attention et cognition

7/27/2019 Attention Et Cognition

1/199


2/199

Attention Is Cognitive Unison


3/199

PHILOSOPHY OF MIND

Series Editor

David J. Chalmers, Australian National University

Thinking without WordsJos Luis Bermdez

The Conscious BrainJesse Prinz

Simulating MindsThe Philosophy, Psychology, and Neuroscience of MindreadingAlvin I. Goldman

Supersizing the MindEmbodiment, Action, and Cognitive ExtensionAndy Clark

Perception, Hallucination, and IllusionWilliam Fish

Phenomenal Concepts and Phenomenal KnowledgeNew Essays on Consciousness and PhysicalismTorin Alter and Sven Walter

Phenomenal IntentionalityGeorge Graham, John Tienson, and Terry Horgan

The Character of ConsciousnessDavid J. Chalmers

The SensesFiona Macpherson

The Contents of Visual ExperienceSusanna Siegel

Attention Is Cognitive UnisonAn Essay in Philosophical PsychologyChristopher Mole


4/199

Attention Is Cognitive UnisonAn Essay in Philosophical Psychology

Christopher Mole

12011


5/199

3

Oxford University Press, Inc., publishes works that furtherOxford Universitys objective of excellence

in research, scholarship, and education.

Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong Karachi

Kuala Lumpur Madrid Melbourne Mexico City Nairobi

New Delhi Shanghai Taipei Toronto

With offices in

Argentina Austria Brazil Chile Czech Republic France Greece

Guatemala Hungary Italy Japan Poland Portugal Singapore

South Korea Switzerland Thailand Turkey Ukraine Vietnam

Copyright 2011 by Oxford University Press, Inc.

Published by Oxford University Press, Inc.

198 Madison Avenue, New York, New York 10016

www.oup.com

Oxford is a registered trademark of Oxford University Press.

All rights reserved. No part of this publication 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 Oxford University Press.

Library of Congress Cataloging-in-Publication Data

Mole, Christopher, 1978

Attention is cognitive unison : an essay in philosophical psychology / by Christopher Mole.

p. cm. (Philosophy of mind)

Includes bibliographical references.

ISBN 978-0-19-538452-9 (alk. paper)

1. Attention. 2. Philosophy of mind. I. Title.

BD418.3.M64 2010128.2dc22 2009053059

9 8 7 6 5 4 3 2 1Printed in the United States of America

on acid-free paper
http://www.oup.com/http://www.oup.com/


6/199

To my parents, with love and thanks.


7/199

vi

Preface

There is a venerable tradition in philosophy, so venerable that evenDescartes satirizes it, of presenting ones work, not as a definitive state-ment, but as a contribution to an ongoing conversation. This book isnot a definitive statement. Nor, however, is the conversation to which

it contributes ongoing. There have been occasional rumblings, but theconversation about the metaphysics of the psychology of attention hasbeen more or less dormant for the century since philosophy and psychol-ogy split. This is not because philosophers and psychologists have beenignoring one another. The two disciplines have been talking, but for thelast century they have not been talking about attention.

The silence has been detrimental to both. Philosophers, by ignoringthe branch of psychology that has attention-related phenomena as its

central explananda, have tended to operate with a false picture of thepsychologists project. According to this false picture psychology gives usa theory in which cognition is thought of as a set of computations, real-ized somewhere in the frontal lobes, while the perceptual processes andthe action-executing processes, located nearer to the back of the head,orchestrate the inputs and outputs for these frontal, cognition-constitut-ing computations. As a picture of psychologys theory of the mind thisis misleading. Except when it borrows philosophys false picture for its

self-conception, psychology does not take the cognitive aspects of themind to be implemented, separately from perception and from actionexecution, in the poorly understood neural centers at the front of thebrain. Certainly the front of the brain is doing cognitive work, but thebits of cognition for which psychology has well-worked out theoriesthe bits to do with attention and memoryare understood to be instinctwith the perception and action processes at every stage.

Psychologists, for their part, have been investigating attention with-

out regard to philosophy, and so have spent the last century blunting eversharper empirical tools against what is fundamentally a metaphysicalproblem. The psychologists mostly know that they have been doing this,but they know also that the underlying problem is somehow not theirsto address. The result is that a great deal of ingenious research succeedsin describing the ways in which various forms of attention behave, andsucceeds in accounting for that behavior by reference to its basis in thebrain, but nobody knows quite how all of this adds up to an explanation

of attention. The literature is loaded with protestations, caveats, andcomplaints. We are told that attention is a family resemblance term, or


8/199

Preface vii

that it is ambiguous, or hopelessly folksy, or ill-defined, or nonreferring.But attention remains the best name that we have for the subject matterof cognitive psychologys central, defining research project.

This book presents a single unified theory of attention, intended toapply to attention in all its forms. According to this theory, attentionis not a family resemblance term, nor is it ambiguous, or folksy, or ill-defined, or nonreferring. Nor, however, is it explicable simply by iden-tifying and describing the processes that constitute it. The theory saysthat the relationship of attention to the cognitive processes executedin various parts of the brain is analogous to the relationship betweenunison and the individual performances of the members of an orchestra.

