, thus making propositions of common sense part of a background that has been handed down culturally). In this group I place Agazzi’s position (which refers to the concept of Lebenswelt) and that of Paolo Piccari (who speaks specifically of "a common core of beliefs"), and probably Mariano Bianca’s as well.5

3

A. Tversky – D. Kahneman, Extensional versus intuitive reasoning: The conjunction fallacy in probability judgment, «Psychological Review», 90 (1983), 297. 4 Agazzi has developed this idea in E. Agazzi (ed.), Valori e limiti del senso comune, FrancoAngeli, Milano 2004. 5 Cf. P. Piccari, Conoscenza ordinaria e senso comune, Franco Angeli, Milano 2011, especially pages 41 to 50.

Common Sense and Truth Conditions

129

In the other group, somewhat solitary, is the foundationalism of Antonio Livi:6 hard foundationalism according to the rigid axiomatic Aristotelian model or Descartes’ Discourse on the Method. Livi has criticized a sociological reading of common sense, speaking rather of basic truths as rationally founded. These truths (things, the self, others, the moral law, God) would be at the base of alethic logic, a logic not reduced to a trivial formalism. These different perspectives, all characterized by a foundationalist concern, might be collected in a sort of positive praesupponendum, as in the theory of Donald Davidson.7 In his theory of radical interpretation, Davidson speaks precisely of a common core of beliefs, a minimal core, but says that you must share in order to start the exercise of interpretation. This common core must at least contain the predicate “is true”. This is a core of beliefs inevitably unrelated with meanings, within a holistic framework. The common core thus becomes a way to access to each other's worlds: the beliefs and meanings of each subject make up the world of the speaker and the interpreter. Davidson’s praesupponendum includes not only sharing the predicate “is true”, but also the consideration of the other as a subject, and also represents the implicit assumption that the other generally does not intend to trick one with his pronouncements, though he may sometimes be mistaken (this does not usually affect the interpretation of his utterances). Radical interpretation theory seems then to invoke, on the one hand, the need for a recognizable and shared platform (similar to a hard foundationalism), but at the same time it tolerates less defined boundaries because the world of the other (Lebenswelt), made of beliefs and meanings, is reached only very partially, which implies the incompleteness of radical interpretation. Moreover, even in ordinary knowledge, as cognitive psychology reveals, the elaboration of knowledge goes through a complex process of interpretation. Cognitive psychology adequately warns us not to expect to see the interpreter's point of view as necessarily fixed or absolute. Starting from a brief investigation of the meaning of meaning, I will attempt to show how the theory of Davidson’s interpretation may represent 6

Cf. A. Livi, Senso comune e logica aletica, Leonardo da Vinci, Roma 2007. Id., Filosofia del senso comune. Logica della scienza e della fede, Leonardo da Vinci, Roma 2010. 7 Cf. on this theme, in the context of the conference in Arezzo, Cristina Amoretti’s speech was very valuable. She is also the author of Il triangolo dell’interpretazione. Sull’epistemologia di Donald Davidson, FrancoAngeli, Milano 2008.

130

Chapter Eight

a valid starting point that allows us to converge on a shared idea of common sense.

2. Common Sense as Sharing of Meanings8 The most common answer to the question about what meaning is is perhaps that which identifies the relationship between a name and its bearer. In other words, it is generally thought that the name means the object. The meaning would then be the relationship that exists between the word and the object. But where would we find this relationship? Is the meaning something mental? Also, when we think of a name in general, for example the word "dog", unattached to an object (a specific dog), we nonetheless attribute it to a mental image, to our concept of a dog. In the end, then, we should recognize that the meaning comes to be a mental relationship between a name and a concept. At this point, the meaning might have almost nothing to do with reality.9 This way of understanding meaning (the so-called representational theory of mind) considers language a way to transfer those mental contents from one mind to another. It is therefore the mental content that ought to be considered, in this perspective, as the meaning of the expression. Not always, however, does an expression that refers to mental content correspond to reality. For example, considering a description like "in 1931, Adolf Hitler made a visit to the United States...", everyone understands the terms used, maybe we even have a mental representation and are willing to believe that what has been said is true, yet it is completely untrue that Hitler made a visit to the United States in 1931.10 This theory, which considers meaning as mental content, has two types of difficulties: the first concerns the communication of mental content, and the second concerns the difference between mental representations that correspond to reality and those that do not. If language needs to communicate mental content, it would follow that a speaker usually has content in his mind and chooses an expression to communicate it. In other words, the mental representation would lead to 8

I have developed more widely the issue of meaning in the third chapter of my book, Cf. G. Piccolo, Significato e interpretazione. Indagine sulla conoscenza, Carocci, Roma 2011, 79-124. 9 Landmark is the inquiry in meaning developed by H. Putnam in various essays, in particular see H. Putnam, Mind, Language and Reality, “Philosophical Papers”, vol. 2, Cambridge University Press, Cambridge (MA) 1975. 10 See W.G. Lycan, Philosophy of Language. A Contemporary Introduction, Routledge, London 2000, 3.

Common Sense and Truth Conditions

131

the linguistic expression. This is not true, however, because the speaker does not choose the expression, but he is "forced" to use the one that expresses that meaning if he wants to be understood. You may also choose to use the word "elephant" to talk about a dog, but it would be really complicated for the listener to understand you. It is rather the object that determines the representation and so if language is "forced", it is forced by reality and not by the mental representation. As shown by the example of the word "elephant", which cannot be arbitrarily used in place of the word "dog", meanings must be intersubjective. It is here that the difference is shown between mental content and meanings. Meanings, in fact, unlike mental content, “can be shared by different participants of communications [...] semantics begins where subjectivity ends hence where psychology ends”.11 For this reason, the fact that the meanings are distinguishable is not a matter of chance; rather the public dimension is a constitutive aspect of language. And the definition of meaning cannot avoid this intrinsically intersubjective dimension. So we come to the following conclusions: meaning is not an object, because sometimes the object is not real, and although it can be something mental, the meaning cannot be defined simply as mental content, because it must be intersubjective. Since the meaning is intersubjective, it is difficult to think that it is hidden somewhere in the mind of individuals. Even in the case of the meanings of objects of fantasy, such as Mickey Mouse or Batman, the intersubjective dimension remains valid. In fact,even a fictional meaning like the one that corresponds to the term "Mickey Mouse" or "Batman" has specific characteristics, to the point that another can correct me, telling me that I'm confusing Mickey Mouse with Batman, if I say, "Mickey Mouse has a partner named Robin". On the second difficulty that we highlighted, recalling the representational theory of mind (i.e. the question relating to the correspondence between mental representation and reality), in Tractatus logico-philosophicus, Wittgenstein attempted to solve the problem, starting from an isomorphism between thought, language and reality. In the middle of the Tractatus, there is indeed the idea that thought is an image of reality.12 In this case, the relationship between language and image would be univocal. Wittgenstein's error, as he himself would admit at the beginning of Philosophical Investigations, was to consider only the 11 J. Peregrin, Meaning and Structure. Structuralism of (Post)Analytic Philosophers, Ashgate, Aldershot 2001, 19. 12 L. Wittgenstein, Tractatus logico-philosophicus, 4,021.

132

Chapter Eight

descriptive use of language. But even within a descriptive use of language, Wittgenstein had to admit that before an image, different descriptions are possible; the meaning therefore cannot be an image of reality either. For example, if I see my friend Mario coming out from the ice cream parlor with a big ice cream, I could think “Mario is gluttonous for ice cream”, or “at this time of day it is a good idea to have an ice cream”, or even “typically, people love ice-cream”. This reduction of language to a descriptive (or physicalist) function is rejected by Wittgenstein, as we said, in Philosophical Investigations, but on that occasion, Wittgenstein ascribes said idea of language not only to the author of the Tractatus, but also to Augustine, misunderstanding a passage from Confessions.13 A final interpretation, which we must mention, is the possibility of understanding meanings as entities that belong to a third kingdom, such as, for example, the Platonic ideas. This idea of a third kingdom (i.e. beyond that of thought and language) has re-emerged in recent times. Frege, for example, brought semantics into the context of logic. According to Frege, while for names the meaning is the object, for statements, the meaning is the function of truth. A thesis evoked and deepened, inter alia, by Wittgenstein in his Tractatus. The meaning of an utterance would, in other words, be its truth table.14 U. Eco, following Ch. S. Peirce, discussed meaning within the broader framework of semiotics. 15 The signifier (or designator) and signified are the two parts that make up a sign. The process whereby we attach meanings to these signs is called semiosis. This process can be of different types: if I attach to the red light of a traffic light the meaning of “stop”, that takes place on the basis of traffic laws; if I attach to red roses the meaning of passion, this is probably based on social history. The mark then comes to be a relationship, in which a signifier is ascribed to the signified.

13

Augustine, Confessiones, i, viii, 13. Augustine is talking about the specific case of learning the mother tongue. About the wrong interpretation given by Wittgenstein about this step of the Confessiones and about the theory of the language of Augustine, see G. Piccolo, Wittgenstein interprete di Agostino. Il caso di Confessiones I, viii, 13, in "Rassegna di teologia", 2 (2008), 309-18. 14 Wittgenstein, Tractatus logico-philosophicus, 4,431. 15 U. Eco, I limiti dell’interpretazione, Bompiani, Milano 1990, 215-228. See also the study entitled Latratus canis on relationship between names and signs in Medieval thought and Stoical and Aristotelian influences on it, in U. Eco, Scritti sul pensiero medievale, Bompiani, Milano 2012, 699-730.

Common Sense and Truth Conditions

133

If language is considered only the giving of names to things, then semantics can easily be regarded as a subset of semiotics. But, given the complexity of the uses of language, as Wittgenstein highlighted in the Philosophical Investigations, a reduction of language to this function appears quite unlikely. The role of semantics, as Quine also clarified, should consist of investigating the structures of language and the logic of the games we play with it.16 Words, therefore, should not be considered labels to fix upon things, but a means to describe the world. It follows that language becomes a kind of toolbox, which reveals a peculiarity: the tools it contains are linked in such a manner that it is impossible to distinguish one from the other. That is why authors like Quine and Davidson place their theories on a holistic background. And that is why semantics can only find its fulfilment in pragmatics: meaning must be interpreted from the situation in which the proposition is uttered.

3. Common Sense and Interpretation To introduce my proposal, I would like to emphasize the word “common”: what makes sense common? The title of this paragraph suggests an answer: it is the possibility of interpretation (among speakers) that makes sense common. For this reason, I will try, using the contribution of D. Davidson’s theory of interpretation, to outline the conditions that make interpretation possible. Common sense therefore consists, in my view, of interpretations that make objective what is already intersubjective. The originality of Davidson, compared to other authors (for example, Dummett), is in developing a theory of interpretation no longer from the point of view of the speaker (i.e. a theory aimed at understanding the pronouncements of the speaker), but from the point of view of the interpreter. Davidson asks: “What are the conditions to ensure that an interpreter can understand the utterances of the speaker?” In the event that the interpreter does not know the language of the speaker, they will have to rely on behavioral evidence. The philosophical problem of interpretation is therefore, according to Davidson, to identify the form of the theory that allows an interpreter to understand the utterances of a speaker in general. Davidson is not searching, then, for the meaning of the sentence, like Frege for example, but the form of the theorem that tells how and why it is possible to 16

J. Peregrin, Meaning and Structure, 31.

134

Chapter Eight

establish a relationship between the utterance of the speaker in his own language and the interpreter's utterance in his language (a relationship between object language and metalanguage). I would say that the form of the theorem developed by Davidson17, being focused on the biconditional connective (or logical equivalence), also provides a definition of common sense. I find it interesting that, for Davidson, this theory of meaning implies the concept of truth18, thus overturning the path of A. Tarski, who built his theory of truth starting from the notion of meaning.19 Tarski formulated his theory with reference to formalized languages. Davidson believes, however, that while Tarski said something fundamental about the concept of truth, he did not say everything, especially because he did not catch the relationship between truth and meanings and belief.20 Tarski's goal was to build a scientifically adequate semantic, giving an axiomatic theory to the concept of truth from which an explicit definition of this concept could be obtained.

17

T-sentences, i.e. the theorems of the theory, should have, according to Davidson, this form (T) s is true-in-L if and only if p. See D. Davidson, In Defence of Convention T, in Id., Inquiries and Interpretation into Truth, Clarendon Press, Oxford 2001,2, 65-75. 18 “It is a mistake to look for an explicit definition or outright reduction of the concept of truth. Truth is one of the clearest and most basic concepts we have, so it is fruitless to dream of eliminating in favor of something simpler or more fundamental.” D. Davidson, Truth and Predication, Harvard University Press, Cambridge (MA) – London 2005, 55. 19 “In this paper I defend a version of the correspondence theory. I think truth can be explained by appeal to a relation between language and the world, and that analysis of that relation yields insight into how, by uttering sentences, we sometimes manage to say what is true. The semantic concept of truth, as first systematically expounded by Tarski, will play a crucial role in the defence.” D. Davidson, True to the Facts, in Id. Inquiries into Truth and Interpretation, Clarendon Press, Oxford, 20012-38 .37. Later, however, Davidson rejected the definition of correspondence theory regarding Tarski's theory of truth, cf. Davidson, Truth and Predication, 41-42. 20 “My own view is that Tarski has told us much of what we want to know about the concept of truth, and that must be more. There must be more because there is no indication in Tarski's formal work of what it is that his various truth predicates have in common, and this must be part of content of the concept [...] The concept of truth has essential connections with the concepts of belief and meaning, but these connections are untouched by Tarski's work.” Cf. Davidson, Truth and Predication, 27-28.

Common Sense and Truth Conditions

135

At the center of the Tarskian theory is the notion “to be true”, expressed by the predicate. According to Tarski, we have a prephilosophical notion of the predicate “to be true”. This notion is expressed by Aristotle’s definition, given in Metaphysics IV, 1001 b 26: “it is false to say of what is that it is not, and of what is not that it is, while it is true to say of what is that it is, and what is not that it is not.” In another passage (Metaphysics ix, 1051b, 6-9), Aristotle writes: “it is not the fact that we really think that you are pale, that you are pale, but rather because you are pale, we that say so, have the truth.21 On this basis, Tarski built the T-Convention: the predicate to be true is expressed by the relationship between reality and the propositional meaning. Inverting the reasoning of Tarski, Davidson finds the function of meaning in the propositional structure, where the predicate to be true relates reality with its description. At this point, I will introduce my argument that common sense is not a thing, but rather a process which allows us to understand one another about things. This process can be exemplified by Davidson’s theory of interpretation.

4. Common Sense and Intersubjectivity The problem of interpretation is not only about the interpretation of a foreign language. Indeed, for Davidson, the problem of interpretation begins at home. Interpretation is the foundation of all linguistic communication. Understanding what a speaker says always assumes an interpretation. Davidson starts from an extreme case, what he calls radical interpretation (the name also inspired by the radical translation of Quine). This is when the speaker speaks a language completely unknown to the interpreter. To check the plausibility of a theory of truth, taken as a theory of interpretation of a certain speaker, the conditions of truth that our theory ascribes to the utterances of L (using T-sentences) must be checked against those which the speaker would retain to be truly said utterances. Unlike Quine, according to Davidson the check of interpretation theories cannot occur only in terms of behavior (a rejection of behaviorism), as a reference to intentional concepts is unavoidable. The check therefore requires sharing the predicate “to be true”, which is a semantic notion. This step was only later developed by Davidson.

21

We must clarify that Tarski gives an analysis of the predicate true-in-L predicate and not of the predicate true in general.

136

Chapter Eight

When interpreting a speaker, we do not limit ourselves to attributing certain meanings to their utterances. Inevitably, we attribute even beliefs to them, and more generally, propositional attitudes. The meaning of an utterance depends directly on the beliefs that the utterance conveys. And the meeting point between meaning and beliefs is in "propositional attitudes". Belief refers to a meaning, since it expresses itself through propositions. The only way to verify the knowledge of the other is to consider its propositional attitudes. And this is also the only way to overcome a cognitive solipsism that would not allow one to go very far. Davidson highlights the close relationship between propositional attitudes and being rational: to be a rational animal simply means to have propositional attitudes, no matter how confused, contradictory, absurd, unjustified or incorrect they may be.22 Propositional attitudes may also be simply called thoughts: I believe, I mean, I think, I desire. Being a rational animal thus means being able to express one’s own inner states. Whenever we deal with knowledge, whatever theoretical approach we take, we always have to deal with the propositional attitudes of individuals — that is, with their beliefs. One of the characteristics of thoughts, as well as propositions, is their logical relationships. The identity of a thought cannot be separated from the logical network that it is inserted into. A thought therefore cannot be relocated to another logical network without becoming another thought altogether ("I think there is a good climate" is an utterance that expresses a belief, whose meaning varies depending on the context, the speaker, the place etc). My idea is that the best way to deal with common sense is by studying meaning. Language, as a starting point, offers guaranteed access to the individual's cognitive dimension. Although the existence of a thought does not necessarily depend on the existence of a statement that expresses it, a creature, according to Davidson, cannot have thought if it does not have a

22

“Neither an infant 1 week old nor a snail is a rational creature. If the infant survives long enough, he or she will probably become rational, while this is not true of the snail. If we like, we may say of the infant from the start that he is a rational creatures because he will probably become rational if he survives, or because he belongs to a species with this capacity. Whichever way we talk, there remains the difference, with respect to rationality, between the infant and the snail on one hand, and the normal adult person on the other.” D. Davidson, Subjective, Intersubjective, Objective, Clarendon Press, Oxford 2001, 95.

Common Sense and Truth Conditions

137

language.23 What is more, a rational creature usually has many thoughts and is able to interpret the words and thoughts of others. The relationship between propositional attitudes and beliefs, then, is fundamental: the first requires the latter. To understand the words of another, I must be thinking the same thing that he is thinking. In other words, I have to be able to share his world. Sharing, of course, does not mean agreeing. But both agreement and disagreement need common representations or a core of common beliefs. This shared world can be called an intersubjective world: “[...] But the concept of an intersubjective world is the concept of an objective world, a world about which each communicator can have beliefs.”.24 So if speakers can understand each other, it is because they share a world. This means that rationality necessarily possesses a social dimension. Individuals who communicate are necessarily rational. Communication, as a social phenomenon, presupposes rationality. Beliefs and meanings seem intertwined in an inseparable way: if I observe a person watching an eclipse and he uses an expression that I do not know, I can interpret that expression on the basis of my statements and attribute to that expression certain conditions of truth; such and such is true if and only if there is an eclipse. Attributing this meaning to the heard expression, I am also attaching a set of beliefs or propositional attitudes to the speaker.25 The interdependence between beliefs and meanings may be problematic for a theory of interpretation. For this reason, Davidson considers it necessary to postulate an indulgence principle (principle of charity), according to which the interpreter postulates a significant degree 23

“My thesis is not, then, that each thought depends for its existence on the existence of a sentence that expresses that thought. My thesis is rather that a creature cannot have a thought unless it has language. In order to be a thinking, rational creatures, the creatures must be able to express many thoughts, and above all, be able to interpret the speech and thoughts of others.” Davidson, Subjective, Intersubjective, Objective, 100. 24 “To understand the speech of another, I must be able to think of the same things she does; I must share her world. I don't have to agree with her on all matters, but in order to disagree we must entertain the same propositions, with the same subject matter, and the same concept of truth. Communication depends on each communicator having, and correctly thinking that the other has, the concept of a shared world, an intersubjective world. But the concept of an intersubjective world is the concept of an objective world, a world about which each communicator can have beliefs.” Davidson, Subjective, Intersubjective, Objective, 105. 25 M. De Caro, Dal punto di vista dell’interprete. La filosofia di Donald Davidson, Carocci, Roma 1998, 37.

138

Chapter Eight

of agreement between his beliefs and those of the speaker. Without this assumption it would be impossible to understand each other.26 Davidson’s principle of charity combines together some necessary conditions for interpretation: first of all, we must assume that whenever the speaker, explicitly or by and large implicitly, says "it is true that", i.e. he asserts something, he does not intend to deceive us. Sometimes he can make mistakes, but generally we have to rely on the fact that he honestly wants us to understand what he means. This in turn implies that the predicate "to be true" must have the same value for both the speaker and the listener who interprets. Within the principle of charity there is also recognition of the other as a rational being who has beliefs. One must also assume, in order that an interpretation can take place, that the speaker and the interpreter have a core of shared beliefs. The first essential belief shared relates to the predicate "to be true". The system to which the belief belongs does not simply consist of beliefs, but is part of a conceptual network that connects that specific intentional state with other mental events. Intentional states are in close dependence with the meanings of the utterances by which those intentional states receive linguistic expression. This relationship of dependence has in turn a holistic character. It is a holism that inextricably binds meanings and beliefs.27

26

“Quine's key idea is that the correct interpretation of an agent by another cannot intelligibly admit certain kinds and degrees of difference between interpreter and interpreted with respect to belief. As a result, an interpreter is justified in making certain assumptions about the beliefs of an agent before interpretation begins. As a constraint on interpretation, this is often called by the name Neil Wilson gave it, the principle of charity.” Davidson, Truth and Predication, 61-62. 27 Davidson proposes to resume a holistic conception of the meaning: “If sentences depend for their meaning on their structure, and we understand the meaning of each item in the structure only as an abstraction from the totality of sentences in which it features, then we can give the meaning of any sentence (or word) only by giving the meaning of every sentence (and word) in the language.” Davidson, Truth and Meaning, 22. Having embraced a radical holism, also as a result of its fidelity to Quine’s approach, Davidson preferred a moderate holism by saying that “he does not argue that the meaning of a sentence depends on the meanings of all sentences”. Cf. M. De Caro, Dal punto di vista dell’interprete. La filosofia di Donald Davidson, Carocci, Roma 1998, 44.

Common Sense and Truth Conditions

139

5. Common Sense and Triangulation Common sense, as Davidson puts it, is the line that binds us, that unites me with the other, as we look upon the same object. Davidson calls this notion “triangular externalism”,28 and it appears for the first time in one of his articles from 1982: “If I were bolted to the earth, I would have no way of determining the distance from me and many objects. I would only know they were on some line drawn from me towards them. I might interact successfully with objects, but I could have no way of giving content to the question of where they were. Not being bolted down, I am free to triangulate. Our sense of objectivity is the consequence of another sort of triangulation, one that requires two creatures. Each interacts with an object, but what gives each the concept of the way things are objectively is the base line formed between the creatures by language.”29 I could say that the word “common” is necessary in order to make the meaning objective: it is common sense that makes sense objective. Only if another can correct me, only if there is the possibility of being corrected, can I experience objective knowledge. Common sense is that objective knowledge, where being objective means sharing a portion of the world, and thus that which is objective can only be so if it is intersubjective as well. The triadic interaction mentioned by Davidson precisely serves to give an account of empirical content and the concept of objectivity. Explaining how we acquire the concept of objectivity is, for Davidson, philosophically relevant, because it highlights the social nature of thought and language. For Davidson, the concept of objectivity is essential in propositional thought and this concept can emerge, according to Davidson, only in an intersubjective context.30 The analogy between triangulation and common sense seems to be supported by Davidson’s thesis: ontogenetic development of thought and speech takes place through the triangulation process. A child learns the names of objects through a process of correction by a third person. You could say that common sense develops itself in the same way. The concept of common sense is therefore closely linked to the concept of objectivity. That said, Davidson also claims that objectivity has to do 28

Externalism is given by the fact that the content of beliefs and the meaning of words are determined by objects that belong to the outside world. 29 Davidson, Subjective, Intersubjective, Objective, 105. 30 M.C. Amoretti, Il triangolo dell’interpretazione. Sull’epistemologia di Donald Davidson, FrancoAngeli, Milano 2008, 18.

140

Chapter Eight

with the awareness of being in error. Developing this knowledge necessarily requires interaction with others: “If we are not in communication with others, there would be nothing on which to base the idea of being wrong, or, therefore, of being right, either in what we say or what we think. […] It is this triangular nexus of causal relations involving the reactions of two (or more) creatures to each other and to shared stimulations in the world that supplies the conditions necessary for the concept of truth to have application.”31

6. Conclusion Common sense can therefore be understood as a necessary and fundamental process, not only for ordinary knowledge, but also for communication in general. Common sense is that process by which it is possible to share knowledge, but also the process by which we learn to know. This process presupposes the existence of another, which allows for the constant possibility of committing errors. If there was not another with whom I shared a space of knowledge, there would be no possibility of error, nor would there be any objectivity. Each would remain in their subjective world and it would be impossible to talk about common sense. Sense is therefore common, firstly and necessarily, between two subjects. In other words, common sense leads inevitably to an intersubjective dimension. Davidson’s triangular externalism is one possible way to describe how, time after time, common sense is realized.

31

Davidson, Subjective, Intersubjective, Objective, 83.

CHAPTER NINE A NON-SUPERFICIAL CONCEPTION OF COMMON SENSE LOURDES VELÁZQUEZ

Serious thinkers who have defended the rights of common sense have certainly not been lacking in modern philosophy; nevertheless it is also without doubt that the most frequent attitude has been that of considering common sense as a “naïve” stage of knowledge and, therefore, as intrinsically unreliable and, at any rate, needing suitable “critical analyses” that are the specific task of philosophy. It is perhaps not arbitrary to say that such an attitude reflects in part the dominance, within our culture, of the paradigm of the exact sciences, whose advancements are often seen as a progressive dismantling of the images of the world accepted by common sense, which are replaced by those, much less intuitive but technically well-grounded, constituted by the various scientific theories. In a similar way, it is often believed that traditional philosophical problems (especially in epistemology and ontology) can be “seriously” treated today only by resorting to the sophisticated tools offered by mathematical logic, semiotics, analysis of language, or the subtle methodologies of phenomenological analysis. Leaving aside the question of why we should consider these more sophisticated methods and instruments more definite and reliable, the point that must be clarified first is that of knowing to what purpose should such instruments be of use, and here the spontaneous and more or less implicit answer is that they should help us in knowing reality, and in such a way, in attaining truth. We know, however, that various trends in contemporary philosophy have maintained precisely that it is impossible to attain such goals by using those methods, and the anti-realist and skeptical positions have broadly expanded in the domains of philosophy of science and more generally in epistemology.

142

Chapter Nine

By investigating the reasons for such an intellectual attitude, it is possible to ascertain that, at least to a great extent, they reside in having confused and almost identified two distinct requirements of our statements: that of truth and that of certainty. Indeed, it is not difficult to agree with qualifying as true a statement that, regarding a certain “state of affairs”, states that things are actually as it describes them. The problem arises when we want to know with certainty whether a well determined statement is true or not. Therefore, it is not possible to get rid of skepticism by qualifying it as the doctrine according to which we are not able to attain truth: the skeptic does not deny that we can make true statements, but says that we can never be certain whether our statement is true or not. At this point, the classical theme of the criteria of truth becomes inextricably linked with that of the criteria of certainty. Indeed, it remains true that a person can be (subjectively) certain of something that is not true (or better, that is not the case), or that she may have no certainty regarding a true opinion she has. Nevertheless, the aspiration of every human being is precisely that of arriving at acquiring with certainty some true knowledge, and this is something more than the simple “tending naturally to truth” that characterizes human reason. This is a characteristic that we could call “anthropologic” rather than “epistemological”, since it has to do with our living, acting, and orienting our own existence more than with the pure openness of our knowledge of the world. On this ground, we rediscover the value of common sense, which we can identify with that knot of fundamental convictions that sustain our being-in-the-world as conscious beings.

1. The Doctrine of J. Balmes Precisely in order not to confuse this anthropologic capacity with the traditional concept of intellectual intuition, the Catalan philosopher of the nineteenth century, Jaime Balmes, called it “intellectual instinct”. It is to his doctrine of common sense (which is little known outside a restricted circle of specialists) that we want to devote some attention now, for we believe it to be topical at a moment when many persons speak, not without reason, of a “pragmatic turn” in epistemology. The “Eagle of Vich”, as his followers called him, did not want to live on the peaks, to remain shut in his own thoughts, and maintained that the philosopher is not someone who – in isolation from the world – devotes himself only to reflection, but places himself within the universe, included in existence, being a man among men and considering other people’s

A Non-Superficial Conception of Common Sense

143

problems his own problems as well. Therefore he believed that the fundamental themes for a thinker may attain the greatest depth but always need to be rooted in reality, the milieu in which operates the very subject who reflects. In keeping with these ideas, Balmes presents various philosophical topics and investigates issues of general interest based on a rational foundation. One of his major tools is the use of examples, which he adopts as a pedagogic method since his concern is not only that of knowing the causes of things, and penetrating the essence of beings, but he also wants to make them known to others, to teach them. Therefore, his language is always simple and his presentations easy to understand. Even when he treats abstract and difficult themes, the form of his discourse is simple, with the view of rendering simple what is complicated and, in such a way, making it accessible to common people. This is why his work is not always as systematic as one might wish; this lack of a rigorous methodology, however, is instrumental to naturalness and clarity. Therefore, if one wants to qualify it more significantly, one should say that Balmes’ philosophy is a convincing example of the methodological value of common sense. One of Balmes’ aims is to oppose rationalism, especially that of Descartes. Therefore he affirms that the human intellect is by itself capable of internalizing what belongs to the external world under material conditions, without the necessity of the intervention of alien factors (such as “animal spirits”). The real truths, he says, refer to facts or things picked up in the external world and are those that our cognitive faculties perceive in transcendent reality. By demonstrating that the human intellect is capable of abstraction, transforming material conditions into concepts without deforming the real contents, he corroborates his intellectual position and formulates arguments in favor of epistemological realism and its consequences. An equally strong criticism is developed by Balmes against the idealistic doctrines inspired by Kant, which were widespread at his time but not of special interest for us now. We only want to stress that a central role in all these kinds of reflections is played by the theme of “intellectual instinct”, a notion that has attracted the interest of many philosophers. Certain authors, such as González Cordero, qualify it as “a-rational” precisely for this reason, arguing that this “instinct”, though it is called “intellectual”, is a blind impulse. Therefore, though one cannot say that it is contrary to reason, one must nevertheless say that it is not grounded in reason. This is possibly the reason – these authors submit – for which Balmes has made use of this notion in his explicit program of fighting against any form of rationalism. Yet it remains a fact that Balmes clearly states that his

144

Chapter Nine

intellectual instinct can never go against the prescriptions of reason and, moreover, reinforces and corroborates them. Taking into account these conclusive remarks we can say that this philosopher clearly defended the thesis that the human intellect has the capability of attaining the certainty that what it perceives is true, and maintains that whoever denies this capability and the possibility of truth contradicts herself and remains in error. Therefore, Balmes’ theory of certainty is propositive and analytic, and its fundamental elements, if properly considered, represent an important contribution to philosophical research into epistemology.

PART III EMPIRICAL AND PERCEPTUAL CONCEPT FORMATION

CHAPTER TEN PERCEPTUAL CONCEPTS IN ORDINARY KNOWLEDGE MARIANO L. BIANCA AND PAOLO PICCARI

Perceptual concepts, which are formed through empirical generalizations, play a key role in the formation of ordinary knowledge. Indeed, the ordinary gnoseological processes, similarly to what happens in the scientific ones, are guided by perceptual concepts. How is the process of generalization which leads to the formation of perceptual concepts carried out? In order to answer this question we must introduce the notion of mental representation and briefly explain its nature and formation.

1. Generalization Process 1.1 Mental Representations Mental representation is one of the functional modalities of knowledge of the world by way of the use of sense organs, or rather sensorial transducers. Perceptual stimulus Ļ Flow of information from the stimulus to the sensorial organs Ļ Change in the state of the sensorial transducers and neurochemical codification Ļ Elaboration in different cortical and non-cortical areas of the information coming from sensorial transducers Ļ Mental representation Figure 1. Mental Representation Process

148

Chapter Ten

The term ‘representation’ refers to the fact that sense data are elaborated and capable of generating one or more mental configurations whose content (information) is, though not always entirely, isomorphic to the structure of the world it represents. Mental representation, relative to every type of perception (visual, auditory, etc.), is a neuro-mental structure that reports the information from an object or event in the world (see Figure 1). It has a double function: firstly, it is cognitive, that is, it generates reliable and adequate knowledge of the world; secondly, it is operative, that is, it triggers various forms of reaction or behavior, useful to acting and operating in the world, and a series of mental configurations such as thoughts, emotions etc. (Bianca 2009: 481-482). Let us consider the case of visual perception. Take, for example, a book on a table. What happens when we perceive it? The photoreceptors of our retina (cones and rods) perceive visual stimuli (photons) emanating from the book (stimulus source). At this point, the codification of the information coming from the stimulus source takes place and the following neurophysiologic process (structural transposition) give rise to a mental representation, in this case the mental image of the book. A structural transposition is a complex neurophysiologic process which allows the transcription of the reception of the stimulus by way of specific neurochemical codes belonging to the structure and functioning of the central and peripheral nervous systems and generates a structure (the representation) which stands in the mind for the structure of the object emitting the stimulus. The mental representation carries in a codified manner, though not always entirely, the attributes of the structure of the source of the stimulus (Bianca 2005: 90-95). Mental images, for example, originating from visual experience, are sufficiently reliable representations of physical objects and as such are those which are prevalently utilized to epistemically access the world and act within it (Bianca 2009: 27). Thus, a mental representation, which is the result of neuronal processes brought about by sensory stimuli capable of triggering the activation of many neuronal networks, isomorphically carries the structure of the objects of the world, generates different mental configurations and can provoke specific behaviors and actions. For example, in the brain of a gazelle the representation ‘lion’, which stands for a lion in the world, is the cerebral configuration that allows her to trigger a neurophysiologic configuration which corresponds to escape, and as such, to the act of escaping. The previous description refers to a cognitive process capable of formulating a representation. This is the final step of an elaboration in

Perceptual Concepts in Ordinary Knowledge

149

which information transmitted by a stimulus from an external source is processed in different areas of the brain. As such, the representation provides knowledge that is reliable and adequate about the object (which it stands for) throughout the elaboration of the information coming to the sense receptors from the stimulus, thereby triggering actions conforming to the state of the subject and the structure of the object. The aggressive stance of the lion, for example, is not mistaken by a gazelle as a gesture of submission, but is taken for what it really is; thus, the adequate action, that of escape, is triggered in the gazelle by the aggressive stance of the lion. In this case, the cerebral representation of the lion in the brain of the gazelle is both reliable and adequate; reliable because it carries the attributes of the lion and adequate because it allows for a certain type of behavioral response, consonant with the received information, to be put into action. As far as the formulating process of mental representations in humans is concerned, it is necessary to observe that a human mental representation, for example a visual perception (a visual image), is not just a reliable and adequate representation of one or more objects of the world, but bearing a collection of attributes is an ‘empirical generalization’ based on perceptual instances of similar objects which are assembled into a class. In which case, it can be considered a conceptual structure (or concept) referring to an object or a class of objects; a concept that might be useful for classifying experienced objects in successive instances. Furthermore, if one considers that the perceptual experience of the human species is prevalently visual in nature, one should conclude that most of the concepts derived from empirical generalization processes are of a visual nature and those derived by auditory, tactile, olfactory and gustative experiences smaller in number.

1.2 Two Kinds of Generalization Process: Generalization e Pluribus and Generalization ex uno The formation of a mental representation is the process preceding that of generalization insofar as the empirically acquired information is elaborated in a representational form. Therefore, for example, it would be possible to inductively form the concept ‘book’ on the basis of different visual perceptions (or representations) of books which referred to single perceptual visual instances of the ‘book’ (see Figures 2 and 3). The inductive process is a generalization of perceptual instances of type X on the basis of the trial of a limited number of perceptual instances of type X. According to Aristotle’s definition in Topics (I, 12, 105 to 11), induction is a process in which details lead to universals, that is, the kind

150

Chapter Ten

of reasoning that formulates a general statement referring to all cases which possess similar characteristics from a group of particular statements referring to single cases. In inductive reasoning, there are three distinguishable parts: a) the base, b) the passage, and c) the conclusion. The first part contains statements that refer to single actually experienced instances; the second, statements that are not formulated but refer to possible other perceptual non-experienced instances of type X; finally, the third contains one or more statements in the form of generalization through which one states or assumes that the attributes evident for a few perceptual instances of type X experienced count for all of the analogous perceptual instances of type X not experienced, and therefore one concludes that they count for all of the possible instances of type X. In the description of the process of generalization, which gives rise to the formulation of perceptual concepts, we mean to make reference to induction in the statistical probability sense, expressed by the following rule: when we have observed a sample of objects Į and found that the frequency of the objects ȕ between them is f, we assume that P (Į, ȕ) = f, that is, the probability that Į. is ȕ is f, calculated on the basis of the statistical frequency of ȕ in Į. In cases in which a specific attribute occurs in a wide proportion of experienced objects of type X, one can assume that attribute is common to all of the other objects of the same type though not experienced, except when proven otherwise. When the proportion is the same in 100% of the objects experienced, that is when the attribute occurs in all of them, there is a uniform or complete generalization. This is the case when we state that ‘all men are mortals’ due to the fact that being mortal has always been found to be a constant in association with humankind. However, when the numeric value of that specific proportion is used as a measurement of the possibility that the attribute in question will occur in a new perceptual instance, there is a judgment of probability. As such, it is adequate to specify that uniform generalization and judgments of probability are aspects of statistical generalization based on the frequency of occurrences of an attribute in a sample of observed cases. The probability p(A) of the occurrence of an attribute identifying a class in a new perceptual instance, like, for example, ‘blackness’ in a class of crows, is always 0 ” p(A) ” 1; p(A) = 0 if and only if A is the impossible event; (A) = 1 if and only if an event is certain. There are two forms of perceptual generalization, even in the statistical sense: e pluribus generalization and ex uno generalization (Bianca 2009: 261-265). E pluribus generalizations contain information that is attributive

Perceptual Concepts in Ordinary Knowledge

151

and referable, following the inductive process, to all of the objects of a given class (the experienced objects that we consider in the inductive process and the objects yet to be experienced). An e pluribus generalization can therefore be considered a concept that contains several attributes possessed by all (or nearly all) of the objects of a class. It is preserved in the memory as a perceptual type and is useful for classifying objects in successive perceptual experiences (Figure 2).

Mental representation of O1

Mental representation of O2

…

Mental representation of On

GENERALIZATION

Concept or perceptual type

Figure 2. E pluribus generalization (O = Object)

However, in the case of ex uno generalizations, certain recurrent attributes are assigned to a determinate object or event that has been experienced many times. This generalization identifies the single object and is accepted as referable in every case in which that single object or event has been or could be experienced or perceived. Ex uno generalization is a neuro-mental process that carries all of the attributes individuated in different perceptual instances of the same object, thus formulating a perceptual type, and allows a new perceptual instance of that object to be subsumed into that perceptual type (Figure 3). Ex uno generalizations refer to a class of perceptual instances of the same object experienced in different instances, for example: ‘my university office yesterday evening’, ‘my university office after the daily

152

Chapter Ten

cleaning’, ‘my university office with a new computer’; instances which are experientially different but refer to the same object.

Perceptual instance1 of Ox

Perceptual instance2 of Ox

…

Perceptual instance n of Ox

GENERALIZATION

Concept or perceptual type

Figure 3. Ex uno Generalization (O = Object)

Thus, we can define this generalization according to the following: given a certain number of perceptual instances of an object/event, a number of attributes get assigned to it which are applicable to perceptual instances of this object, although not all of them are applicable to every instance. In other words, ex uno generalization identifies the single object and makes reference to all of the perceptual instances of that single object/event, those which have been verified and those which are yet possible. Generalizations of both forms, which give rise to perceptual types or perceptual concepts, are particularly important to everyday knowledge because they allow for the identification, recognition and thereby the classification of the objects in the world, and for their groupings into classes, consisting of different objects in the case of e pluribus generalizations and of different perceptual instances of the same object in the case of ex uno generalizations.

Perceptual Concepts in Ordinary Knowledge

153

2. Conceptual Framework The classical theory of concepts states that the classes of objects are defined by necessary and sufficient criteria of belonging. All members of a class should therefore possess an equal degree of belonging to that class, that is, they should have all the attributes that typify that class and make up the necessary and sufficient conditions for membership. Actually, that may hold for some artificial and conventional concepts, but not for most perceptual concepts. Indeed, classes are not logical entities defined by a set of necessary and sufficient conditions, but are determined by a set of identitive attributes most frequent (in the statistical sense) in members of the same class; from this set it is possible to obtain a typical exemplar by abstraction, one that does not possess the average value or the highest number of characteristics shared by members of the class as expected by prototype theory (Rosch 1977), but the identitive attributes common (most frequent) to those members. With regards to the attributive structure of perceptual concepts, it is essential to underline that, contrary to what many scholars maintain, a perceptual concept does not consist solely of attributes referable to the physical properties of objects, but also of other attributes, assigned to objects by the individuals who experience them. For this reason we will be considering perceptual concepts as conceptual frameworks, within which it is possible to distinguish two structures: a) an identitive attributive nucleus (IAN) that refers to a class of objects or to different perceptual instances of the same object and is defined according to statistical criteria, which consists of empirical generalization, be it e pluribus or ex uno, operating on the basis of information contained in the mental representations corresponding to specific phenomenic objects, and b) a specifying attributes set (SAS), referable to single objects belonging to a class, or in denotative terms, the set of specifying attributes or characteristic predicates of given members of a class. Furthermore, the IAN, made up of identitive attributes (or predicates), represents the intention of the concept, i.e. the set of predicates that, from a connotative point of view, define it. The SAS, on the other hand, is the set of attributes specifically characterizing each single object of one class, and also possesses one, some or all of the attributes of the IAN. For example, in the case of the concept ‘table’, the IAN is made up of identitive attributes (or predicates) like ‘having a rigid panel which is parallel to the ground’, and ‘having one or more supporting legs’, whereas the SAS unites all of the specifying attributes (or predicates) referable to single members of the

154

Chapter Ten

class of ‘tables’, such as ‘having drawers’, ‘having fold-out sections’, ‘being foldable’, etc. Therefore, we consider concepts conceptual frameworks (CONF), which are represented by a five-dimensional vector: CONF = . These dimensions, examined in the sections below, correlate with one another and allow a concept, according to precise modalities and a determinate state of mind, to perform the function of conceptual operator, capable of triggering the processes of classification of objects and subsequently triggering various thoughts and actions. At this point, it is necessary to formulate an analytical description of each dimension of the vector CONF.

2.1 The Identitive Attributive Nucleus (IAN) The IAN is the result of an empirical generalization, which might be, as previously stated, either e pluribus or ex uno. In the first case, the attributes of the IAN are individuated on the basis of their empirical statistical frequencies as revealed in different objects, namely, they are most often found in perceptual instances of objects and, by induction, may be assigned to objects experienced at a later time. The identitive attributes perceived are assigned both a series of possible values, that is, the range of possible values of an attribute, and a value of default, which is the value given to an attribute in a typical exemplar. For example, in the concept ‘table’, the attribute ‘number of legs’ might be ‘4’ – the typical table has 4 legs – while the possible values are many: 1, 2, 3, 4, 5, etc. Analogously, a human being, who generally has two eyes (the default value being ‘2’), may, due to a physical accident or a genetic abnormality, have only one or even none (possible values 1 and 0). Therefore, the set of default values configures the IAN, that is, describes an object that can be immediately recognized as belonging to a given class. Let us consider, for example, the class ‘crows’: an identitive perceptual attribute of this class is indisputably represented by ‘blackness’, the color whose high frequency characterizes the plumage of single crows. However, a white colored passerine with all of the identitive attributes of corvids, though lacking the identitive attribute of ‘blackness’, can regardless be considered a crow, even if the attribute of ‘whiteness’ found in that exemplar has a low statistical occurrence within the class of ‘crows’ and furthermore, can be considered a specific attribute of that corvid. Thus, this bird might be placed by all rights in the class of corvids, with its

Perceptual Concepts in Ordinary Knowledge

155

color (white) being considered not as an identitive attribute of the class it belongs to, but rather as a specifying attribute (an SAS attribute). The same can be said with regards to a human being who, due to an accident or genetic abnormality, is without hands. In this case, the perceptual instance refers to an individual who possesses all of the other identitive attributes of the class ‘human beings’, thereby rendering the absence of the identitive attribute ‘hands’ irrelevant from the point of view of statistical occurrence, because the coexistence of other identitive attributes legitimizes their classification among the members of the class ‘human beings’. In the second case, that of ex uno generalizations, an analogous process develops, but in reference to a single object which, as we have seen, has been exposed to several perceptual instances. It follows, then, that in ex uno generalization, the frequency of the attributes refers to the perceptual instances of the same object. It is therefore possible to define the IAN as the set of identitive attributes empirically checked on the basis of their statistical frequency with respect to a sample of objects or to the identitive attributes checked with the highest frequency related to different perceptual instances of the same object.

2.2 The Specifying Attributive Set (SAS) The SAS is the set of perceptual specifying attributes (added to the identitive ones) pertaining to the individual members of a class or the individual perceptual instances of the same object. In other words, in the case of an e pluribus generalization, it is the set of all of the predicable attributes of objects belonging to the same class (these objects at same time share one, some or all the predicates of the IAN) or in the case of an ex uno generalization, to different perceptual instances of single objects. In the case, for example, of the class ‘chairs’, the IAN would be made up of identitive attributes such as ‘having a seat’, ‘having four legs’, ‘having a back’, with which specifying attributes which make up the SAS, such as ‘being foldable’, ‘having arms’ etc., can be associated. Analogously, the SAS can also refer to many perceptual instances of one single object: in this case, it is made up of the attributes of the experienced object from several perceptual instances (see Section 3.2.1). Furthermore, if we take into consideration ‘my watch’, the corresponding SAS will be made up of all the specifying attributes found in each perceptual instance of ‘my watch’, among which will be ‘my watch this morning’, ‘my watch on my desk’, ‘my watch with the glass broken’, etc.

156

Chapter Ten

2.3 Significances (SIG) SIG consist of the different and possible significances assigned to the objects of one class, or a class formed by different perceptual instances of the same object. For example, our concept of ‘watch’ is not constituted by its features alone (the image of the watch as structurally isomorphic to experienced watches), but includes many significances that we have assigned to the watch that we use, that is, for example to ‘our watch, in relation to a particular moment in our existence in which we bought it’ or ‘our watch with the sentimental significance referring to the person gave it to us’, etc. Thus, as an example, we can assign many significances to the watch that we own: the significance of ‘mine’ which distinguishes it from other watches, making it unique with respect to the entire class to which it belongs; ‘a gift from the woman I love’, which bears an emotionalexistential content referring to the giver; ‘precious’ or ‘beautiful’, which define the specific qualities of that particular watch, both from a material or aesthetic point of view and from an emotional one (precious because it is made of gold or is from a loved one).

2.4 The Reference (SEM) SEM is a concept’s reference to a class of objects or to one object. Thus, the concept ‘cat’ is also made up of its reference, that is, of all of the animals that possess the identitive attributes pertaining to the class of cats, and as such are classified as cats, or even a single cat that has been experienced repeatedly, as would be true of the concept ‘my cat’, resulting in an ex uno generalization.

2.5 The Name (NOM) A ‘name’ as a linguistic expression is assigned to every concept, thereby allowing for the nominability of the class of objects or the single object to which the concept refers. The name ‘cat’, for example, indicates the concept that refers to a specific class of animals. The NOM brings the related concept into awareness and as such can be used for different purposes: to formulate different cognitive functions, to evoke emotions or mood states, or to recall other concepts linked to the sharing of one or more attributes.

Perceptual Concepts in Ordinary Knowledge

157

3. Visual concepts In this section we will briefly consider the structure of what we call visual concepts, because in the human perceptual experience, these are the most common and most relevant. Here we are referring in particular to the visual concepts formulated by e pluribus generalization. Analogously, we can analyze the visual concepts derived from ex uno generalizations. The notable relevance of visual concepts has a neurophysiologic basis. The part of the CNS specialized in the elaboration of the visual stimuli (the primary and secondary visual areas in the occipital lobes) is more vast and complex than the cortical areas which elaborate information coming from the other sense organs. Perceptual concepts are the cognitive result of the constant and intense perceptual activity of humans. Thus, there are perceptual concepts of a tactile origin, such as ‘smooth’ or ‘soft’, others that are olfactory such as ‘stench’ or ‘fragrant’, others that are auditory like ‘sound’ or ‘noise’, and others still that are gustatory such as ‘sweet’ or ‘salty’. Finally, there are concepts derived from visual perception, which represent the greater portion of perceptual concepts and are made up of figural attributes (but also of non-figural attributes) predicable of an object. Indeed, a visual predicate refers to the figural attributes of the object (dimension, color etc.), while the non-figural attributes are related to the non-figural structure of an object or to its functional attributes: ‘being high’ in reference to a mountain is a figural predicate, while ‘being scrupulous’ in reference to a human being is a non-figural predicate. Often in a concept’s connotative definition both types of predicates are included, for example, in the cases of the concepts ‘dog’, ‘airplane’ and ‘house’, although concepts that do not have figural predicates are also not rare, like the concepts of ‘tasty’ or ‘salty’. For example, regarding a dog one could predicate figural attributes like ‘quadruped’, or ‘having a tail’, and nonfigural attributes like ‘being loyal’ or ‘being domestic’. At any rate, in daily life we rely heavily both on visual concepts which refer to single objects, like ‘my car’, and on classes of objects, like ‘works of visual art’, which are predominately figural. These concepts distinguish themselves from non-visual concepts in that figural attributes are those that define them and make up the identitive attributive nucleus which individuates them and distinguishes between them. Hence, their IAN is mainly made up of figural attributes. Visual concepts consist of identitive figural attributes which can be applied to a class of objects (even if the class consists of only one object). On the one hand, they can be ex uno and

158

Chapter Ten

e pluribus visual generalizations, and on the other they can be formulated independently of visual generalization based on visual perceptions. In the first case, a visual concept derives from a visual generalization, which can be codified in figural or non-figural language and is stored without any further direct reference to cases which may have allowed for its formulation. As such, visual concepts play a role of figural types which refer to a class in accordance with the principle of cognitive economy, indispensable for the correct and efficacious functioning of the mind. In the second case, there are visual concepts which are independent of generalizations and have no reference to cases or occurrences, and their formulation derives from the use of visual material stored or formulated ex novo, although it utilizes material which has been stored previously. This is the case, for example, with visual concepts from the mythological matrix, such as the unicorn, phoenix, or sphinx, which utilize visual perceptual material derived from experience referring to animals such as the horse and the lion, albeit re-elaborated by our mind or through the visual perception of illustrations or photos. Therefore, a visual concept consists of a set of entirely or primarily figural attributes and originates from the elaboration of information both visual and stored in the long term memory as schemes to be recalled (sometimes, though not always, consciously) in the moment of recognizing an object and assigning it to the class to which it belongs. In ordinary thought, visual concepts are rarely made up entirely of figural attributes, because the objects are often perceived in a perceptual space that is broad and ‘contaminated’ by non-visual attributes. A ‘neutral’ space, the ‘pure’ background upon which objects can be seized and perceived by the subject in their empirical actuality exists only from a theoretical point of view. The analysis of mental images (perceptual and non-perceptual) and of the processes involved aids in revealing the structure of perceptual concepts and the specific dimensions (IAN, SAS, SIG, SEM, NOM) much more clearly, and consequently, in particular, the identitive and specifying attributes. Indeed, as is clear from current neurophysiologic research on vision, after having been elaborated by visual apparatuses, including areas of the visual cortex, visual stimuli are elaborated with information from different cortical and non-cortical areas and thus the image which gets formulated is the result of different and complex elaborations which involve cortical and non-cortical information. Hence, images cannot be reduced to the mere elaboration of information from visual stimuli (Bianca 2009:109-158).

Perceptual Concepts in Ordinary Knowledge

159

As has already been shown, at the same time, different visual mental representations are further elaborated in the formulation of generalizations (i.e. sharing predicates related to different representations), which are visual concepts whose IAN, contrary to other concepts, is made up primarily of figural attributes. The analysis of visual concepts, therefore, can corroborate the thesis on the structure of perceptual concepts that we formulated in the previous sections.

4. Concluding remarks This article has examined the nature of perceptual concepts, describing their structure through the formulation of a theoretical model based on statistical criteria which, if researched further and verified, could indicate a new perspective for the study and analysis of the formation of conceptual structures based on perceptual instances. The formation of perceptual concepts is a fundamental process for knowing and describing the physical world, which is useful for ‘ordering’ our experience, distinguishing one object from another, recognizing respective differences and acting in our world. In human beings, evidently, perceptual concepts possess a complex structure because they are influenced by the cultural environment and their formation includes information which comes from different cortical, sub-cortical, and noncortical areas. To this end, we have examined the structure of perceptual concepts not only from the point of view of their perceptual content, but also taking into consideration other information of a non-perceptual nature (for example, the different significances that a subject can attribute to one or more objects belonging to a class) generally overlooked in studies and research carried out to date, both in the field of philosophy and cognitive psychology. Perceptual concepts are undoubtedly the means of formulating awareness of the physical world. They speak of this world and allow for its description by virtue of their perceptual foundation and hence lend a reliable and adequate (as discussed) empirical understanding, whose end is to live and thrive within it. In light of these considerations, a model was formulated in which concepts are considered in their complexity as conceptual frameworks and each one represented by a five-dimensional vector in which, in addition to the merely perceptual contents (the identitive and specifying perceptual attributes), non-perceptual contents (semantic references and various significances assigned to concepts by single individuals) are analyzed. This model does not consider perceptual concepts the simple result of a

160

Chapter Ten

generalization conducted on the basis of different perceptual instances, but highlights the relevance of non-perceptual contents involved in their formation process. The model outlined in this paper tries to overcome the contrast between perceptual and non-perceptual mental processes, since many perceptual concepts contain non-perceptual attributes resulting from an autonomous neuro-mental elaboration carried out by every single individual in relation to her/his experience with the information contained in the different neuro-mental areas; attributes assigned on the basis of statistical frequency and on the two types of generalization. Therefore, contrary to what is generally claimed, there are not only concepts that refer to a class of objects (e pluribus generalizations), but also concepts that refer to a class of perceptual instances related to single objects (ex uno generalizations). Perceptual concepts possess a complex informative structure, like that indicated in Section 4, and cannot be reduced to the mere elaboration of information from perceived stimuli. This model requires further in-depth analysis, including experimental trials, in order to clarify aspects in this article that have not been dealt with or that have only been touched upon briefly.

References Armstrong, S., Gleitman, L., Gleitman, H. (1983). What some concepts might not be. Cognition, 13, 263-308. Bianca, M. L. (2005). Rappresentazioni mentali e conoscenza. Un modello teorico-formale delle rappresentazioni mentali, Milano, FrancoAngeli. —. (2009). La mente immaginale. Immaginazione, immagini mentali, pensiero e pragmatica visuali, Milano, FrancoAngeli. Bianca, M.L., Lucia Foglia (2006). Non-Perceptive Mental Image Generation: a Non-Linear Dynamic Framework. Anthropology & Philosophy, 7, 28-63. Bianca, M.L., Piccari, P. (2007). Inherent Logic: Isotopic and Inherent Bonds in Argumentation. Anthropology & Philosophy, 8, 1-2, 9-31. Carey, S. (2009). The Origin of Concepts, New York, Oxford University Press 2009. Gelman, S.A., Coley, J.D. (1991). Language and categorization: The acquisition of natural kind terms. In Gelman, S.A., Byrnes, J.P. (eds). Perspectives on Language and Thought. Interrelations in Development, Cambridge, Cambridge University Press.

Perceptual Concepts in Ordinary Knowledge

161

Hampton, J.A. (1995). Testing the Prototype Theory of Concepts. Journal of Memory and Language, 34, 686-708. Keil, F.C. (1989). Concepts, Kinds, and Cognitive Development, Cambridge (MA), The MIT Press. Kneale, W.C. (1949), Probability and Induction, Oxford, Oxford University Press. Lakoff, G. (1987). Women, Fire, and Dangerous Things: What Categories Reveal About the Mind, Chicago, University of Chicago Press. Laurence, S., Margolis, E. (2002). Concepts and Concept Analysis. Philosophy and Phenomenological Research, LXVII, 2, 253-282. Machery, E. (2009). Doing without Concepts. New York, Oxford University Press. Margolis, E. (1998). How to Acquire a Concept. Mind & Language, 13, 347-369. Margolis E., Laurence, S. (1999). Concepts and Cognitive Science, in Idd. (eds.), Concepts. Core Readings, Cambridge (MA), The MIT Press, pp. 3-81. Margolis E., Laurence, S. (2007). The Ontology of Concepts. Abstract Objects or Mental Representations? Noûs, 41, 561-593. Murphy, G. L. (2002). The Big Book of Concepts, Cambridge (MA)London, The MIT Press. Murphy, G.L., Medin, D.L. (1985). The Role of Theories in Conceptual Coherence. Psychological Review, 92, 289-316. Piccari, P. (2011), Conoscenza ordinaria e senso comune, Milano, FrancoAngeli. Prinz, J.J. (2002). Furnishing the Mind. Concepts and Their Perceptual Basis, Cambridge (MA), The MIT Press. Rips, L. J. (1989). Similarity, tipicality, and categorization. In Vosniadou, S., Ortony, A. (eds). Similarity and Analogical Reasoning, Cambridge, Cambridge University Press, 21-61. Rosch, E. (1975a). Cognitive Reference Points. Cognitive Psychology, 7, 532-547. Rosch, E. (1975b). Cognitive Representation of Semantic Categories. Journal of Experimental Psychology: General, 104, 192-233. Rosch, E. (1977). Human Categorization, in Warren, N. (ed.) Studies in Cross Cultural Psychology, I, London, Academic Press. —. (1978). Principles of Categorization, in Rosch, E., Loyd, B.B. (eds.), Cognition and Categorization, Hillsdale (NJ), Erlbaum. Rosch, E., Simpson, C., Miller, R.S. (1976). Structural Bases of Tipicality Effects. Journal of Experimental Psychology: Human Perception and Performance, 2, 491-502.

162

Chapter Ten

Smith, E.E., Medin, D.L. (1981). Categories and Concepts, Cambridge (MA), Harvard University Press.

CHAPTER ELEVEN COMMON SENSE CONCEPTS BETWEEN COMPOSITIONALITY AND TYPICALITY EFFECTS: A DUAL PROCESS APPROACH MARCELLO FRIXIONE AND ANTONIO LIETO

1. Introduction Typicality effects are a well-established phenomenon in the field of cognitive research on common sense concepts. However, typicality effects are hard to reconcile with compositionality, a crucial characteristic of conceptual systems. In order to face this problem, we propose an approach to conceptual representation inspired by the so-called dual process theories of reasoning and rationality. According to these theories, the existence of two different types of cognitive systems is assumed. The systems of the first type (type 1) are phylogenetically older, unconscious, automatic, associative, parallel and fast. The systems of the second type (type 2) are more recent, conscious, sequential and slow, and are based on explicit rule-following. We advance the hypothesis that conceptual representations should consist of (at least) two different kind of components,1 each responsible for different processes: type 2 processes, involved in complex inference tasks, and fast and automatic type 1 processes, which perform categorization, taking advantage of the prototypical information associated with concepts. The chapter is organized as follows: in section 2 we expose the problem of reconciling compositionality and typicality effects. This same problem emerges in the field of concept-oriented knowledge 1

For a similar approach, see Piccinini (2011). A way to split the traditional notion of concept along different lines has been proposed by Machery (2005).

164

Chapter Eleven

representation systems developed in artificial intelligence (section 3). In sectsions 4 and 5 we illustrate the dual process approach to cognition and its relevance for concept representation. Some conclusions follow.

2. Prototypical Effects vs. Compositionality in Concept Representation In the field of cognitive psychology, most research on concepts moves from critiques to the so-called classical theory of concepts, i.e. the traditional point of view according to which concepts can be defined in terms of necessary and sufficient conditions. The central claim of the classical theory of concepts is that every concept c can be defined in terms of a set of features (or conditions) f1, ..., fn which are individually necessary and jointly sufficient for the application of c. In other words, everything that satisfies features f1, ..., fn is a c, and if anything is a c, then it must satisfy f1, ..., fn. For example, the features that define the concept bachelor could be human, male, adult and not married; the conditions defining square could be regular polygon and quadrilateral. This point of view was unanimously and tacitly accepted by psychologists, philosophers and linguists until the middle of the 20th century. Chronologically, the first critique of classical theory was due to a philosopher: in a well known section from Philosophical Investigations, Ludwig Wittgenstein observes that it is impossible to identify a set of necessary and sufficient conditions to define a concept such as GAME (Wittgenstein 1953, § 66). Therefore, concepts exist which cannot be defined according to classical theory, i.e. in terms of necessary and sufficient conditions. Concepts such as GAME rest on a complex network of family resemblances. Wittgenstein introduces this notion in another passage in Investigations: “I can think of no better expression to characterise these similarities than “family resemblances”; for the various resemblances between members of a family: build, features, color of eyes, gait, temperament, etc. etc.” (ibid., § 67). Wittgenstein's considerations were corroborated by empirical psychological research, starting with the seminal work by Eleanor Rosch (1975), with psychological experiments that showed how common sense concepts do not obey the requirement of the classical theory:2 common sense concepts cannot usually be defined in terms of necessary and sufficient conditions (and even if for some concepts such a definition is 2

On the empirical inadequacy of the classical theory and the psychological theories of concepts see Murphy (2002).

Compositionality and Typicality Effects

165

available, subjects do not use it in many cognitive tasks). Concepts exhibit prototypical effects: some members of a category are considered better instances than others. For example, a robin is considered a better example of the category of birds than, say, a penguin or an ostrich. More central instances share certain typical features (e.g. the ability of flying for birds, having fur for mammals) that, in general, are neither necessary nor sufficient conditions. Prototypical effects are a well established empirical phenomenon. However, the characterisation of concepts in prototypical terms is difficult to reconcile with the compositionality requirement. In a compositional system of representations, we can distinguish between a set of primitive or atomic symbols and a set of complex symbols. Complex symbols are generated from primitive symbols through the application of a set of suitable recursive syntactic rules (generally, a potentially infinite set of complex symbols can be generated from a finite set of primitive symbols). Natural languages are the paradigmatic example of compositional systems: primitive symbols correspond to the elements of the lexicon, and complex symbols include the (potentially infinite) set of all sentences. In compositional systems, the meaning of a complex symbol s functionally depends on the syntactic structure of s as well as the meaning of the primitive symbols in it. In other words, the meaning of complex symbols can be determined by means of recursive semantic rules that work in parallel with syntactic composition rules. This is the so-called principle of compositionality of meaning, which Gottlob Frege identified as one of the main features of human natural languages. Within cognitive science, it is often assumed that concepts are the components of thought, and that mental representations are compositional structures recursively built up starting from (atomic) concepts. However, according to a well-known argument by Jerry Fodor (1981), prototypical effects cannot be accommodated with compositionality. In brief, Fodor's argument runs as follows: consider a concept like PET FISH. It results from the composition of the concept PET as well as the concept FISH. However, the prototype of PET FISH cannot result from the composition of the prototypes of PET and FISH. For example, a typical PET is furry and warm, a typical FISH is greyish, but a typical PET FISH is neither furry and warm nor greyish. Therefore, some strain exists between the requirement of compositionality and the need to characterise concepts in compositional terms.

166

Chapter Eleven

3. Representing Concepts in Computational Systems The situation outlined in the section above is, to some extent, reflected by the state of the art of AI and, in general, in the field of computational modelling of cognition. This research area often seems to vacillate between different (and hardly compatible) points of view (Frixione and Lieto 2011). In AI, the representation of concepts is faced mainly within the field of knowledge representation (KR). Symbolic KR systems (KRs) are formalisms whose structure is, broadly speaking, language-like. This usually entails assuming that KRs are compositional. In their early development (historically corresponding to the late 1960s and the 1970s), many KRs oriented to conceptual representations attempted to take into account suggestions from psychological research. Examples are early semantic networks and frame systems. Frame and semantic networks were originally proposed as alternatives to the use of logic in KR. The notion of frame was developed by Marvin Minsky (1975) as a solution to the problem of representing structured knowledge in AI systems.3 Both frames and most semantic networks allowed for the possibility of characterising concepts in terms of prototypical information. However, such early KRs were usually characterised in a rather rough and imprecise way. They lacked a clear formal definition, with study of their meta-theoretical properties being almost impossible. When AI practitioners tried to provide a stronger formal foundation to concept oriented KRs, it turned out to be difficult to reconcile compositionality and prototypical representations. As a consequence, they often chose to sacrifice the latter. In particular, this is the solution adopted in a class of concept-oriented KRs which were (and still are) widespread within AI, namely the class of formalisms that stem from so-called structured inheritance networks and the KL-ONE system (Brachman and Schmolze 1985). Such systems were subsequently called terminological logics, and today are usually known as description logics (DLs) (Baader et al. 2010). From a formal point of view, description logics are subsets of first order predicate logic that, if compared to full first order logic, are computationally more efficient. In more recent years, representation systems in this tradition (such as formal ontologies) have been directly formulated as logical formalisms (the above mentioned description logics), in which Tarskian compositional semantics is directly associated with the syntax of the language. This has 3

Many of the original articles describing these early KRs can be found in Brachman and Levesque (1985), a collection of classic papers from the field.

Compositionality and Typicality Effects

167

been achieved at the cost of not allowing exceptions to inheritance and, in so doing, we have forsaken the possibility of representing concepts in prototypical terms. From this point of view, such formalisms can be seen as a revival of the classical theory of concepts, in spite of its empirical inadequacy in dealing with most common sense concepts. Nowadays, DLs are widely adopted within many application fields, in particular within that of the representation of ontologies. This state of affairs constitutes a problem for KR systems, since prototypical effects in categorisation and, in general, in category representation, are of the greatest importance in representing concepts in both natural and artificial systems. Several proposals have been advanced to extend concept-oriented KRs, and DLs in particular, in such a way that non-classical concepts can be represented. Various fuzzy extensions of DLs (Bobillo and Straccia 2009) and ontology-oriented formalisms have been proposed that can represent vague information in semantic languages. However, from the standpoint of conceptual knowledge representation, it is well known (Osherson and Smith 1981) that approaches to prototypical effects based on fuzzy logic encounter difficulties with compositionality. (In short, Osherson and Smith show that approaches to prototypical effects based on fuzzy logic are vulnerable to the problem of compositionality mentioned at the end of section 2, above; for a discussion of Osherson and Smith's argument see section 5.) A different way to face the representation of non-classical concepts in DL systems are DL extensions based on non-monotonic logic. We shall return later (section 4) to the plausibility of non-monotonic extensions of DL formalisms as a way to face the problem of representing concepts in prototypical terms.

4. Concepts and Dual Process Theories In the field of cognitive psychology, good reasons emerged to suspect that the very notion of concept is somewhat heterogeneous. Various positions and theories on the nature of concepts have been proposed. These are usually grouped into three main classes: prototype views, exemplar views and theory-theories (see again Murphy 2002). These approaches turn out not to be mutually exclusive. Rather, they seem to succeed in explaining different classes of cognitive phenomena, and many researchers hold that all of them are needed to explain psychological data. This consideration led Edouard Machery (2005, 2009) to claim that concept is not a natural kind, hypothesising that three different natural kinds exist, corresponding respectively to prototypes, exemplars and theories. In the following passages, we advance the hypothesis that there

168

Chapter Eleven

are other reasons to suspect that concepts are not a homogeneous class of entities from the standpoint of cognitive science. In particular, taking into account the considerations emerging from section 3, above, we shall consider suggestions from the so-called dual process theories of mind. According to the dual process theories (Stanovich and West 2000, Evans and Frankish 2008, Kahneman 2011) two different types of cognitive processes and systems exist, which have been called respectively system 1 and system 2. System 1 processes are automatic. They are phylogenetically older, and are shared by humans and other animal species. They are innate, and control instinctive behaviors, so they do not depend on training or specific individual abilities and, in general, are cognitively undemanding. They are associative, and operate in a parallel and fast way. Moreover, system 1 processes are not consciously accessible to the subject. System 2 processes are phylogenetically more recent, and are specific to the human species. They are conscious and cognitively penetrable (i.e. accessible to consciousness), and are based on explicit rule-following. As a consequence, system 2 processes are sequential and, if compared to system 1, are slower and cognitively more demanding. Performances that depend on system 2 processes are usually affected by acquired skills and differences in individual capabilities. The dual process approach was originally proposed to account for systematic errors in reasoning: systematic reasoning errors (consider the classical examples of the selection task or the conjunction fallacy) should be ascribed to fast, associative and automatic system 1 processes, while system 2 is responsible for the slow and cognitively demanding activity of producing answers that are correct with respect to the canons of normative rationality. In our opinion, the distinction between system 1 and system 2 processes may also be plausibly applied to the problem of conceptual representations, as emerges in the above section (for a similar position in this respect, see Piccinini 2011; we shall briefly return to Piccinini’s stance in the conclusions of this paper – section 5). For example, categorization based on typical traits (either prototype based or exemplar based) is, presumably, in many cases a fast and automatic process which does not require any explicit effort, and which therefore could presumably be attributed to a type 1 system. On the contrary, there are types of inference that are usually included within "conceptual competence" which are slow and cognitively demanding and which should be attributed to processes that are more likely to be ascribed to type 2. An example could be

Compositionality and Typicality Effects

169

categorization processes based on complex classical definitions given in terms of necessary and sufficient conditions. Consider, for example, the well-known Linda problem and the aforementioned conjunction fallacy (Tversky and Kahneman 1983). In a well-known experiment from the psychology of probabilistic reasoning, subjects are given a description of a person named Linda who perfectly fits the stereotype of a feminist activist. Then they are asked to judge if it is more likely that Linda is (a) a bank teller or (b) a bank teller and a feminist. The majority of subjects choose (b), without realizing that being a feminist bank teller is in any case more demanding (and therefore less probable) than simply being a bank teller. Indeed, the conjunction fallacy consists in failing to realize that a conjunction is always less probable than its conjuncts. From the standpoint of a theory of reasoning, this is a paradigmatic case of a systematic error that, according to the dual process approach, can be explained in terms of the system 1 vs. system 2 distinction: the conjunction fallacy is an exemplary case of the effects of a system 1 process. However, if we consider the problem from the point of view of a theory of concepts, the conjunction fallacy can be interpreted as an example of the strong tendency of human subjects to resort to prototypical information in categorization, even when this is not appropriate (at least from the point of view of those who devised the experiment): Linda is categorized as a feminist because she perfectly fits the prototypical traits of a feminist (while she does not in the least fit the prototypical traits of a bank teller). So, we could see the “error” as being originated by a process of categorization based on prototypical knowledge. Typicality effects in categorization and, in general, in category representation are crucial for the human cognition. Under what conditions should we say that somebody knows the concept DOG (or, in other terms, that (s)he possesses an adequate mental representation of it)? It is not easy to say. However, if a person does not know that, for example, dogs usually bark, that they typically have four legs and that their bodies are covered with fur, that in most cases they have a tail and that they wag it when they are happy, then we should probably conclude that this person does not grasp the concept DOG. Nevertheless, none of these pieces of information are either necessary or sufficient conditions for being a dog. In fact, they are traits that characterize dogs in typical (or prototypical) cases. The concept DOG is not exceptional from this point of view. The majority of everyday concepts behave in this way. For most concepts, a classical definition in terms of necessary and sufficient conditions is not available (or, even if it is available, it is unknown to the agent). On the other hand, it may be that we know the classical definition of a concept,

170

Chapter Eleven

but typical/prototypical knowledge still plays a central role in many cognitive tasks. Consider the following example: nowadays, most people know the necessary and sufficient conditions for being WATER: water is the chemical substance whose formula is H2O, i.e., the substance whose molecules are formed by one atom of oxygen and two atoms of hydrogen. However, in most cases of everyday life, when we categorize a sample of something as WATER, we do not take advantage of this piece of knowledge. We use such typical traits such as the fact that (liquid) water is usually a colorless, odorless and tasteless fluid. The use of typical knowledge in cognitive tasks such as categorization has to do with the constraints that concern every finite agent that has limited access to the knowledge relevant for a given task. In most cases, cognitive processes based on typical knowledge are fast and automatic, cognitively undemanding, and are presumably homogenous to the processes employed in similar tasks by non-human animals. Consider, for example, the following variant of the Linda problem. Let us suppose that a certain individual, Pippo, is described as follows. He weighs about 200kg, and he is approximately two meters tall. His body is covered with a thick, dark fur, he has a large mouth with robust teeth and paws with long claws. He roars and growls. Now, given this information, we have to evaluate the plausibility of the two following alternatives: a) Pippo is a mammal b) Pippo is a mammal, and he is wild and dangerous. Which is the “correct” answer? According to the dictates of the normative theory of probability, it is surely a. But if you encountered Pippo in the wilderness, it would probably be best to run. So, many aspects of the psychology of concepts have presumably to do with the fast, type 1 system and processes, while others can be more plausibly ascribed to type 2. In particular, the ability to make explicit, high level inferences, and, moreover, the ability to justify them, which some philosophers consider to be constitutive of concept possession (see, for instance, Peacocke 1992; Brandom 1994), can be plausibly considered type 2 processes. Problems arise from the fact that the dual process approach is not monolithic. Different dual process theories exist – a detailed review is provided in Evans (2008). Moreover, the dual process approach has also received a number of criticisms (many of which have recently been reviewed and answered by Evans and Stanovich 2013a, b; for a skeptical attitude towards the possibility of applying dual process theories to the

Compositionality and Typicality Effects

171

study of concepts see section 8 of Machery 2011). In any case, it is likely that many kinds of type 1 systems and processes exist, with partially different properties. Consider, for example, expertise: an art historian can distinguish a painting by, say, Rubens from one by van Dyck at a glance, without the need for any form of conscious, sequential application of explicit rules. But surely this ability is not innate and depends on specific training (and, presumably, it is not shared with other animal species). A recurring distinction in dual process literature concerns the interaction of the two types of system. A first possibility (Sloman 1996) is that they operate in parallel. In this case, adopting the terminology proposed by Evans (Evans 2008), the two systems are assumed to be parallelcompetitive. A second possibility (Kahneman and Frederick 2002, Kahnemann 2011, Evans and Stanovich 2013a) is the so called defaultinterventionist approach. According to this view, deliberative type 2 reasoning processes can inhibit the possibly biased default responses of type 1 systems, and replace them with the outputs of reflective reasoning. This second perspective better fits our approach. In any case, our concern here is not to take a stand for any one specific version of the theory. In particular, we do not make any claim about how many systems involved in the two types of process effectively exist. Rather, our claim is that dual process theories can supply some useful suggestions for understanding and classifying the wide range of cognitive phenomena that pertain to human conceptual abilities. Compositionality, which is often considered to be an irrevocable trait of conceptual systems, could turn out to be more akin to system 2 abilities. Compositionality has to do with higher cognition and with complex inferential tasks: paradigmatic examples of compositional systems are natural languages and logical formalisms. And compositionality is somewhat at odds with typicality effects: the characterization of concepts in typical terms is difficult to reconcile with the requirement of compositionality (consider again the pet fish example from section 2). A possible solution could be to hypothesize that compositional representations and representations in (proto)typical terms depend on different cognitive components, based on different types of representation. In this respect, it may be interesting to again take into account what happens in the field of computational simulations of cognition (Frixione and Lieto 2012). Consider classification, one of the most characteristic forms of inference of description logics (section 3, above): classifying a concept amounts to individuating its more specific superconcepts and its more general subconcepts, or, in other words, identifying the implicit superconcept-subconcept relations in a taxonomy. For human subjects

172

Chapter Eleven

such a process is far from natural, fast and automatic; it is usually slow, it can require great effort and it is facilitated by specific training. So, in terms of dual process theories, the inferential task of classifying concepts in taxonomies is prima facie a type 2 process, qualitatively different from the task of categorizing items as instances of a certain class on the basis of typical traits (e.g. the task of categorizing Fido as a dog because he barks, has fur and wags his tail). Therefore, it is plausible that conceptual representation in computational systems should be assigned to (at least) two different kinds of components, responsible for different tasks: type 2 processes, involved in complex and cognitively demanding inference tasks, and fast and automatic type 1 processes (such as those involved in categorization based on prototypical information). In the next section we shall explore some aspects of this hypothesis. Note that, from this perspective, the approach to prototypical representation of concepts based on non-monotonic extensions of some DL formalisms (see again sect 3, above) seems to be particularly implausible. The idea that is the basis of such an approach is that prototypical representation of concepts should be obtained by augmenting DLs with non-monotonic constructs, which should allow the representation of defeasible information. In such a way, categorization based on prototypical traits results in a process homogeneous to classification, but still more demanding, and carried out with a still more complex formalism (it is well known that, in general, non-monotonic formalisms have the worst computational properties if compared to their monotonic counterparts).4

5. Compositionality, Typicality and Conceptual Spaces In order to clarify the relationship between compositionality and typicality effects, let us consider the argument that was proposed by Osherson and Smith (1981), which is a version of the pet fish argument mentioned above. Osherson and Smith’s original aim was to show that fuzzy logic is inadequate for capturing typicality. However, as we shall see, their argument has a wider range of application. At first sight, fuzzy logic seems to be a promising approach to facing the problem of typicality. Indeed, one consequence of typicality effects is that some members of a category C turn out to be better (i.e. more typical) 4

This does not amount to a claim that, in general, non-monotonic extensions of DLs are useless. Our claim is simply that they seem to be unsuitable (and cognitively implausible) for the task of representing concepts in prototypical terms.

Compositionality and Typicality Effects

173

instances of C than others. For example, a robin is a better example of the category of birds than, say, a penguin or an ostrich. More typical instances of a category are those that share a greater number of typical features (e.g. the ability to fly for birds, having fur for mammals, and so on). The fuzzy value of a predicate (say, F) could be interpreted as a measure of typicality: given two individuals h and k, it seems natural to assume that F(h) > F(k) iff h is a more typical instance of F than k is. However, let us consider the zebra in figure 1 (and let us suppose that her name is Pina).

Fig. 1

Pina is presumably a good instance of the concept POLKA DOT ZEBRA;5 therefore, if such a concept were represented as a fuzzy predicate, then the value of the formula polka_dot_zebra(Pina) should be close to 1, say: (1) polka_dot_zebra(Pina) = .97 On the other hand, Pina is a rather poor (i.e. atypical) instance of the concept ZEBRA; therefore the value of the formula zebra(Pina) should be low, say: (2) zebra(Pina) = .2 (Of course, the specific values are not relevant here; the point is that Pina is more typical as a polka dot zebra than as a zebra.) But POLKA DOT ZEBRA can be expressed as the conjunction of the concepts ZEBRA and POLKA DOT THING; i.e. in logical terms, it holds that: 5

Osherson and Smith's original example was not a polka dot zebra but a striped apple.

174

Chapter Eleven

(3) x (polka_dot_zebra(x) ļ zebra(x) š polka_dot_thing(x)) Now, the problem is the following: if we adopt the simplest and most widespread form of fuzzy logic, the value of a conjunction is calculated as the minimum of the values of its conjuncts, and this makes it impossible for the value of zebra(Pina) to be .2 and that of polka_dot_zebra(Pina) to be .97 at the same time. Of course, there are other types of fuzzy logic, in which the value of a conjunction is not the minimum of the values of the conjuncts. But in no case can a conjunction exceed the value of its conjuncts. Worse still, in logic, once a suitable order has been imposed on truth values, it generally holds that: val(A š B) d val(A) and val(A š B) d val(B) So, the problem pointed out by Osherson and Smith does not seem to concern only fuzzy logic. Rather, Osherson and Smith’s argument seems to show that, in general, logic-based representations are unlikely to be compatible with typicality effects. And, as mentioned before, logic-based representations are paradigmatic examples of compositional systems, which fully embody the Fregean principle of compositionality of meaning. (Note also that this is exactly the same problem which, in the context of probabilistic reasoning, gives rise to the conjunction fallacy mentioned in section 4, above.) Indeed, the situation seems to be more promising if, instead of logic, we face typicality by adopting some other form of representation, such as, for example, conceptual spaces, a geometric representation proposed by Peter Gärdenfors (2000). A conceptual space (CS) is a space in a certain number of quality dimensions. Concepts are represented in terms of such dimensions, and correspond to regions in CSs. CS dimensions can be more or less directly related to perception (such as temperature, weight, brightness, pitch), or more abstract in nature. Each quality dimension is associated with a geometrical (topological or metrical) structure. The central idea behind this approach is that knowledge representation takes advantage of the geometrical structure of conceptual spaces. For example, similarity should be calculated in terms of some suitable distance measure. So, if we represent a concept as a (convex) area in a suitable conceptual space CS, then the degree of typicality of a certain individual can be measured as the distance of the corresponding point from the center of the area. The conjunction of two concepts is represented as the intersection of the two corresponding areas, as in figure 2.

Compositionality and Typicality Effects

175

POLKA DOT ZEBRA

Pina

POLKA DOT

Fig. 2

According to the conceptual space approach, Pina should presumably turn out to be very close to the center of POLKA DOT ZEBRA (i.e. to the intersection between ZEBRA and POLKA DOT THING). In other words, she should turn out to be a very typical polka dot zebra, despite being very eccentric with respect to both the concepts ZEBRA and POLKA DOT THING (that is to say, she is an atypical zebra and an atypical polka dot object). This seems to better capture our intuitions about typicality. We should conclude that the treatment of compositionality and that of (some forms of) typicality require rather different approaches and forms of representation, and should therefore presumably be assigned to different components of cognitive architecture. The above considerations can be reconciled with both a hybrid approach to concepts (Miller and Johnson-Laird 1976, Osherson and Smith 1981) and with the so-called heterogeneity hypothesis (Machery and Seppälä 2010, Machery 2014). Here we do not take a stand on this point. Our concern is simply to stress that some aspects of typicality effects have features that are closer to those usually associated with type 1 systems and processes (they are fast, automatic, and so on), while compositional representations seem to better fit type 2 tasks. This favors the hypothesis of adopting different formalisms for representing different aspects of conceptual knowledge.

6. Some Conclusions In conclusion, different aspects of “conceptual competence” seem to involve different types of cognitive processes (e.g. type 1 vs. type 2 processes), and to require different kinds of representation (e.g. compositional, “linguistic” representations vs. other types of representation, such as conceptual spaces). We do not maintain that typicality pertains exclusively to fast, lowlevel type 1 processes. Certain forms of categorization, which are certainly more akin to type 2 processes, do not rest on classical definitions given in terms of necessary and sufficient conditions. Such processes are slow, sequential, penetrable to consciousness, and they depend on high-level,

176

Chapter Eleven

acquired pieces of knowledge. But neither do they resemble the application of classical definitions; they rather depend on forms of typicality-based reasoning, on abductions and on “theory-based” inferences. As stated before, Machery (2005, 2009) claims that concept is not a natural kind; he rather hypothesizes three different kinds, which correspond respectively to prototypes, exemplars and theories. Piccinini (2011) argues for an alternative point of view, according to which only two kinds of concept exist: implicit and explicit. He correlates implicit and explicit concepts to system 1 and system 2 processes respectively. In our opinion, it is likely that, in some respects, both Machery and Piccinini are right, in that they both individuate important discontinuities in the traditional notion of concept. However, it is also likely that the distinction between system 1 and system 2 processes is in part orthogonal to Machery's tripartition, and there are good reasons to suspect that they are not mutually exclusive (Frixione and Lieto 2012, Frixione 2013). In conclusion, the situation is rather complex. It is plausible that, in light of the distinction between type 1 and type 2 systems, certain aspects of conceptual knowledge pertain to type 1 systems, and that “type 1 concepts” are likely to be closer to perceptual processes. But, without a clear specification of what concepts are, the problem of the relationships between concepts and perception remains somewhat ill-posed; a satisfactory solution would presuppose a better understanding of the notion of concept itself. Recently, an implementation of the presented dual process conceptual proposal has been realized (Ghignone, Lieto and Radicioni 2013; Lieto, Minieri, Piana and Radicioni in press) and preliminarily tested in a knowledge-based system involved in a question-answering task. In this system, imprecise and common sense natural language descriptions of a given concept were provided as queries. The task designed for the evaluation consisted of individuating the appropriate concept that fits a given description by exploiting the inferential capability of the proposed hybrid conceptual architecture. According to the assumption presented by Frixione and Lieto (2013), the S1 component is based on the conceptual spaces framework (Gärdenfors 2000) and the classical S2 component on standard Description Logics and ontology-based formalisms. An example of such common sense descriptions is: “the big carnivore with black and yellow stripes”, denoting the concept of tiger. The obtained preliminary results are encouraging and show that the identification and retrieval of concepts described in terms of typical features is considerably improved by such hybrid architecture compared to the classical case, based simply

Compositionality and Typicality Effects

177

on the use of ontological knowledge. Furthermore, this result is obtained with relatively limited computational effort if compared with the other, logic-based, approaches. These results suggest that a dual process approach to conceptual representation of concepts can also be beneficial for enhancing the performance of artificial systems in tasks involving nonclassical conceptual reasoning. .

References Baader, F., D. Calvanese, D. McGuinness, D. Nardi, and P. PatelSchneider (2010). The Description Logic Handbook: Theory, Implementations and Applications, 2nd edition. Cambridge, UK: Cambridge University Press. Brachman, R. and H. Levesque (1985). Readings in Knowledge Representation. Los Altos, CA: Morgan Kaufmann. Brachman, R. and J.G. Schmolze (1985). An overview of the KL-ONE knowledge representation system, Cognitive Science, 9, 171-216. Brandom, R. (1994). Making it Explicit. Cambridge, MA: Harvard University Press. Evans, J.St.B.T. (2008). Dual-processing accounts of reasoning, judgment, and social cognition. Annual Review of Psychology, 59, 255-278. Evans, J.St.B.T. & K.E. Frankish (eds.) (2008). In Two Minds: Dual Processes and Beyond. New York, NY: Oxford UP. Evans, J.St.B.T. & K.E. Stanovich (2013a). Dual-Process Theories of Higher Cognition: Advancing the Debate. Perspectives on Psychological Science, 8, 223-241. —. (2013b). Theory and Metatheory in the Study of Dual Processing: Reply to Comments, in Perspectives on Psychological Science, 8, 263271. Fodor, J. (1981). The present status of the innateness controversy. In J. Fodor, Representations. Cambridge, MA: MIT Press. Frixione, M. (2013). Concepts and Fat Plants: Non-Classical Categories, Typicality Effects, Ecological Constraints. Concepts – Contemporary and Historical Perspectives, ProtoSociology, 30, 152-166. Frixione, M. and A. Lieto (2011). Representing Concepts in Artificial Systems: A Clash of Requirements. Proceedings of the 4th International HCP11 Workshop, 75-82. —. (2012). Representing Concepts in Formal Ontologies. Compositionality vs. Typicality Effects. Logic and Logical Philosophy, 21(4), 391–414

178

Chapter Eleven

—. (2013). Dealing with Concepts: from Cognitive Psychology to Knowledge Representation. Frontiers in Psychological and Behavioural Science, Vol. 2, Iss. 3, Jul. 2013, 96-106. —. (2014). Towards an Extended Model of Conceptual Representations in Formal Ontologies: A Typicality-Based Proposal. Journal of Universal Computer Science, 20(3), 257-276. Gärdenfors, P. (2000). Conceptual Spaces: The Geometry of Thought. MIT Press, Bradford Books. Ghignone, L., A. Lieto and D. Radicioni (2013). Typicality-based Inference by Plugging Conceptual Spaces into Ontologies. In Proceedings of AIC 2013, International Workshop on Artificial Intelligence and Cognition, 68-79. Kahneman, D. (2011). Thinking, fast and slow. New York, NY: Farrar, Straus and Giroux. Kahneman, D. & S. Frederick (2002). Representativeness revisited: Attribute substitution in intuitive judgment. In T. Gilovich, D. Griffin & D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment, 49–81. Cambridge, MA: Cambridge University Press. Lieto, A., A. Minieri, A. Piana and D.P. Radicioni (in press). A Knowledge-Based System for Prototypical Reasoning. To appear in Connection Science, DOI: 10.1080/09540091.2014.956292 Machery, E. (2005). Concepts are not a natural kind. Philosophy of Science, 72, 444–467. —. (2009). Doing without Concepts. Oxford, UK: Oxford University Press. —. (2011). Replies to Lombrozo, Piccinini, Poirier and Beaulac. Dialogue, 50(1), 195-212. —. (2014). Concepts: Investigating the Heterogeneity Hypothesis. In Justin Sytsma (ed.), Advances in Experimental Philosophy of Mind, London, UK: Bloomsbury Publishing. Machery, E., and S. Seppälä (2010). Against hybrid theories of concepts. Anthropology & Philosophy, 10, 97–125. Miller, G.A. and P.N. Johnson-Laird (1976). Language and perception. Cambridge, MA: Harvard University Press. Minsky, M. (1975). A framework for representing knowledge. The Psychology of Computer Vision. P. H. Winston (ed.), McGraw-Hill. Murphy, G.L. (2002). The Big Book of Concepts. Cambridge, MA: The MIT Press. Osherson, D.N. and E.E. Smith (1981). On the adequacy of prototype theory as a theory of concepts. Cognition, 9(1), 35-58.

Compositionality and Typicality Effects

179

Peacocke, C. (1992). A Study of Concepts. Cambridge, MA: The MIT Press . Piccinini, G. (2011). Two Kinds of Concept: Implicit and Explicit, Dialogue, 50(1), 179-193. Rosh, E. (1975). Cognitive representation of semantic categories. Journal of Experimental Psychology 104: 573-605. Sloman, S.A. (1996). The empirical case for two systems of reasoning. Psychological Bulletin, 119, 3-22. Stanovich, K. and R. West (2000). Individual differences in reasoning: Implications for the rationality debate? The Behavioural and Brain Sciences 23, 5: 645- 65. Tversky, A. and D. Kahneman (1983). Extension versus intuitive reasoning: The conjunction fallacy in probability judgment. Psychological Review, 90 (4): 293–315. Wittgenstein, L. (1953). Philosophische Untersuchungen. Oxford: Blackwell.

CHAPTER TWELVE AN EMBODIED AND GROUNDED PERSPECTIVE ON CONCEPTS ANNA M. BORGHI

1. Concepts According to Embodied and Grounded Views In the last few years, embodied and grounded (hereafter EG) theories have become more and more popular in a variety of disciplines, ranging from philosophy to linguistics, from psychology to cognitive science and neuroscience, computer science and robotics (for a recent overview, see Borghi and Caruana, 2015). There has been a marked increase in publications on embodied cognition, as noticed by some scholars (e.g. Gentner, 2010), and some recent special issues testify to how lively the field is (e.g. Borghi and Pecher, 2011) and start to ask where research on embodiment is going (e.g. Davis and Markman, 2012; Dove, in press). Even if they can be quite different - for example, some of these theories assume the notion of representation, others do not think it is necessary these views share the idea that the cognitive processes cannot be studied without referring to the brain and the body as a control system. Furthermore, they propose that perception, action and cognition are deeply interconnected. This is in contrast with classical cognitive science views, which posited a linear relationship between perception, cognition and action, implicitly assuming that cognition can be seen as a sandwich, with perception and action considered as the external and less tasty slices, compared to the tasty and crucial part: cognition (Hurley, 1998). From the idea of a strict link between perception, action and cognition derives the way in which EG views look at concepts, i.e. the minimal units of our knowledge on a given object or entity. Concepts consist of the reactivation of the neural pattern activated during previous experience with the social and physical environment, aimed at facilitating the

182

Chapter Twelve

interaction with objects and entities which are present in the current situation/context (Barsalou, 1999; Borghi, 2005; Gallese and Lakoff, 2005; Prinz, 2002). For example, the concept of “telephone” would imply the multimodal activation (Gallese and Lakoff, 2005) of the acoustic, visual, tactile etc. properties and corresponding brain areas we typically experience and activate while seeing and using telephones. Specifically, Barsalou (1999), one of the most influent proponents of EG views, argued: "In this theory, a concept is equivalent to a simulator. It is the knowledge and accompanying processes that allow an individual to represent some kind of entity or event adequately. A given simulator can produce limitless simulations…Whereas a concept represents a kind generally, a conceptualization provides one specific way of thinking about it"(Barsalou, 1999, p. 587). The chapter will focus on concepts as simulators. We will discuss the concepts of objects and of actions; then we will turn to how the concepts of objects and actions are mediated by language; finally, we will briefly discuss so-called "abstract concepts".

2. Simulation The notion of simulation is complicated and controversial. As there are many embodied theories, so there are many views of simulation (for discussion, see Borghi and Cimatti, 2010; Gallese, 2009; Jeannerod, 2006; Decety and Grezes, 2006). Here we will regard simulation as composed of two aspects: reenactment and prediction. Simulating means reactivating offline the same neural network used while perceiving and interacting with objects and entities, and while experiencing emotions. Simulating has a predictive function, as it prepares us to interact with the objects and entities which are present in the current context/situation. We simulate when we perceive objects; for example, seeing an apple activates a grasping action, and it also pre-activates our gustatory and olfactory systems. We simulate when we perceive others performing an action and this action is part of our motor repertoire; for example, if we observe and hear a dog barking, we cannot simulate, since we cannot bark (Buccino et al., 2004). Since grasping objects is one of our motor abilities, we simulate when we observe monkeys grasping an object. Finally, we simulate when we comprehend language; for example, when hearing the sentence "he kicked the ball" we internally reproduce the situation described by the sentence, activating our legs. It is an open issue whether we also simulate when we understand abstract words, such as "truth". The neural basis of this simulation is likely to be found in the activation of the canonical and

An Embodied and Grounded Perspective on Concepts

183

mirror neuron systems, i.e. the systems activated while observing objects and people interacting with them (for reviews, see Rizzolatti and Craighero, 2004; Gallese, 2008). How does this simulation work during language comprehension? As an example, the Action Based Language model (Glenberg and Gallese, 2012) emphasizes the predictive aspects of the simulations we form during language comprehension. For example, upon hearing the verb “kick”, the mirror neuron system will be activated and possible outcomes of the action being performed will be generated. In the following part of the paper we will describe experiments performed in our lab which help exemplify how simulation occurs.

3. Simulating during Object Observation: Affordances Affordances are an important notion for EG theories, as they provide a bridge between perception and action. With affordance, Gibson (1979) originally referred to the fact that the environment offers us invitations to act; for example, a pen invites us to hold it. The notion of affordance has had a lot of success within psychology and cognitive neuroscience, at least as an inspiration source. In particular, Ellis and Tucker (2000) have proposed using the notion of microaffordance, to indicate the similarities but also the differences between their own and Gibson's view. Microaffordances are rather specific and pertain to components of actions, for example grasping and reaching affordances. Furthermore, activating the microaffordances of a given object, e.g. a cup, implies having recognized it. Finally, the authors proposing the notion of microaffordance are also interested in how they are represented in the brain, as the product of visuomotor associations developed through experience. This interest in the neural correlations of affordances (for a recent review see Maranesi et al., 2014) was not present in Gibson's view. Until some years ago, the majority of studies focused on how affordances were activated, independently from the task. For example, in some elegant experiments, Tucker and Ellis (1998; 2001) asked participants to categorize objects into natural objects and artifacts, mimicking either a precision or a power grip, and found a compatibility effect between the object size and the grip to execute to respond: the categories of larger objects were determined faster and more accurately when executing a power grip, while the opposite was true for the smaller objects. The results of this study indicate that, even if the task - a categorization one - did not require access to object size, microaffordances related to grasping were automatically activated. Very recent studies on affordances tend to highlight, rather than their automaticity, their

184

Chapter Twelve

flexibility and contextual dependency (e.g. van Elk et al., 2014; for an overview of recent work in our lab see Borghi, 2014). For example, Borghi et al. (2012) presented participants with objects and with a hand. The objects could be either linked by a spatial relationship (objects typically located in the same context, e.g., knife-butter) or by a functional relationship (objects typically used together, e.g. knife-coffee-mug), or by no relationship. The hand could be displayed grasping the object in order to move/manipulate it, in order to use it, or it could simply be near the object. In one condition no hand was present. Participants were required to press two different keys on the keyboard to determine whether the two objects were related or not. When the objects were functionally related, manipulation postures were the slowest, while when the objects were spatially related, functional postures were inhibited. This interaction was not present when participants were required to perform foot responses by pressing a pedal. This suggests that observing hands in potential interaction with objects activates affordances, which differ depending on the context. Furthermore, the results indicate that observing a hand in potential interaction with an object elicits a motor simulation which is specific for a given effector (here the hand, not the foot). A further study (Scorolli et al., 2014) investigates the role of social context in affordance activation. We again had pairs of objects linked by a spatial relation (e.g., can-knife), by different kinds of functional relations or by no relation. The functional relations could either imply an individual action (functional–individual: e.g., can-straw) or an action to perform with somebody else (functional–cooperative: e.g., can-glass). Participants were asked to refrain from responding when the two objects were not related; when they were related they had either to move toward the experimenter, as if to give something to him (giving condition) or to move away from him, as if taking something for themselves (getting condition). We also manipulated the presence and lack of presence of eye contact communication. Results revealed that participants were sensitive to the information given by the hand posture and by eye contact: responses were faster when the other was engaged in individual rather than in cooperative actions, and when the other used a manipulative hand posture more open to cooperation, rather than a functional hand posture, used for individual actions. This study clearly shows how affordances are deeply modulated by the social and physical context, and indicates that we possess a precise and sophisticated ability to predict others' actions, as well as their social vs. individual character. We have briefly reported these studies to show how initial work on affordances was focused on single objects, while it is now becoming more

An Embodied and Grounded Perspective on Concepts

185

important to show the effects of the physical context (see also Kalenine et al., 2014) or the social context (see also Ellis et al., 2013) on their activation.

4. Simulating during Action Observations Motor Resonance In recent years, a number of studies have revealed the existence of motor resonance processes (e.g. Aglioti et al., 2008). Our mirror neuron system responds more when we observe others performing actions that are part of our motor repertoire, compared to actions we are not able to perform. In the last few years, the literature has gone even further, revealing the existence of motor resonant processes when we observe others similar to us - for example sharing our culture, or our ideology. In our lab we have performed some studies showing motor resonance processes when we observe actions of people whose bodily characteristics are similar to our own. Liuzza et al. (2010) have demonstrated this with a priming paradigm. Children were shown pictures of adults’ hands and children’s hands in a grasping posture or in a control posture (fist), followed by images of light vs. heavy objects (e.g. a block of notes vs. a dictionary). They had to decide whether the target objects they saw were light or heavy and press a different key on a keyboard accordingly. We found that response times were faster when children observed children’s hands in a grasping posture than when they observed adults’ hands or children’s hands in a control posture. These results suggest that children simulate the action of lifting the object, thus their response times differ depending on the object's weight. More crucially, they resonate more when they observe hands similar to their own, as adults possess a completely different body schema. Ranzini et al. (2011) provided evidence of motor resonance processes with a line bisection paradigm, in which participants were required to bisect a line with two flankers. In one experiment the flankers consisted of images of a hand displaying a precision vs. a power grip; they flanked a line which could be thin or thick. We found that people tended to bisect a line more toward the precision grip flanker when the line was thin, more toward the power grip flanker when the line was thick. This suggests that we are sensitive to the affordances provided by the line in combination with the action evoked by the hand posture. In further experiments, we also found an effect of motor resonance: participants tended to bisect the line more toward the precision grip, as if preparing themselves for a skilled action, such as the kind of action which is typically associated to a precision posture. Finally, the

186

Chapter Twelve

effect was modulated by motor resonance: it was stronger with human hands than with fake hands or with robotic hands. Further motor resonance processes were observed during observation of a hand grasping dangerous objects. In a recent study (Anelli et al., 2013), adults and children observed human and robotic hands of their own/a different gender followed by graspable neutral and dangerous objects (e.g., tomato vs. cactus). Neutral graspable objects produced a facilitation of the motor response, while dangerous objects inhibited the response, probably due to an aversive effect. Both children and adults responded faster to human than to robotic hands, revealing a motor resonant mechanism. Specific resonant mechanisms related to the gender of the hand emerged only in adults. Overall, these studies show stronger motor resonance processes when we observe actions performed by someone possessing a body similar to ours, in terms of body schema (children vs. adults), gender and body parts (hands).

5. Simulating during Language Comprehension What happens when concepts are expressed through words? We simulate the processes described by means of linguistic expressions (Barsalou, 2008; Borghi and Cangelosi, in press; Gallese, 2008; Glenberg and Gallese, 2013). Many studies in various labs, and in our lab as well, have demonstrated that this simulation is fine-grained: it is sensitive to object characteristics, such as shape, orientation, and weight (Scorolli et al., 2009), as well as to characteristics of actions, such as the effectors (hands, mouth or feet) used to perform them: for example, the arm/hand is activated while reading or hearing “throw a ball”, the foot while reading or hearing “kick a ball” (for an overview of the studies of our lab see Borghi, 2012). We recently demonstrated that the simulation differs when we read sentences referring to a different social context, such as "give the object to a friend” vs. “to an enemy” (Lugli et al., 2012). In a recent kinematics work (Gianelli et al., 2013), we investigated how the simulation of a social situation formed during language comprehension was modulated by a real social situation. Participants were required to evaluate whether sentences made sense or not by moving the mouse away from or toward their own bodies. The sentences referred to objects characterized by positive or negative valence to bring or to give to someone else (e.g., “The object is ugly/smooth. Bring it toward yourself/Give it to another person”). In one condition, participants performed the task individually; in another an experimenter observed the performance, and in a further condition the experimenter interacted with participants, relocating the mouse where they

An Embodied and Grounded Perspective on Concepts

187

had left it. The simulation formed during the comprehension of the social context illustrated in the sentences was enhanced by matching it with the real social context, given by the presence of the experimenter. Specifically, we found an increase in accuracy and a slowing down of RTs when the sentence stimuli referred to the "another person" target and the experimenter acted as a confederate. This is consistent with kinematics data showing that, in social situations, responses become slower to allow for an increase in accuracy (e.g. Ferri et al., 2010). Our results thus suggest that the simulation formed is sensitive both to the fictitious (linguistically expressed) and to the real social context.

6. Abstract Concepts So far we have discussed concepts of objects and of actions, as well as concepts mediated by words. The notion of simulation can be applied in a rather straightforward way to concrete concepts, but it is much more complicated to defend the view that abstract concepts, such as "fantasy" and "truth", activate simulations. We have recently proposed the Words As social Tools (WAT) view on abstract concepts, and have collected some evidence supporting it (Borghi and Cimatti, 2009; Borghi and Binkofski, 2014; see also Borghi, 2013). According to traditional views, abstract concepts are represented solely on the basis of linguistic information: while concrete concepts and words such as "book" activate visual information, this is not the case for abstract words (e.g. Paivio, 1986). According to standard embodied views, abstract concepts also activate sensorimotor experience, hence simulations: for example, the concept of justice would reactivate the experience of a tribunal (for a review, see Pecher et al., 2011). Our view is an EG view, hence it proposes that abstract concepts activate the perception, action and emotional systems. At the same time, however, we maintain that this is not the whole story, and that to fully account for abstract concept representation, EG views need to be extended. We propose, indeed, that not only sensorimotor, but also linguistic experience plays a major role in the acquisition and representation of abstract concepts. In our view, the modalities of acquisition of concrete and abstract concepts profoundly differ, and this influences their neural representation. To form concrete categories such as “dogs” or “chairs”, the influence of language is certainly of importance, but not crucial, since the category members share many perceptual similarities. To form abstract categories, however, hearing a common label that holds together sparse and varied experience, or hearing someone explaining to us the meaning of a given word, might be crucial. We

188

Chapter Twelve

therefore predict that for both concrete and abstract concepts, sensorimotor and linguistic experience count, but that their distribution differs, since for abstract concepts linguistic information would be more relevant. Importantly, we speak of linguistic “experience”. We do not intend to deny the role of distributional statistical information in word meaning (see distributional views, according to which meaning of a word is given by its co-occurrence with other words; for an influential example see Landauer and Dumais, 2007). However, neither do we limit the contribution of language to this; rather, by linguistic “experience” we refer to the whole social and emotional experience of being taught a word by somebody in a context, and of starting to use it to refer to a particular object, event, situation or event. We have collected some evidence in support of this view. In two acquisition studies with adults (Borghi et al., 2011; Granito et al 2015), we have demonstrated that learning a novel label and receiving a meaning explanation facilitates the acquisition of novel abstracts more than that of concrete words, and that it leads to an activation of the mouth. Novel abstract words lead to faster responses with the mouth (using a microphone), while novel concrete words evoke faster manual responses. The result is further supported by ratings we obtained on a subset of abstract words (from the database of Barca et al., 2002). Participants were asked to rate to what extent an effector (hand or mouth) is involved in a possible action with an item. We found that, while the hand was preferentially activated for concrete words, both the hand and the mouth were activated with abstract words. This suggests that activation of the mouth underscores abstract concept representation, be it due to the enhancement of past linguistic comprehension or linguistic production experiences (see Borghi and Binkofski, 2014, for a detailed discussion). Further work on the neural basis of abstract concept representation supports this view: in a TMS study (Scorolli et al., 2012) we found that, compared to phrases with a concrete verb (e.g., to grasp a flower/a concept), phrases with an abstract verb (e.g., to describe a flower/a concept) implied a late activation of the motor manual system, probably due to the fact that they activated the mouth first, and that this lead to a cascade activation of the hand. Furthermore, in an fMRI study (Sakreida et al., 2012) we found that sensorimotor networks in the brain are activated by both concrete and abstract phrases, while linguistic networks are activated mainly by abstract ones. Finally, in a recent study on Italian sign language (LIS), we found and illustrated examples of signs used to express abstract concepts (e.g. linguistics, truth) that exploit linguistic information, taken either from the same sign language or from a different language, be

An Embodied and Grounded Perspective on Concepts

189

it spoken or signed (Borghi, Capirci, Gianfreda and Volterra, 2014). Overall, the evidence described supports the WAT proposal on abstract concepts.

7. Conclusion Concepts reactivate past experiences to help us deal with current experiences. This paper presents an EG view of concepts and outlines recent evidence supporting it. The view that concepts of objects and of actions are grounded in sensorimotor systems and evoke simulations is supported by much evidence. Accounting for abstract concepts, however, represents a major challenge for the future of EG cognition, one not easy to deal with. We have recently proposed that, to explain abstract concept representation, not only sensorimotor experiences should be taken into account, but also linguistic experiences, considered in their social and emotional complexity. Further research is needed to address this fascinating challenge.

References Aglioti, S. M., Cesari, P., Romani, M., & Urgesi, C. (2008). Action anticipation and motor resonance in elite basketball players. Nature neuroscience, 11(9), 1109-1116. Anelli, F., Borghi, A.M., Nicoletti, R. (2012). Grasping the pain: Motor resonance with dangerous affordances. Consciousness & Cognition, 21, 1627-1639. Barca, L., Burani, C., & Arduino, L. S. (2002). Word naming times and psycholinguistic norms for Italian nouns. Behavior Research Methods, Instruments, & Computers, 34(3), 424-434. Barsalou, L.W. (1999). Perceptual symbol systems. Brain and Behavioural Sciences, 22: 577e660. —. (2008). Grounded cognition. Annual Review of Psychology, 59, 617645. Borghi, A.M. (2014). Affordances and contextual flexibility. Physics of Life Reviews. doi: 10.1016/j.plrev.2014.01.009. Commentary. —. (2013). Embodied cognition and word acquisition: The challenge of abstract words In: C. Müller, A. Cienki, E. Fricke, S.H. Ladewig, D. McNeill & J. Bressem (Eds.) Body-Language-Communication: An International Handbook on Multimodality in Human Interaction. Berlin, Boston: De Gruyter: Mouton.

190

Chapter Twelve

—. (2012). Action language comprehension, affordances and goals. In Yan Coello, Angela Bartolo (Eds). Language and action in cognitive neuroscience. Contemporary topics in cognitive neuroscience series (pp. 125-143). Psychology Press. Borghi, A.M. & Binkofski, F. (2014). Words As social Tools: An embodied view on abstract concepts. SpringerBriefs in Cognition series. Springer: New York. Borghi, A.M., Cangelosi, A. 2014). Action-language integration: from humans to cognitive robots. Topics in Cognitive Science, 6,3, 344-358 ( special issue edited by A.Cangelosi- A.M.Borghi, Action-language integration: from humans to cognitive robots, 6.3.344-558). Borghi, A.M., Capirci,O., Gianfreda, G., Volterra, V., (2014). The body and the fading away of abstract concepts and words: A sign language analysis. Frontiers in Psychology 5.811.doi:10.3389/fpsyg.2014.00811 Borghi, A.M., Caruana, F. ( 2015). Embodiment theories. In: James D.Wright (editor-in-chief), International Encyclopedia of the Social and Behavioral Sciences, 2nd. Edition, Vol.7.Oxford: Elsevier. 420426. Borghi, A.M., Cimatti, F. (2009). Words as tools and the problem of abstract words meanings. In N. Taatgen & H. van Rijn (eds). Proceedings of the 31st Annual Conference of the Cognitive Science Society (pp. 2304-2309). Amsterdam: Cognitive Science Society. —. (2010). Embodied cognition and beyond: Acting and sensing the body. Neuropsychologia, 48, 763-773. Borghi, A.M., Flumini, A., Cimatti, F., Marocco, D. & Scorolli, C. (2011).Manipulating objects and telling words: A study on concrete and abstract words acquisition. Frontiers in Psychology, 2:15. doi: 10.3389/fpsyg.2011.00015, Borghi, A.M., Pecher, D. (2011). Introduction to the special topic Embodied and Grounded Cognition. Frontiers in Psychology, 2, 187. doi: 10.3389/fpsyg.2011.00187. Buccino, G., Lui, F., Canessa, N., Patteri, I., Lagravinese, G., Benuzzi, F. & Rizzolatti, G. (2004). Neural circuits involved in the recognition of actions performed by nonconspecifics: An fMRI study. Journal of cognitive neuroscience, 16(1), 114-126. Davis, J. I., Markman, A.B. (2012). Embodied Cognition as a Practical Paradigm: Introduction to the Topic, The Future of Embodied Cognition. Topics in Cognitive Science, 4: 685–691. Decety, J., Grèzes, J. (2006). The power of simulation: Imagining one’s own and other’s behavior. Brain Research, 1079, 4-14.

An Embodied and Grounded Perspective on Concepts

191

Dove, G. ( 2015). How to go beyond the body: An introduction. Frontiers in Psychology, 6, 660. Ellis, R., Swabey, D., Bridgeman, J., May, B., Tucker, M., & Hyne, A. (2013). Bodies and other visual objects: The dialectics of reaching toward objects. Psychological Research, 77(1), 31–39. Ellis, R., Tucker, M., 2000. Micro-affordance: The potentiation of components of action by seen objects. British Journal of Psychology. 91, 451-471. Ferri, F., Campione, G. C., Dalla Volta, R., Gianelli, C., & Gentilucci, M. (2010). To me or to you? When the self is advantaged. Experimental brain research, 203(4), 637-646. Gallese, V. (2008). Mirror neurons and the social nature of language: The neural exploitation hypothesis. Social Neuroscience, 3, 317-333. —. (2009). Motor abstraction: A neuroscientific account of how action goals and intentions are mapped and understood. Psychological Research, 73(4):486-98. Gallese, V., Lakoff, G. (2005). The brain’s concepts: the role of the sensory-motor system in conceptual knowledge. Cogn. Neuropsychol. 22, 455–479. Gentner, D. (2010). Psychology in Cognitive Science: 1978–2038. Topics in Cognitive Science 2, 328-344. Gianelli, C., Lugli, L., Baroni, G., Nicoletti, R., Borghi, A.M. (2013). The impact of social context and language comprehension on behaviour: A kinematic investigation. PLoS ONE, Volume 8(12): e85151. doi: 10.1371/journal.pone.0085151. Gibson, J.J. (1979). The ecological approach to visual perception. Boston, MA: Houghton Mifflin. Glenberg, A.M., Gallese, V. (2012). Action-based language: A theory of language acquisition, comprehension, and production. Cortex, 48, 905922. Granito,C., Scorolli, C., Borghi, A.M. (2015). Naming a Lego world. The role of language in the acquisition of abstract concepts. PLos ONE 10(1): e0114615.doi: 10.1371/journal.pone.0114615. Hurley S., (1998). Consciousness in Action. Cambridge: Harvard University Press. Jeannerod, M. (2006). Motor Cognition: What actions tell the self. Oxford: Oxford University Press. Kalenine, S.,Shapiro, A., Flumini, A., Borghi, A.M., Buxbaum, L. (2014). Visual context modulates potentiation of grasp types during semantic object categorization. Psychonomic Bulletin & Review, 21(3):64551.doi:10.3758/s13423-013-0536-7.

192

Chapter Twelve

Landauer, T. K., & Dumais, S.T. (1997). A solution to Plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge. Psychological review, 104(2), 211. Liuzza, M.T., Setti, A., Borghi, A.M., 2012. Kids observing other kids’ hands: Visuomotor priming in children. Consciousness. & Cognitio, 21, 383-392. Lugli, L., Baroni, G., Gianelli, C., Borghi, A.M., Nicoletti, R. (2012). Self, others, objects: How this triadic interaction modulates our behavior. Memory and Cognition, 40, 1373-1386 Maranesi, M., Bonini, L., Fogassi, L. (2014). Cortical processing of object affordances for self and others’ action. Frontiers in Psychology,5. Paivio, A. (1986). Mental Representations: A Dual-Coding Approach, Oxford University Press, New York. Pecher, D., Boot, I., Van Dantzig, S., 2011. Abstract Concepts: Sensorymotor Grounding, Metaphors, and Beyond. In B.H. Ross (Ed.), The Psychology of Learning and Motivation. Academic Press: Burlington. Prinz, J.J. (2002). Furnishing the mind. Concepts and Their Perceptual Basis. MIT Press. Ranzini, M., Lugli, L., Anelli, M., Carbone, R., Nicoletti, R., Borghi, A.M. (2011). Graspable objects shape number processing. Frontiers in Human Neuroscience, 5, art.147, doi: 10.3389/fnhum.2011.00147. Rizzolatti, G., Craighero, L. (2004). The mirror neuron system. Annual Review of Neuroscience, 27, 169-192. Sakreida, K., Scorolli, C., Menz, M.M., Heim, S., Borghi, A.M., & Binkofski, F. (2013) Are abstract action words embodied? An fMRI investigation at the interface between language and motor cognition. Frontiers in Human Neurosciences 7:125. doi: 10.3389/fnhum.2013. 00125. Scorolli, C., Borghi, A.M., Glenberg, A.M. (2009). Language-induced motor activity in bimanual object lifting. Experimental Brain Research, 193, 43-53. Scorolli, C., Jacquet, P., Binkofski, F., Nicoletti, R., Tessari, A., Borghi, A.M. (2012). Abstract and concrete phrases processing differently modulates cortico-spinal excitability. Brain Research, 1488, 60-71. doi: 10.1016/j.brainres.2012.10.004. Scorolli, C., Miatton, M., Wheaton, L., Borghi, A.M. (2014). I give you a cup, I get a cup: A kinematics study on social intention. Neuropsychologia, 57, 196-204. doi: 10.1016/j.neuropsychologia. 2014.03.006.

An Embodied and Grounded Perspective on Concepts

193

Tucker, M., Ellis, R.(1998). On the relations between seen objects and components of potential actions. Journal of Experimental Psychology: Human Perception and Performance, 24(3), 830-846. Tucker, M., Ellis, R.(2001). The potentiation of grasp types during visual object categorization. Visual Cognition, 8(6), 769-800. van Elk, M., van Schie, H., & Bekkering, H. (2014). Action semantics: a unifying conceptual framework for the selective use of multimodal and modality- specific object knowledge. Physics of life reviews, 11(2), 220-250.

CHAPTER THIRTEEN ON THE SOCIAL DIMENSION OF BELIEFS AND CONCEPTS MARIA CRISTINA AMORETTI

1. Triangulation as an Argument for the Sociality of Thought Donald Davidson has strongly argued for the sociality of thought. In his view, a creature cannot have any concept or belief whatsoever unless he or she has ‘triangulated’ with other creatures sufficiently similar to itself on the background of a shared external world. Among other things, the intrinsic social dimension of concepts and beliefs plays a key role in the epistemology of our ordinary knowledge of the external world, the minds of others, and ourselves, as it guarantees that, despite idiosyncratic perceptions and causal histories, our concepts and beliefs are still ‘aligned’ and that we cannot be largely mistaken about our image of the world, other minds, and our own mind (Amoretti 2008a, 2008b, 2011). In this paper, however, I will not discuss the epistemological consequences of the sociality of thought, but instead the very argument in its support. To begin, Davidson agrees with the general idea that concepts are the constituents of thoughts, and conceives thoughts as the contents of propositional attitudes. In his view, having concepts and beliefs is not a matter of mere discrimination but rather a matter of classification: ‘Being able to discriminate cats is not the same thing as having the concept of a cat. You have the concept of a cat only if you can make sense of the idea of misapplying the concept, of believing or judging that something is a cat which is not a cat’ (Davidson 1997: 123). In discussing concepts and beliefs, Davidson refers to full-blown propositional thought, and believes that there is no middle ground between a total absence of thought and fullblown propositional thought. Among ordinary concepts and beliefs, those about worldly objects and events, which can be (and typically are)

196

Chapter Thirteen

acquired directly through the process of triangulation, play an exceptionally important role, as they ‘anchor’ our whole thought to the outside world and guarantee a ‘veridical’ common ground, which can prevent both conceptual relativism and philosophical scepticism (Davidson 1983, 1990a). Briefly, Davidson describes the process of triangulation as ‘The basic situation […] that involves two or more creatures simultaneously in interaction with each other and with the world they share; […] each is interacting simultaneously with the world and with the other agent. […] Each creature learns to correlate the reactions of other creatures with changes or objects in the world to which it also reacts’ (Davidson 1997: 128). This process is supposed to stand for the ‘simplest’ kind of social triadic interaction (Davidson 1990a, 1990b, 1991, 1992, 1997). In this regard, it is worth noting that the process of triangulation has the potential to support the social dimension of concepts and beliefs only as long as such a process is fully social and intersubjective; that is, only as long as the creatures involved recognise each other as intentional subjects mutually interacting on the background of a shared world. The process of triangulation ‘is essential to the existence, and hence to the emergence, of thought. For without the triangle, there are two aspects of thought for which we cannot account: […] [a] the objectivity of thought and [b] the empirical content of thoughts about the external world’ (Davidson 1997: 129). Here, the two conditions [a] and [b] are taken to be necessary for thought, as without them there can be no thought at all. Regarding the former condition, Davidson argues that, as having concepts and beliefs is a matter of classification, ‘one cannot believe something, or doubt it, without knowing that what one believes or doubts may be either true or false and that one may be wrong’ (Davidson 1997: 129). This means that in order to have concepts and beliefs, a creature must be aware of the objectivity of thought, which in turn requires an understanding of the possibility of error. But this understanding needs a triangular situation to emerge: a creature must first learn to associate certain reactions of other creatures to some relevant external stimuli and then be able to anticipate the same reactions when the same external stimuli occur; once some basic associations are established, a particular expectation may occasionally fail, thus creating the space to understand the possibility of error and the objectivity of thought. Moving to the latter condition, concepts and beliefs must have content, and what determines content is, at least in part, their causal relationship to the outside world (Davidson 1988: 51). Specifically, Davidson argues that the content of concepts and beliefs is (partially) determined by their ‘typical’ cause; that is, the worldly object that, in basic

On the Social Dimension of Beliefs and Concepts

197

situations, triggers relevantly similar responses by the subject. The problem is establishing the typical cause, circumscribing ‘what it is in the world to which we are responding’ (Davidson 1997: 129). The typical cause is in fact ambiguous with respect to distance (it may be located at the proximal level of our sensory receptors, at the distal level of the external world, or somewhere in between) and to width (it is not clear how much of the outside world is relevant). Again, such ambiguities can only be solved through a triangular process: due to the presence of two creatures, the typical cause is forced to be at the distal level of the external world, which is common to both creatures; triangulation can narrow down the right cause by locating it at the intersection of the two lines connecting, respectively, the first creature to the world, and the second creature to the world. Davidson considers the process of triangulation necessary but not sufficient to account for the two above aspects of thought (the objectivity of thought and the empirical content of thoughts about the external world), as language is necessary as well. Unless the lines of the triangle connecting the two creatures implements language, they can neither communicate the contents of their common experiences nor share their judgements about the world. Thus, they cannot grasp the objectivity of thought; moreover, there is no way to wholly solve the underdetermination of the typical cause. Now, it is possible to summarise Davidson’s argument for the sociality of thought as follows: [P1] [a] Grasping the objectivity of thought and [b] fixing the empirical content (of beliefs about the external world) is necessary to have beliefs and concepts; [P2] The process of triangulation is necessary to obtain both [a] and [b]; [C1] The process of triangulation is necessary to have beliefs and concepts.

As [C1] alone does not support the social dimension of thought, it should be made explicit that the process of triangulation is a fully social and intersubjective interaction. [P3] The process of triangulation is a fully social and intersubjective interaction; [C2] A fully social and intersubjective interaction is necessary to have beliefs and concepts.

Though the argument seems valid, it is still flawed in many respects (Amoretti 2012, 2013).

198

Chapter Thirteen

2. Some Problems with the Argument Starting with [P1], Davidson talks about two aspects of thought that cannot be accounted for unless the process of triangulation takes place; these two aspects are pivotal, as without them there is no thought. To begin, the idea of grasping the objectivity of thought can be characterised in different ways. In general this expression means having the concept of objectivity, which in turn also has various senses. In most cases, the concept of objectivity is equated or related to that of truth: ‘To have the concept of truth is to have the concept of objectivity, the notion of a proposition being true or false, independent of one’s beliefs or interests’ (Davidson 1995: 10). Here Davidson refers to a concept of ‘strong’ objectivity that involves other concepts such as belief, proposition, error, truth, and falsity, and demands the ability to grasp the difference between true and false beliefs. Even if the concept of strong objectivity needs a triangular situation to emerge (and thus [P2] is not in question), it also presupposes the very presence of thought and language (Wimmer and Perner 1983). Thus, if thought can go before it, then strong objectivity cannot be necessary for concepts and beliefs. In a few cases, however, the concept of objectivity seems somewhat weaker: ‘the concept of objectivity, the concepts of objects and events that occupy a shared world, of objects and events whose properties and existence is independent of our thought’ (Davidson 1990a: 202). Here it is better to refer to a concept of ‘weak’ objectivity, as it can be identified with the minimal understanding that worldly objects and events are different from us and external to our own mentality, and that there are other animate creatures somehow similar to us, with whom we share emotions and an independent outside world. Even if the concept of weak objectivity does not presuppose thought itself, and thus may be truly necessary for it, it would make [P2] false, as it needs no process of triangulation to emerge; rather, it is a precondition for triggering any triangular process. Triangulation can only get off the ground when a creature can see worldly objects and other creatures as part of the external world, as opposed to elements in his or her own subjective world, and can grasp a certain difference between physical objects and animate beings (Amoretti 2009; Roskies 2011). In other circumstances, Davidson (1982) also identifies the idea of grasping the objectivity of thought with the concept of belief or the ability to entertain second order beliefs. As these alternatives are both related to the concept of objectivity (strong or weak), they face the same problems discussed above (Amoretti 2009). Moving to the fixation of the empirical content, some scholars have observed that the problem of establishing the typical cause, or solving its

On the Social Dimension of Beliefs and Concepts

199

underdetermination, is much more complex than that described by Davidson, as the ambiguities associated with the identification of the typical cause do not only involve distance and width (Føllesdal 1999; LePore and Ludwig 2005; Verheggen 1997, 2007). If so, in order to fix empirical content in an appropriate and unambiguous way, a certain amount of thought (and maybe language) should already be there. The fixation of empirical content, then, cannot be necessary for concepts and beliefs. In order to evaluate the plausibility of [P2] and [P3], it is compulsory to better specify the process of triangulation. Davidson (2001) recognises three different triangular situations. First, triangulation is regarded as a primitive situation, where two creatures lacking any concepts or beliefs simply react to the same external stimuli, and correlate each other’s reactions with the observed external stimuli. Second, triangulation describes a learning situation: one creature has not yet developed concepts or beliefs, but interacts with another creature, which already has concepts and beliefs. Third, triangulation illustrates an interpretative situation, where two cognitively mature human beings, already equipped with thought and language but lacking a common language, learn to understand one another thanks to the intentional sharing of a joint cause. The first case can be dubbed ‘basic’ triangulation; it is a non-cognitive, non-intentional, non-linguistic interactive situation in which two creatures simply react to common external stimuli and to each other’s reactions to those stimuli. Basic triangulation must be regarded merely as a responsivediscriminatory interactive situation, and the creatures involved as unaware of being part of any triangle. Thus, it seems reasonable to claim that ‘the other subject’s behaviour is simply part of the surrounding “passing show” – something that can be correlated (or not) with items in the world (objects, events, state of affairs)’ (Bar-On and Priselac 2011: 125-126). Skipping to the third case, which may be dubbed ‘full-fledged’ triangulation; it is a fully cognitive, intentional, and linguistic intersubjective situation where two cognitively mature human beings consciously and intentionally react to common external stimuli and to each other’s reactions to those stimuli. The creatures involved in this process are totally aware of being part of a triangle: they appreciate each other as intentional subjects who share a common external world and way of perceiving it, and have subjective takes on the world that can either match or fail to match the way things actually are. Back to the second case; it cannot be identified with a different kind of triangulation, as Davidson sees no middle ground between a merely responsive-discriminatory situation and a fully cognitive and linguistic one. Thus, we should

200

Chapter Thirteen

understand the learning situation as twofold: while the child is engaging in basic triangulation, the adult is participating in full-fledged triangulation. With the above distinctions in mind, let us analyse [P2], beginning with basic triangulation. With regards to [a], grasping the objectivity of thought, some critics have argued that there are other situations, different from basic triangulation, that can equally grant the space for the concept of error, and thus of (strong) objectivity, to emerge. For instance, a single creature that compares its past and present experiences, or its current experiences from different perspectives, would be enough (Briscoe 2007; Montminy 2003; Verheggen 1997). Therefore, basic triangulation cannot be considered necessary for grasping the objectivity of thought. Davidson may eventually answer these charges, stressing that, without some intersubjective standards, a creature which is not already equipped with a conceptual and doxastic apparatus, cannot distinguish a misapplied concept or a mistaken belief from a concept or belief that is merely different. Thus, a creature in isolation does not have enough information to grasp the possibility of error or the objectivity of thought (Amoretti 2008b). There is a more substantial problem, though. In basic triangulation, the two creatures see each other not as intentional subjects, but merely as a part of the surrounding passing show. Thus, it is difficult to understand why two basically triangulating creatures should be more advantaged than a single creature in acquiring the concepts of error and objectivity. Basic triangulation, then, should not be considered truly necessary for grasping the objectivity of thought, as it does not represent a significant advantage for the creatures involved (Pagin 2001; Sinclair 2005). Similarly, with regards to [b], fixing the empirical content, a creature involved in basic triangulation is no more advantaged than a single creature in narrowing down the typical cause and fixing the empirical content. It seems reasonable to agree that, as far as basic triangulation is concerned, the causes narrowed down are still too coarsely grained to be useful for content determination (Føllesdal 1999; LePore and Ludwig 2005). Again, then, basic triangulation should not be considered truly necessary for fixing empirical content, as it is not a significant advantage for the creatures involved. Moving to full-fledged triangulation, it could be equally argued that there are other situations which grant the space for the concept of error to emerge, but as we have seen, Davidson may reply to that. In full-fledged triangulation, the interacting creatures consciously and intentionally react to common external stimuli and to each other’s reactions to those stimuli, and are able to communicate to each other the content of their common experiences and beliefs about the world. Hence, this kind of triangulation can easily create the right space for the

On the Social Dimension of Beliefs and Concepts

201

concepts of error and (strong) objectivity to emerge. However, this would be possible only because concepts and beliefs, as well as language, are already in place. Similarly, even if the creatures involved in full-fledged triangulation were entirely able to narrow down the typical cause in an appropriate and unambiguous way, this would only be possible because, again, concepts and beliefs, and language itself, are already in place. To sum up: either basic triangulation makes [P2] false or full-fledged triangulation makes the entire argument circular and useless. Finally, basic triangulation would make [P3] false because, even if two intentional subjects are involved, they do not recognise each other as such, nor are they aware of being part of a triangle. This means that basic triangulation cannot be seen as a fully social and intersubjective interaction. Full-fledged triangulation, on the contrary, is wholly compatible with [P3], as the creatures involved are totally aware of being part of a triangle and recognise each other as intentional subjects who share a common world.

3. A Novel Argument for the Sociality of Concepts and Beliefs Even if Davidson’s argument for the sociality of concepts and beliefs is flawed, its general structure is still promising. To begin, I will identify two plausible preconditions of thought; that is, two elements that are necessary to have concepts and beliefs, but that can also potentially be in place even when there is no thought (Amoretti 2013). Then, as basic triangulation is not a fully social intersubjective interaction and fullfledged triangulation can hardly be seen as a precondition of thought, I will consider a third variety of triangulation, a sort of middle ground, which is fully social and intersubjective but does not presuppose thought itself (Amoretti 2013; Bar-On and Priselac 2011; Brink 2004; Elian 2005). Finally, I will briefly show the effectiveness and advantages of this novel argument. The main difficulty related to the two aspects of thought considered by Davidson is that they are not truly necessary for concepts and beliefs, as a certain amount of thought can reasonably be present even without them. Putting aside weak objectivity, which is a precondition of any kind of triangulation, strong objectivity already requires a good mastery of thought and language to emerge. Also, narrowing down a typical cause in an appropriate and unambiguous way so as to fix empirical content demands a certain amount of thought. However, these conditions point to better ones. On the one hand, let us assume that having concepts and beliefs

202

Chapter Thirteen

means being able to classify objects and events, to make sense of the idea of misapplying them. In order to do that, a creature must be able to understand that there might be either agreement or disagreement between one’s own current intentions and subjective takes on the world and other creatures’ current intentions and subjective takes. This understanding is connected to agency-related abilities and to present activities and circumstances, but it does not need the mastery of abstract concepts such as truth and falsity. We can dub it ‘proto-objectivity’ because it is a necessary step toward a fully developed concept of objectivity. Though necessary, making sense of proto-objectivity is not sufficient for thought. On the other hand, let us admit that the content of concepts and beliefs is determined, at least in part, by their typical cause. Solving the underdetermination of the typical cause is not a straightforward process, but a promising starting point is establishing some common domains of current relevance. As a single creature is not even able to ascertain whether the relevant cause is at a proximal level, a distal one, or somewhere in between, circumscribing a shared background is a necessary step in starting the narrowing down process and possibly solving the underdetermination. Once a common domain of current relevance has been created, it might become the right space for fixing the typical cause and determining content. Again, though necessary, this condition is not sufficient for thought. As basic and full-fledged triangulation are unfit to support the sociality of thought, it is important to make explicit what requirements a different kind of triangular process should have (Bar-On and Priselac 2011). First, it should be a process resting crucially on the psychological infrastructure of human beings, an infrastructure that originated evolutionarily. Second, it should be part of our best theories about child development and language acquisition. (These two requirements are compulsory if one aims to offer a scientifically plausible argument.) Third, in opposition to basic triangulation, it should be something more than a merely responsivediscriminatory interactive situation and represent a genuinely social intersubjective situation. The creatures involved should express communicative intentions, recognise other creatures’ communicative intentions, and see each other as intentional subjects. Fourth, it should be an authentic triadic interaction, not a combination of dyadic relations. Fifth, in opposition to full-fledged triangulation, it should not describe a fully cognitive and linguistic situation. Sixth, it should be necessary and sufficient to obtain the two preconditions discussed above: [a*], grasping the concept of proto-objectivity, and [b*], defining some common domains of current relevance. All these requirements considered, the best

On the Social Dimension of Beliefs and Concepts

203

candidate to fill the gap between basic and full-fledged triangulation is the psychological process of joint attention (Elian et al. 2005; Moore and Dunham 1995; Tomasello 1999, 2008). According to developmental psychologists, human infants have a rough understanding of an independently existing physical world by three or four months of age. They display an awareness of worldly objects as independent entities which continue to exist even when they are not currently being observed, and they show the ability to grasp some basic features of worldly objects, such as that they cannot be in two places at one time or that they cannot pass through one another. Likewise, human infants have an understanding of their ecological selves. Finally, probably due to the presence of a simulation mechanism in which the child understands other people by analogy with the self, human infants show an ability to recognise other human beings as animate creatures, as opposed to physical objects: they engage in protoconversations, share basic emotions and mimic adults’ bodily movements. These features are compatible with grasping weak objectivity and are preconditions for triggering a fully social and intersubjective triangular process. However, at this stage, relationships with other human beings are not fully social and intersubjective, as the other creatures are not yet understood as intentional subjects with their own experiences. Moreover, human infants are not engaging in triadic interactions, as they only perform two different kinds of dyadic interactions: they relate dyadically with worldly objects and events, and dyadically with other human beings. Between nine and twelve months of age, human infants incur in a sort of cognitive revolution, represented by the process of joint attention, which rests crucially on the psychological infrastructure of human beings and is part of our current best theories about child development and language acquisition. According to Michael Tomasello, this process is genuinely social and intersubjective, as the creatures involved recognise each other as intentional subjects: ‘At nine months of age human infants begin engaging in a number of so-called joint attentional behaviours that seem to indicate an emerging understanding of other persons as intentional agents like the self whose relations to outside entities may be followed into, directed, or shared’ (Tomasello 1999: 61). This ability is probably due to a combination of simulation mechanisms, already present during the first few months of age, with the child’s new understanding of his or her own intentionality, which becomes more complex and sophisticated by eight to nine months of age. Thus, when human infants come to understand themselves as intentional subjects, they also recognise other people as such. Also, the process of joint attention represents an

204

Chapter Thirteen

indisputable kind of triadic interaction, consisting of two creatures that, as a result of their mutual attention to each other’s states of attention, intentionally coordinate their attention on a common worldly object or event, so that there is a mutual understanding that they are both attending to the very same object or event. It is a triadic process as it involves ‘a coordination of [the] interactions with objects and people, resulting in a referential triangle of child, adult, and the object or event to which they share attention. […] Infants for the first time begin to flexibly and reliably look where adults are looking (gaze following), to engage with them in relatively extended bouts of social interaction mediated by an object (joint engagement), to use adults as social reference points (social referencing), and to act on objects in the way adults are acting on them (imitative learning). […] Infants for the first time begin to “tune in” to the attention and behaviour of adults toward outside entities’ (Tomasello 1999: 62). This means that the child and the adult, through pointing, holding up an object, using eye contact, and alternating their gazes between each other and a particular external object, attend to each other’s states of attention and eventually appreciate that those states of attention are pointing to a shared worldly object. Here, the child’s pointing to an object is not imperative but declarative, having the sole aim of sharing attention to that particular object with the adult. The process of joint attention, then, is quite demanding: human infants must be able to understand themselves as intentional subjects, to recognise adults as intentional subjects, to perform primitive forms of communicative intention through pointing and gazes (that is, intentions to attend to worldly shared objects and events, and to direct other creatures’ attention to those objects and events), and to identify primitive communicative intentions among other human behaviours (that is, those behaviours that are intentionally directed at worldly shared objects or events with the communicative aim to jointly focus the attention on a specific target). This means that, in contrast to basic triangulation, the creatures engaging in joint attention have the cognitive resources to be mutually aware of being part of a triangle composed of oneself, the other creature and a worldly object or event. But thus far such a mutual awareness depends on agency-related abilities and present circumstances only, and does not require the introduction of propositional thought, or concepts such as those of belief, proposition, error, truth, and falsity, or to master a language. The process of joint attention, then, compared to full-fledged triangulation, is not fully cognitive and is definitely not linguistic; it presupposes neither thought nor language.

On the Social Dimension of Beliefs and Concepts

205

To conclude, it is important to evaluate whether joint attention is necessary and sufficient to obtain [a*] and [b*]. Grasping the concept of proto-objectivity means to have an understanding that there might be either agreement or disagreement between one’s own current intentions and subjective takes on the world and other creatures’ current intentions and subjective takes. Such an understanding cannot emerge from distinct couples of dyadic interactions (child-world and child-adult) or from a merely social triadic interaction (in which the creatures do not recognise each other as intentional subjects), but requires a fully social and intersubjective triadic interaction. Thus, it is a feature that can only emerge through the process of joint attention: human infants must first engage in successful joint attentional interactions in order to grasp the idea that other people may be attending to worldly objects and events different than those to which they are attending. Consequently, they must understand that their intentions and subjective takes on the world might fail to meet their success conditions and sometimes conflict with those of others. Hence, the concept of proto-objectivity emerges through the very process of joint attention. Likewise, some common domains of current relevance cannot emerge from distinct couples of dyadic interactions, nor from a merely social triadic interaction; rather they effortlessly appear and become more and more precise through joint attentional processes, as they need a fully social and intersubjective triadic interaction, in which the mutual awareness of being and attending to the same worldly object or event can emerge. To sum up, the process of triangulation as joint attention satisfies all the requirements stated above: it rests on the evolved psychological infrastructure of human beings; it is part of our best theories about child development; it is a fully social and intersubjective situation; it is a triadic interaction; it is not fully cognitive or linguistic; and it is necessary and sufficient to grasp the concept of proto-objectivity and define some common domains of current relevance. So we can modify the argument for the sociality of thought: [P1*] (a*) Grasping the concept of proto-objectivity and (b*) defining some common domains of current relevance are necessary conditions for having beliefs and concepts; [P2*] Triangulation as joint attention is necessary and sufficient to obtain (a*) and (b*); [C1*] Triangulation as joint attention is necessary to have beliefs and concepts; [P3*] Triangulation as joint attention is a fully social and intersubjective interaction;

206

Chapter Thirteen [C2] A fully social and intersubjective interaction is necessary to have beliefs and concepts.

This novel argument is effective: [a*] and [b*] are truly necessary for concepts and beliefs (no amount of propositional thought can reasonably be present without them); triangulation as joint attention is necessary and sufficient for the emergence of [a*] and [b*]; joint attention is a fully social and intersubjective triadic process, most likely the simplest one. (Even if there were other processes sharing relevant characteristics with joint attention, it would still remain sufficient for [a*] and [b*]. More importantly, these alleged alternatives should be fully social and intersubjective as well, and thus the general efficacy of the argument would not be threatened.) At the same time, the reasoning is not circular because triangulation as joint attention does not require propositional thought or language. Another advantage is its scientific plausibility: the capability of engaging in joint attentional interactions is a biologically inherited skill and an integral part of our best theories about the origin and development of the human cognition. Besides, as joint attentional skills come in degrees, it is possible to account for various cognitive deficits, such as autism. Despite its scientific plausibility, however, the argument does not automatically imply a reduction of thought (and language), because [a*] and [b*] are necessary but not sufficient to have beliefs and concepts. Finally, it might have some interesting consequences for the epistemology of our ordinary knowledge, especially with regard to various forms of scepticism (about the external world, other minds, and our own mind) and relativism.

References Amoretti, M.C. (2008a). Davidson, self-knowledge, and scepticism. In M.C. Amoretti & N. Vassallo (Eds.), Knowledge, language, and interpretation: On the philosophy of Donald Davidson. Frankfurt: Ontos. —. (2008b). Il triangolo dell’interpretazione: Sull’epistemologia di Donald Davidson. Milano: FrancoAngeli. —. (2009). Comunicazione preverbale e razionalità. In D. Gambarara & A. Givigliano (Eds.), Origine e sviluppo del linguaggio, tra teoria e storia,(291-301). Roma: Aracne. —. (2011). Triangulation between externalism and internalism. In M.C. Amoretti & G. Preyer (Eds.), Triangulation: From an epistemological point of view (47-68). Frankfurt: Ontos.

On the Social Dimension of Beliefs and Concepts

207

—. (2012). Sul carattere sociale di pensiero e linguaggio. RIFL. Rivista Italiana di Filosofia del Linguaggio, Special Number SFL, 4-16. —. (2013). Concepts within the model of triangulation. ProtoSociology, 30, 49-62. Bar-On, D. & Priselac, M. (2011). Triangulation and the beasts. In M.C. Amoretti & G. Preyer (Eds.), Triangulation: From an epistemological point of view (121-152). Frankfurt: Ontos. Brink, I. (2004). Joint attention, triangulation and radical interpretation: A problem and its solution. Dialectica, 58, 179-206. Briscoe, R.E. (2007). Communication and rational responsiveness to the world. Pacific Philosophical Quarterly, 88, 135-159. Davidson, D. (1982). Rational animals. In Subjective, intersubjective, objective ( 95-106). Oxford: OUP. —. (1983). A coherence theory of truth and knowledge. In D. Henrich (Ed.), Kant oder hegel (pp. 423-438). Stuttgart: Klett-Cotta. —. (1988). The myth of subjective. In Subjective, intersubjective, objective (39-52). Oxford: OUP. —. (1990a). Epistemology externalised. In Subjective, intersubjective, objective (193-203). Oxford: OUP. —. (1990b). Turing’s test. In Problems of rationality (77-86). Oxford: OUP. —. (1991). Three varieties of knowledge. In Subjective, intersubjective, objective (205-218). Oxford: OUP. —. (1992). The second person. In Subjective, intersubjective, objective (107-122). Oxford: OUP. —. (1995). The problem of objectivity. In Problems of rationality (3-18). Oxford: OUP. —. (1997). The emergence of thought. In Subjective, intersubjective, objective (123-134). Oxford: OUP. —. (2001). Externalisms. In P. Kotatko, P. Pagin & G. Segal (Eds.), Interpreting Davidson (1-16). Stanford: CSLI. Elian, N. (2005). Joint attention, communication and mind. In N. Elian, C. Hoerl, T. McCormack & J. Roessler (Eds.), Joint attention: Communication and other minds (1-33). Oxford: OUP. Elian, N., Hoerl, C., McCormack, T. & Roessler, J. (Eds.) (2005). Joint attention. Communication and other minds. Oxford: OUP. Føllesdal, D. (1999). Triangulation. In L. E. Hahn (Ed.), The philosophy of Donald Davidson ( 719-728). La Salle: Open Court. LePore, E. & Ludwig, K. (2005). Donald Davidson. Meaning, truth, language, and reality. Oxford: Clarendon Press.

208

Chapter Thirteen

Montminy, M. (2003). Triangulation, objectivity and the ambiguity problem. Critica, 35, 25-48. Moore, C. & Dunham, P.J. (1995). Joint attention: Its origins and role in development. Hillsdale: Lawrence Erlbaum Associates. Pagin, P. (2001). Semantic triangulation. In P. Kotatko, P. Pagin & G. Segal (Eds.), Interpreting Davidson (199-212). Stanford: CSLI. Roskies, A.L. (2011). Triangulation and objectivity. Squaring the circle? In M.C. Amoretti & G. Preyer (Eds.), Triangulation: From an epistemological point of view (97-102). Frankfurt: Ontos. Sinclair, R. (2005). The philosophical significance of triangulation: Locating Davidson’s non-reductive naturalism. Metaphilosophy, 36, 708-727. Tomasello, M. (1999). The cultural origins of human cognition. Cambridge: Harvard University Press. —. (2008). Origins of human communication. Cambridge: MIT Press. Verheggen, C. (1997). Davidson’s second person. The Philosophical Quarterly, 47, 361-369. —. (2007). Triangulating with Davidson. The Philosophical Quarterly, 57, 96-103. Wimmer, H. & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, 13, 103-128.

PART IV ORDINARY REASONING AND HISTORICAL KNOWLEDGE

CHAPTER FOURTEEN SUBTLETIES OF NAÏVE REASONING: PROBABILITY, CONFIRMATION, AND VERISIMILITUDE IN THE LINDA PARADOX GUSTAVO CEVOLANI AND VINCENZO CRUPI

1. The Linda Paradox Here is a character sketch that has become famous in the study of human reasoning: Linda is 31 years old, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.

In seminal investigations by Tversky and Kahneman (1982, 1983), most people faced with Linda’s description ranked the conjunctive statement “Linda is a bank teller and is active in the feminist movement” (bšf from now on) as more probable than the isolated conjunct “Linda is a bank teller” (b). In a particularly neat demonstration of the phenomenon, 142 university students were simply asked to choose the more probable state of affairs from b and bšf: 85% of them chose the latter. This pattern of judgments is puzzling in that it conflicts with a basic and uncontroversial principle of probability theory, known as the “conjunction rule”, prescribing that a conjunction of statements can not be more probable than any of its conjuncts. The Linda paradox – as we will call it here – is a prominent instance of a widespread and well-documented phenomenon, known in the literature as the conjunction fallacy (or the conjunction effect). Indeed, Tversky and Kahneman themselves, along with many others in subsequent studies, were able to replicate this phenomenon in a variety of experimental scenarios, including real-life

212

Chapter Fourteen

settings such as medical prognoses (see Tversky and Kahneman 1983, p. 301, and Wedell and Moro 2008, for a survey). In the last few decades, scholars interested in the analysis of human reasoning and decision making under uncertainty have devoted a great deal of work to the Linda paradox. Interestingly, ever since Levi’s (1985) insightful review of Kahneman, Slovic, and Tversky’s (1982) influential work, the issue has also attracted the attention of a number of epistemology scholars (see Levi 1985, 2004; Bovens and Hartmann 2003; Hintikka 2004; Crupi, Fitelson, and Tentori 2008; Hartmann and Meijs 2012; Shogenji 2012; Kuipers 2012). In any case, providing a satisfactory account of the phenomenon has proved rather challenging. Presumably, the only established fact about the Linda paradox is that people’s responses cannot be accounted for by assuming both (i) that participants indeed mean to provide plain judgments of the probabilities of b and bšf, and (ii) that they are elaborating their judgments in a rational fashion.1 From here on in, agreement gives way to open controversy. One reaction to the Linda paradox has been to retain assumption (ii) and challenge (i), thus claiming that the experimental evidence has not demonstrated the occurrence of a reasoning error after all. As instantiated in the psychological literature, this line of argument was inspired by recurrent concerns about the pragmatics of communication in experimental settings: in the Linda problem, participants might have in fact interpreted the isolated conjunct b as bš¬f, or they might have read the ordinarylanguage conjunction “and” as a disjunction. From this perspective, it is submitted, the “conjunction fallacy” would reflect nothing more than “intelligent inferences” which only “look like reasoning errors” (see Hertwig and Gigerenzer 1999, and Tentori and Crupi 2012a for a recent review and criticism). In the epistemological field, assumption (i) has been attacked, too, but on different grounds. Scholars like Bovens and Hartmann (2003, pp. 85–88) and Hintikka (2004), for instance, independently proposed that experimental subjects may not be simply assessing P(b) and P(bšf), but the relative reliability of two different sources of information which are assumed to report the hypotheses b and bšf. Again, the case was made that, under this reconstruction of the paradox, “there need not be anything fallacious or otherwise irrational about the conjunction effect” (Hintikka 2004, p. 30).

1

Recently, advocates of the so-called quantum probability approach to cognition have tried to retain both (i) and (ii) (Pothos and Busemeyer 2013). See Tentori and Crupi (2013) for criticism of this attempt.

Subtleties of Naïve Reasoning in the Linda Paradox

213

Those who attempted to explain (as opposed to question) the Linda paradox as a reasoning error have instead retained (i), at the expense of (ii). A possible approach in this vein is that participants may have degrees of belief in the conjuncts b and f, but then combine them in an assessment of the conjunction bšf by some rule that departs from the principles of classical mathematical probability (see, e.g., Nilsson, Rieskamp, and Mirjam 2013, and Yates and Carlson 1986). A related and popular perspective is illustrated by Tversky and Kahneman’s original view of human judgment under uncertainty (Tversky and Kahneman 1974). Here, human reasoners are assumed to make intuitive judgments of chance and probability on the basis of the assessment of “heuristic attributes”. As these attributes do not generally fit the principles of mathematical probability, such a process may yield biases and lead to outcomes conflicting with compelling standards of rationality in cases such as Linda’s. Briefly put, “the answer to a question can be biased by the availability of an answer to a cognate question — even when the respondent is well aware of the distinction between them” (Tversky and Kahneman 1983, p. 312; also see Kahneman and Frederick 2002). This immediately raises a key question: what attributes (other than standard probability) are guiding participants’ prevailing responses? Traditional answers to this question involving attributes such as representativeness or typicality have been effectively criticized as suggestive but ultimately too vague (see Tentori, Crupi, and Russo 2013 for references and discussion). In what follows, we explore two different ideas. Both can be seen as novel attempts to flesh out the otherwise esoteric statement that “feminist bank teller is a better hypothesis about Linda than bank teller” (Tversky and Kahneman 1983, p. 311). To this purpose, we present two independently motivated and formally definable epistemological notions relying on which many judges would rank “feminist bank teller” over “bank teller” in the Linda problem. Such notions are verisimilitude (or truthlikeness) and confirmation (or inductive support), introduced in section 2. In sections 3 and 4, respectively, confirmation-based and verisimilitude-based accounts of the Linda paradox are presented. Finally, some general remarks concerning the roles of truth, probability and information in reasoning and cognition are presented in section 5.

214

Chapter Fourteen

2. Probability, Verisimilitude, and Confirmation A statement or hypothesis is more probable than another if it is more likely true, or closer to being true. For instance, lacking any other information about Linda, “Linda is not a bank teller” is more probable than “Linda is a bank teller” (in symbols, P(¬b) > P(b)) simply because bank tellers are only a small fraction of the general population. As already mentioned, probability calculus prescribes that a conjunction like bšf can never be more probable than any of its conjuncts, b and f. This is so because the class of feminist bank tellers is necessarily smaller than (is a subset of) both the class of bank tellers and the class of feminist activists. In this sense, probability is a decreasing function of logical strength: if a statement (e.g., bšf) logically entails another statement (e.g., b), then the former cannot be more probable than the latter. On the other hand, logically stronger statements provide more information about the relevant domain than weaker ones, and this fact alone shows that high probability cannot be the only guide in choosing among competing hypotheses. This last point was famously made by Popper (1934/1959; 1963), who forcefully argued that science progresses by putting forward highly informative and hence improbable hypotheses. While likely false, scientific hypotheses can approximate the truth in the sense of providing much true information about a given domain and so being close to “the whole truth” about that domain. In an effort to ground this theoretical framework, Popper (1963, ch. 10) proposed the notion of “verisimilitude” or “truthlikeness” as a “mixture of truth and information” (Oddie 1986, p. 12). This notion has to be carefully distinguished from probability, as Popper (1963, 236) clearly highlighted: The differentiation between these two ideas [verisimilitude and probability] is the more important as they have become confused; because both are closely related to the idea of truth, and both introduce the idea of an approach to the truth by degrees. [...] [P]robability [...] represents the idea of approaching logical certainty, or tautological truth, through a gradual diminution of informative content. Verisimilitude, on the other hand, represents the idea of approaching comprehensive truth. It thus combines truth and content.

An important consequence of the distinction between verisimilitude and probability is that a hypothesis can be regarded as quite close to the truth, i.e., as highly verisimilar, but still not be expected to be true, i.e., regarded as highly probable. In particular, while bšf can not be more probable than b (or f), it may well be expected to be more verisimilar, i.e., a better

Subtleties of Naïve Reasoning in the Linda Paradox

215

approximation than b (or f) to the whole truth about Linda. In turn, this may explain why experimental participants judge bšf a better hypothesis about Linda as compared to b: while less likely to be true, the hypothesis “feminist bank teller” may well be evaluated as more verisimilar than “bank teller” alone (Cevolani, Crupi, and Festa 2010, 2012). Interestingly, Tversky and Kahneman themselves (1983, 312) – even without mentioning the idea of verisimilitude, of which they were probably unaware – made a very similar point: The expected value of a message can sometimes be improved by increasing its content, although its probability is thereby reduced. [...] Consider the task of ranking possible answers to the question “What do you think Linda is up to these days?” The maxim of value could justify a preference for bšf over b in this task, because the added attribute feminist considerably enriches the description of Linda’s current activities at an acceptable cost in probable truth.

This kind of preference may be based on a notion that, unlike probability, can favor a stronger hypothesis over a weaker one (cfr. Levi 1985, 338). Verisimilitude is such a notion; another is that of inductive support or confirmation, as construed in the Bayesian approach to reasoning and (scientific) inference. The difference between probability and (inductive) confirmation has long been known and discussed in the logical analysis of inductive reasoning (for a very useful survey, see Fitelson 2005). Roughly, a hypothesis is confirmed or supported by some piece of evidence or information when its probability increases once this information is taken into account. In Carnap’s (1962, xv-xx) telling terminology, probability is related to credibility, construed as “firmness”, while confirmation concerns “increase in firmness”. For instance, when Linda’s description (e) is provided, the “posterior” probability of “Linda is a feminist” in the light of e may well be greater than its “prior” probability (in symbols P(f|e) > P(f)). Intuitively, this is so, since, among people with Linda’s background, one expects to find more feminist activists than among the general population. Thus, the hypothesis “Linda is a feminist” is confirmed or supported by Linda’s description. In this case, f is both highly probable based on e and strongly supported by e. In general, however, assessments of (posterior) probability and confirmation can point in opposite directions: as John Irving Good (1968, 134) once effectively

Chapter Fourteen

216

remarked, “if you had P(h|e) close to unity, but less than P(h), you ought not to say that h was confirmed by e”.2 As far as the Linda paradox is concerned, the crucial point of interest is that confirmation assessments may favor stronger hypotheses over weaker ones. In the present case, this means that the description provided by e may confirm “feminist bank teller” as a hypothesis about Linda more than “bank teller” alone (Crupi, Fitelson, and Tentori 2008). Indeed, the probability of b is surely not increased by e (and it is perhaps lowered, i.e., P(b|e) ” P(b)); while the probability of b ƒ† f may well increase when e is given, due to the increased probability of f from e. Thus, while bšf remains less probable than b – i.e., P(bšf|e) ” P(b|e) – it may well be more confirmed, since P(bšf|e) > P(bšf). The foregoing observations suggest that the Linda paradox can be accounted for on the basis of both verisimilitude and confirmation considerations. The resulting verisimilitude-based and confirmation-based accounts of paradox are presented, starting with the latter, in the next two sections.

3. Increasing Probability: A Confirmation-Based Account of the Linda Paradox A confirmation-theoretic framework for the conjunction fallacy has been presented in full detail by Crupi, Fitelson and Tentori (2008) and Tentori and Crupi (2012b), who also discuss a number of earlier contributions that are more or less strictly related (see Lagnado and Shanks 2002; Levi 1985, 2004; Sides et al. 2002; Tenenbaum and Griffiths 2001). In the philosophical and statistical literature, a number of (partially conflicting) confirmation measures have been proposed, expressing the degree c(h,e) to which hypothesis h is supported by evidence e, i.e., how much the probability of h increases (or decreases) by learning e (see Festa 1999 and Crupi 2014). As an example, the well-known “difference measure” of confirmation is defined as follows (Carnap 1962, p. 361): (1) c(h,e) = p(h|e) í p(h)

2

As neat as it is, the distinction between posterior probability and inductive confirmation has proved the recurrent need for theoretical clarity in philosophy (Popper 1954; Peijnenburg 2012), artificial intelligence (Horvitz and Heckerman 1986), and the psychology of reasoning alike (Sides et al. 2002; Crupi, Fitelson, and Tentori 2008).

Subtleties of Naïve Reasoning in the Linda Paradox

217

Note that c(h,e) is positive, expressing the fact that h is confirmed by e if and only if p(h|e) > p(h), i.e., if the probability of h is increased by e. According to the confirmation-based account of the Linda paradox, when judging probabilities, participants tend to rely on assessments of confirmation; thus, their preference for bšf over b as a hypothesis about Linda ultimately depends on the fact that the former hypothesis, unlike the latter, is confirmed by Linda’s story. More precisely, Crupi, Fitelson, and Tentori (2008, p. 188) proved that for a wide class of confirmation measures – including the difference measure defined in (1) – the following holds: (2) if c(b,e) ” 0 and c(f,e|b) > 0, then c(bšf,e) > c(b,e).

In other words, if “bank teller” is not supported by the evidence, while “feminist” is, even conditional on b, then “feminist bank teller” is more confirmed than “bank teller” by this same evidence. The second condition of the above theorem expresses the crucial fact that being a feminist is confirmed by Linda’s story, even if being a bank teller is concurrently assumed to hold true. The confirmation-based account identifies in the relatively high confirmation of the “added” conjunct (being a feminist activist) the crucial determinant of the Linda paradox. As mentioned in section 2, however, it is clear that Linda’s description not only supports the added conjunct, but also makes it fairly probable. Thus, the issue arises as to which of these two variables – the perceived probability of the added conjunct vs. the perceived support (or inductive confirmation) that the added conjunct receives from the relevant information – is critical for the conjunction fallacy. In connection with this, virtually all current explanations of the paradox have identified the relatively high perceived probability of the added conjunct as the main factor on which the phenomenon depends. As a consequence, beyond their specificities, these explanations share the prediction that the tendency to commit the fallacy should increase when the perceived probability of the added conjunct increases. On the contrary, the confirmation-theoretic account suggests that this widespread implication, as plausible as it may seem at first sight, is ultimately unsound. How can one discriminate between the role of the probability of the added conjunct, on the one hand, and its degree of confirmation, on the other, as determinants of the conjunction fallacy? As an illustration, consider a modified version of the Linda scenario. In this enriched setting, a new hypothesis about Linda is included, i.e., the statement “Linda owns

218

Chapter Fourteen

an umbrella” (u from now on), so participants are asked to pick the most likely statement among b, bšf, and bšu. It seems clear that u is extremely likely to be true (since almost everybody owns an umbrella) and at least as probable as f, even in the light of the rest of Linda’s story. However, only f is confirmed by the evidence, in the sense that its probability is increased by e (Linda’s description), while we can safely assume that the probability of u remains unchanged once e is learned. Now if the perceived probability of the added conjunct is the key determinant of the conjunction fallacy (as is implied by most competing psychology models, such as averaging rules), then fallacious responses would largely target bšu more than bšf. If, on the other hand, the perceived confirmation of the added conjunct is the key factor, the opposite would occur, with bšf as the prevalent choice among conjunction fallacy responses. In a series of experimental studies, behavioral data consistently yielded the latter pattern of results instead of the former (Tentori, Crupi, and Russo 2013). These findings support the confirmation-theoretic analysis of the paradox against alternative accounts by showing that the high probability of the added conjunct alone is not sufficient to trigger the conjunction effect.

4. Approaching the Truth: A Verisimilitude-Based Account of the Linda Paradox According to the confirmation-based account of the Linda paradox, “feminist bank teller” is a better hypothesis about Linda than “bank teller” alone since the former is regarded as more confirmed than the latter by Linda’s story. Cevolani, Crupi, and Festa (2010, 2012) proposed a different, verisimilitude-based account, according to which “feminist bank teller” is judged a better approximation to the whole truth about Linda than “bank teller” alone. Within this account, the participants’ relative preferences for different hypotheses about Linda is explained by assuming that, in assessing probabilities, they actually rely on evaluations of verisimilitude of the competing hypotheses. Philosophers of science have proposed a number of different measures of verisimilitude, expressing the closeness of a given hypothesis to the “whole truth” about a target domain (see Niiniluoto 1987, 1998, Zwart 2001 and Oddie 2014 for overviews). Intuitively, a hypothesis is highly verisimilar if it says many things about the domain, and if many of these things are (almost exactly) true. For instance, suppose that, as a matter of fact, Linda is indeed a bank teller who is active in the feminist movement (i.e., bšf is true of Linda). If this were the case, then bšf would be clearly

Subtleties of Naïve Reasoning in the Linda Paradox

219

more verisimilar than b alone, since both hypotheses are true, but the former is more informative than the latter.3 Of course, since the whole truth about the relevant domain is typically unknown, the crucial point of interest is usually the estimated verisimilitude of competing hypotheses, not their actual verisimilitude. For instance, in the Linda scenario, participants simply do not know whether Linda is actually a bank teller or is active in the feminist movement. Thus, they can only estimate the relative verisimilitude of two hypotheses like b and bšf on the basis of the available evidence e provided by Linda’s story. Such comparisons of estimated verisimilitude can be made by introducing the notion of the expected verisimilitude Evs(h|e) of hypothesis h based on evidence e: Evs(h|e) is high when, given the available evidence e, h appears as probably very close to the truth (see the appendix for relevant definitions). According to the verisimilitude-based account of the Linda paradox, the participants’ preference for “feminist bank teller” over “bank teller” as a hypothesis about Linda can be explained assuming that they assess the expected verisimilitude of bšf as higher than b in the light of Linda’s story, i.e., that Evs(bšf|e) > Evs(b|e). More precisely, Cevolani, Crupi and Festa (2010) proved that, for a wide class of verisimilitude measures vs:4 (3) Evs(bšf|e) > Evs(b|e) iff p(f|e) > ı

where ı is a threshold value characterizing the specific measure vs used to assess the verisimilitude of bšf and b. In other words, if “Linda is a feminist” is sufficiently probable given Linda’s story (i.e., p(f|e) > ı), then “feminist bank teller” is estimated as more verisimilar than “bank teller” alone. This makes the former a better hypothesis about Linda than the latter, thus explaining people’s preferences. Theorem 3 shows that the verisimilitude-based account agrees with most analyses of the Linda paradox in identifying the relatively high probability of f given e as the main factor underlying the phenomenon. As 3

As far as hypotheses are expressed as conjunctions of logically independent atomic statements (as in the Linda scenario), virtually all verisimilitude measures agree on such an evaluation (cfr. Cevolani, Crupi, and Festa 2011). More generally, according to most verisimilitude theorists (an exception being Oddie 1986), if h and g are both true, and h logically entails g, then h is more verisimilar than g (cfr. Niiniluoto 1987232 ff., for a discussion of this and other relevant properties of verisimilitude measures). 4 For proofs of the results in the present section see the appendix at the end of the paper.

220

Chapter Fourteen

a consequence, it is at odds with the experimental findings, described in section 3, of Tentori, Crupi, and Russo (2013). In particular, it seems unable to appropriately address the “umbrella” scenario, in which a further conjunctive statement occurs involving the new hypothesis “Linda owns an umbrella” (u). In fact, from theorem 3, bšu is estimated as more verisimilar than b whenever P(u) > ı. Since we are assuming that P(u) • P(f), inequality P(u) > ı holds whenever P(f) > ı, it follows that conjunction fallacy responses should amount to choosing bšu for its higher expected verisimilitude against both b and bšf, contrary to the experimental results. As a way to address this problem, one can deploy the notion of “verisimilitudinarian confirmation” (vs-confirmation for short) as introduced by Cevolani, Crupi, and Festa (2012), namely as how much the expected verisimilitude of a hypothesis h increases (or decreases) in the light of evidence e. A simple definition, corresponding to definition 1 of probabilistic confirmation, is the following: (4) cvs(h,e) = Evs(h|e) í Evs(h)

where vs is a verisimilitude measure. Thus, h is vs-confirmed by e if and only if cvs(h,e) is positive, i.e., if and only if Evs(h|e) > Evs(h). As far as the Linda paradox is concerned, “feminist bank teller” would be preferred to “bank teller” alone insofar as cvs(bšf,e) was greater than cvs(b,e), i.e., to the extent that the expected verisimilitude of bšf increases more than that of b once e is learned. In this perspective, the umbrella case may be accounted for by noting that only bšf, but not bšu, is vs-confirmed by Linda’s story. Indeed, one can prove that, under suitably defined conditions, the following holds: (5) cvs(bšf,e) > cvs(b,e) iff c(f,e) > 0

In other words, the conjunction bšf is vs-confirmed by evidence e more than b only in the case of the added conjunct f being (probabilistically) confirmed by e. Theorem 5 provides an interesting link between vs-confirmation and standard (probabilistic) confirmation, which in turns allows the accommodation of the umbrella case within a verisimilitude-based account of the Linda paradox. In fact, this case was originally introduced as implying that f is confirmed by e while u is not, i.e., that one has c(f,e) > 0 • c(u,e). Given theorem 5, this implies that bšf, but not bšu, is more vs-confirmed by e than b, i.e., that cvs(bšf,e) > cvs(b,e)

Subtleties of Naïve Reasoning in the Linda Paradox

221

• cvs(bšu,e). In turn, this provides an explanation for the participants’ preferences based on the notion of verisimilitudinarian confirmation.

5. Concluding Remarks: Cognitive Utilities and “Naïve” Reasoning In the foregoing sections, we have presented two accounts of the Linda paradox. Both rely on notions explored in philosophical discussions about reasoning, rationality, and the scientific method to explain the participants’ prevailing responses in the Linda task. According to these accounts, when judging probabilities, people may rely on assessments of the confirmation and expected verisimilitude, respectively, of the relevant hypotheses at issue, and this may explain their deviation from the rules of probability calculus. Thus, both confirmation and (expected) verisimilitude can be construed as defining new relevant heuristic attributes along the general lines of Tversky and Kahneman’s “cognate question” quote. In support of this idea, one may suggest that probability, confirmation, and verisimilitude can all be seen as distinct formal explicata of a presystematic notion of “plausibility”, whose meaning is not necessarily exhausted by the rules of mathematical probability. In this connection, it is interesting to note, for instance, that the terms “probability” and “verisimilitude” are closely related both in ordinary and scientific language. Indeed, in many western languages, the word for “probable” derives from the concept of “truth” and either “likeness” or “appearance”, like for the German wahrscheinlich (cf. Hertwig and Gigerenzer 1999, 277). Moreover, the Latin words probabilis and verisimilis, introduced by Cicero as a translation of the Greek pithanos (said of a view which is plausible or persuasive), were still being used, more or less interchangeably, by the founders of probability calculus in the seventeenth century and beyond (Niiniluoto 1987, 160 ff.). Thus, the concepts of “probability”, “verisimilitude”, and “likelihood” are strongly related, at least etymologically, and it does not seem unreasonable to assume that in human intuitive judgment they may overlap in one way or another. On the other hand, the fact that the notion of probability can be easily confused with that of an increment of probability (i.e., of confirmation) is not surprising, given that such confusion is still quite widespread today, even in the scientific literature (see Peijnenburg 2012 for a recent discussion). The foregoing remarks suggest that probability, confirmation, and verisimilitude may be viewed as three ways of rendering different aspects of the notion of a “plausible” hypothesis. In connection with this, this paper has exploited a crucial formal feature separating probability, on the

Chapter Fourteen

222

one hand, from verisimilitude and confirmation on the other hand. In fact, both confirmation and verisimilitude allow for a more informative hypothesis (like “feminist bank teller”) to rank higher than a weaker one (like “bank teller”), while the former is necessarily less probable than the latter. Starting as early as Popper (1934/1959), this has been a central point in the philosophical discussion concerning the “cognitive” or “epistemic” goals guiding the rational choice between competing hypotheses in scientific and ordinary reasoning. As philosophers like Levi (1967) and other proponents of cognitive decision theory (cf. Niiniluoto 1987, ch. 12) made clear, neither probability, construed as an indicator of the truth of a hypothesis, nor truth itself can be the only relevant virtue in comparing different hypotheses and choosing among them. In fact, if truth were the only cognitive utility at issue, one should simply prefer the most probable hypotheses given the available evidence. Such a choice, however, would ignore the role of information (or content) as another relevant feature in evaluating different cognitive options, since highly probable hypotheses are bound to be quite uninformative (cf. Popper’s quotation in section 4). Table 1

Static

Change

Truth

Verisimilitude (informative truth)

probability

Expected verisimilitude

(probabilistic) confirmation

Verisimilitudinarian confirmation

The main outcome of this ongoing discussion is that a trade-off between truth and information seems necessarily involved in reasoning about competing hypotheses. This insight plays a central role both in verisimilitude-based and confirmation-based accounts of hypothesis assessment, as the present paper has shown. In this respect, both these approaches improve on a purely probabilistic account. This comes as no surprise in the case of verisimilitude, which was introduced precisely to capture a notion of “informative truth” that, contrary to probability, is positively associated with high content. Remarkably – as recent work has concurrently highlighted (see, for instance, Huber 2008 and Kuipers 2012) – confirmation, construed as increasing probability, may also sometimes favor logically stronger (and hence more informative) hypotheses over weaker ones. Table 1 classifies the notions discussed in this paper along

Subtleties of Naïve Reasoning in the Linda Paradox

223

two conceptual dimensions: the cognitive utility at issue (truth vs. informative truth or verisimilitude) and how this utility is considered, i.e., as a “static” virtue of the hypotheses or as changing in response to new incoming evidence. A general implication of our discussion above is that so-called naïve reasoning may not be as naïve as it seems. This claim has multiple meanings, though, for so does the term “naïve” itself, including the distinct nuances of “simple, unsophisticated” along with “poor, inaccurate”. According to a vocal school of thought, for instance, naïve reasoning is not naïve, in that it allegedly achieves high degrees of accuracy despite relying on a limited array of basic cognitive tools in an adaptive way (see Gigerenzer and the ABC Group 1999 for a now classic statement). At least as far as the Linda paradox is concerned, we suggest that a fairly different reading prevails, which is, however, no less interesting; namely, that even when biased in its final outcome, naïve reasoning may still be tracking subtle notions of remarkable epistemological interest.

6. Acknowledgments We acknowledge financial support from Grant CR 409/1-1 to Vincenzo Crupi from the Deutsche Forschungsgemeinschaft (DFG) as part of the priority program “New Frameworks of Rationality” (SPP 1516) and from the Italian Ministry of Scientific Research within the FIRB project Structures and dynamics of knowledge and cognition (Turin unit: D11J12000470001).

Appendix In the following, we introduce the “basic feature” approach to verisimilitude (Cevolani, Crupi, and Festa 2011; Cevolani, Crupi, and Kuipers 2013) and prove the results presented in section 4. Given a finite propositional language with n logically independent atomic statements, let us focus on its so called conjunctive statements. A conjunctive statement or hypothesis h is a finite conjunction of k different “basic statements”, with k ” n, where each basic statement is an atomic statement or its negation. Within such a language, the possible states of affairs of the underlying domain (the relevant “possible worlds”) are described by the (propositional) constituents c1, …, cq, which are the strongest (most informative) conjunctive statements of the language. There are q = 2n such constituents and only one of them is actually true, describing “the whole truth” about the domain within the language. As an example, if b and f are

Chapter Fourteen

224

the only atomic statements of the language, the constituents are the four conjunctions b š f, b š ¬f, ¬b š f and ¬b š ¬f, one of which will be the true complete description of Linda. A conjunctive hypothesis h is verisimilar if it is close to the (whole) truth. Thus, the (degree of) verisimilitude vs(h) of h can be defined in terms of the similarity or closeness of h to the true constituent. In turn, the (degree of) similarity s(h,ci) between h and a constituent ci expresses how close h is to ci, i.e., how much of what ci says is correctly entailed by h. Let us call each member of the set t(h,ci) of conjuncts of h which are also conjuncts of ci (i.e., which would be true if ci were the truth) a “match” for h; similarly, the set f(h,ci) contains the “mistakes” of h, namely the conjuncts of h which are not conjuncts of ci (i.e., which would be false if ci were the truth). Intuitively, h will be the closer to ci the greater the number of matches and the smaller the number of mistakes h makes with respect to ci. More formally, a feature contrast measure of the similarity of conjunctive hypotheses h is defined as follows (Cevolani, Crupi, and Festa 2011, 188): (1’) ‫ݏ‬ఝ ሺ݄ǡ ܿ௜ ሻ ൌ

ȁ௧ሺ௛ǡ௖೔ ሻȁ ௡

െ߮

ȁ௙ሺ௛ǡ௖೔ ሻȁ ௡

with ij > 0. If c* denotes the true constituent, then vsij(h) = sij(h,c*) measures the verisimilitude of h expressed as a weighted sum of the (normalized) number of true and false conjuncts of h. Intuitively, ij expresses the relative weight assigned to truth and falsity: the greater ij, the farther from the truth h will be due to its mistakes. If ij = 1 we can say that sij (and hence vsij) is a “balanced” measure, in the sense that truth and falsity are equally weighted. Two features of definition 1’ are worth noting. First, vsij is “monotonic” in the following special sense (cf. Cevolani, Crupi, and Festa 2011, 187): if hypothesis g is identical to h except for making one more match (i.e., entailing one more true conjunct) then g is more verisimilar than h, while if hypothesis g is identical to h except for making one more mistake (i.e., entailing one more false conjunct) then g is less verisimilar than h. As unproblematic as this property of “monotonicity” may seem, it is in fact violated by some of the verisimilitude measures proposed in the literature (most notably, Niiniluoto’s favored “min-sum” measure violates the second of the above conditions). Second, it is easy to check that vsij is “additive” in the (again, special) sense that the verisimilitude of h is the sum of the verisimilitude of its conjuncts. More formally, suppose that h is a conjunction of the form h1 š … š hk, with k ” n, then:

Subtleties of Naïve Reasoning in the Linda Paradox

225

(2’) ‫ݏݒ‬ఝ ሺ݄ሻ ൌ σ௛೔ ‫ݏݒ‬ఝ ሺ݄௜ ሻ Since the truth c* is typically unknown, comparing the actual verisimilitude of different hypotheses is usually impossible. One can instead assess their estimated verisimilitude, by applying the standard idea of mathematical expectation to vsij. If P is an epistemic probability distribution over the set of constituents c1, …, cq, then the (degree of) expected verisimilitude of h given evidence e is defined as: (3’) ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൌ σ௖೔ ‫ݏݒ‬ఝ ሺ݄ǡ ܿ௜ ሻ ൈ ܲሺܿ௜ ȁ݁ሻ i.e., as the average of the verisimilitude degrees of h in each possible world weighted by its corresponding probability given the evidence. Given two hypotheses h and g, one can then assess whether h is estimated based on e as more, less or equally verisimilar than g by comparing the values of Evsij(h|e) and Evsij(g|e). A noticeable property of measure Evsij is that it is also additive, given the additivity of vsij: (4’) ‫ݏݒܧ‬ఝ ሺ݄ሻ ൌ σ௛೔ ‫ݏݒܧ‬ఝ ሺ݄௜ ሻ Proof. By definition 3’, ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൌ σ௖೔ ‫ݏݒ‬ఝ ሺ݄ǡ ܿ௜ ሻ ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ. By the additivity of vsij (see 2’):

‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൌ σ௖೔ σ௛ೕ ‫ݏݒ‬ఝ ሺ݄௝ ǡ ܿ௜ ሻ ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ iff ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൌ σ௛ೕ σ௖೔ ‫ݏݒ‬ఝ ሺ݄௝ ǡ ܿ௜ ሻ ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ iff (by definition 3’) ‫ݏݒܧ‬ఝ ሺ݄ሻ ൌ σ௛ೕ ‫ݏݒܧ‬ఝ ሺ݄௝ ሻ. Now let h be a conjunctive hypothesis and x a basic statement; then one can prove the following result concerning the comparison between the expected verisimilitude of hypotheses hšx and h: (5’) Evsij(h š x|e) > Evsij(h|e) iff p(x|e) > ij/(ij+1) Proof. ‫ݏݒܧ‬ఝ ሺ݄š‫ݔ‬ȁ݁ሻ ൐  ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ iff, by additivity 4’,

‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൅  ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ ൐  ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ iff ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ ൐ Ͳ iff, by definition 3’, σ௖೔ ‫ݏݒ‬ఝ ሺ‫ݔ‬ǡ ܿ௜ ሻ ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ ൐ Ͳ iff (splitting the class of constituents in those entailing x and those not entailing x, i.e., entailing ¬x)

σ ௖೔

௫ ‫ݏݒ‬ఝ ሺ‫ݔ‬ǡ ܿ௜ ሻ

ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ ൅ σ௖೔

൓௫ ‫ݏݒ‬ఝ ሺ‫ݔ‬ǡ ܿ௜ ሻ

ൈ

Chapter Fourteen

226

‫݌‬ሺܿ௜ ȁ݁ሻ ൐ Ͳ iff (by definition 1’) σ௖೔ σ ௖೔

ఝ

൓௫ െ ௡

ଵ

ଵ ௫ ௡

ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ ൅ ఝ

ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ ൐ Ͳ iff ൈ ‫݌‬ሺ‫ݔ‬ȁ݁ሻ െ ൈ ‫݌‬ሺ൓‫ݔ‬ȁ݁ሻ ൐ ௡

௡

Ͳ iff, given ‫݌‬ሺ൓‫ݔ‬ȁ݁ሻ ൌ ͳ െ ‫݌‬ሺ‫ݔ‬ȁ݁ሻ, ‫݌‬ሺ‫ݔ‬ȁ݁ሻ ൐ since ij > 0,

ఝ ఝାଵ

Now let ߪ ൌ

ఝ

. Note that,

ఝାଵ

varies between 0 and 1. ఝ

; then theorem 3 in section 4 is an immediate

ఝାଵ

application of theorem 5’, above, by letting b be h, and f be x. As far as the expected verisimilitude of a basic statement x is concerned, one can prove that it only depends on the probability of x given the evidence; more precisely: (6’) ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ ൌ ‫݌‬ሺ‫ݔ‬ȁ݁ሻ

ఝାଵ

ఝ

െ ௡ Proof. By definition, ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ ൌ σ௖೔ ‫ݏݒ‬ఝ ሺ‫ݔ‬ǡ ܿ௜ ሻ ൈ ‫݌‬ሺܿ௜ ȁ݁ሻ, ௡

which, by using the relevant steps in the proof of theorem 5’, is equal to

ଵ ௡

ఝ

ൈ ‫݌‬ሺ‫ݔ‬ȁ݁ሻ െ ൈ ‫݌‬ሺ൓‫ݔ‬ȁ݁ሻ,

‫݌‬ሺ‫ݔ‬ȁ݁ሻ

ఝାଵ ௡

ఝ

௡

i.e,

by

calculations,

to

െ . ௡

In turn, theorem 6’ allows for a straightforward proof of result (5) concerning verisimilitudinarian confirmation. If h is a conjunctive hypothesis and x a basic statement: (7’) cvs(h š x,e) > cvs(h,e) iff c(x,e) > 0 Proof. cvs(h š x,e) > cvs(h,e) iff, by definition 4 of verisimilitudinarian ‫ݏݒܧ‬ఝ ሺ݄š‫ݔ‬ȁ݁ሻ െ  ‫ݏݒܧ‬ఝ ሺ݄š‫ݔ‬ȁ݁ሻ െ confirmation,

‫ݏݒܧ‬ఝ ሺ݄š‫ݔ‬ሻ ൐ ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ െ  ‫ݏݒܧ‬ఝ ሺ݄ሻ, iff, by additivity 4’, ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ ൅ ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ െ ሺ‫ݏݒܧ‬ఝ ሺ݄ሻ ൅ ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ሻሻ ൐ ‫ݏݒܧ‬ఝ ሺ݄ȁ݁ሻ െ  ‫ݏݒܧ‬ఝ ሺ݄ሻ iff, by calculations, ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ȁ݁ሻ െ ‫ݏݒܧ‬ఝ ሺ‫ݔ‬ሻ, i.e., by definition 4, cvs(x,e), is greater than 0 iff, by theorem 6’ ‫݌‬ሺ‫ݔ‬ȁ݁ሻ െ ‫݌‬ሺ‫ݔ‬ሻ, i.e., c(x,e), is greater than 0. References Bovens, L., and S. Hartmann (2003). Bayesian epistemology. Oxford: Oxford University Press.

Subtleties of Naïve Reasoning in the Linda Paradox

227

Carnap, R. (1962). Logical Foundations of Probability. Chicago: University of Chicago Press. Second edition; first edition: 1950. Cevolani, G., V. Crupi, and R. Festa (2010). “The whole truth about Linda: Probability, verisimilitude, and a paradox of conjunction”. In SILFS New Essays in Logic and Philosophy of Science, edited by M. D’Agostino, G. Giorello, F. Laudisa, T. Pievani and C. Sinigaglia, pp. 603–615. London: College Publications. Cevolani, G., V. Crupi, and R. Festa (2011). “Verisimilitude and belief change for conjunctive theories.” Erkenntnis 75 (2): 183–202. Cevolani, G., V. Crupi, and R. Festa (2012). “A verisimilitudinarian analysis of the Linda paradox”. In VII Conference of the Spanish Society for Logic, Methodology and Philosphy of Science, edited by C.M. Vidal, J.L. Falguera, J.M. Sagüillo, V. Verdejo, and M. PereiraFariña, pp. 366-373. Santiago de Compostela (Spain): USC Press. Cevolani, G., R. Festa, and T. A. F. Kuipers (2013). “Verisimilitude and belief change for nomic conjunctive theories”. Synthese, 190, pp. 3307–3324. Crupi, V. (2014). “Confirmation”. In The Stanford Encyclopedia of Philosophy (Spring 2014 Edition), edited by E. N. Zalta, URL = . Crupi, V., B. Fitelson, and K. Tentori (2008). “Probability, confirmation and the conjunction fallacy”. Thinking and Reasoning, 14, 182–199. Festa, R. 1999. “Bayesian confirmation”. In Experience, Reality, and Scientific Explanation, edited by M.C. Galavotti and A. Pagnini. Dordrecht: Kluwer Academic Publishers. Fitelson, B. (2005). “Inductive logic”. In Philosophy of Science. An Encyclopedia, edited by J. Pfeifer and S. Sarkar, 384–393. New York: Routledge. Gigerenzer, G. and the ABC group (1999). Simple Heuristics That Make Us Smart. Oxford University Press. Good, I.J. (1968). “Corroboration, explanation, evolving probability, simplicity, and a sharpened razor”. British Journal for the Philosophy of Science, 19,. 123–143. Hartmann, S. and W. Meijs (2012). “Walter the banker: The conjunction fallacy reconsidered”. Synthese, 184, 73–87. Hertwig, R., and G. Gigerenzer (1999). “The ‘conjunction fallacy’ revised: How intelligent inferences look like reasoning errors”. Journal of Behavioral Decision Making 12, 275–305. Hintikka, J. (2004). “A fallacious fallacy?” Synthese, 140, 25–35. Horvitz, E. and Heckerman, D. (1986). “The inconsistent use of measures of certainty in artificial intelligence research”. In Uncertainty in

228

Chapter Fourteen

Artificial Intelligence, edited by L.N. Kanal and J.F. Lemmer, 137– 151. Amsterdam: North-Holland. Huber, F. (2008). “Assessing theories, Bayes style”. Synthese, 161, 89– 118. Kahneman, D. and Frederick, S. (2002). “Representativeness revised: Attribute substitution in intuitive judgment”. In Heuristics and Biases: The Psychology of Intuitive Judgment, edited by T. Gilovich, D. Griffin, and D. Kahneman,. 49–81. New York: Cambridge University Press. Kuipers, T.A.F. (2012). “A realist partner for Linda: Confirming a theoretical hypothesis more than its observational sub-hypothesis”. Synthese, 184,. 63-71. Lagnado, D.A. and Shanks, D.R. (2002). “Probability judgment in hierarchical learning: A conflict between predictiveness and coherence”. Cognition, 83, 81–112. Levi, I. (1967). Gambling with truth. New York: Alfred A. Knopf. Levi, I. (1985). “Illusions about uncertainty.” British Journal for the Philosophy of Science, 36, 331–340. Levi, I. (2004). “Jaakko Hintikka”. Synthese, 140, 37–41. Niiniluoto, I. (1987). Truthlikeness. Dordrecht: Reidel. Niiniluoto, I. (1998). “Verisimilitude: The third period”. British Journal for the Philosophy of Science, 49, 1-29. Nilsson, H., Rieskamp, J., & Jenny, M.A. (2013). Exploring the overestimation of conjunctive probabilities. Frontiers in Psychology, 4. DOI:10.3389/fpsyg.2013.00101. Oddie, G. (1986). Likeness to truth. Dordrecht: Reidel. Oddie, G. (2014). “Truthlikeness”. In The Stanford Encyclopedia of Philosophy (Spring 2014 Edition), edited by E. N. Zalta. URL = http://plato.stanford.edu/archives/spr2014/entries/truthlikeness/ Peijnenburg, J. (2012). “Case of confusing probability and confirmation”. Synthese, 184, 101–107. Popper, K.R. (1934/1959). The Logic of Scientific Discovery. London: Routledge. Popper, K.R. (1954). “Degree of Confirmation”. British Journal for the Philosophy of Science, 5, 143–149. Popper, K.R. (1963). Conjectures and Refutations. London: Routledge and Kegan Paul. Pothos, E.M. and Busemeyer, J.R. (2013). “Can quantum probability provide a new direction for cognitive modeling?” Behavioral and Brain Sciences, 36, 255-274.

Subtleties of Naïve Reasoning in the Linda Paradox

229

Samuels, R., S. Stich, and M. Bishop. (2002). “Ending the rationality wars: how to make disputes about human rationality disappear”. In Common Sense, Reasoning and Rationality, edited by R. Elio, 236– 268. New York: Oxford University Press. Sides, A., Osherson, D., Bonini, N., and Viale, R. (2002). “On the reality of the conjunction fallacy”. Memory & Cognition, 30, 191–198. Tenenbaum, J.B. and Griffiths, T.L. (2001). “The rational basis of representativeness”. In Proceedings of 23rd annual conference of the Cognitive Science Society, edited by J.D. Moore and K. Stenning, pp. 1036-1041. Hillsdale: Erlbaum. Tentori, K. and Crupi, V. (2012a). “How the conjunction fallacy is tied to probabilistic confirmation: Some remarks on Schupbach (2009)”. Synthese, 184, 3-12. Tentori, K. and Crupi, V. (2012b). “On the conjunction fallacy and the meaning of and, yet again: A reply to Hertwig, Benz, and Krauss (2008)”. Cognition, 122, 123-134. Tentori, K. and Crupi, V. (2013). “Why quantum probability does not explain the conjunction fallacy”. Behavioral and Brain Sciences, 36, 308-310. Tentori, K., V. Crupi, and S. Russo (2013). “On the determinants of the conjunction fallacy: Probability vs. inductive confirmation”. Journal of Experimental Psychology: General, 142, 235-255. Tversky, A. and Kahneman, D. (1974). “Judgment under uncertainty: Heuristics and biases”. Science, 185, 1124–1131. Tversky, A. and Kahneman, D. (1982). “Judgments of and by representativeness”. In D. Kahneman, P. Slovic, and A. Tversky (eds.), Judgment under uncertainty: Heuristics and biases, 84-98. New York: Cambridge University Press. Tversky, A. and Kahneman, D. (1983). “Extensional vs. intuitive reasoning: The conjunction fallacy in probability judgment”. Psychological Review, 90, pp. 293–3l5. Shogenji, T. (2012). “The degree of epistemic justification and the conjunction fallacy”. Synthese, 184, 29–48. Wedell, D. H., and R. Moro. (2008). “Testing boundary conditions for the conjunction fallacy: effects of response mode, conceptual focus, and problem type”. Cognition 107, 105–136. Yates, J.F. and Carlson, B.W. (1986). “Conjunction errors: Evidence for multiple judgment procedures, including ‘signed summation’”. Organizational Behavior and Human Decision Processes, 37,. 230– 253. Zwart, S. (2001). Refined verisimilitude. Dordrecht, Springer.

CHAPTER FIFTEEN THE ORDINARINESS OF REASONING IN ARGUING PAOLO LABINAZ

Over the last forty years, an increasing number of psychological experiments have been conducted to study the cognitive processes involved in human reasoning. On the basis of their results, several psychological theories have been developed to describe how reasoning actually works. Beyond their differences, most of these theories converge on a similar way of conceiving of reasoning, giving rise to what can be called the standard approach to it. While this approach has been a cornerstone of psychological studies on reasoning, the picture of reasoning emerging from it has also been very influential in philosophical debate. The paper has a twofold objective: on the one hand, to question the picture of reasoning fostered by the standard approach and, on the other, to propose an alternative one which aims to provide a deeper and more complete understanding of human reasoning by focusing on its often neglected interpersonal function. The outline of the paper is as follows: in Section 1, I first introduce what I hold to be the two basic assumptions characterizing the standard approach and then show how these two assumptions constrain the picture of reasoning emerging from today’s most prominent psychological theories of reasoning. In Section 2, I criticize this picture because it represents reasoning as a self-centered cognitive activity, unfolding outside the reasoner’s awareness, whose function is fundamentally instrumental. In Section 3, I turn to Hugo Mercier and Dan Sperber’s argumentative theory, according to which reasoning has an interpersonal function, being evolved so as to evaluate others’ arguments and to justify our own conclusions and decisions (see Mercier and Sperber 2011a). In Section 4, starting from some ideas about human rationality presented (in a non-systematic way) by Paul Grice (1991; 2001), which Marina Sbisà (2006; 2007) has recently presented under the label argumentative

232

Chapter Fifteen

rationality, I first distance myself from Mercier and Sperber’s theory, as they too emphasize the rhetorical (and so instrumental) function of reasoning, and then outline an alternative picture of it which focuses not only on our ability to justify our conclusions and decisions, but also on our motivations (that is, our concern) to do so. I then conclude by making some suggestions as to the importance of studying reasoning in natural interactional situations where people ordinarily take care to justify what they say and do.

1. The Standard Approach to Reasoning Over the last thirty years, studies on human reasoning and decisionmaking have significantly increased. Particularly, findings from Wason’s and Kahneman and Tversky’s trailblazing experimental investigations, showing that people are not so rational as we think they are when reasoning and making decisions (see respectively Wason 1966; Tversky and Kahneman 1974) have stimulated the development of a variety of different experimental paradigms, all aimed at understanding the cognitive processes involved in human reasoning and decision-making, such as mental logic, mental models, and the heuristics and biases approach, which were among the first theories proposed (see respectively Rips 1994; Johnson-Laird and Byrne 1991; Kahneman, Slovic and Tversky 1982). However, while the proponents of these theories have limited their attention to the workings of reasoning, the most recent psychological theories, particularly evolutionary, ecological and dual-system theories (see respectively Cosmides 1989; Cosmides, and Tooby 1996; Gigerenzer 2000; Evans and Over 1996; Stanovich 1999), broaden their investigations to the wider cognitive context in which reasoning occurs. By focusing not only on the workings of reasoning, but also on its cognitive function, these theories are considered to be more complete than previous ones were intended to be. While differing substantially among themselves, evolutionary, ecological and dual-system theories all agree (at least implicitly) on the two following theses, which characterize what can be called “the standard approach to reasoning” in cognitive psychology: (i) reasoning is instrumental; (ii) reasoning is cognitive in nature. In what follows I will focus on the above-mentioned three theories, or more precisely families of theories, so as to understand how (i) and (ii) constrain the picture of reasoning emerging from them.

The Ordinariness of Reasoning in Arguing

233

1.1 Reasoning is Instrumental Claiming that reasoning is instrumental means that it is a cognitive tool for achieving one’s goals; that is, reasoning is goal-directed: ways of reasoning typical of the human mind are as they are because they have been efficient and effective in achieving specific goals that were relevant for our ancestors’ lives and continue to be relevant for us to this day. The fact that certain ways of reasoning are included in the human cognitive repertoire is due, then, to their effectiveness for achieving valuable goals (for example, promoting one’s well-being) that might be too difficult or even impossible to achieve with other cognitive tools. So, only ways of reasoning that have been conducive to achieving valuable goals are passed, genetically or culturally, from generation to generation: relying on ineffective ways of reasoning may indeed lead humans to extinction in the long run! To see this more concretely, consider what the above-mentioned three families of theories maintain about the kinds of goal at which reasoning is directed and how, they assume, it proceeds to achieve them: – evolutionary theories: reasoning as a cognitive tool has evolved on the basis of its effectiveness in increasing, directly or indirectly, the probability of survival and reproductive success of the human species. It occurs when certain mental modules among those composing the human mind are activated to solve restricted classes of problems for which they are well adapted. So what today is called “reasoning” is no more than a system of cognitive adaptations, each of which evolved in response to specific evolutionary problems posed by the social and physical environments which our ancestors were faced with in the Pleistocene era (see Cosmides 1989; Cosmides and Tooby 1992; 1996); – ecological theories: reasoning works well not only when applied to solve problems that were relevant in the evolutionary history of the human species, but also when it applies to situations that go beyond its original domain of application, provided that the structure of the problem to which it applies resembles that of its original domain of application. When this is the case, the most relevant of the heuristic procedures underlying reasoning will be activated, leading quickly and without wasting cognitive effort to good results. Our reasoning activity indeed depends on a suite of heuristic procedures that are characterized as “fast and frugal”; being fast and frugal, however, does not preclude them from leading reliably to good results when applied to situations whose structures mirror those for which they are well adapted (see Gigerenzer 2000); – dual-system theories: because of its complexity, the human mind should be seen as composed of two sub-systems, known as System 1 and System 2. Reasoning activity is the result of the cognitive mechanisms

234

Chapter Fifteen

belonging to these two systems: while the mechanisms of System 1 are dedicated to solving problems that are in need of quick and fast responses, those of System 2 are activated to solve problems whose complexity requires a higher level of cognitive processing. Accordingly, each system gives rise to different forms of reasoning that can be differentiated from one another on the basis of their effectiveness in solving specific classes of problems (see Evans and Over 1996; Stanovich 1999; Evans 2006). While relying on the same empirical data in supporting their claims, these psychological theories conflict as to the goals at which, they assume, reasoning is directed when applied to specific classes of problems. Depending on the goals one attributes to the subjects engaged in a reasoning problem, evaluations of their responses, and therefore evaluations of their reasoning abilities, can evidently vary: subjects’ performances that seem to be irrational can be judged as perfectly rational, on the condition that they are evaluated as aiming at a different goal from the one previously attributed. This happens, for example, with the selection task. Although it has been traditionally claimed that, when approaching this task, subjects should aim to falsify the conditional rule on the basis of a deductive understanding of it, only 10% of the subjects give the “logically correct” response (see Wason and Johnson-Laird 1972). However, insofar as we acknowledge the task’s inductive nature, and so attribute to the subjects a goal different from that usually attributed to them, their way of approaching it turns out to be the most effective way to solve it (see Oaksford and Chater 1994; Nickerson 1996). In any case, any attempt to provide an explanation of the subjects’ performances other than the one inspired by classical logic rests on the assumption that reasoning serves a specific goal or set of goals, be they evolutionary or personal ones, that is, that reasoning is to be explained in instrumental terms.

1.2 Reasoning is Cognitive in Nature None can deny that reasoning is a cognitive activity. Since, however, “cognitive activity” refers to the whole spectrum of conscious and unconscious cognitive processes, this assumption is ambiguous unless we can tell which of these two types of cognitive process prevails in reasoning. While reasoning has been traditionally considered to be a conscious cognitive activity, characterized as aimed at enhancing one’s knowledge and maximizing one’s personal utilities, according to the most recent psychological and neurological findings, much of its workings are accessible only in occasional glimpses, as held also by evolutionary, ecological and dual-system theories:

The Ordinariness of Reasoning in Arguing

235

– evolutionary theories: our cognitive systems consist of many mental modules which, among other things, determine (and so are responsible for) our reasoning and decision-making. Since mental modules are programmed to automatically produce responses to the social and physical environmental challenges our ancestors were frequently faced with in the Pleistocene era, our reasoning and decision-making are the results of automatic processes inherited from our evolutionary past, which, while very effective (at least in the environment of evolutionary adaptedness), are totally inaccessible to our awareness. That is, we are conscious of the outputs of the mental modules underlying reasoning and decision-making, but not of their inner workings (see Cosmides 1989; Cosmides and Tooby 1992; 1996); – ecological theories: fast and frugal heuristics operate according to the principle of least effort. In light of the paucity of cognitive resources, our cognitive system naturally “chooses” the inferential paths of least resistance, that is, those which provide the best trade-off between cognitive effort and desired outcomes. As in the case of mental modules, conclusions and decisions from these heuristics’ processes surface in consciousness without the reasoner’s having any awareness of the inferential paths that have led to them (see Gigerenzer 2000); – dual-system theories: the two sub-systems composing the human mind operate simultaneously. While in certain circumstances they work together, the mechanisms belonging to System 1 providing “rough” information (i.e., perceptual inputs) to the mechanisms of System 2, in other cases System 2 is activated to block the automatic responses produced by the mechanisms of System 1, replacing them with more complete and accurate responses. In any case, cognitive mechanisms belonging to System 1 always take priority over those of System 2 in the sense that, requiring relatively little cognitive effort, they are activated automatically upon encountering a relevant stimulus, while those of System 2 intervene only when specific conditions are met (e.g., the importance of the goal at which reasoning is directed, the motivation of the reasoner in dealing with certain reasoning problems, and so on) (see Evans and Over 1996; Stanovich 1999; Evans 2006). While, according to evolutionary and ecological theories, both mental modules and heuristics operate independently from the reasoner’s will, dual-system theories, characterizing System 2 as conscious and deliberate (see, e.g., Evans and Over 1996: 153-161), seem to leave open the possibility of intervention from the reasoner. However, they fail to provide a plausible explanation of how the reasoner may intervene in the ongoing process of reasoning, switching from one system to the other. Since,

236

Chapter Fifteen

according to dual-system theories, when starting to reason we naturally rely on the automatic cognitive processes of System 1, it is not clear how we can block them, as we have no conscious access to their operations. Moreover, since we have no awareness of how System 1 works, how can we come to know whether a particular conclusion or decision has been drawn correctly? It can be argued, therefore, that to a certain extent, all these theories converge on the idea that we have no control over how reasoning operates. Thus, reasoning appears to be a cognitive activity mainly because it consists of unconscious cognitive processes.

2. Against the Picture of Reasoning Emerging from the Standard Approach The picture of reasoning emerging from the above-described theories is that of a self-centered cognitive activity, unfolding outside the reasoner’s awareness, whose function is fundamentally instrumental. However, as I will argue in what follows, when people reason, they are conscious, at least to a certain extent, of what is happening in their minds. While it is true that most of the inferential processes occurring in our minds proceed unconsciously, there is significant evidence from experimental studies in developmental psychology that our ability to reason is strictly related to metacognitive control.

2.1 To what Extent is Reasoning Unconscious? If one looks at the experimental literature on reasoning, she will notice that almost all the experiments are carried out in non-natural settings, that is, in situations detached from the subjects’ ordinary lives. Some of the non-natural settings to which I am referring here are experimental labs, small cabins, university offices and so on, which are unusual, nonrepresentative situations for most people. However, cognitive psychologists maintain that this type of setting does not influence the “natural” course of reasoning. Since reasoning is considered to be an individualistic enterprise which is already located in its natural context, that is, the cognitive context of one’s mind, they assume that the instructions for reasoning tasks, if well delivered, are enough to activate the “right” inferential processes. According to this view, whether reasoning is located in an experimental setting or in an ordinary context does not influence its way of proceeding; reasoning needs only the appropriate stimuli to be activated, regardless of where it takes place. The automatic inferential processes underlying reasoning are activated by

The Ordinariness of Reasoning in Arguing

237

specific sets of inputs, that is, those to which they are well adapted, and are not subject to variation: what they do is to generate output from input in accordance with predetermined mechanical inferential paths. By contrast, many “pragmatic” experimental studies suggest that subjects’ solutions to reasoning tasks depend on their interpretations of these tasks (see, e.g., Schwartz 1996; Politzer and Macchi 2000). Regardless of whether the material is abstract or content-specific, subjects may interpret the task they are asked to undertake in many ways. In particular, different subjects’ understanding of the same reasoning task can vary depending on how they interpret the question and the information available in the instructions. As a consequence, subjects frequently understand the reasoning tasks they are faced with in ways that differ significantly from what the experimenter assumes to be the case, giving responses that in no way conform to the expected solutions. But if people can give different responses to the same reasoning problem, depending in part on their past experience and knowledge, it is questionable to assume that reasoning works as if in response to the same stimuli and always follows the same inferential path. Rather, there seems to be an active involvement on the part of the subjects when faced with reasoning problems, which suggests that they have some control over their own ways of reasoning.

2.2 From Cognitive Reaction to Metacognitive Control In opposition with the pragmatic approach to reasoning studies, some supporters of the theories described above hold that differences between subjects’ responses and experimenters’ expected solutions are not due to the former’s understanding of the reasoning task but depend on the latter’s misidentification of what the “real” function of reasoning is. For example, in explaining why subjects only solve certain variants of the selection task correctly, evolutionary psychologists have made reference to the proper function of a specific form of reasoning: according to them, the correct responses given by the subjects are due to a specific module dedicated to detecting cheaters, which activates under specific conditions (see Cosmides 1989; Cosmides and Tooby 1992). So, based on this interpretation, subjects solve the task correctly because of the cheaterdetection module’s activation, and not on the basis of their way of approaching it. But if reasoning works like this, what kind of goal-directed activity is it? Since, according to the above-described theories, as reasoning people are conscious of the conclusions, but not of the inferential paths that have

238

Chapter Fifteen

led to them, reasoning can be represented as a cognitive reaction, something which happens in our minds subconsciously most of the time, people usually come to conclusions or make decisions on the basis of what Gerd Gigerenzer (2007) calls “gut feelings”; emotions, instincts and intuitions. If reasoning is to be taken as nothing more than a cognitive reaction, the picture of reasoning emerging from psychological theories suggests that, when reasoning, we do not actually aim at goals, because the cognitive processes underlying it proceed independently from our intention or will. But can we reduce reasoning to an unreflective activity of thinking, on a par with emotions, instincts, intuitions? The answer is NO. Our minds consist not only of automatic processes, but also of stronger processes of self-reflection capable of coordinating one’s inferences (see Moshman 2004). As explained by the developmental psychologist Deanna Kuhn (2000: 320), this metacognitive ability is related to “the achievement of increasing awareness, understanding, and control of one’s own cognitive functions, as well as awareness and understanding of these functions as they occur in others”. While the inferential processes performed by one’s mind carry out specific functions, reasoning serves the metacognitive function of coordinating these inferential processes to achieve specific purposes, about which I will say more in the coming sections. To the extent that reasoning carries out this metacognitive function, it cannot be equated to a mere cognitive reaction, nor be represented as proceeding completely outside the reasoner’s consciousness. As suggested by David Moshman (1994: 245), to be a creature of reason means that […] one’s thoughts and actions are not entirely random, arbitrary, reflexive, or conditioned. To the extent that reason rules, one’s beliefs and behavior are not innately fixed, nor are they completely molded by experience, nor are they mechanically caused by some interaction of genetic and environmental factors.

3. Beyond the Standard Approach: The Argumentative Theory of Reasoning If developmental psychologists such as David Moshman and Deanna Kuhn are right, reasoning cannot be exhaustively described or explained away by reference to unconscious, automatic cognitive processes. In order to provide a more comprehensive understanding of it, one has then to look beyond this alleged identification. To this end, a very useful starting point comes from the editors of a recent special issue of Thinking & Reasoning that deals with the relationship between reasoning and argumentation: in

The Ordinariness of Reasoning in Arguing

239

presenting the special issue, Jos Hornikx and Ulrike Hahn (2012: 229) point out that there is ample evidence […] that attempts to understand our ‘reasoning’ ability—that is, our ability to evaluate individual premise–conclusion connections—must take into account that, in our everyday lives, such reasoning is typically embedded in broader argumentative contexts […].

According to them, reasoning is typically done interpersonally rather than privately, taking place in “broader argumentative” contexts where people interact with each other, and cannot be detached from such contexts. This is consistent with the following remark by Moshman (2004: 233): Reflection and coordination [...] often take place in the context of social interaction, and especially peer interaction. In social contexts we may find ourselves challenged to justify our conclusions, and thus to recognise and justify our inferences. We may also be challenged to understand the inferential paths that led others to alternative views, and to coordinate those inferences and conclusions with ours.

We can do so thanks to our ability to reason, which allows us to make explicit and evaluate individual premise–conclusion connections. However, most psychological studies have focused on the responses subjects give when faced with reasoning tasks, irrespective of the reasons on the basis of which they give their responses. So, empirical data collected in these experimental studies seem to be only partially relevant to understanding how reasoning actually works in what Hornikx and Hahn call “broader argumentative contexts”. A theory of reasoning which takes seriously into account the link between our ability to reason and its interpersonal function is Hugo Mercier and Dan Sperber’s argumentative theory of reasoning (see Mercier and Sperber 2011a). On the basis of an evolutionary model of human cognition, Mercier and Sperber (2011a: 59-61) argue that reasoning has evolved so that we can achieve two different but complementary goals when communicating with others: (a) to judge others’ arguments and (b) to provide effective arguments in support of our claims. These complementary activities are carried out through two different types of inferences, which Mercier and Sperber (2011a: 58-59) call “intuitive inferences” and “reflective inferences” respectively. Intuitive inferences are the result of those unconscious cognitive processes that the theories discussed above identify as “reasoning” (heuristic procedures, modules dedicated to solving specific tasks etc.): what characterizes these

240

Chapter Fifteen

inferences is the reasoner’s awareness of their outputs, but not of the inferential paths from which they proceed. According to Mercier and Sperber, reflective inferences amount instead to “reasoning proper”: what characterizes these inferences is “the awareness not just of a conclusion but of an argument that justifies accepting that conclusion” (Mercier and Sperber 2011a: 58). From this point of view, reasoning has evolved as it is today just because it facilitated our ancestors in arguing for their claims and assessing each other’s arguments (Mercier and Sperber 2011a: 71-72). But if reasoning operates this way, people should reason better when engaged in argumentative contexts than when they are alone or with an experimenter in experimental labs. The data Mercier and Sperber (2011a: 61-63) have collected, which concern the results of past and more recent experimental studies on reasoning and argumentation, suggest, indeed, that people reason better in situations where they are part of groups which reason, discuss or argue collectively. Reasoning tasks in which subjects taken in isolation tend to give wrong answers are more frequently solved correctly when they are approached in groups and discussed collectively. Consider the following two experimental cases. Although it is empirically demonstrated that people find it very difficult to recognize the modus tollens (if p then q, not-q, so not-p), when engaged in argumentative dialogues, in order to criticize the claims of their opponents, they recognize and easily apply this argumentative form (Thompson, Evans and Handley 2005). If they discuss how to solve it with others, even subjects involved in the selection task for the first time give the correct response in about 70% of cases (Moshman and Geil 1998).

4. Argumentative Rationality The argumentative theory, however, rests on a picture of reasoning that is not too different from that fostered by the standard approach, as emerging from evolutionary, ecological and dual-system theories of reasoning. Insofar as reasoning is considered to be not so much a real monitoring system over one’s inferences as an “argumentative” device serving the twofold goal of prompting others to change their minds by giving them reasons to do so and defending oneself from others’ attempts to do the same (see Mercier and Sperber 2011b: 95), the argumentative theory attributes to reasoning a basically instrumental function. As to the thesis that reasoning is cognitive in nature, the position expressed by Mercier and Sperber is unclear about the role of conscious vs. unconscious cognitive processes in reasoning. Although Mercier and Sperber emphasize the role of reflective inferences, they hold that these inferences

The Ordinariness of Reasoning in Arguing

241

are so strongly determined by intuitive inferences that when reasoning people have “[...] intuitive judgements that given conclusions follow from given premises” (Mercier and Sperber 2011a: 59), which they accept spontaneously, without calling their validity into question. In other words, when people are conscious of the argument supporting the conclusion, they still have little conscious access to the inferential path that has led to the representation of the relationship between reasons and conclusions; they only perceive that some arguments are stronger than others. As a consequence, Mercier and Sperber’s argumentative theory does not differ from the above-described theories of reasoning in taking it to be determined by unconscious cognitive processes: reflective inferences amount to becoming aware of the outputs of intuitive inferences that deliver intuitions about the relationship between reasons and conclusions and its strength (see Mercier and Sperber 2011a: 59). While distancing myself somewhat from Sperber and Mercier’s argumentative theory, I accept their distinction between intuitive and reflective inferences, characterizing it in a different way. It is a fact that there are plenty of automatic and routinized inferential processes, that is, intuitive inferences, that give rise to intuitive forms of thinking; however, none of them can be considered full-fledged reasoning. These forms of thinking may serve reasoning, but they are not, properly speaking, “reasoning”. Automatic inferential processes exhibit a specific form of rationality that is necessarily instrumental: on the basis of the few cognitive resources at one’s disposal, cognitive processes that lead to good solutions for many types of problems without expending too much time or effort take precedence over those which instead require a high amount of cognitive resources and time to be applied. People possess this type of rationality unconsciously: we can be called rational, in this sense, insofar as our cognitive systems aim to achieve the best trade-off between mental effort and processing efficiency, but not because we, as individuals, engage in thinking and acting rationally. When we think or act rationally, we are instead aware of and responsible for what we say and do. What is at the core of this kind of rationality is our providing reasons and/or arguments in support of what we say and do; to reason properly is not to accept conclusions and decisions proceeding from intuitive inferences uncritically, although sometimes it may be more effective to act this way, but to be able to provide good reasons for supporting our conclusions and decisions. In opposition with the instrumental conception supported by the argumentative theory, I maintain that reasoning occurring in argumentative contexts exhibits a different form of rationality. We do not

242

Chapter Fifteen

always justify what we say or do because we want to gather new information or to influence our interlocutors, as the argumentative theory holds; most of the time we engage in argumentative interactions without necessarily aiming to achieve personal profits. In such cases, we are still said to be rational, although we cannot justify our behavior in instrumental terms. At this point, we can rely on the notion of argumentative rationality which Marina Sbisà (2006; 2007), elaborating the ideas and reflections of Paul Grice (1991; 2001) on human rationality, proposed. As Sbisà describes it, using Grice’s words, “rationality is a concern that one’s moves are justified and a capacity (to some degree) to give effect to that concern” (Sbisà 2006: 241-242; see Grice 1991: 82-83). I believe that the framework I am trying to develop here could suit such a definition of rationality. In my view, this definition states two conditions for human rationality: to be acknowledged as rational one must (i) care about justifying her moves, be they linguistic or of a different nature, and (ii) give effect to her concern by providing reasons and/or arguments supporting these moves. While condition (i) is concerned with one’s intention or motivation to justify what she says and does, condition (ii) refers to her ability to do so. Condition (i) prevents us from considering unreflective systems, such as computers, which do not care their own moves to be justified, rational. Condition (ii) refers instead to our ability to make explicit and evaluate individual premise–conclusion connections, thanks to which, as suggested in 2.2, we coordinate the great variety of automatic inferential processes that our cognitive systems continuously perform to give justificatory support to what we say and do. According to David Moshman (2004: 227), “awareness of inference allows one to consider the possibility that some inferences are better than others, in the epistemic sense that the conclusions they generate are more justifiable”. As metacognitive reasoners, we can be held responsible for most of our reasoning activity. In particular, we engage in two qualitatively different ways of reasoning, which can be called forward and backward reasoning respectively. The first way in which we can give effect to our concern for justifying our moves is by reasoning forward. Forward reasoning, in accordance with the received way of conceiving of reasoning, amounts to a conscious mental activity which aims at drawing conclusions, decisions and judgments from a set of initial information, premises or assumptions, and consists of several steps, of which some may be performed unconsciously while others consciously. Since when reasoning forward we move from explicit premises or assumptions to a conclusion or decision, even if we

The Ordinariness of Reasoning in Arguing

243

are not conscious of all the steps needed to complete the argument, both conditions (i) and (ii) can be said to be respected. As to the second way, sometimes we accept conclusions and decisions proceeding from intuitive inferences uncritically. As Mercier and Sperber (2011a: 59) hold, we have a natural inclination to accept the output of intuitive inferences: we rarely question whether they are right or not, since we consider what comes from our cognitive system perfectly reliable. So, what we usually perceive as well-grounded conclusions are nothing but the result of automatic cognitive processes, i.e. they are intuitive conclusions. But when called upon to justify these intuitions, we may want to construct post-hoc justifications for them. In such cases, we engage in ex-post facto reasoning aimed at supporting our intuitive conclusions. As to the two conditions for rationality I have proposed above, when we reason backward they are both respected, because, although we accept being guided by our intuitions, when needed we engage ourselves in reasoning with the aim of justifying our acceptance of our unreflective pieces of thinking. In doing so, we often mislead ourselves by thinking that our intuitive conclusions are the products of conscious reasoning, while they are not. Indeed, the connection that we establish between these conclusions and the reasons and/or arguments which we suppose justify them may reflect the actual inferential path that has led to them, but most of the time is likely to be completely arbitrary.

5. Concluding Remarks I began by presenting and discussing the picture of reasoning emerging from today’s most prominent psychological theories of reasoning, such as evolutionary, ecological and dual-system accounts. As to this picture, I have questioned its way of representing reasoning as a self-centered cognitive activity, unfolding outside the reasoner’s awareness, whose function is fundamentally instrumental. I then turned to developmental psychological studies, particularly those of David Moshman, which suggest that reasoning is strictly related to metacognitive control. Conceived in this way, reasoning consists of coordinating the great variety of automatic inferential processes that our cognitive systems continuously perform to make explicit reasons and/or arguments supporting our conclusions and decisions. A similar picture of reasoning is supported by Hugo Mercier and Dan Sperber’s argumentative theory, according to which reasoning has evolved so as to evaluate others’ arguments and to give arguments in support of one’s own contentions. However, since according to them reasoning is more of an “argumentative” device, aimed

244

Chapter Fifteen

at changing others’ minds by providing reasons to do so, than a real monitoring system for one’s inferences, the argumentative theory relies on an instrumental conception of rationality, after all, and so is not too different from the evolutionary, ecological and dual-system theories. By distancing myself from the instrumental conception of rationality, which is assumed by the standard approach to reasoning, I have argued for a different conception of it, which involves one’s awareness of and responsibility for her reasoning performances. This way of conceiving rationality fits well with Grice’s argumentative rationality, as recently presented by Marina Sbisà. According to Grice, what characterizes human rationality is our deep concern for justifying what we do and say, supported by an ability to give effect to that concern, that is, an ability to make explicit individual premise–conclusion connections. If our rationality can be described in this way, it could be interesting to study reasoning in the context of social interaction where people are asked to provide reasons and/or arguments in support of what they say and do. While thanks to the standard approach we have discovered a lot of important things about the cognitive processes involved in reasoning, the difficulties we have in keeping them under control and so on, a further step in reasoning studies could be that of taking into account how reasoning is performed in natural interactional situations where, unlike in experimental labs, people take care to justify what they say and do, having in front of them “real” interlocutors.

References Cosmides, Leda. 1989. “The Logic of Social Exchange: Has Natural Selection Shaped How Humans Reason? Studies with Wason Selection Task.” Cognition 31: 187-276. Cosmides, Leda, and John Tooby. 1992. “Cognitive Adaptations for Social Exchange.” In The Adapted Mind, ed. John Barkow, Leda Cosmides and John Tooby, 163-228. New York: Oxford University Press. Cosmides, Leda, and John Tooby. 1996. “Are Humans Good Intuitive Statisticians after all? Rethinking Some Conclusions from the Literature on Judgment under Uncertainty.” Cognition 58: 1-73. Evans, Jonathan St.B.T. 2006. “The Heuristic-Analytic Theory of Reasoning: Extension and Evaluation.” Psychology Bulletin Review 13: 378-395. Evans, Jonathan St.B.T, and David E. Over. 1996. Rationality and Reasoning. East Sussex, UK: Psychology Press.

The Ordinariness of Reasoning in Arguing

245

Gigerenzer, Gerd. 2000. Adaptive Thinking. Rationality in the Real World. New York: Oxford University Press. —. 2007. Gut Feelings: The Intelligence of the Unconscious. New York: Viking Press. Grice, Paul H. 1991. The Conception of Value. Oxford: Oxford University Press. —. 2001. Aspects of Reason. Oxford: Oxford University Press. Hornikx, Jos, and Ulrike Hahn. 2012. “Reasoning and Argumentation: Towards an Integrated Psychology of Argumentation.” Thinking & Reasoning 18: 225-243. Kahneman, Daniel, Paul Slovic, and Amos Tversky (eds). 1982. Judgment under Uncertainty: Heuristics and Biases. New York: Cambridge University Press. Kuhn, Deanna. 2000. “Theory of Mind, Metacognition, and Reasoning: A Life-span Perspective.” In Children’s Reasoning and the Mind, ed. by Peter Mitchell and Kevin Riggs, 301-326. Hove, UK: Psychology Press. Johnson-Laird, Philip N., and Ruth M.J. Byrne. 1991. Deduction. Hillsdale, NJ: Erlbaum. Mercier, Hugo, and Dan Sperber. 2011a. “Why Do Humans Reason? Arguments for an Argumentative Theory.” Behavioral and Brain Sciences 34: 57-74. Mercier, Hugo, and Dan Sperber. 2011b. “Argumentation: Its Adaptiveness and Efficacy.” Behavioral and Brain Sciences 34: 94111. Moshman, David. 1994. “Reason, Reasons, and Reasoning: A Constructivist Account of Human Rationality.” Theory & Psychology 4: 245-260. Moshman, David. 2004. “From Inference to Reasoning: The Construction of Rationality.” Thinking & Reasoning 10: 221-239. Moshman, David, and Molly Geil. 1998. “Collaborative reasoning: Evidence for Collective Rationality.” Thinking & Reasoning 4: 231– 248. Nickerson, Raymond S. 1996. “Hempel’s Paradox and Wason’s Selection Task. Logical and Psychological Puzzles of Confirmation.” Thinking & Reasoning 2: 1-31. Oaksford, Michael, and Nick Chater. 1994. “A Rational Analysis of the Selection Task as Optimal Data Selection.” Psychological Review 101: 608-631. Politzer, Guy, and Laura Macchi. 2000. “Reasoning and Pragmatics.” Mind and Society 1: 73-93.

246

Chapter Fifteen

Rips, Lance J. 1994. The Psychology of Proof. Cambridge, MA: MIT Press. Schwarz, Norbert. 1996. Cognition and Communication: Judgmental Biases, Research Methods and the Logic of Conversation. Hillsdale, NJ: Erlbaum. Sbisà, Marina. 2006. “Two Conceptions of Rationality in Grice’s Theory of Implicature.” In Rationality of Belief and Action, ed. by Elvio Baccarini and Snježana Prijiü-Samaržija, 233-247. Rijeka: University of Rijeka. Sbisà, Marina. 2007. “On Argumentative Rationality.” Anthropology & Philosophy 8: 89-99. Stanovich, Keith E. 1999. Who is Rational? Studies of Individual Differences in Reasoning. Mahwah, NJ: Erlbaum. Thompson, Valerie A., Jonathan St.B.T. Evans, and Simon J. Handley. 2005. “Persuading and Dissuading by Conditional Argument.” Journal of Memory and Language 53: 238–57. Tversky, Amos, and Daniel Kahneman. 1974. “Judgment under Uncertainty: Heuristics and Biases.” Science 185: 1124-1131. Wason, Peter C. 1966. “Reasoning.” In New Horizons in Psychology, ed. by Brian M. Foss, Harmondsworth, UK: Penguin. Wason, Peter C., Philip N. Johnson-Laird. 1972. The Psychology of Reasoning: Structure and Content. Cambridge, MA: Harvard University Press.

CHAPTER SIXTEEN REGIMENTING THE ORDINARY NOTIONS OF KNOWLEDGE AND JUSTIFICATION AFTER GETTIER MARIO ALAI

1. Introduction Ordinary knowledge about given subjects often consists of clusters of strong, widely shared (but not always consistent), particular intuitions or judgments. Philosophy typically tries to subsume those particular claims under a coherent concept, definition, or general principle. If necessary, to this end, it rejects as few as possible of the weaker and more peripheral intuitions. Once the right concept or principle is found, however, philosophers wonder if it is useful, well-grounded or otherwise acceptable, and if not, how it might be amended. Of course, ordinary knowledge can also have its own concepts, definitions or principles from the start, but they may be vague or uncertain, so philosophy not only works bottom-up from particular to general ideas or claims, but also top-down, starting from concepts and principles and checking them vis-à-vis particular intuitions. Goodman pointed out that often one of these processes launches the other, in a (hopefully) virtuous spiral, until some sort of reflective equilibrium is reached.1 Thus philosophy analyses, evaluates, regiments and possibly reforms ordinary knowledge. This is what Plato, Gettier and many recent authors have done with the ordinary notions of knowledge and justification, bringing out the relevant intuitions with a wealth of examples and counterexamples. In Theaetetus, Plato argues that knowledge is (i) a belief which is (ii) true and (iii) “with reasons”, or as we now say, justified (163a-201d). It 1

Goodman 1955, ch. III § 2.

Chapter Sixteen

248

must be a belief, not merely a sensation, since examples show that one can have sensations without knowing anything; besides, sensations are neither true nor false, while knowledge must be true. Finally, knowledge must be justified, for it is supposed to be a safe and sufficiently certain grasp of the truth, grounded on the facts rather than on accidental and unrelated factors (200d-201d). But Plato regrets that he fails to complete his analysis with an adequate, non-empty and non-circular definition of justification (206c210b). Edmund Gettier’s famous paper “Is Justified True Belief Knowledge?” showed that these basic problems are still open. As the title says, Gettier thought his work questioned Plato’s tripartite definition, but as we shall see, it can also be read as a challenge to complete Plato’s work, by explaining the nature of justification. As it soon became clear, Gettier’s two original examples did not bring out all the difficulties of that complex problem: in fact, their paradoxical features could be accounted for by partial and contingent solutions, which however, did not solve the basic difficulties,2 brought out by a wealth of thought experiments in the subsequent literature. Therefore I shall now modify one of his examples, and discuss two more from later works, none of which are liable to quick solutions.

2. Challenging the Definition of Knowledge or of Justification? Example I I am aware that (a) since I know him, Jones has always owned a Ford, and (b) just now he is giving me a ride in that car. However, unknown to me, (c) one hour ago he sold it; in fact he is taking it to its new owner. Moreover, unknown to me and to himself, (d) half an hour ago he won another Ford in a lottery. Therefore I believe that (1) Jones owns a Ford, and this belief is true. From Gettier’s point of view, it is also justified, since I have reasons (a) and (b) for it. Nonetheless, as explained in Theaetetus, one would not say it is knowledge, but merely a casually true 2

See Alai 2005, 103.

Regimenting the Ordinary Notions of Knowledge

249

belief: for all I actually know, (1) might well be false, and that it is true is purely accidental, due to facts which have nothing to do with reasons (a) and (b), on which my belief is based. Hence, according to Gettier, this shows that being (i) a belief, (ii) true, and (iii) justified, is not enough for something to count as knowledge, and a fourth condition is needed. But what might the missing condition be? Since justification should make safe and non-accidental the truth of our beliefs, if our ordinary notion of justification cannot accomplish this task, should we not probe our intuitions and look for a better concept of justification, so accomplishing the task Plato was unable to perform? And if we found a condition X warranting the non-accidental truth of our beliefs, should this not be called a justification (or part thereof)? Thus, unlike Gettier, we might claim that Plato’s tripartite definition is correct, but belief (1) is not knowledge, since the pieces of evidence (a) and (b) are not sufficient justification for it: in fact, they cannot make it sufficiently safe, for they are compatible with its falsity. We might then require stronger reasons to justify our beliefs, but it is easy to realise that for any set of reasons which do not make the truth completely certain, one can think of cases in which a belief so justified is true but only accidentally so. On the other hand, since we are not infallible, we can never have enough evidence to make our beliefs completely certain. So, if this was a requirement for a justification, there could be no justifications, hence no knowledge at all. But this is absurd. Thus we are right back to Plato’s unsolved question: what exactly is justification? And how can it be fallible, yet ensure that our beliefs are well founded and not just casually true?

3. The Objective and External Character of Justification In the Meno (97d-98b), Plato likens justifications to strong bonds: supposedly, Daedalus’ statues were so realistic that they would move and run away from their owners, unless they were tightly fastened to the ground. Equally, says Plato, true opinions must be fastened by strong ties. He conceives such ties as reasonings, but perhaps beliefs can be objectively secured to their grounds even by bonds which escape the subject’s awareness. For Alvin Goldman’s causalism, for instance, what matters is that a belief is actually brought about by the fact which makes it true (its truthmaker) through a tight chain of causes, although the latter is partly unknown to the subject (Goldman 1967; Swain 1998). For example, my belief that there is an apple on the table is justified, hence it is knowledge, because it is the ultimate effect of a chain of causes initiated

250

Chapter Sixteen

by the apple, roughly consisting of light radiations, retinal stimulations, nervous impulses, neurons firing, and my perception of the apple). That a belief is thus caused by its truthmaker implies that it is true; on the other hand, the subject can presume the existence of such a chain (e.g., the apple, the light waves, the retinal stimulations, etc.), but never be absolutely certain of it. Hence, while the chain objectively warrants the truth, preventing fortuitous truths like (1), the evidence directly available to the subject cannot make her subjectively certain: fallibilism is thus compatible with objectively certain justifications. Causalism, therefore, conceives justification as objective and external to the subject, not entirely transparent to her. Other externalist approaches are reliabilism3 and functionalism (Plantinga 1993), which we cannot discuss here. According to these views, a belief is justified when it is brought about by cognitive processes or faculties which are reliable, i.e. whose function is producing true beliefs. Externalism has not been adopted only by epistemologists,4 but also by philosophers of language and of mind, who have stressed that the meanings and references, as well as the content, of thoughts and perceptions are not determined exclusively by internal mental states, but crucially depend also on the external context.5 This is how causalism solves the problem of Example I: my belief that (1) Jones owns a Ford is not knowledge because it is not causally grounded in its truthmaker. It is brought about by the facts that (a) since I know him, Jones has always owned a Ford, and (b) just now he is giving me a ride in that car. But these facts have nothing to do with what makes (1) true, viz., the lucky coincidence that (d) half an hour ago Jones won another Ford in a lottery. Belief (1) is not justified because its truthmaker (d) has no role in causing it.

3

See Goldman 1986, Talbott 1990, Papineau 1993. See BonJour 1980, 2002. 5 See for instance Kripke 1980; Putnam 1975; Davidson 1986; Dretske 1993. 4

Regimenting the Ordinary Notions of Knowledge

251

4. Productive Chains Various objections have been raised against causalism, which is now less popular or even abandoned altogether.6 But I think at least its basic intuition can be saved and usefully employed in understanding the nature of justification. One objection is that beliefs about mathematical and other non-physical facts cannot be caused by their truthmakers, for those facts cannot cause anything. Even beliefs about the future cannot be caused by their truthmakers, which have not happened yet. A more radical objection is that even if there are chains from facts to beliefs about them, usually the last links are non-physical factors, like perceptions or further beliefs, and they produce the beliefs by non-causal processes, like inferences. Sure enough, we do not experience our beliefs as produced by causes, but by reasons. 7 These problems, however, can be overcome by generalising the notion of the causal chain: in order to be objectively justified, we can simply require a belief to be brought about by its truthmaker through a productive chain, whose links can also be non-causal. For instance, in the simplest cases, those of direct perceptual beliefs, we can simply require a chain like the one just mentioned, in which an apple causes light radiations, which cause retinal stimulations, which cause nervous impulses and neuronal activities, which bring about the perception of an apple, which brings about belief: there is no need to decide whether perceptions and beliefs are physical states or not, or whether they are produced causally or not. In the more complex case of non-perceptual beliefs, we will have chains like this: a fire causes smoke, which causes stimulations of sense organs, which cause neural impulses, which produce sensations and (perhaps non-causally) the perception of smoke. The latter produces the belief that there is smoke, which in turn inferentially (so perhaps noncausally) brings about the belief that there is fire. Next, if it is granted that there are mathematical facts, that they are non-physical, and that they can be known, then wholly non-causal chains can be enough: for instance, a mathematical fact brings about an intellectual intuition, which produces one or more beliefs. For beliefs about the future, it is sufficient that there is a past fact A which produces both the future truthmaker B and the belief that A is the case, from which one forms the belief that B is the case by inference from A to B. For instance, the belief that it will rain is objectively justified by a 6 7

See Swain 1998, § 3. More difficulties, with possible solutions, are mentioned in Swain 1998, 3.

252

Chapter Sixteen

chain beginning with big black clouds, which on one hand eventually cause the rain, and on the other hand emit light waves, which, as usual, cause related neuronal activities, which produce perceptions of clouds, which bring about the belief that there are big black clouds. This belief, together with the background belief that clouds cause rain, inferentially produces the belief that it will rain. This final belief is justified because its truthmaker, even if as of yet non-existent, will eventually be produced (in fact, caused) by the same fact which produces it.8

5. Normal productive chains But this approach runs into troubles in cases like the following: Example II Immediately after landing at Tokyo airport, at a newsstand, Steve sees a newspaper with the headline: “Earthquake hit Japan yesterday.” So, he comes to believe that (2) there was an earthquake in Japan yesterday. Now, there actually was an earthquake, so (2) is true. However, that newspaper happens to be an old one, referring to an earthquake that happened 10 years ago. It has been forgotten on some high shelf for all this time, and yesterday’s earthquake caused it to fall down to where Steve noticed it.9 So, not only is belief (2) true, but it is also brought about by a chain originating in its truthmaker. Yet, we could hardly consider it justified, for its correctness is due to aleatory coincidences. It would seem, therefore, that productive chains cannot provide adequate justification (or, if they can, truth and justification are not sufficient conditions for knowledge, as Gettier thinks). It might be replied that the chain in Example II cannot justify belief (2) because it is not of the right kind, but somehow deviant. But what is a chain of the right kind, and how can it be distinguished from one of the wrong kind? Any event is connected to thousands of others by a tangled web of multiple threads, so that a belief might be connected to its 8

See Goldman 1967, 362-365. Reported in University of Reading: http://www.reading.ac.uk/AcaDepts/ld/Philos/jmp/Theory%20of%20Knowledge/G ettier&CausalTheory.htm

9

Regimenting the Ordinary Notions of Knowledge

253

truthmaker in many ways, all accidental and equally irrelevant to our intuitive idea of justification. This problem also besieges causalism in philosophy, of language and of mind.10 My proposal is that a productive chain is of the right kind when it is normal, i.e., when each link is of a type which is normally directly or indirectly produced only by facts of the same type as the truthmaker.11 In other words, the chain must include no connections which are merely accidental, only links which can normally be assumed to obtain. This codifies the intuition that a belief should derive from its truthmaker in a non-aleatory way, for if an event that in this particular circumstance eventually produced a true belief normally would not do so, the belief, although it is now true, would be false in most similar circumstances; hence, the subject does not have a secure grasp of the truth. For instance, the belief that there is an apple on the table is justified by the perception of an apple on the table, because such a perception is usually brought about only by apples on the table. If such a perception were also normally produced by different facts (like illusions, magic tricks, or recurrent hallucinations), it would no longer be a sufficient justification. Now, belief (2) in Example II is not justified, although produced by a chain initiated by its truthmaker, because some links in the chain are not normally produced by the truthmaker; so, from the point of view of the evidence actually available to Steve, (2) is true only by a lucky coincidence. In fact, the chain is twice abnormal: it was utterly abnormal and fortuitous that yesterday’s earthquake brought to Steve’s attention a 10 year old newspaper; and it was utterly abnormal and fortuitous that a 10 year old newspaper produced a belief in an event of the previous day. Hence, although the belief was ultimately brought about by its truthmaker, the chain has, so to speak, first accidentally deviated from its normal course, and then accidentally resumed it. The chain is thus too aleatory to provide a justification. One might worry that the notion of normality, which plays a key role in this characterisation of justification, is context-relative: what is normal in one context might not be in another. For instance, at a museum of daily newspapers, copies from 10 years ago might normally be displayed at museum stands. But we can consider a chain to provide justification if it is normal in the particular context in which the productive chain takes place.12 10

See, for instance, Putnam 1981, 51, 65-66. Alai 2001, 498 ff. 12 Granted, a chain is always located at many concentrical contexts at once: what is normal in my neighbourhood might not be normal in my town, but normal again in 11

254

Chapter Sixteen

Normality is also vague and gradual; there is no sharp distinction between being normal and being abnormal, and a connection may be more or less normal. So, whether we consider a chain sufficiently normal to provide justification, hence knowledge, may depend on the interests at stake (either for the subject, in deciding whether to believe or not, or for us, in deciding whether to attribute to her knowledge or not). In this sense, justification and knowledge are relative to the context, as argued by contextualists:13 when deciding whether a belief is sufficiently justified to count as knowledge, we may be more or less demanding, depending on whether the subject is engaged in a coffeehouse discussion or in scientific research, in a court case or in a medical diagnosis, etc. If I ask a passerby directions for a street address, her belief that the right direction is North may be sufficiently justified from the point of view of her desire to be kind and helpful, but not equally justified form my point of view, if by taking a wrong route I risk missing an extremely important appointment. But this vagueness, graduality and relativity is not a problem for my approach; quite the opposite, the notions of justification and knowledge do not come from an exact science, but from ordinary knowledge, where they prove extremely useful, no doubt also thanks to their flexibility and adaptability to the context. Preserving these characteristics is thus an advantage for a philosophical explication of these notions.

6. The Subjective and Internal Character of Justification However, against causalism, and more generally against the objectivist and externalist approaches, stand some strong intuitions, which are at the basis of various internalist conceptions, such as deontologism14 and evidentialism.15 According to these intuitions, a belief is justified if and only if the subject has an epistemic right to hold it, by virtue of her reasons; that is, if and only if she possesses evidence which directly or through suitable inferences makes it certain or sufficiently probable.16 So, if her belief were later shown to be false, the subject’s good reasons would be a good excuse for her mistake, i.e., a justification in the etymological sense of proving her epistemic justice, or innocence. In other words, this is the Cartesian idea that justification is an internal epistemic state: in fact, if my region, yet not normal in my country, etc. But what determines normality, here, is the narrowest of these contexts. 13 See DeRose 1999, Lewis 1996, Cohen 1999, Vassallo 2003,. 102-121. 14 Steup (2010), § 2.1, Pappas (2005) § 5. 15 Steup (2008), § 4, Steup (2010), § 2.2. 16 See BonJour 2002, Pappas 2005.

Regimenting the Ordinary Notions of Knowledge

255

the reasons for her belief were not directly accessible to the subject, how could she be justified in holding it, and should she not withdraw her judgment? Actually, the previous examples show that an external objective justification is necessary to knowledge, not that it is also sufficient. They actually suggest it is not, for as we saw, our beliefs are normally produced by reasons of which we are introspectively aware, such as perceptions, perhaps even intellectual intuitions, and inferences from further beliefs. These are normally the last links in the productive chains; hence, a belief would not be produced by a normal chain, so not even objectively justified, unless it were immediately grounded on some internal evidence. So, justification also requires internally available reasons. Moreover, consider the following: Example III Let us suppose that clairvoyance is possible, and Helen is reliably clairvoyant. One day she “sees” that Barack Obama is in New York, where he actually is. Therefore Helen has the true belief that (3) Barack Obama is in New York. Moreover, this belief is produced by a normal productive chain, since saying that she is reliably clairvoyant is saying that she normally “sees” that P is the case only when P is the case. Helen, however, does not know about her own clairvoyance; in fact, like most of us, she believes that clairvoyance is impossible. Moreover she has strong empirical evidence that Obama is in Washington (she heard it on the news, etc). Therefore, we would not say that Helen’s belief (3) is justified, in spite of having an objective justification. If one were to think that Example III is invalidated by the assumption of an impossible hypothesis, one might imagine the unlikely but in principle possible case of an eyewitness, against whom a criminal organisation sets up a conspiracy in order to suggest that nothing of what she saw could possibly have happened, and that she was a victim of hallucinations, drugs, etc.17 Now, if this conspiracy is well arranged, the witness herself may eventually come to doubt her perceptual beliefs. This, of course, is an instance of the more general principle that occasionally, even a direct perceptual belief can be rejected if it conflicts with a number of other empirical or theoretical beliefs. So, in a case like this, the 17

A similar case is staged in Hitchcock’s movie North by Northwest.

256

Chapter Sixteen

eyewitness’s doubts about her perceptions could be entirely rational: we do not have the epistemic right to believe unless (a) our pieces of evidence are sufficiently strong, hence non-contradictory, and (b) we believe that they are the last links of a chain started by the truthmaker. Now, both these conditions are lacking in Example III, for not only is Helen’s clairvoyant perception contradicted by what she hears on the news, etc., but she does not believe that it is produced by the truthmaker (for she does not believe in her own clairvoyance). Internalism, then, is partly right: internal reasons are necessary to knowledge, although they are not sufficient, as shown by Examples I and II, where the subject does not know, in spite of having good internal reasons. This is because, due to human fallibility, no internal reason can possibly prevent our beliefs from being either false, or only accidentally true.

7. Normality and Right to Believe Thus, both the objective-external and the subjective-internal condition are required for justification, or at least knowledge; besides, each of them presupposes and is partly explained by the other. First, the internal condition presupposes, and is partly explained by, the external one: when and why does a perception constitute a good reason, and provide the right, to believe P? When, and precisely because, it is normally produced only by facts of the type of P, hence the subject is legitimised in assuming that this is the case. And a belief Q is a good reason to believe P when (a) it is itself normally produced only by facts of the type of Q, and (b) from it one can correctly infer P. Of course, what is normally produced only by facts of a certain type can occasionally be produced by different facts; hence, unlike productive chains, reasons are compatible with the falsity of beliefs: one may be subjectively entitled to hold a belief even if it is false. Second, the external condition presupposes and is partly explained by the internal one: why do we require that a belief is produced by its truthmaker non-accidentally, but via a normal chain? Deontological intuition is still at work here: the subject is entitled to her belief if she can assume that her evidence is produced by the truth-maker; of course, she cannot ever be certain about this, since she cannot ever be directly aware of all the links of the chain. But our intuition requires at least that the subject could still assume that the chain was started by the truthmaker (hence be entitled to her belief), even if she became aware of more and more links of the chain. That is, her assumption must not depend on ignorance. Now, only normal chains fulfil this requirement. For instance,

Regimenting the Ordinary Notions of Knowledge

257

in Example II, if Steve learned even a little bit more about the causal history of the paper he saw, he would cease to believe in yesterday’s earthquake (or at least, he would cease to be entitled to that belief): that is why his belief is unstable, just like Daedalus’ statues. These considerations explain the rationale of some solutions proposed for Gettier’s paradox. For instance, it has been claimed that beliefs must not have false presuppositions.18 In fact, what the subject always presupposes is that her evidence is produced by a normal chain; so, excluding false presuppositions is excluding abnormal links. For instance, Steve is not justified because he falsely presupposes that what he sees is today’s newspaper. It has also been suggested that there must be no defeaters, i.e., facts which, if known, would prevent the subject from legitimately holding her belief.19 But the existence of an abnormal link in the chain is precisely such a fact, which makes the presupposition that the chain is normal false. In Steve’s case, the defeater is the fact that the newspaper is an old one. An objection to the “no defeaters” proposal is that there may be also vindicators, i.e. further facts which, if known, would entitle the subject to her belief in spite of learning about the defeater; there can also be defeaters of vindicators, and so on. So, the question of whether a belief counts as knowledge or not ends up turning into an inextricable maze.20 But this is no problem, for a vindicator is simply the fact that a particular abnormal link of the chain is itself (although abnormally) produced by the truthmaker (i.e., that one deviation from the normal chain is luckily compensated for by another). Now, our requirement for knowledge is that the subject would be entitled to her belief even if she knew more than she does. So, if she learned that there is a defeater, she would no longer be entitled; but if there is a vindicator, and she knows it, then even if she learned about the defeater she would still be entitled to her belief. On the other hand, if she ignores the vindicator, the day she found out about the defeater she would lose her right to believe. So, knowledge simply requires that there is a vindicator known to the subject for any defeater (in other terms, that there is no abnormal link in the chain unless the subject knows that it is (abnormally) produced by a normal link). For instance, if Steve knows that the newspaper fell because of yesterday’s earthquake, he would still be entitled to believe (2) even if he discovered that the paper is 10 years old. 18

Armstrong 1973, Moser 1992,. 158 See Lehrer and Paxon (1969), 225-237; Swain, (1974), 15-25; Moser, 1992, 157-159; Sudduth 2008. 20 Moser 1989; 1992, 158. 19

258

Chapter Sixteen

Finally, it has been proposed that the subject must distinguish the relevant alternatives in which her true belief would be false.21 This, again, is saying that when a defeater is around, the subject must know its vindicator. In Example II, the actual case is that the vision of an old newspaper, which abnormally produced Steve’s belief (2), is abnormally caused by the truthmaker (yesterday’s earthquake), so belief (2) is true; but Steve could be said to know that (2) is the case only if he could distinguish the actual case from the relevant (i.e. a priori more probable) alternative in which the vision of the old paper would not have been caused by yesterday’s earthquake, and so the belief would be false.

8. Conclusion To sum up, the vague ordinary notion of justification has both an internal-subjective-deontological connotation, and an external-objective connotation; so, there are two candidates for its theoretical explication: first, an objectively existing link to the truthmaker, warranting that the belief is not only true, but non-accidentally so; second, some internally available evidence and reasons, which entitle the subject to believe. But as we have seen, both conditions are required for knowledge, and both are provided by normal productive chains. This is why normal productive chains reconcile fallibilism with the requirement of a justification so strong as to entail truth: the chain itself entails truth, since it originates in the truthmaker, but human subjects are fallible, for they are directly aware only of the last links of the chain, which provide subjective justification, but do not entail truth. This explains why we can have knowledge (an optimally established belief) even without certainty. Initially, I asked whether Gettier-like counterexamples challenge Plato’s tripartite definition of knowledge, requiring a fourth necessary condition, or our current conception of justification. This question can now be seen as partly terminological, depending on which of the two connotations of the ordinary notion of justification one wishes to keep: the internal-subjective connotation, the objective-external one, or both. One may keep only the former, understanding justification as what gives the subject a right to believe, even without making her belief completely certain; justification will thus be identified with internally available reasons, but Plato’s definition of knowledge will have to be changed by introducing a further condition besides truth and justification. By now, however, we know what that further condition must be: that the 21

Vassallo 2003, 44; Goldman 1976.

Regimenting the Ordinary Notions of Knowledge

259

subjective reasons be the last links of a normal productive chain. Moreover, in spite of this added condition, the definition of knowledge will still be tripartite, for truth is implied by the existence of a productive chain: knowledge will thus be (i) a belief (ii) justified by reasons which (iii) are produced by a normal chain. Otherwise, one may keep exclusively the external-objective connotation of justification, identifying it with the section of a normal chain to which the subject does not have access. I think this would be an unnecessarily radical departure from the ordinary notion; however, if one takes this option, she should change Plato’s definition, characterising knowledge as a (i) justified (ii) belief (iii) which one is entitled to hold on the basis of subjective evidence or reasons. Truth would now be implicit in justification. In my view, however, the best option is keeping both the intuitive connotations of justification. This can easily be done using the notion of the normal productive chain, which encompasses both the external and the internal condition, so providing the most faithful formalisation of the ordinary notion. Thus, justification becomes what warrants both subjective reasons and objective grounds for one’s belief. In this way, Plato’s tripartite definition of knowledge is still correct, but pleonastic, since truth is implied by justification; hence, knowledge can simply be defined as a (i) justified (ii) belief.

References Alai, M. (2005) “Conoscenza in Platone e giustificazione oggi. Spunti teoretici e didattici dal paradosso di Gettier”, in M. Mengozzi, F. Strocchi (eds.) Il Liceo e la città 199/95 – 2004/05, Stilgraf: Cesena, pp. 99-117. Alai, M. (2011) “Subjective and Objective Justification in the Solution of Gettier’s Problem”, in Selected Proceedings of the SILFS 2010 International Congress, edited by Salvatore Roberto Arpaia, L&PS Logic and Philosophy of Science, IX, 1, pp. 493-501. Online: http://www2.units.it/episteme/L&PS_Vol9No1/L&PS_Vol9No1_2011 _48_Alai.pdf Armstrong, D.M. (1973) Belief, Truth and Knowledge, Cambridge University Press, Cambridge. BonJour, L. (1980) “Externalist Theories of Empirical Knowledge”, Midwest Studies in Philosophy, 5: 53–73.

260

Chapter Sixteen

BonJour, L. (2002) “Internalism and Externalism”, in P.K. Moser (ed.), The Oxford Handbook of Epistemology, Oxford University Press: Oxford, pp. 234-286. Cohen, S. (1999) “Contextualism, Skepticism, and the Structure of Reasons”, Philosophical Perspectives 13, 57-89. Dancy, N., Sosa, E. (eds.) (1992) A Companion to Epistemology, Blackwell: Oxford [etc.] Davidson, D. (1986) “Knowing One’s Mind”, in Proceedings and Addresses of the A. P. A. 60, 441-458, tr. it. in A. Paternoster (cur.), 1999, 175-198. DeRose, K. (1999) “Contextualism: an Explanation and a Defence”, in J. Greco, E. Sosa (cur.) Epistemology, Blackwell, Oxford, 187-205. Dretske, F. (1993) “The Nature of Thought”, Philosophical Studies 70, 185-199, tr. it. in A. Paternoster (cur.) 1999, 199-212. Gettier, E.L. (1963) “Is Justified True Belief Knowledge?” Analysis 23: 121-123. Goldman, A.I. (1967) “A Causal Theory of Knowledge”, The Journal of Philosophy 64: 357-372, now in A.I. Goldman Liasons. Philosophy Meets the Cognitive and Social Sciences, MIT Press: Cambridge MA., 1992. Goldman, A. (1976) “Discrimination and Perceptual Knowledge”, The Journal of Philosophy 73: 771-791. Goodman, N. (1955) Fact, Fiction, and Forecast, Cambridge, MA: Harvard University Press. Kripke, S. (1980) Naming and Necessity. Cambridge: Harvard University Press. Lehrer, K., T. (1969) “Knowledge: Undefeated Justified True Belief”, The Journal of Philosophy 66: 225-237. Lewis, D. (1996) “Elusive Knowledge”, Australasian Journal of Philosophy, 74, 549-567. Moser, P.K. (1989) Knowledge and Evidence, Cambridge University Press: Cambridge [etc.] Moser, P.K. (1992) “Gettier Problem”, in N. Dancy, E. Sosa (eds.) 1992, pp. 157-159.

Regimenting the Ordinary Notions of Knowledge

261

Papineau, D. (1993) Philosophical Naturalism, Blackwell, Oxford. Pappas, G. (2005) “Internalist vs. Externalist Conceptions of Epistemic Justification”, in E.N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, URL: http://plato.stanford.edu/entries/justepintext/#GolArgForInt Plantinga, A. (1993) Warrant and Proper Function, Oxford U.P., Oxford. Putnam, H. (1975) “The meaning of ‘Meaning’”, in K. Gunderson ed., Language, Mind, and Knowledge, Minnesota Studies in the Philosophy of Science, 131-193, later in Putnam Mind, Language and Reality, Philosophical Papers, Vol. 2, Cambridge University Press, Cambridge 1975, 215-271. Putnam, H. (1981) Reason, Truth and History, Cambridge University Press: Cambridge [etc.]. Steup, M. (2008) “The Analysis of Knowledge”, in E. N. Zalta (ed.) The Stanford Encyclopedia of Philosophy (Fall 2008), URL: http://plato.stanford.edu/archives/fall2008/entries/knowledge-analysis/ Steup, M. (2010) “Epistemology”, in E. N. Zalta (ed.) The Stanford Encyclopedia of Philosophy (Spring 2010), URL: http://plato.stanford.edu/archives/spr2010/entries/epistemology/ Swain, M. (1974) “Epistemic Defeasibility”, American Philosophical Quarterly 11, 15-25. Swain, M. (1998) “Knowledge, Causal Theory of”, in Islamic Philosophy. From the Routledge Encyclopedia of Philosophy, Routledge, http://www.muslimphilosophy.com/ip/rep/P004.htm Sudduth, M. (2008) “Defeaters in Epistemology”, Internet Encyclopedia of Philosophy, http://www.iep.utm.edu/ep-defea/ Talbott, W.J. (1990) The Reliability of the Cognitive Mechanism, Garland, New York, N.Y. University of Reading, Philosophy Dept. “Gettier Counterexamples and the Causal Theory”, http://www.reading.ac.uk/AcaDepts/ld/Philos/jmp/Theory%20of%20K nowledge/Gettier&CausalTheory.htm Vassallo, N. (2003) Teoria della conoscenza, Laterza: Roma-Bari.

CHAPTER SEVENTEEN MEMORY, HISTORY AND ORDINARY KNOWLEDGE1 PIER LUIGI LECIS

Years ago, John Searle maintained that realism does not belong to the field of theories, opinions and conceptions but rather to the Background of thought and language. The Background precedes specific opinions on events in the world, as well as common sense. When we act, think and speak, we take several things for granted, among which the existence of external reality is one of the most important. Although it does not have the status of an opinion, and even less a theoretical position, it is in any case an indispensable assumption in order to have opinions and theories. It is part of a complex network of abilities, some of them innate and some acquired, that the experiences, linguistic and non-linguistic, of human beings presuppose, as they are inherent in their form of life.2

1

Abbreviations: TR I = P. Ricœur, Temps et récit I. L’intrigue et le récit historique, Paris, Seuil, 1983. TR II = P. Ricœur, Temps et récit II. La configuration dans le récit de fiction, Paris, Seuil, 1984. TR III = P. Ricœur, Temps et récit III. Le temps raconté, Paris, Seuil, 1985. HR = P. Ricœur, Histoire et rhétorique, «Diogène», n. 168, 1994, p. 9-26. MP = P. Ricœur, La marque du passé, «Revue de métaphysique et de morale», n. 1, 1998, p. 7-31. EHRP = P. Ricœur, L'écriture de l'histoire et la représentation du passé, «Annales. Histoire, Sciences Sociales», 2000, p. 731-747. MHO = P. Ricœur, La mémoire, l’histoire, l’oubli, Paris, Seuil, 2000. 2 References to the notion of Background are taken from J. Searle, Mind, Language and Society, Phoenix, London, 2000 (I ed. 1999), 10-12; for the theme of realism as part of a framework, in part innate and in part acquired, cf. Ibid.i, 32, 108-109).

264

Chapter Seventeen

Can we also apply this interpretive scheme to the sense of the reality of the past? And what acquisitions would this line of analysis permit as regards the epistemological and ontological problems of historical knowledge? Can one speak of spontaneous realism on the part of historians, whose intuitions and cognitive practices are characterized by a reference to something that no longer exists, which escapes any experimental reproducibility? In this regard, interesting arguments emerge from examining the relationships between history and memory, in the perspective developed by Paul Ricœur, certainly anti-naturalistic and thus oriented differently as concerns the abilities and methods required for the representation of the past but, on this specific point, not lacking convergences with that of Searle. Memory seems to be an ability/faculty, a form of experience and pre-theoretical ability to recognize something as the trace of an event, whose role cannot be clarified if not by admitting an original psychic impact caused by an event touching consciousness. The method of conceiving of the link between history and memory opens up various epistemological and ontological strategies for the representation of the past.3 Ricœur systematically worked on this subject, finally making it the nevralgic center of his theoretical research. Today I would like to present these themes as indications of a significant change in perspective taking place between his two great works, Temps et récit and La memoir, l’histoire, l’oubli.

1. Historical Imagination In Temps et récit, we are dealing with historiographical constructivism, contaminated by a robust esthetic - literary inspiration. In history, a primary role is played by imagination, understood as the invention of schemes of ordering material capable (by means of producing connecteurs) of placing and re-depicting events according to links and meanings not immanent to the “datum”. The guiding value of Ricœur ’s analysis is its figurative power, the ability of its historical writing to render facts no longer directly observable visible, without underestimating the question of the reality of the past and the representational dimension of history, focalized by introducing the concept of représentance. 3

Ricœur speaks of mémoire naturelle (MHO,. 69), in the sense of an everyday, unspecialized, not clearly differentiated ability; in any case the matrix of specialized practices, like for example history: a cognitive ability to maintain traces and evoke memories of what is past, anchored to an individual psychic dimension, however closely connected with collective, identitarian experiences and memories.

Memory, History and Ordinary Knowledge

265

The theoretical framework is that of “a common space for exchanges between history and fiction” (TR III,280). Ricœur reformulates the concept of mimesis, attacking its interpretation in terms of copy-imitation, and criticizes realism understood as the theory of mirroring, emphasizing the elements of artistic-poetic-historical continuity by means of analyzing the metaphorical and figurative resources of language. In respect to the question of reference to the past, Ricœur writes, it is a matter of seeing how “the imaginary is integrated into the perspective of having been, without attenuating the ‘realistic’ perspective”. The closer the approximation to what was, the more extensive the role played by the imaginary. Ricœur follows the line of an Aristotelian interpretation of rhetorical tradition, with an attentive eye on the dimension of the ȜȒȟȚȢ (elocutio) in the third book of Rhetoric III;4 he conceives of the metaphorical suspension of literal reference as the basis for a redescription of reality, exploiting the results of La Métaphore vive (Seuil, Paris, 1975, with direct references to Max Black, Mary Hesse, and the seeing as of Wittgenstein). It is well known that in this constructive figurative aspect, he was indebted to the narrative approach of Hayden White, whose relativistic solution, however, he always rejected. It must be pointed out that he gives prominence to the role of imagination; emphasizing the figurative dimension relegates the polarity of memory to the background, in a role deprived of theoretical impact. Analyzing the structures and functions of expert, professional historical knowledge, Ricœur writes acute, evocative pages (in a methodological sense as well), relating the formation of historical calendars to more elementary experiences of measuring time. Three aspects are common to all calendars: a founding event, acting as the starting point for calculating time and the order of events; the bilateral practicability of time in respect to an axis of reference pastĺpresent; presentĺpast), and a repertory of units of measurement to indicate constant intervals between recurring historical phenomena.5 These intervals are determined on the basis of astronomy: dawn and sunset for the day, lunar phases for the month, seasons for the year.6 Calendars lie at 4

At the beginning of the third book of Rhetoric (ī, 1, 1403 b 15), the treatise on the ȜȒȟȚȢ is introduced in this way: “We must deal with elocution; indeed, it is not sufficient to have subjects on which to expound, but it is necessary to expound on them in the proper way” (my English translation). 5 The suggestion comes from Benveniste; Ricœur refers to É. Benveniste, Problèmes de langage, Gallimard. Paris, 1966, 6. 6 TR III, Le temps raconté, II Poétique du récit. Histoire, fiction, temps, 5, L'entrecroisement de l'histoire et de la fiction, 157-158.

266

Chapter Seventeen

the vertex of a network of connectors (connecteurs), whose function is to link cosmic time and experienced time; in living experience, a sense of an extremely rich, complex route is outlined, in which the present is not clearly separable from the past (term of retention) or the future (protention, expectations) (TR III, p. 159). This intermingling, Ricœur maintains, clearly emerges even if one chooses, as it were, the most realistic path to the historical past, linked to the re-inscription of narrated time in the independent and objective time of the universe, the only scale in common with the history of the earth, living species, the solar system and the galaxies (TR III, p. 267). This unique, universal scale supplies the reference system for placing the events investigated by historians; although elaborated in astronomical or biological knowledge, it has to do with the most common perceptions and experiences of everyday knowledge.7 For a deeper understanding, we may proceed like Ricœur, starting from a simpler inventive process, the one guiding the construction of the gnomon (TR III, 267). Its rules of technical planning go back to imagination as the source of operations of schematizing synthesis, in the (freely) Kantian sense of providing an image for a concept, harmonizing categories and phenomena, and creating a representation rendering multiple perceptible concepts compatible. Relying on Fraser’s analysis,8 the text singles out the epistemically relevant point in the transposition and 7

To define the field of ordinary knowledge, I refer here to the theoretical proposals of M. Bianca. Differing from scientific knowledge in its structure and method of formation, ordinary knowledge shares with the former its reference to the phenomenal world. It contains “a minimal ontology” linked to a perceptive nucleus, permitting the singling out and recognition of objects. Within it comes together the biological knowledge characterizing every animal, the “knowledge of oneself, his body and the world surrounding him” required to survive. However, in human beings, the difference lies in the “neocortical complexity of the CNS and the presence of a social life generating informative structures above individual ones (a collective mind)”. The most apt dimension of ordinary human knowledge is not biological, but constitutes “common sense”, as it is “a set of beliefs, mental attitudes, conceptions and values peculiar to a culture and present in the mind of every man belonging to it”. One may say that it derives in part from the sensorialperceptive field (and thus has biological roots) and in part from a “noetic basin”, which elaborates perceptive information by means of semiotic activity. I have based these allusions on M.L. Bianca, Introduzione to P. Piccari, Conoscenza ordinaria e senso comune, FrancoAngeli, Milano, 2011 (9, 11-12, 15). Cf. M.L. Bianca, Rappresentazioni mentali e conoscenza. Un modello teorico-formale delle rappresentazioni mentali), FrancoAngeli, Milano, 2005, in particular 45-48. 8 Julius Thomas Fraser, The Genesis and Evolution of Time: A Critique of Interpretation in Physics, University of Massachusetts Press, Amherst, 1982.

Memory, History and Ordinary Knowledge

267

homologation of two different languages, with the possibility of considering two different processes at the same time: the movement of the sun and the life of those who act according to the solar quadrant. This is part of the human universe but also of the astronomical one, as the movement of the shadow is independent of human will. The creativity of imagination is exercised in different spheres, contributing to connecting two heterogeneous processes to formulate a general hypothesis on nature and finally to design an instrument capable of utilizing the shadow of things to regulate social life (information on the time and season are supplied by devising a quadrant with appropriate divisions and concentric curves). According to Ricœur, here we see the constituting principle of calendars in general at work. Naturally, the creation of a calendar (TR III, 267) requires quite complex mathematical operations to homologize the various rhythms in play, and also refers to a fundamental institutional and political dimension. Perceptive dimensions, of direct experience of the physical and biological environment, technical dimensions, cognitive dimensions with great conceptual content and social dimensions interact at the origin of these processes in a considerably articulated explanatory model; however, as an effect of polemics against ingenuous realism (the ideas of knowledge as a copy or resemblance), the focus of the matter lies in the survival of structures constructing the representation of time, in a perspective charged with anti-associationistic, anti-empiristic and antinaturalistic motivations. The role of imagination, mediating between empirical aspects and aspects of mathematical construction, is also expressed at this level, in assigning a place to a present event within a system of measurement of unified time.9 By means of rewriting all recollections accumulated by collective memory on the calendar, they become events dated in some way, orienting life by means of ordering time (TR III, 268). Similar results emerge from an analysis of other connectors, like the succession of generations, in part a biological datum and in part a prosthesis of remembering in a Husserlian sense, remembrance that is extended by means of ancestral memories and imagination, until its own temporality is placed on the generational chain with the help of the calendar. The network of contemporaries, successors and predecessors creates a figurative scheme, halfway between a biological phenomenon and an “intellectual” phenomenon, for the reconstruction of the reign of 9

In principle, any living present will have an objective position, while in a practical sense, to any instant of objective time can be attributed a present as if, in the Husserlian sense of recalling. Dates are assigned to potential or imaginary presents (TR III, 268).

268

Chapter Seventeen

contemporaries, successors and predecessors (Ibid). Using Ricœurian analysis, the structuring of an event is viewed as the product of the original activity of schematizing synthesis, with imagination playing a decisive role, much more than the reproduction of an objective pre-constituted order of events based on the preponderant role of data collection. It is also worth noting that the theme of the trace,10 to which Ricœur attributes great importance as a material characteristic of history, is treated in this perspective in Temps et récit. In the phenomenon of the trace, Ricœur writes, culminates the imaginative nature of the connectors that institute historical time. The trace is a mixed structure of effects and signs, in which complex synthetic activity is expressed, including both causal inferences on the “marque laissée” and “activités d'interprétation liées au caractère de signifiance de la trace” (Ibid), as “chose présente valant pour une chose passée”. It may become an operator of historical time only if duly elaborated by models linked to archives and documents; it requires a process of thought that interprets the remains, the fossil, the heap of ruins, the museum piece, the monument. These thought processes assign the material its value as a trace, representing its life-context, the surrounding world (in a Heideggerian sense) it is lacking. A few pages later, the text poses the problem of the relationship between the controlled fiction of history and memory, in particular the collective memory of culturally significant events for the life of a community. The great question of La mémoire, l’histoire, l’oubli is presented in nuce. Let us see how. Fiction, the controlled illusion of historical narration, may certainly create tension with the critical vigilance of the historian, which has to do with the effect of writing, with the figurative power present as a theme in Aristotle (Retorica III, 10, 1410 b 10 30; 11, 1411b 25): the showing, the placing before the eyes using words, “as if it were happening now”, “representing things in action”; it may, however, also integrate the perspective of representation and confer that “filling which is lacking”. That happens, for example, with events considered decisive by a community “as it perceives in them an everliving origin or source”; their meaning lies in the power of founding or reinforcing the “awareness of community identity”, the collective narrative identity of the members of the community, founded on events producing particularly intense ethical sentiments (TR III, 272). Fiction helps to create an “illusion of presence” and puts itself at the service of what is

10

In large part inspired by M. Bloch, Apologie pour l’histoire ou métier d’historien, Armand Colin, Paris, 1949, Chap. II (L’observation historique).

Memory, History and Ordinary Knowledge

269

unforgettable;11 it places the historian on the same plane as the power and requirements of memory, aimed at singling out past events on the basis of vital needs and pragmatic experiences. Historiography without memory is simple curiosity, exoticism. Mixed with history, fiction returns to its origins, reactivating something of the epic. A characteristic Ricœurian thesis is already present in Temps et récit: after Auschwitz, we are in the era of the negative epic, of the memory of suffering on the dimension of entire peoples. The suffering of the victims is a request for narration, more than for revenge (TR III,275).12

2. Memory, History and Surroundings Now we must point out the theoretical distance separating Temps et récit from La mémoire, l’histoire, l’oubli. In Mémoires: approches historiennes, approche philosophique,13 Ricœur himself speaks of the passage from poetics of the narrative (shared between literature and history, centered on the skepticism of temporality) to an epistemology of 11

Ricœur is well aware of the theme of narration as the revelation of who, of the subject of the action in Hannah Arendt, in particular in the fifth chapter (The Action) of The Human Condition, University of Chicago Press, Chicago, 1958. 12 When we are dealing with recent events, as Auschwitz is for us, the professional neutralization of sentiments is not possible or desirable. Ricœur frames the question of the Holocaust in a broader ethical and epistemological perspective, investigating the relationship between the pre-theoretical roots of memory and the historical representation of the past. Following the Biblical train of thought (Zakhor, remember!), he focalizes on the theme of events which must not be forgotten in the life of a community, endowed with the value of identity, both in a positive sense, as the object of admiration and veneration, and in a negative sense, as a source of loathing and horror (the ethical motivation of the story of the victims). Horror and admiration operate in the historical conscience with the specific function of individualization. There is no reciprocal antithesis between historical explanation and individualization by means of horror/admiration. From memory as a vital experience arises the question referring to the explanatory function of history. If this circuit is interrupted, a ruinous dichotomy arises between a history that dissolves the event, that is its emotional substance, significance, vital energy in the objective explanation on one hand and, on the other “une riposte purement émotionnelle, qui dispenserait de penser l’impensable » (TR III, 273). The more we explain, the more indignant we become; the more indignant we are, the more we want to understand. This is the dialectics of historical explanation. 13 The essay appeared in 2002, in «Le Débat» 2002/5 (n° 122), 41-61, as a response to the intervention of four historians: Roger Chartier, Alexandre Escudier, Pierre Nora and Krysztof Pomian.

270

Chapter Seventeen

history, in a dialogue with the phenomenology of memory (p. 42). As far as we are concerned, here we will examine a series of texts illustrating the change in perspective induced by the centrality of memory. The dialectics between the notions of trace, image, memory and testimony adds an interesting piece to the “reckoning” of Ricœur’s exegetics with realism in history. The guiding significance now becomes that of faithfulness to the past, which projects its shadow, in spite of differences, on the role of truth in history. One might say that, as memory is in the foreground as a matrix of history, it pushes back the role of imagination. The accent is inevitably placed on the sense of the extralinguistic anteriority of the past, as opposed to any idealistic and relativist temptation. All of Ricœur’s work, before and after La mémoire…, faces the problem of the reality of the past, redesigning the relationships between the rhetorical-literary, epistemological and ontological dimensions of history, well beyond the confines of the discussion on reality and fiction in Temps et récit. The ontological obligation is developed starting from an analysis of the historical condition, from the way of being of the major characters in history. An especially clear formulation, to test this change, can be found as early as the 1994 essay Histoire et rhétorique. History has as its ultimate object men like us, acting and suffering in circumstances they have not produced (HR, 23). The theory of history combines with the theory of action (HR, 24). The basis for this convergence is set down by Ricœur in this way: « l’agir humain est tel qu’il appelle pour se comprendre lui-même le récit qui en restitue les articulations fondamentales » (Ibid). The past is champ praxique, structured in the forms illuminated by the theory of action. These observations confirm, Ricœur writes, the realism ingenuously practiced by historians which, in hindsight, offers an opening to escape from the blunt alternative, which seems to block theoretical debate, between the idea of a story ready to be told (a ready-made reality, already formed in a certain sense but not yet pronounced) and that of a formless chaos of historical material. The text goes beyond the aspect of formally belonging to the same spatial-temporal scheme to postulate a level of material belonging to the action of historians au même champ praxique of historical agents;14 they face the past above 14 In the replies to Chartier emerges a specifically ontological aspect of these characteristics, an immanent return to the reality of the past as an extralinguistic reference. The historical being is certain of the existence of the past. The expressions avoir-été, n’être plus refer to an analysis of the historical being. Certitude is an assumption both of the real action of agents, of the man who is making history, and the historiographical operation of faire de l’histoire, an anthropological characteristic belonging to his way of being and acting in the

Memory, History and Ordinary Knowledge

271

all as héritiers, as heirs of something which se perpétue dans le présent et ainsi l’affecte. Ricœur seems to explicitly allude to a relationship of causal dependence, inseparable from the common belonging to a single chain of interactions. Historiography must be considered an operation which is itself anchored to other human activities and characterized by a complex relationship with the experiences of men from the past. From the very beginning, a passive aspect of the inheritance received comes into play, which can be better expressed by the idea of debt (dette), a theme that will become essential in La mémoire…. We are indebted to men of other eras. “Avant la représentation vient l’être affecté par le passé”, Ricœur forcefully writes (HR, 25).

3. Memory In the layout of the important work dated 2000, memory, as the power of retaining and reactivating recollections, in addition to a naturalistic psychology of cognitive faculties and abilities,15 must be investigated by a phenomenology of intentional forms typical of the experiencing consciousness, internally articulated in time. Its cognitive dimension is interwoven with the pragmatic dimension of the individual and the collective use made of it; like every actual exercising of a faculty, it is exposed to the risk of abuse. This renders the study of its layers and its intermingling with various powers very complex, to use Ricœur’s lexicon of the homme capable.16 Ricœur operates within a cultural setting, above world. The pact of interpretation between the historian and reader reflects this way of being, and thus it is not simply an epistemological matter or a deontological rule. It has very deep roots in the type of experience and the way men behave in the world (Mémoires, …, cit.. 44) and in their essential relationship with the past. 15 For a naturalistic approach, the past tends to be reified, treated as an entity, a place where memories dwell, forgotten experiences to be extracted; this misunderstanding is fostered by spatial metaphors, beginning with those of empreinte, left by a ceau sur la cire (MP, 9), with the inflexible consistency of a fait accompli. 16 The essay by M. Breviglieri, L’espace habité que réclame l’assurance intime de pouvoir, in “Études Ricœuriennes / Ricœur Studies”, Vol 3, No 1 2012 (online), speaks of Ricœur’s “anthropologie capacitaire”, recalling its explicit formulation in Soi-même comme un autre. However, this depends on a more profound, preceding level of precise analysis, defined in a very interesting way, in Philosophie de la volonté. “La question des capacités n’est plus envisagée sous l’angle de leur affirmation et de leur reconnaissance qui implique autrui en mettant en jeu l’estime de soi. Elle est prise très en amont de cela, au niveau où s’exercent des expériences corporelles primitives et où peut s’établir une assurance intime de pouvoir se

272

Chapter Seventeen

all the French culture of the 1990s, marked by the debate on historical revisionism and the representability of the Holocaust, a debate linked to the question of the testimony of the victims of historical traumas; to the duty of memory he tends to oppose the work of memory, a way of coming to terms with the past without removal and without being constrained to repeat it. We are dealing with the most characteristic and well-known motifs of La mémoire…, relating to the ethical-political sphere and centered on the formative impact of history on psychic and cultural memory (the problems, says Ricœur, of a memory avisée et instruite). However, I do not intend to dwell on this point; on the contrary, I would like to confine myself to the cognitive and ontological profile of the problem. Thus I will attempt to explore the dialectics between two important themes, that of the trace and that of testimony, in unstable equilibrium in Ricœur’s analysis of the memory-history relationship. The two categories refer to opposing and complementary perspectives; one marked by the enigma of the relation de ressemblance (indemonstrable for reasons similar to those which, in other traditions, are opposed to metaphysical realism), the other inseparable from the no less enigmatic relation fiduciaire (MP, 14), the believing in, rather than believing that. On one hand, the theme of testimony seems to emancipate historical knowledge from that of the trace, conferring on it a clear dialogical, linguistic, communicative significance, in which the fiduciary, not strictly cognitive dimension plays a decisive role lying at the basis of social bonds. On the other hand, the question of the trace, with its perceptive and causal conditions, again arises, says Ricœur, at the root of testimony. We may once again use the essay Mémoire: approches historiennes to clarify this point. Ricœur maintains that the original, privileged characteristic of memory emerges if one distinguishes a fundamental capacity for remémoration (in reference to the past) from the plurality of the collective and individual memories exercising it (Ibidemi, 56). Stated from a phenomenological, not empirical, viewpoint (aimed at defining the typical, essential forms of memory), the research singles out several components of the power of remembering, in each of which an aspect of reference to the past is evident. Ricœur distinguishes at least three of them: rendre capable. Cette assurance ne tient pas à la confiance de prendre à son compte un pouvoir, mais à la fondation d’un espace pratique usuel intégré par la familiarité au schéma corporel.” We must also recall that Ricœur dedicates a chapter (II in the II study) of Parcours de la reconnaissance, Stock, Paris, 2004, to the phenomenology of the homme capable.

Memory, History and Ordinary Knowledge

273

1) the ability to receive a memory, understood as the simple appearance and presence of something from the past in an image. An experience naturally referring to anything in the past appears spontaneously, as in the Greek mnéme (ȝȞȒȝȘ), focused on in particular by Aristotle (Ibidem, 46).17 The memory as an apparition, question du quoi, is a mental disposition, pathos (MHO, 4). The feeling of temporal distance is fundamental here: due to this paradoxical capacity, we may say of an event that it took place before we narrate it. Memory has a temporalizing function. Ricœur observes that this is another theme put into perspective by Aristotle18). However, the ideas of imprint and trace increase rather than solve the paradox; they are effects in the present of something that is not present and has significance (Mémoire: approches…, cit., 57); 2) the ability to recall the past by means of research (rappeler le passé), to the extent that forgetfulness or other obstacles permit it. The memory loses its characteristic of a simple, spontaneous apparition and is now the object and fruit of research brought about by moving away from the principal impression and having as the objective its recovery and recognition; it is Aristotle’s anamnesis (ĮȞȐȝȞȘıȚȢ) (Ibidem, 46), placed at the end of a specific effort and intentional process.19 The pragmatic approach to anamnesis complicates the picture and favors the passage from the question of quoi to that of here: remembering is not only receiving, but doing something that surely interferes with the pretext of the fidelity of memory (MHO, 4); what Bergson calls survivance and Husserl the retention and reproduction (Wiedererinnerung) of the memory (Mémoire: approches…, cit., 57)20 also lead back to this aspect; 17

See MHO 6-7 e EHRP732. Ricœur refers to Parva naturalia, Perì mnemes kai anamneseos, I, 15; in the Latin version by St. Thomas, the distinction between memorari and reminisci is presented. 19 Cf. MHO,. 6-7, EHRP, 732. 20 Ricœur presents this scenario: Husserl took a decisive step by interweaving memory and internal awareness of time in a series of writings (1898-1925), now collected in Volume XXIII of Husserliana, Phantasie, Bildbewusstsein, Erinnerung: zur Phänomenologie der anschaulichen Vergegenwärtigungen: Texte aus dem Nachlass (1898-1925), E. Marbach ed., M. Nijhoff Publishers, HagueBoston-London,1980; with Heidegger, memory is incorporated into temporality and attracted to the area of the being’s gravitation towards death (EHRP, 734). The reference is obviously to Sein und Zeit, in Jahrbuch für Philosophie und phänomenologische Forschung, VIII, p.1-438, Niemeyer, Halle 1927. Regarding the theme of the survivance of images in Bergson, we may certainly recall Matière et mémoire : essai sur la relation du corps à l’esprit, whose Chapter III is entitled De la survivance des images. La memoir et l’esprit Œuvres, PUF, Paris, 1970, 276 18

274

Chapter Seventeen

3) the essential belonging of the memory to a subject with direct, privileged access to it, mienneté, we might say the phenomenic subjectivity of memory. The memory is the object of appropriation in the first person (singular or plural); the memory of an individual or collective is not reducible to that of any other. Memories are not transferable. This theme may be linked to an important characteristic of testimony, the assertion of reality, inseparable from the self-designation of the witness and clearly expressed by the formula I was there. It is a decisive characteristic, the true root of the plurality of memories and their conflict. Appropriation is expressed in the form of vindication and prétension, in a scenario in which the contents of one are contested by another memory; there is competition because there is appropriation. In addition, each subject sees his own presumed fidelity compared with the presumed truth, but on the critical bases of the truth of historical discourse (Ibid.). This aspect is directly interwoven with that of testimony as the principal and ultimate source of historical knowledge and the fiduciary foundation of the overall social cohesion of a community. As is evident, the theme of the trace is unavoidable at this level of analysis, even if, for Ricœur, it constitutes in a certain sense the weak point in the reconstruction, the one creating centuries-old, irresolvable controversies in philosophical tradition, without arriving at an adequate resolution of the difficulties. Indeed, following the evolutionary thread of the complex structure of La mémoire…, we see that Ricœur aims at a strategic line that re-evaluates its role, taking the dispute to a different, more complex, pragmatic, communicative, clearly linguistic level, that of testimony. Analyzed according to classic categories (image, imprint), memory indeed seems, in the layers emerging from a long tradition, indissolubly linked to the enigmas of the presence of the absent and resemblance between the original and the image, rendering fragile, controversial and problematic its very need for fidelité. It seems insufficient in itself, and even more inadequate face-to-face with the narrower bonds and control standards of historiographical truth. We may refer to La marque du passé to better understand the nature of these enigmas. The first enigma seems natural to the metaphor of the marque, the image, İੁțȫȞ, as imprint (tupos). Something present (for Aristotle a form, a trace, like a group of drawn, painted strokes) that refers back to something not present. It is, however, necessary to distinguish two forms

foll.), and in addition, L’évolution créatrice, in the section La durée, in the first chapter (in uvres, cit. 494 foll).

Memory, History and Ordinary Knowledge

275

of absence: that of what is unreal, imagined, and that of something real but past. The distinction, Ricœur notes, is quite uncertain and returns on a grand scale in modern discussions linked to the category of portrayal: what does it mean to portray? To present once again, as in a second encounter? Or to reconstruct? (MHO, 12) The second enigma (already clearly present in the Platonic Sophist) is that of the similarity between the portrait and the original, which opens up the dangerous path of mimesis as similitude.21 Here Ricœur uses a Wittgensteinian-type argument against realism: from what do we recognize the truthful character of an image? In what sense are memories similar to events? How does a reconstruction differ from a fantastic construction? In absence of a tertium, from whose perspective to make a comparison, how can we evaluate the similarity between the image and its cause? In any case, Ricœur gives several reasons supporting the thesis that memory, above and beyond the epistemological breach, is and remains the matrix of history, anchored to the original, preliminary certainty of memory and testimony. On one hand, the lack of the petit miracle of recognition triggers the energy of historical research, which is motivated by the effort to compensate for the irrecuperable absence of the past with its chain of reconstructive mediations (EHRP, 736, 747). On the other hand, the very search for truth, organized epistemically, derives from the pragmatic substratum of memory, directly linked to the experience of recognition. With its need for fidelité, memory, as an individual faculty and collectively-shared resource, is the great matrix of history. La mémoire… insists on the aspect of pragmatic rootedness of the theory; the source of the request for truth itself is, as it were, pragmatic and vital. It is not a representation, but “dans l’expérience vive du ‘faire histoire’ tel qu’elle est diversement affrontée par les protagonistes” (MHO, 337). The extreme event Auschwitz lies first in individual or collective memory and then in the historian’s narrative.

21

Cf. EHRP, 733. The question of mimesis is always poised between the imagecopy and the phantom of the imagination; it is never completely free of the sentiment d’un lien d’adéquation, de convenance de l’image-souvenir à la chose souvenue.

276

Chapter Seventeen

4. Testimony In the perspective of La mémoire…, testimony plays a crucial role as a moment of transition from memory to history,22 above all in the passage from the oral to the written form of recollection. As we have said, memory (due to its third aspect) has a dialogical, not solipsistic dimension; we must speak of mémoire à plusieurs. As indicated above, this point is particularly important for theoretical emancipation from ingenuous realism. In the dialogical comparison of testimony, the question of ressemblance loses importance (MP, 7) and there appears to be a way to avoid the aporias traditionally linked to the primary dimension of memory (mneme and anamnesis). There are fundamentally three stages marking the transition from memory to testimony and to autonomous historical discourse: declarative memory, story, and inscription. With the story, the witness reveals himself in a public space, accompanying the assertion of reality with his selfdeclaration as a believable author of the testimony. His willingness to repeat/compare his testimony and its inscription outside himself using any material support constitute the successive, decisive step. It is not only a matter of writing in a strict sense. The field of inscription (including masks, tattoos, drawings, the use of colors on clothing, etc.) is much broader than that of the scripturalité in literary form. Writing may compete with and may also fail to do justice to memory. When based on this material, history begins its epistemological course of separation from memory, also leading to a radical overturning in which memory becomes the object of history (history of the images of the past, representations of the past and memory).23 22

In Mémoire: Approches historiennes (56), testimony, as it is open to comparison with the remarks of other witnesses, places itself in a strategic position as an element of transition between the regime of the fidelité and that of the vérité of history. The originality and privilege of memory, on the plane of reference to the past, as a felicitous moment of immediate recognition, is what makes it the matrix of history, regardless of the degree of autonomy of history regarding it. 23 The dimension of involuntary testimony (secret Chancery documents, confidential reports, archive documents) emerges, finally constituting the documentary dimension of the archive. The objectification of memory and its radical confrontation with history are themes typical of Pierre Nora. Cf. Pierre Nora, (sous la direction de), Les Lieux de mémoire, (3 tomes:), 1. La République, 2. La Nation, 3. Les France, Gallimard, Paris, 1984-1992. Along the same lines, edited by M. Isnenghi, I luoghi della memoria, 3 vols., Roma-Bari, Laterza, 199697, investigates the symbols, myths, events, structures and characters of unified Italy. Its approach derives from the pioneering studies by Maurice Halbwachs in

Memory, History and Ordinary Knowledge

277

Let us see at close range how history, by means of testimony, comes to depend on memory. Ricœur follows this line of reasoning: we have nothing better than memory to assure the reality of our recollections. We have nothing better than testimony and its criticism to confirm the historical representation of the past (MHO, 364). Sooner or later, analysis must go back to the testimonial dimension of the document; on the linguistic level, one may not go beyond the triple declaration of the witness: J’y étais, Croyez-moi, Si vous ne me croyez pas, demandez à quelqu’un autre. Testimony transmits to history the power of declarable memory (MHO,. 647); beyond this level lies only the passage to the extralinguistic, with its final support in ce qui se donne, sous le nom de trace, comme l’effet-signe de sa cause. Historiography can freely give, correct or refute testimony, but not eliminate it. We may take some other cues from a paragraph in La mémoire…24 which supplies an analytical description of the pragmatic and performative aspects of testimony and of the highest levels of the discursive elaboration which its content undergoes after the perceptive event. A type of tripolar structure of the formation of testimony is delineated: the three poles are the attestation of reality based on perception, the presence of the narrator and finally the inevitable pragmatic component of trust of the recipientbeneficiary of the testimony. In this tripartite evolutionary scheme, the perception-memory-attestation of oral testimony, the initial problem lies in the linguistic transcription of the perceptive foundation. Attestation of reality is inseparable from the self-designation of the testifying subject (“I was there”), which always renders it controversial and problematic. Its verbal description is a description of the “scene experienced” by means of a declaration in which a narration is inserted. In any case, testimony has a fundamental function, as it establishes, on an initial level, a frontier (frontière nette) between réalité and fiction by means of the factualité attestée. The original perceptive nucleus marks the difference from simple imagination (referring to something absent, but not attested [auparavant] by perception, a distinction, as we have seen, emerging from antiquity with the notions of eikòn and tupos). Ricœur uses all the instruments of the pragmatics of testimony to show that this constituting feature lacks sufficient cognitive solidity. The paradigm of registration and the impartial observer must be abandoned. The phenomenology of memory, he observes, has shown the problematic character of this frontier, the 1925 - Les cadres sociaux de la mémoire (1925), Albin Michel, Paris, 1994 and La mémoire collective (written in the 1930’s), Albin Michel, Paris, 1997. 24 MHO, Épistemologie, Chapter I (Phase documentaire : la mémoire archivée, par. III, Le témoignage) 201-224.

278

Chapter Seventeen

source of theoretical shadows appearing on the stage of the historiographical representation of the past, which has an ambivalent nature. What is attested to inseparably fuses the reality of what is past with the presence of the narrator on the sites of the event. On one hand, memory is the guardian (le gardien) of the ultimate constitutive dialectics between two moments in the past; the relationship with the no longer (ne plus), which reveals its characteristic of completeness (le caractère révolu, aboli, dépassé) and l’ayant-étè, which reveals its original anteriority without rift from the present and the future (MHO, 648). On the other hand, it must be admitted that there is something obscure in this subjective imprint of the event, capable of impressing the witness in a way that does not necessarily coincide with the importance given to it by the receiver of the testimony (MHO, 204-205). Therefore the fiduciary and thus fundamentally intersubjective dimension of testimony is set at the center of analysis. The most appropriate point of view to understand fiabilité starts from everyday experience, an approach in profound agreement with the theory of action implied by the explanatory and representative phase of historiography (MHO, 203). Testifying is an activity with a range as broad as that of narrating, related to the activity of promising. Self-designation opens up a dialogical space and is a request for trust, requiring acceptance on the part of the recipient. The witness asks to be believed: I was there, believe me, is the implicit sense of his action. Certification is not complete without the response of the listener. The witness’s accreditation opens up the dialectics between trust and suspicion. Confiance and soupçon are the new values put into play; testimony acquires its true meaning when it explicitly leaves the monological sphere of a solipsist conscience to transpose its pretext of validity to a public space. Ricœur himself specifies the three grounds on which the witness is evaluated: doubts may concern erroneous perception, memory, restitution or verbal formation (MHO, 205). It is important to consider the great pervasive strength of this fiduciary bond, extending to all social exchanges; accepting the word of others lies at the origin of social bonds, a habitus at the foundation of an intersubjectively shared world. Trusting the word of others is the composing principle of the sens commun (MHO, 207). Initially trusting and then doubting, if strong reasons incline one to do so, is an essential ability of the homme capable. One may speak, Ricœur writes, using an oxymoron from the sociologist Renaud Dulong,25 of natural institution, a useful expression to distinguish 25

Cf. Renaud Dulong, Le Témoin oculaire : les conditions sociales de l’attestation personnelle, Éditions de l’EHESS, Paris, 1998. According to Dulong, testimony is not the simple transmission of information but is a complex social object, an

Memory, History and Ordinary Knowledge

279

this institution, linked to everyday conversation, from its technical, artificial uses, like placing in archives or the regulated deposition in a court of law (Ibid. 206-7). Following Ricœur’s train of thought, it thus appears that the very establishment of the difference between reality and fiction can fully unfold only at this level, although it could not be posed without its perceptive base. Without doubt, emphasis on the pragmatic aspects of testimony, showing the theoretical advantages of establishing historical consciousness (independent, as far as possible, from the aporias of the initial level of memory, conditioned by the metaphors of the image and the imprint), tends to place the role of the original impression in the background. We may now ask ourselves: does the communicative projection of the contents of testimony in a dialogically controlled public space really dispel the enigmas of the trace and image and completely exorcize the phantom of realism? Ricœur’s antennae are too sensitive in a theoretical sense to avoid posing the problem, as he actually does in the essay La marque du passé and in La mémoire…. We must therefore concentrate on these texts. As we have said, testimony depends, in the final analysis, even in its written and archival form, on memory; this opens up various very interesting scenarios. However, along this line, the original impression comes to the forefront as an irreducible component of the phenomenological description of memory and appears as an original datum of past experience, a type of natural evidence which any social player and historical agent has at his disposal during his everyday actions. This line of reasoning, undoubtedly secondary, survives and acts as a counterpoint to the principal line in La mémoire…. In my opinion, there are very clear texts favoring this interpretive approach. We may focus on this question as regards three themes: 1. the passivity common to the memory, image, trace and oral testimony, at the origin of every “inscription”; 2. the receptive character of the psychic trace as the persistence of a primary impression, as the experiencing of an impact, a feeling caused by an event that made an impression on us; 3. the pre-predicative, natural belief in the reality of the past, in the fact that something actually happened, is the basis for recognition institution grafted onto the mechanisms of natural oral communication, in the form of a narrative report of events the witness directly observed. The story may lead to written testimony. The verbal expression of experience implies a quantum leap, as it belongs to a public space in which the bond between the witness and the event is socially recognized.

280

Chapter Seventeen

of images of the past in oral testimony. For example, attestation, interwoven with protest (in the memory of traumatic events), is irrefutable; they both belong to the order of belief (we are, in some aspects, in a theoretical “humean” area).26

5. The Profound Nucleus of the Trace Ricœur considers the depth of the connections emerging in our philosophical tradition (in a sort of eidetic necessity) between memory and image, which resist total resolution of the classic themes of the imprint and image into that of testimony. It seems that the return of memory cannot occur if not in the form of becoming an image. It is reasonable to believe that in the living experience of memory lies an unshakable characteristic reflected in the image-souvenir nexus (MHO, 54); it is a genuine trait, although the source of a snare, in turn irredeemable, due to the veracious (cognitive) function of memory and its pretension to fidelité (MHO, 7). Indeed, memory remains the primary resource for assuring that something has already happened (before the moment when we are speaking of it); historiography itself, with its methodical elaborations, does 26

I have found no significant references to Hume in the texts consulted; this fact merits further consideration, keeping in mind that, in any case, from Ricœur ’s perspective, the humean one would appear to be compromised by the paradigm of the image-copy and resemblance. Excluding this difference, there are in any case interesting parallelisms as regards the themes of belief, assent and the connection between testimony and original impression. In a passage from Treatise on Human Nature (I, III, sect. 4, Of the Component Parts of our Reasonings concerning Cause and Effect), humean reasoning is developed as follows: let us accept the belief that Caesar was killed on the Ides of March on the basis of unanimous testimony. In the face of this testimony, we are dealing with characters and letters present in our senses and memory, which we know are signs of other ideas. There are two possibilities: either these ideas were present in the mind of the one present at the event, receiving them directly from the existence of that event, or they derived from other testimony, and so on, until we once again arrive at those who were eyewitnesses and spectators of the event. Without the authority of memory and the senses, there would be nothing definite at one end of the chain of testimony; thus there would be no evidence and no credibility. Today, there are quite animated discussions on the epistemology of testimony, and we know how, as early as Hume’s time, an analysis by Thomas Reid (Inquiry into the Human Mind on the Principles of Common Sense, 1764) revealed an alternative line of thought, extremely aware of the fiduciary, social, pragmatic dimension of testimony. Cfr. N. Vassallo, Per sentito dire. Conoscenza e testimonianza, Feltrinelli, Milano, 2011.

Memory, History and Ordinary Knowledge

281

not modify the certainty that, whatever the past has been, it remains the ultimate referent of memory and the objective correlate of our “reconstructive” actions (Ibid). In the essay La marque du passé, this point had already been considered, developing the referential dimension of history and memory by means of the concept of representation, lieutenance, previously introduced in Temps et récit. In this perspective, the motif always linked to the perceptive root of memory takes on all its importance; a moment of passivity, a receptive aspect, pathique, joins the icon and the trace. Both refer to something that happened, an impact, the inscription of the imprint, a passage which has left its mark. Testimony conceals at its origin an enigma of the same type: before narrating, the witness has seen, heard, experienced, or believed he seen, heard, experienced; what is important here is that he is affected, struck, wounded, impacted by the événement he witnessed. We are on the same level as the feeling of temporal distance inherent in the memory which comes to mind. It is what testimony transmits with its saying c’est quelque chose de cet être-affecté par…l’empreinte; it preserves the energy (énergie) and even the violence (violence) of the previous event, which from the first-level witness is transmitted to the listener, a second-degree witness. This theme, Ricœur recalls, is clearly present in our philosophical culture, from the Aristotelian definition of memory to awareness of its historical effect (Wirkungszusammenhang) in the Dilthey-Gadamer interpretive tradition (MP, 17).27 We find this theme in numerous places and with varying nuances in Ricœur’s later texts, beginning with the expression of the notion of trace in several passages of La mémoire…. In the section L’oubli in part III (La condition historique), with its interesting correlations between forms of forgetfulness and forms of memory, three types of trace are distinguished, relevant for investigating the memory-history relationship (539). First of all, Ricœur speaks of trace mnésique, maintaining that the appropriation of what we know in our brain is complex. The cerebral, cortical trace, the domain of neurosciences, is external (p. 540); it is known only from the exterior, scientifically, without corresponding to what one feels and experiences live; in this sense, we say we see with our eyes and grasp with our hands (while we do not say we think with our brain). Ricœur brings into focus, from his angle, that which, in 27

Cf. A. MARINI Alle origini della filosofia contemporanea: Wilhelm Dilthey. Antinomie dell’esperienza, fondazione temporale del mondo umano, epistemologia della connessione, La Nuova Italia, Florence, 1984 (205); for a previous translation as dynamic connection, it substitutes the translation concrete/real connection, with specific reference to the structures of the historical-social world.

282

Chapter Seventeen

controversies over consciousness in philosophy of the mind, is referred to as the explanatory gap. The written trace, becoming a documentary trace in historiographical operations (the source of its initial “distancing” from memory; EHRP 737), is a material trace (like the cortical one); thus it may be altered and destroyed. The archive is also established to respond to this threat of destruction (MHO, 539). The psychic trace, defined as an impression more than an imprint, in the sense of feeling left in us by an event, “marquant” or “frappant” (Ibid), is presented by Ricœur as an interior trace. For our discussion, we are principally interested in the characterization of the psychic trace, corresponding to the experience of recognition, peculiar to memory as anamnesis. It must be observed that Ricœur lays claim (aware of the demanding implications of mental ontology) to the autonomy and irreducibility of this level of analysis, raising the problem of the relationship between mnestic and psychic traces. The direction in which he moves is that of safeguarding the epistemological gap between the two levels, a discussion of the neuronal and a discussion of the psychic, shielding it from any form of ontological, spiritualistic or materialistic reductionism. He states he is abstaining from the dispute on the union of soul and body; to face the tangle of psychophysical relationships, he considers adequate the use of notions like sine qua non causes and organization/function correlation. The formation of psychic traces is not a process that can be explained simply in neuroscientific terms, as the trace has a non-eliminable semiotic component. I have already quoted the analysis of memory as the recognition of past experiences; the nature of that recognition renders access to these experiences extremely diverse with respect to the third-person one in regard to mnestic traces. We must now go back and examine this theme more closely to highlight an aspect, surprising in a certain sense, regarding the main line of the Ricœurian interpretive approach, based on the idea that one cannot understand memory without starting from the antagonistic form of forgetfulness. In the case of the psychic trace, we are dealing with an oblie de réserve, rather than for effacement, and this requires notions like endurance, persistence (rémanence), revivification, and survival (MHO, 541), of the type elaborated by phenomenological tradition and by Bergson.28 The point is that, analyzing Ricœur’s lexicon, it is easy to find 28

The phenomenology of memory, following the thread of living experience, will speak of the persistence of the original impression, or, with Bergson, of reviviscence des images at the moment of recognition (cf. above, note 18).

Memory, History and Ordinary Knowledge

283

a realist substratum. Images from the past can be recognized. We discover that, for example, many childhood memories have not actually been destroyed, but only rendered unavailable (Ibid. 540, 541). At the moment of recognition, the present image is considered faithful to the affection première, au choc de l’événement. Other expressions, converging in this nuance, are found in the section L’oubli et la persistance des traces. Here is a short list: persistance des impressions premières en tant que passivités: un événement nous a frappés, touchés, affectés et la marque affective demeure en notre esprit (Ibid. 554). The problem of the recognition-memory relationship is extremely intricate, but its contradictory aspects may be summarized in this way: recognition is defined as the superimposition of the image present in the mind and the psychic trace left by the first impression (Ibid.556), or even as the persistence (demeurer) of an impression (Ibid. 557) constituting the original datum of the whole process of remembering, and then of telling stories and explaining events, with cognitive intentionality. One may not be aware of it without going back to a precise causal antecedent, to an impact of a perceptive nature, which left its impression (internal trace). In the first pages of La mémoire… it is pointed out that Aristotle shifted the metaphor of the seal to the graphic plane, to the portrait, highlighting the assumption of an agent that has imprinted a trace, a sign of his passage. Behind the shape imprinted in the wax lies the act of imprinting the seal. We have an implicit reference to the exterior cause of the mark-imprint. To have value as a mark, something must include a dimension of alterity concerning its origin (MHO, 12). In a text appearing in a digital version in 2002, Entre la mémoire et l’histoire, which supplies a brief but penetrating presentation of the themes in La mémoire…, three enigmatic aspects of memory are singled out: a) presence (of an image or imprint in the mind); b) absence (of the thing in the past which the image refers to), and c) priority of the event (feeling of temporal distance, expressed by the tense of the verb or by adverbs: something that existed auparavant is recalled). The initial, provisional solution, which analysis of the memory proposes for the enigma of the presence of what is not present, is the small miracle of recognition or experience of the order of certitude, what is obviously lacking in history but which history in any case is seeking. Regardless of how suspect it is, it is the only form of experience that in some way refers to what no longer exists in the present. Recognition includes the moment of conforming,

284

Chapter Seventeen

concurring, adapting the present image to the absent thing of which memory has protected and preserved (gardé) the trace.29 Memory depends on an original perceptive base, on the impact of an event referring to past actions on the part of agents similar to us. History is bound to the perceptive nucleus of memory (from which it inherits aporias). This is particularly evident in the case of traumatic events, of the living experience of a wound sedimented in the faire histoire, of events at the limit of the representable, of the ictus with which real history strikes collective and private memory. Something that demands to be told, narrated and not forgotten (it is not important whether it is terrible or worthy of admiration, as in archaic history).30 The source of the request for truth is, as it were, pragmatic and vital; it lies not in the representation, but dans l’expérience vive du “faire historire” tel qu’elle est diversement affrontée par les protagonistes (MHO,. 337). Remaining with the example which, not only in Ricoœur, takes on a paradigmatic value, the extreme event of Auschwitz remains in individual and collective memory before it becomes part of the historian’s discourse. We may now draw some conclusions on the potentialities of the memory-history relationship as regards Ricœur’s interpretative itinerary and, more generally, the fruitfulness of these realistic themes. The picture that I have attempted to paint is extremely complex and laden with 29

Here is the text of Entre la mémoire et l’histoire: “Aussi douteux que soit le souvenir dans le moment de la reconnaissance, nous n’avons pas mieux que lui pour nous faire éprouver, croire, dire, raconter, que quelque chose a eu lieu auparavant tel que nous en faisons mémoire. Tel? C’est toute la question que la mémoire transmet à l’histoire.” No other primary experience will give us certainty of the past, of the current presence of absence, better than the recognition of memory. For further references, see also (in MHO, 45-48) the analysis dedicated to memory as anamnesis and recognition, the small miracle, inspired by Recognizing by Edward Casey, Remembering, Indiana UP, Bloomington, 1987. Recognizing a present memory as corresponding to the impression première visée comme autre, or the presence of an absent previously encountered, cloaking (enrober) with presence the alterity of what is finished: these are in the same way aspects of the re-presentation of the past. The recognized past tends to demand respect as the perceived past, due to which reconnaissance has the strange destiny of appearing in the phenomenology of perception as well as in that of memory (MHO, 47). Ricœur ’s text also relies on Kant, as an author who would have demonstrated how recognition provides coherence to what is perceived (Ibid.. 48) 30 Great events like the Shoah and the great crimes of the twentieth century, although lying at the limits of representation, rise up, like all events which, having left their traumatic mark on hearts and minds, demand to be articulated, narrated, understood (648, final section Le pardon difficile).

Memory, History and Ordinary Knowledge

285

variations, conditioned by the uncertainty of the passage from perception to language. This uncertainty is emphasized by the dominant interest in its structural and interpretative nature, always linked to individual and collective subjectivity, of the acts of memory and even more those of historiography. The original impact offers no fundamental guarantee, no support for firm objectivism which does not take into account epistemological descriptions and the “subjective” assumptions of firstperson experience, the intrinsic bond between the attestation of reality in the perspective of the witness and his interests, emotions, needs, etc. However, in his later writings, as we have attempted to clarify, he persistently returns to the essential role of the passive aspect of experience. He proposes purifying the subject of the subjectivistic values of the concept of representation; he does this by recognizing an original pretheoretical competence inserted in the constituent structure of experience of the past. In its meaning and typical structures, a memory includes reference to the extralinguistic and extraperceptive reality of the past, as it is distinct from any mere construction, fiction or imagination. The reality of the past is the correlation of the act of memory; the certainty that something actually took place is an implicit element in our everyday way of seeing things. We might say, in the sense clarified above, that we are dealing with background aspects of everyday knowledge. Ricœur speaks of croyance antéprédicative - et même prénarrative, on which are based the fundamental matrixes of historical knowledge, recognition of images of the past and oral testimony (MHO, 648). This theme leads us back, in addition to an ontological analysis of the historical condition, to the question of compréhension pratique already focused on in Temps et récit; it is an intuitive mastery, a preliminary competence of the conceptual network that, according to Ricœur, characterizes human actions and interactions (articulation over time, the difference between action and movement, the purposes, motivations, means and symbolic mediations of the action).31 The phenomenological analysis of the pre-theoretical, intuitive, pragmatic roots of the experience of memory converges in the same direction. We have arrived at a particular chapter of philosophical anthropology, of the homme capable, agissant et souffrant, enriching the list of his “powers”: I can remember is added to I can speak, act, narrate, hold myself responsible for my actions, etc. (note 1, MHO, 68). The act of remembering involves historiography as a theoretical practice. The historian begins to faire de l’histoire just as each of us begins to faire 31

Cfr. TR I, 88-95 su Mimésis I, Préfiguration.

286

Chapter Seventeen

mémoire; dealing with an experience that unravels in the course of time and has a constitutive reference to the past is essential for the formation of individual and collective identity, on whose fundamental narrative structure Ricœur focuses. Just as the poetics of the story emphasized the active aspect of the structural procedures characterizing the historical operation and the role of imagination in representing time, in the same way, the phenomenology of memory brings to light the realistic values of past experience, linking them to pragmatic skills. The faithfulness of memory always refers to something certain and which has already happened, to an event that has left its original trace on consciousness; it is an imprint received passively and then recognized by means of its memory. The real-past (unlike what is simply imagined) is intuitively not something produced by the mind; by means of abilities which, going back to the point from which we started, we can also compare to Searlean background abilities, we are naturally led to consider the psychic trace as the effect of an external cause. It is part of our knowing how to treat memories in this way and using them in our everyday actions, even before having precise awareness of them or developing any reflective elaboration of them. This background is the basis of testimony, and testimony, organized and registered, forms the basis of history as a specialistic activity. The critical area around which Ricœur works, with authentic intellectual distress, is that of the passage from perception to linguistic expression and then to the cultural formation of a world of common sense, supported by shared interpretations of past facts. If the aporias of the relation of resemblance, characterizing memory, can be confronted by giving the proper weight to testimony, the aporias of the fiduciary relationship, characterizing testimony, can be confronted by taking into consideration the perceptive roots of the memory, its immediate correlation in temporal terms with a previous event endowed with causal power (an action having left a trace), certainly not produced by consciousness itself, which is “impacted” by it. The result is an ambivalent image filled with the tension of experience and knowledge of the past, of which from time to time analysis favors different moments, oscillating between the two poles of construction and reconstruction (of doing or enduring), of creative autonomy and dependence on unyielding data. Ricœur meticulously explores them during various stages of his research, with different, sometimes contrasting results. At a certain point in its development, La mémoire, l’histoire, l’oubli presents a type of vindication of the theoretical role of the category of the trace, generally considered with suspicion, as the source of aporias more than philosophical solutions. This is a theme that flows, fortifying it, in the direction of what Ricœur

Memory, History and Ordinary Knowledge

287

himself has defined critical realism in history. Called into question in order to release it from the obligatory result of a form of ingenuous realism, the concept of the trace proves to be resistant and steadfast; it is reinforced precisely by confronting the communicative dimension of testimony and, in a certain sense, finally takes its “revenge” by finding its specific theoretical space. It restores a “triangular” image, as it were, of experience and knowledge, not reducible to the intersubjective dimension of the processes of reciprocal communicative control. If the actors in the process of communication and control of testimony occupy two vertexes of the triangle of representation of the past, the trace indicates the third vertex as an insuppressible, extralinguistic referential pole of experience, without which the model would certainly be inadequate and misleading, or unbalanced in an idealistic or relativistic sense. Neither the fidelity of memory nor the epistemic truth of history belong to an encounter played out only in the communicative meeting place by a plurality of cognitive agents reciprocally exchanging, controlling and sharing their respective experiences. The construction of memory, individual and collective, depends both on the dialogue between witnesses and the power of the original impression, the impact and the passivity which primarily characterize the experience of memory from its origins; in a parallel sense, the referential aspect of discourse regarding the past is not endangered by the irrecoverable absence of the ultimate referent of memory. It is not a question of illusion référentielle or effet de réel, in the sense of structuralistic semiotics, nor a tropological-rhetorical effect without criteria for distinguishing it from literary fiction, as White believed.

CHAPTER EIGHTEEN GRASP OF CONCEPTS: COMMON SENSE AND EXPERTISE IN AN INFERENTIALIST FRAMEWORK PIETRO SALIS

Inferential role semantics is the idea that conceptual content is determined by its role in reasoning: a concept C is given by the premises we use to draw C, and by the conclusions we draw from C. This theory entails that concepts are rich and stratified, and that conceptual competence involves the mastery of a wide range of inferential transitions. A problem shows up about the grasp of concepts. Should we conceive grasp of C as a full understanding and mastery of C? Should we understand grasp as an all-or-nothing problem? Should we be able to master all the inferential transitions determining C to grasp it? An affirmative answer to these questions would involve problems both for learning concepts and communication. I propose, in this context, a distinction between a common sense version of the grasp of concepts and an expert one. The first idea is that grasp does not entail the mastering of all the inferences which are constitutive of C, but just a few. Expert grasp, on the contrary, is a full and qualified mastery of the inferential transitions involving C. I present and discuss cases that support this distinction, and then try, on this basis, to shed light on the connection between common sense knowledge and specialist knowledge in terms of conceptual (inferential) matters.

1. Inferentialism and Conceptual/Linguistic Competence According to inferential role semantics, concepts are determined by their inferential roles, the distinctive roles they play in reasoning. Such a conceptual role is identified, for the concept C, by the set of inferences

290

Chapter Eighteen

comprehensive of the premises and conclusions where C is involved in the right way. A theory that directly appeals to these ideas is that of Robert Brandom (1994).1 The basic idea is based on the generalization of Gentzen’s style introduction and elimination rules for logical connectives to conceptual contents (and to linguistic meanings as well).2 The circumstances of application (both linguistic and non-linguistic) of C play the role of introduction rules for the concept; consequences of the application of C will play the role of elimination rules for the concept.3 An inferential role thus encompasses all the premises that infer C and all the conclusions we can draw from C. These inferences give substance to the wittgensteinian idea that meaning is use, and inferential role semantics is explicitly endorsed as our best account of use (in reasoning). Indeed, introduction and elimination rules – circumstances and consequences of application – play the roles of the peculiar rules of use for our concepts/words. The inferences that determine content are those that Brandom and Sellars call materially good.4 Their validity should not be intended just as logical (or formal), but also as material: it (also) depends on the nonlogical concepts involved. For instance, if I infer from “whales are cetaceans” that “whales are mammals”, the inference is valid in virtue of the content of “cetacean” and “mammal” (and not just in virtue of its logical form). Thus, the inferences responsible for the content of our concepts are not the standard inferences of deductive logic (this idea is entailed by default by inferential role semantics – since concepts/meanings

1

Brandom provided a pragmatic account of how these inferences determine the content of speech acts within what he called the social practice of “giving and asking for reasons”. This game determines what the correct inferential moves are and which beliefs are correct in the light of their consequences and of the best reasons at hand (from the perspective of the participants in the practice). The score-keeping activity implicit in such a practice determines the normative statuses of our linguistic utterances and behavior: we should distinguish between “commitments” and “entitlements”. This social activity is supposed to enable us to make explicit, through inference, the normative difference between undertaking a particular commitment and being entitled to something. 2 A basic example is conjunction: given A and B, we can introduce š, and then we have AšB (introduction rule for š–Iš); AšB is the basis to infer both its conjuncts A and B by eliminating š (elimination rule for š–Eš). 3 See Casalegno (2004), Fodor and Lepore (2007), and Williamson (2003, 2007, 2012) for different kinds of criticisms of this idea. See Boghossian (2003, 2012) for a defense. 4 See Brandom (1994: 97-105) and Sellars (1953).

Grasp of Concepts

291

are determined by inferences, these turn out to be good on the basis of the concepts involved).5 A particularly controversial aspect of this theory is its holistic structure. The content of a concept C is a function of the contents of other concepts (the inferences constitutive of C). This entailment appears to be troublesome, at least prima facie, since an inferentialist theorist is pressed to find out the specific criteria to circumscribe the scope of this conceptual holism: without the individuation and adoption of these criteria, semantic holism can be expanded as the view that the content of the concept C is determined by almost all the other concepts in the conceptual system CS to which C belongs (the linguistic counterpart of this point would involve almost an entire language L as determining the meaning of an expression E). Let us call this view SSH (strong semantic holism), which is the most radical option we can adopt to understand semantic holism. An inferentialist should weaken such a radical option to avoid a list of wellknown problems, and then she must try to diminish the scope of the peculiar holism generated by the adoption of inferential semantics. Let us first have a look at the list.

2. Semantic Holism: Prima Facie Problems with Grasp, Learning, and Communication The list of difficulties, highlighted especially by Fodor and Lepore (1992, 1993), Jackman (1999) and Whiting (2008), is as follows: (Grasp of concepts)–First of all, grasping a concept C seems to entail the complete mastery and understanding of all the inferences (relevant for C). But one can go too far in following the long chain of inferences: where does the entailment in the overall language/conceptual system stop? We require the mastery and the understanding of all the other concepts to grasp C. This is a problem, since the conditions for grasping concepts become very hard to meet. Here we need an explanation of the way in which such a grasp could be accommodated holistically. (Concept learning)–SSH also faces problems in explaining concept learning. Imagine a child learning her first concept C: how can it be possible to learn C if its conceptual content is determined by certain inferences that the child is ignoring? Our insights about learning as a cumulative process would be impossible to accommodate (we would in 5

Material inferences also differ from standard deductive inferences because they are generally non-monotonic (new premises can affect the goodness of an inference). See Brandom (2000: 87-89, 2008: 106).

292

Chapter Eighteen

fact need to learn concepts before acquiring the preliminary conditions that allow us to learn them). If we want to preserve these insights about concept learning, then we should avoid SSH or provide a quite different account suitable to accommodate both semantic holism and concept learning. (Content sharing and stability)–SSH faces further problems when it comes to explaining our ability to share contents. These are usually considered the biggest worries about holism: 1) (Constitutive instability)–The most controversial aspect of SSH is the underlying thesis of instability of meaning and of content of belief. Every form of SSH systematically generates this instability. This means that what is unacceptable in holism is that it seems obvious that we have different beliefs (or commitments) and, at the same time, that we share the same contents/meanings. SSH, in fact, entails that semantic contents vary depending on the beliefs of the speakers (Jackman, 1999). This means that holistic contents cannot be stable; rather, they must be something shifting depending on people’s beliefs.6 Therefore, we have reason to suppose that this instability is a feature strictly related to SSH. 2) (Communication)–If we endorse SSH, the inferences and beliefs that we take to be constitutive of the concept C are often different, and this entails, not only that C can be slightly different for you and me but, in principle, also that it can turn out to be radically different. This entailment of SSH no longer seems enough to warrant successful communication between speakers (there is, in fact, the possibility of meaning different things by using the same words). Here again, to explain communication, we would need to avoid SSH or to provide a theory of how it can be possible to explain communication holistically.7 6

See Jackman (1999: 362): “Unless two people shared all their beliefs, they wouldn’t share any of them. For instance, if Peter believes that elephants are afraid of mice, and Mary doesn’t, then what may have initially appeared to be shared beliefs – both claim ‘elephants are big animals’–would turn out to have different contents in virtue of the two believers meaning different things by ‘elephant’”. See also Whiting (2008) for different examples. 7 Brandom developed a personal holistic account of communication and claimed that communicating would be better understood in terms of “cooperating in joint activity” and that it should involve “coordinating social perspectives by keeping deontic score [the argumentative score between discussants] according to common practices” (Brandom, 1994: 479).

Grasp of Concepts

293

3. Grasp of Concepts. An Anti-Individualist Suggestion Grasp of concepts is the special problem on the list that will be the focus of this paper. At least in a sense, it seems to have consequences for all the other problems related to SSH. In fact, the peculiar holism entailed by inferentialism seems to depend strongly on the preliminary conception of grasp that one endorses. A further problem is that inferentialism lacks an official account of which inferences are constitutive of conceptual contents. If we cannot say which these inferences are, we find ourselves pressed to admit, somehow, that all are constitutive. This point is quite sloppy for our main theme, since content constitutiveness widely overlaps with the grasp of concepts: it is a prima facie reasonable view that to grasp a concept C we have to know the inferences which are constitutive of C. Otherwise, our grasp will be at best partial and incomplete. This becomes even more evident as one thinks of the grasp of concepts as an all-or-nothing matter: since to grasp a concept, one has to master its complete inferential role, grasp will entail the kind of holism that is conceptually disastrous (everything being constitutive). Thus, this conception of grasp (together with inferential semantics and the missing story about which inferences are content constitutive) seems to lead to the holism we should avoid in semantics. Should we understand grasp of C as a full understanding and mastering of C? Should we conceive grasp as an all-or-nothing problem? Should we be required to master all the inferential transitions determining C to grasp C? This line of reasoning can already be undermined by proposing a conception of grasp that does not commit us to these holistic entailments (but without answering the “which inferences determine content” question). I believe, and it seems to me that I am supported by good evidence on this, that grasp is not (and cannot be) an all-or-nothing matter; rather, it is a matter of degree. It seems actually sound and coherent to distinguish between at least these two basic (anti-individualist)8 levels of grasp:

8

See Burge (1979) for an introduction to the very idea of anti-individualism, and Putnam (1975) for the ground-breaking idea of “division of linguistic labor”. Very roughly, the idea is that common speakers do not entirely master the concepts they use, and that in case of doubt and controversy, they “defer to experts” (who actually master the relevant concepts). See Salis (2012) for a different route relating inferentialism and anti-individualism.

294

Chapter Eighteen

1. “Minimum grasp”, regarding a neophyte’s possession and use of concepts, does not require the mastery of inferential roles, but just a few inferences. 2. “Full grasp” is equivalent to expertise; here, the expert has a qualified mastery of the inferential transitions involving a concept. We should also consider a medium level of grasp, (it does not matter how vague), exemplified by those who are working to become experts (like graduate students). A first example, which is worth discussing to defend such a distinction, is the following: while it is strongly plausible that linguistic children are people we can talk with, we do not consider them experts. We presuppose in practice that grasp is a matter of degree, since there are radical cases where our interlocutors are not experts (in any controversial sense of “expert”). But that alone amounts to admitting that one can understand us without having a full grasp of the relevant concepts. A first moral tells us that speakers need not be experts to communicate (and to understand) with each other. A second example worthy of a little discussion is slightly different and involves the very idea of inferential roles. If we think about the inferential roles of many descriptive or technical concepts, we easily realize that we do not master all the constitutive inferences, but just a few, and, to know more about these, we have to ask experts (think about legal, clinical, physical, and biological concepts and so on and so forth).9 Thus, common sense grasp and expertise are different things, since we can grasp concepts without being experts; grasp, then, concerns the first basic judgments and inferences we learn to perform with certain words (and not the whole framework of implications).10 Another feature of this suggestion is that our common sense mastery of inferential roles then has to be understood as minimum grasp. We should not (and actually do not) require expertise about anything to attribute linguistic/conceptual competence to speakers, and so minimum grasp is enough for language users. Speakers need not be experts to be decent language users (by ordinary standards). This default competence is also the degree zero, the a priori condition, for eventually becoming an expert 9

In the same way, someone who is not a doctor can even come to think that “arthritis” can concern one’s thigh. See Burge (1979). 10 Brandom seems to explicitly endorse this kind of view, even if maybe a little bit quickly, in (2000: 64), where he talks about grasping concepts: “the metallurgist understands the concept tellurium better than I do, for training has made her master of the inferential intricacies of its employment in a way that I can only crudely approximate.”

Grasp of Concepts

295

about certain conceptual systems/items. Think about the concept “copper”: we can have a common sense inferential competence about it, but then we can, over a further period of time, become experts about the peculiar physical properties of copper (and so we can become competent with regards to all the relevant specialist inferences involving “copper”). This is a coherent account about which inferences constitute common sense grasp: it is not an account of particular inferences, but rather a matter of statistics about the things we know (there are merely things that most of us know as common speakers). Vice versa, there is a set of specialist inferences that most of us ignore as common speakers. Far from being a brute fact about knowledge, this is something related to our (common, actual) practices, the relevant epistemic and cultural context, the effective way in which we learn language. If this is our conception of grasp, then the only version relevant to common speakers will be, from a semantic perspective, the minimal one. Grasp is minimum grasp. Grasp of concepts and common sense knowledge seem to be somehow intertwined by this view: the inferential roles mastered by the layman and those mastered by the expert reveal, in fact, two distinct realms of knowledge. Somehow, the criteria adopted to circumscribe minimal grasp of concepts are helpful in trying to spell out what common sense knowledge amounts to. From an inferentialist point of view, common sense knowledge appears to be a certain amount of information that every one of us possesses and shares as a simple speaker, as a mere user of a natural language. This is the generic amount of information that comes together with learning and using a language. This does not mean that all people have the exact same amount of information: there is a core of information (and indeed of inferences) that is shared by, and distributed throughout, a community of speakers and that is mastered in an approximate and flexible way. Common speakers do not equally possess the information they share (possession of knowledge/information, as obvious as it may be, admits degrees). Another important dimension that distinguishes common sense knowledge is that it is not perfectly explicit and transparent to its users and owners (it is not a form of knowledge that). This is perfectly fine with Brandom when he says that language is a social practice where norms are implicit; speakers, as participants in the practice, have only implicit knowledge of the peculiar rules they obey when talking with each other. The explicit dimension requires meta-vocabularies, hence theory, and this is a genuine property of expertise (though experts may represent this dimension in many different ways, and not just as a set of inferences – theories are many).

296

Chapter Eighteen

On the contrary, specialist knowledge amounts to a wide set of specialist inferences that are, in a more general and generous way, the inferences that experts are trained to master by their education and work (the specific epistemic profile, which is not included in common sense knowledge). Even for expertise, we can develop an analogue point about the possession of this specialist core of knowledge within a community of experts. Again, there are in fact degrees of knowledge (otherwise, the very assessment of scientific work would simply be impossible).11 Finally, this kind of knowledge possesses the explicit dimension of theory by default; this aspect could, in principle, be developed in a further inferentialist direction. The core discursive practice of “giving and asking for reasons” is, in fact, the basic practice that can put simple speakers – trying to cope with other speakers and with reality – on the trail of the explicit specialist dimension: disagreement in argument and curiosity often work together in pressing us to improve our basic common sense knowledge and understanding of the issue(s) at stake.

References Boghossian, P. (2003) “Blind Reasoning”, Aristotelian Society, suppl. vol. 77: 225-48. —. (2012) “Inferentialism and the Epistemology of Logic: Reflections on Casalegno and Williamson”, dialectica, 66: 488-97. Brandom, R. (1994) Making It Explicit. Cambridge (MA): Harvard University Press. —. (2000) Articulating Reasons. Cambridge (MA): Harvard University Press. —. (2008) Between Saying and Doing. Oxford: Oxford University Press. Burge, T. (1979) “Individualism and the Mental”, Midwest Studies in Philosophy, Vol. 4: 73-121. Casalegno, P. (2004) “Logical Concepts and Logical Inferences”, dialectica, 58: 395-411. Fodor, J. and Lepore, E. (1992) Holism. A Shoppers’ Guide. Oxford: Oxford University Press. —. (eds.) (1993) Holism. A Consumer’s Update. Amsterdam: Rodopi. —. (2007) “Brandom Beleaguered”, Philosophy and Phenomenological Research, Vol. LXXIV, N°. 3: 677-91. 11

Furthermore, there is a constructive section to these practices: vocabularies and skills that tend to expand knowledge in new directions, the realm of new hypotheses and the growth of the amount of empirical data.

Grasp of Concepts

297

Jackman, H. (1999) “Moderate Holism and the Instability Thesis”, American Philosophical Quarterly, Vol. 36, No. 4: 361-69. Putnam, H. (1975) “The Meaning of ‘Meaning’”, Minnesota Studies in the Philosophy of Science, Vol. 7: 131-93. Sellars, W. (1953) “Inference and Meaning”, Mind, vol. 62, n° 247: 31338. Salis, P. (2012) “Inferenzialismo, pratiche argomentative e oggettività”. RIFL–Rivista Italiana di Filosofia del Linguaggio, vol. 6, n° 3:108-20. Whiting D. (2008) “Meaning Holism and De Re Ascription”, Canadian Journal of Philosophy, 38: 575-99. Williamson, T. (2003) “Understanding and Inference”, Aristotelian Society, suppl. vol. 77: 249-93. —. (2007) The Philosophy of Philosophy. Oxford: Oxford University Press. —. (2012) “Boghossian and Casalegno on Understanding and Inference”, dialectica, vol. 66: 237-47.

CHAPTER NINETEEN THE EPISTEMOLOGY OF TESTIMONY: A HERMENEUTICAL KEY VINICIO BUSACCHI

I am totally in agreement with Robert Brandom's point of view (quoted by Paul Faulkner): “If one individualises the space of reasons, forgetting that it is a shared space within we adopt attitudes towards each other – and so does not think about standings in the space of reasons as socially articulated, as potentially including the social difference of perspective between attribution and undertaking commitments, that is between your standing and mine – then one will not be able to understand knowledge as a standing in the space of reasons.”1 Strong is the correlation (or the “dynamic”, according to Faulkner) between individual knowers and communities of knowledge. So close that we can say that there is no knowledge without community. Why? Because most of what we know comes from transmission. Right here arises the question of testimony. This aspect is widely recognised today: the fact of considering, not just that a certain amount of knowledge are beliefs that we get from what others tell us, but that “your knowledge depends pervasively on the word of others”.2 Knowledge is a common good, and accepting/believing the testimony of others (orally, in writing, direct, indirect, present, past) is a way to acquire essential knowledge (not only about the world, but also about ourselves). There are different types of evidence. However, the schema of acquisition and development of knowledge is the same in every area. Without an immense heritage of knowledge and without indirect experienced evidence, socially mediated, historically acquired, there would not even be 1

P. Faulkner, “The Social Character of Testimonial Knowledge”, The Journal of Philosophy, 97 (2000): 601. 2 P. Lipton, “The Epistemology of Testimony”, Studies in History and Philosophy of Science, 29 (1998): 1.

300

Chapter Nineteen

scientific knowledge. (Many accidents in chemical laboratories can be prevented by others’ negative experiences, accepting the testimony that, for example, if you pour water into concentrated sulfuric acid you may get a nasty acid burn.) Medical knowledge, too, is largely gained through personal experience; much of it is acquired by others, and can be found in published reports, or indirectly in new forms of medical treatment, in new approaches; even in terms of ordinary life and ordinary knowledge, the testimony of others can provoke a transformation of our stable habits, our understanding, our behaviour (for example, a certain accident may persuade us that it is just not safe to pass too close in front of or behind a Pit Bull). As says Peter Lipton, we live in a “sea” of other people’s assertions and “little if any of our knowledge would exist without it”.3 On the one hand, the theme of testimony can be considered a classic of philosophy - it is already present in St. Augustine and large speculative development happened with Locke, Hume, and Thomas Reid and consequently with Kierkegaard and many other Christian philosophers. On the other hand, only recently, in our contemporary era, has epistemology rehabilitated the role of testimony, of belief, and the role of the “communities of knowledge” in the formation of individual knowledge and community knowledge, whether ordinary or scientifically specialised. Thus, not many sociological notions have been acquired in the conceptual vocabulary of the philosophy of science and epistemology (such as testimony, confidence, authority, deference...). So today there is no epistemology without social epistemology.4 But there are many differences between the issue of the social impact of ordinary knowledge and the issue of the social impact of scientific knowledge. In the latter area, which is very wide and varied, the role of testimony comes in a better regulated territorium, under a much more restricted chain of conceptualisation, understanding, technique and praxis – as is already focused upon in thinkers such as Descartes, Locke, Hume, and Kant. If Descartes distinguishes between metaphysical certainty (not testimonial) and moral certitude, where testimony plays a certain role, for Locke, the assumption of a belief must pass an examination that is both procedural (epistemological) and ethical (I am referring to the principles of his ethics of belief5). Hume, not so differently from Locke, does not confer special 3

Ibid. Cf. G. Origgi, “Peut-on être anti-réductionniste à propos du témoignage?”, Philosophie, 88 (2005): 47-57. 5 A passage from the Essay (Locke, Essay, IV, 15, 5): “«Probability wanting that intuitive evidence which infallibly determines the understanding and produces certain knowledge, the mind, if it will proceed rationally, ought to examine all the 4

The Epistemology of Testimony

301

status to testimony: in the end, its epistemic validity does not go beyond our ordinary inductive inferences.6 Even Kant tends to give it a limited role – a neutral, if not negative, role (in relation to the scientific acquisition of knowledge). In fact, the cause of a belief may not exactly be a reason for it: Every prejudice is to be regarded as a principle of erroneous judgments, and from prejudices arise not prejudices, but rather erroneous judgments. Hence one must distinguish the false cognition that arises from prejudice from its source, the prejudice itself. (…) Sometimes prejudices are true provisional judgments; what is wrong is only that they hold for us as principles or as determining judgments. (…) For even if we can accept some cognitions, e.g., immediately certain propositions, without investigating them, i.e., without examining the conditions of their truth, we still cannot and may not judge concerning anything without reflecting, i.e., without comparing a cognition with the power of cognition from which it is supposed to arise (the sensitivity or the understanding). If we accept judgments without this reflection, which is necessary even when no investigation occurs, then from this prejudices arise, or principles for judging based on subjective causes that are falsely held to be objective grounds.7

In different ways, these authors established a certain hierarchy between information gathered directly (perceptually, cognitively) and information obtained from others’ experiences. All of them somehow opposed the methodological approach to the indirect acquisition of knowledge (acritical information, testimonies, information orally communicated). But with these modern thinkers, a close critical relationship between the socioepistemological and ethical question of reliability and the scientific question of fallibility was constituted. Pursuant to Stéphane Chauvier, two problems arise as radical in the epistemology of testimony – both clearly determined from the assumption (now large) of testimony being an grounds of probability, and see how they make more or less for or against any proposition, before it assents to or dissents from it; and, upon a due balancing the whole, reject or receive it, with a more or less firm assent, proportionably to the preponderancy of the greater grounds of probability on one side or the other».” Cf. N. Woltersdorff, John Locke and the Ethics of Belief (Cambridge: Cambridge University Press, 1996). 6 Cf. D. Hume, An Enquiry Concerning Human Understanding, ed. Eric Steinberg (Indianapolis: Hackett Publishing Company, 1977 [1748]), Section 10, Of Miracles, P. I, 74-75. 7 I. Kant, Lectures on Logic (Cambridge: Cambridge University Press, 1992, 2004), 578-579.

302

Chapter Nineteen

element of great importance in the field of the epistemology of ordinary knowledge, as in the field of scientific knowledge. The first and most profound advice is that of the comparative value of knowledge from testimony (being the mode or form of knowledge among other possible ways). The second relates to the procedural aspect of the identification and evaluation of the different operations which establish/determine the degree of credibility of certain acquisitions.8 I totally agree with the critique of Chauvier, although the perspective that I intend to develop here deviates somewhat from his (while aiming to achieve a valid solution for both problems). I would say that the work of the historian and historical knowledge are the best fields for the critique of the role of the testimony in the creation of scientific knowledge and the relationship between ordinary knowledge and historical knowledge.

1. The Mixed Epistemology of the Historical Explanation The French philosopher Paul Ricœur rejects the radical dichotomy of reason and cause: rather, reasons and causes are placed at the bottom of a series of modulations and forms, including one in which reasons are causes. Historical knowledge must consider, taking in his apparatus and procedural knowledge, the whole “conceptual network of action”: essentially, action, intention, motivation and agent. In the Discours de l'action, Ricœur, taking into consideration the work of von Wright, points out that “in the field of action, the opposition between explanation and understanding, reproduces the same aporias and the same contradictions that, on the one hand, are present in the theory of the text (...), and on the other hand, in the theory of history (in which the German and the English traditions converge in the work of authors such as Collingwood, Dray, Peter Winch etc)”.9 As far as Ricœur is concerned, von Wright “introduces a whole range of possibilities between the strictly causal explanation and explanation purely teleological (internalized in the consciousness of the agent in the form of an “intention to do”). In his view, between these two extremes you have to include quasi-teleological explanations, how the process of homeostatic regulation are (that simulate the teleology without adding anything to the causality). But it must also be considered a quasicausal explanation, as the explanation by the impulses, irresistible desires

8

Cf. Ibid. Cf. P. Ricœur, “Le discours de l’action”, in D. Tiffeneau, La sémantique de l’action (Paris: CNRS, 1977), 104.

9

The Epistemology of Testimony

303

etc., which are actually teleological explanations dissimulated”.10 I think that the analysis developed here by Ricœur offers substantial confirmation and an improvement to his work of the nineties, especially in the great synthesis La mémoire, l'histoire, l'oubli. In this work, in particular, the notion of trace is very relevant, along with the complex issues related to it, to retaking and re-expressing the problematic question of the historical explanation – or, rather, “historical knowledge”. In La mémoire, l’histoire, l’oubli, Ricœur confirms its epistemological perspective – explaining and understanding – on the construction of historical knowledge, always centred on a particular focus on human action. Historical knowledge builds its confidence on a double movement, parallel and circular: [1] the referential credibility of history is given by truthfulness, from the connection between real/factual facts and specific claims within narrative constructions, logically structured; [2] historical facts are not only products made from speeches, but “if the historical explanation has mastered the facts as producing them, then the event remains the hypothetical support of the reorganisation of the events on the time axis”. An important passage (containing a reference to von Wright): The great mass of historical works unfold in a middle region where disparate modes of explanation alternate and sometimes combine in an unpredictable way. (...) In this regard, we can take the quarrel that arose at the beginning of the twentieth century around the terms “explanation” and “understanding”, taken as antagonistic to each other, as surpassed. Max Weber, in combining explanation and understanding from the start, was perspicacious in elaborating the leading concepts of his social theory [ref., M. W., Economy and Society: An Outline of Interpretive Sociology; V. B.]. More recently G. H. von Wright, in Explanation and Understanding, constructed a mixed model of explanation for history that made causal (in the sense of law-like regularity) and teleological (in the sense of motivations capable of being rationalized) segments alternate. In this regard, the correlation (...) between the type of social fact taken as determining the scale of description and reading, and temporal rhythm can offer a good guide in the exploration of differentiated models of explanation in their relation to understanding.1111

Testimony is now submitted to the discipline (explicativecomprehensive) of the historiographical process shown here; a discipline that only a mixed epistemology can incorporate. Testimony with the value 10

Ibid., 109. P. Ricœur, Memory, History, Forgetting, trans. K. Blamey and D. Pellauer (Chicago & London: The University of Chicago Press, 2004), 185. 11

304

Chapter Nineteen

of “knowledge scientifically relevant” and with the value of “knowledge historically established” is ordinary knowledge subjected to the critical scrutiny of a rigorous process. How far are we from the principles of the Lockian “ethics of belief”? I would say, “not very”. In some cases, the question is not about “ethics” but about procedural operations; in other cases, however, it is ethical in every way. But this is not the exception. All scientific knowledge, as a procedure that declares itself rigorous, constantly puts forward the ethical dilemma of its acquisitions. The historian may be biased in its reconstruction, as the witness in his story; the scientist can distort the empirical data that he uses as proof of verification, as the historian can produce false documents, false testimony, or omit some elements, some donations...

2. The Epistemology of Ricœur's Hermeneutic Arc The connection, on the one hand between ordinary knowledge and scientific knowledge (contextually, historical knowledge) and, on the other, between testimony in the ordinary communication sense and testimony in the specialised knowledge sense, raises a problem with the theory of knowledge similar to that considered by John Locke (regarding the social role of testimony), with its ethical propositions of belief. However, rather than an ethical dilemma, here, I speak of a procedural dilemma – although I agree with Pascal Engel, that it is not entirely correct to consider a radical opposition between the two terms (because the acceptance of a testimony to faith, and the same presumptive right on which it is based, is by no means incompatible with the fact that the testimony is strengthened by data and evidences that you have).12 In fact, given the vastness, now fully recognised, the massive and incisive presence of testimony in every sphere of human knowledge, it seems that no epistemological model can exempt itself any longer from a certain hermeneutical dimension. With different modulation, and greater or lesser hermeneutical emphasis (depending on who you talk to about hard sciences or soft sciences), we must understand that all knowledge works (to different degrees) between explanation and understanding. Here comes into play the full address of the epistemological structure of the Ricœurian philosophical process. His theory of the hermeneutic arc, in fact, is not only attributable to the theory of text and action, and to disciplines like history, psychoanalysis and sociology; his theory is the hub of his own

12

Cf. P. Engel, “Faut-il croire ce qu'on nous dit?”, Philosophie, 88 (2005), 59.

The Epistemology of Testimony

305

philosophy (understood not in terms of speculative contents and questions, but from the point of view of the metodus, of logic). Ricœur gave the following determination, methodological and philosophical, of his work: (a) a “reflective philosophy” angled (b) “from the perspective of Husserl's phenomenology” as its (c) “hermeneutical variation”.13 While it is undeniable that Ricœur has remained faithful to the phenomenology, on the other hand, the development of his epistemological model of the hermeneutic arc has given to his philosophy much more of the critical hermeneutics than a description-based reflection. The idea of “critical hermeneutics” refers to the philosophical project of the young Habermas (a sort of evolution of the Frankfurt Kritische Theorie) and the debate between Habermas and Gadamer in the seventies (that is, the Critique of ideology vs. Hermeneutics of tradition).14 In basic terms, this critical hermeneutics is characterised by mixed epistemology and a strong interdisciplinary configuration. Here, there is not the simple element (Socratic?) of a philosophy constantly in dialogue with the sciences, but the project of a true interdisciplinary philosophy. Critical hermeneutics operates between different kinds of knowledge (scientific and non-scientific, specialist and non-specialist), and between models, theories and discourses fragmented (if not disintegrated); models, theories and discourses resistant to all comprehensive synthesis, requiring a flexible and transverse approach, capable of governing tensions and aporias. The methodological and theoretical character of the process of critical hermeneutics can be summarised as follows: (1) the ideal of research and dialogue within the “community of philosophers” (borrowed from the master, Jaspers); (2) the philosophical procedure according to which “all the books are open simultaneously”;15 (3) an interdisciplinary work, (4) focusing on “philosophical argument” (the philosophy of Ricœur claims a full disciplinary autonomy); (5) the reflexive-hermeneutical dynamic between philosophical and non-philosophical dimensions; (6) a search for connection with analytic philosophy; (7) a philosophical engagement in relation to the reality of life and in relation to political and social issues; (8) the placement of philosophy in the dialectic theory-praxis (philosophy, like science, is a theoretical practice); (9) the articulation of the philosophical process in degrees, in philosophical, methodological and thematic registers. 13

Cf. P. Ricœur, From Text to Action, trans. K. Blamey and J. Thompson (Evanston, Ill.: Northwestern University Press, 1991). 14 See P. Ricoeur, Hermeneutics and the Critique of Ideology (1973), in Ibid., 270307. 15 Ricœur, Memory, History, Forgetting, xvii.

306

Chapter Nineteen

3. Conclusion It seems clear that Ricœur’s philosophical model responds better than others to the evidence of the close relationship between ordinary knowledge and scientific knowledge, not only in that it is more suitable and effective than others at implementing the contents of ordinary knowledge in the context of a knowledge specialised and rigorous. My opinion is that ordinary knowledge has a constitutive familiarity with scientific knowledge, because I think scientific knowledge is no more than a sophisticated form of ordinary knowledge. The main difference lies in the high degree of organisation and rationality of scientific knowledge, and differences in purpose, sense, and meaning between scientific knowledge and ordinary knowledge. Issues such as those related to the discourse of testimony in the construction of historical knowledge, in fact, cannot be solved through the means of rigorous and empirical verification alone. This strategy is successful and appropriate in certain cases or moments, but not in others. The hermeneutic critique is versatile in response to the ethical principles of belief, in response to the requests of ethical knowledge, and in response to occurrences of a type exquisitely hermeneutic. .



LIST OF CONTRIBUTORS

Evandro Agazzi, Full Professor of Philosophy, Panamerican University, Mexico City Mario Alai, Associate Professor of Theoretical Philosophy, University of Urbino Maria Cristina Amoretti, Research Fellow, University of Genoa Mariano Bianca, Full Professor of Theoretical Philosophy, University of Siena Anna Maria Borghi, Associate Professor of General Psychology, University of Bologna Vinicio Busacchi, Associate Professor of Theoretical Philosophy, University of Cagliari Gustavo Cevolani, Research Fellow, University of Turin Vincenzo Crupi, Associate Professor of Philosophy of Science, University of Turin Massimo Dell’Utri, Full Professor of Philosophy of Language, University of Sassari Maurizio Ferraris, Full Professor of Theoretical Philosophy, University of Turin Marcello Frixione, Full Professor of Philosophy of Language, University of Genoa Paolo Labinaz, Research Fellow, University of Trieste Elisabetta Lalumera, Assistant Professor of Philosophy of Language, University of Milan-Bicocca



308

List of Contributors

Pier Luigi Lecis, Full Professor of Theoretical Philosophy, University of Cagliari Antonio Lieto, Research Fellow, University of Turin Antonio Livi, Emeritus Professor of Epistemology, Pontifical Lateran University, Rome Fabio Minazzi, Full Professor of Theoretical Philosophy, University of Insubria Paolo Piccari, Assistant Professor of Theoretical Philosophy, University of Siena Gaetano Piccolo, Professor of Metaphysics, Gregoriana University, Rome Pietro Salis, Research Fellow, University of Cagliari Lourdes Velázquez, Mexico

   



Professor of Bioethics, University of Anàhuac,