Just as there is no place in the orchestra where unison sits, so, accordingto this theory, there is no place in the brain where attention is located.A similar point applies to particular cognitive activities: Just as there areno particular melodies that all instances of unison involve, so there areno particular patterns of neural activity that we should expect to find inall instances of attention. The fact that a person is paying attention doesnot, by itself, entail anything specific about which cognitive processesare taking place or where.

Nor is there an entailment running in the other direction. If attentionis cognitive unison, then facts about specific cognitive processes do not,by themselves, entail facts about attention. Just as the execution of aparticular activity on the part of a musician may be unison constitutingon one occasion and not on another, depending on how it is related tothe things that his fellow musicians are doing, so a cognitive process thatplays a role in attention in one context may, in a different context, beinstantiated without attention. In this way the cognitive unison theory

entails that no particular cognitive activity is sufficient for attention, oreven a very reliable indicator of it.

The contribution that this theory is intended to make to our under-standing of attention is a contribution to our understanding of its meta-physics. It is the metaphysical relationship between attention and itsconstituent processes that psychologists have failed to get clear on, andit is the failure to get clear on this that has led to confusion about thenature of attention and about the way in which its explanation is to be

accomplished.There is more than one reason why metaphysical theories are a hard

sell when the audience is composed of psychologists. The theme of meta-physics avoidance recurs throughout psychologys history as a science,often with some justification. Even the most enthusiastic practitioners ofmetaphysics would admit that the period when psychology broke fromphilosophy was a period in which metaphysics had got itself into a mess.It is therefore understandable that psychologists have been steering clear

of metaphysics ever since. But the happy fact is that over the course ofthe twentieth century the mess that metaphysics had got into was sorted


9/199

viii Preface

out. Metaphysics, as we have it now, gives us tools that can usefullybe deployed on foundational issues in psychology. The development ofthese tools has, perhaps inevitably, upped the entry requirements for the

metaphysical debates. Those debates now tend to be incomprehensibleto those not familiar with the issues and with the accompanying vocabu-lary. And by now few psychologists are interested enough in what hasbeen going on to think that the entry price to the metaphysical debatesis worth paying. The result is that most psychologists think that, oncewe have all agree to be physicalists, there are no interesting metaphysicalquestions about mental entities to be asked. Metaphysical, they think,is just a show-offs word for spooky.

I would like to think that these misconceptions will have been aban-doned by the time the reader gets to chapter 7 of this book, but I amalso aware that the chances of her ever getting there can be increased bysaying something now. Metaphysical does not mean spooky. It is nota word to say in a hushed voice when the door is shut. By metaphysi-cal all that I mean here is basic in the sense of basic even from thepoint of view of physics. The physicist, most of the time, can take it forgranted that there are distinctions between existence and nonexistence,

between objects and events, intrinsic and extrinsic properties, actualityand mere possibility, and so on. The job of the metaphysician, insofaras we are concerned with him here, is to give an account of what thesebasic categories amount to. My contention is that questions pertainingto such basic categories arise in psychology too, and that they matter forthe business of building explanations. This has the consequence that thepsychologist cannot avoid questions about the metaphysical category ofthe items that she seeks to explain.

There is much in the cognitive unison theory that creates a nuisancefor the project of studying attention with the methods characteristic ofcurrent cognitive psychology, but it would be quite wrong to think thatthe theory is intended as part of an attack on that science. The workof cognitive psychologists is indispensably valuable, and it informs theargument of this book at almost every step. The lesson for the cognitivepsychologists is not that their work is no good, but that there is a need todeploy the resources of analytic metaphysics in order to understand what

that work tells us. For some psychologists this will come as unwelcomenews, but it cannot be avoided. The giant leaps forward that were takenby experimental psychology and by analytic metaphysics in the twenti-eth century were leaps in one anothers direction: Psychology steppedout into territory where metaphysical questions were unavoidable, and,at much the same time, philosophy got straight on the matter of how toaddress those questions.

The cognitive unison theory of attention was first presented in the dis-sertation that I defended at Princeton University in September 2005.My fellow graduate students at Princeton, both in philosophy and out of


10/199

Preface ix

it, were unfailingly inspiring and supportive throughout my time there.Huge sections of this book contain ideas that I could only have workedout in their company. I owe huge thanks to every one of them. The dis-

sertation was supervised by Sean Kelly, who patiently guided me out ofthe blind alleys that I repeatedly charged down. Several other facultymembers also read and commented on drafts of chapters and on vari-ous half-baked thoughts. They, too, made huge differences, and with-out them there are several confusions from which I would never haveescaped. I am very grateful to them. My time at Princeton was madepossible by the William Alexander Fleet Fellowship. I owe an enormousdebt of gratitude to Miss Julia Fleet for her patronage. I am sad that she

died without seeing this book.Between 2005 and 2006 I had the good fortune to be a postdoctoral

fellow in the Philosophy-Neuroscience-Psychology (PNP) Program atWashington University in Saint Louis. There, again, I benefited frominnumerable conversations with an astonishingly collegial and stimulat-ing group of faculty and students.

The most important stages of the theorys development into a booktook place between 2006 and 2009, during which time I was teaching

in the School of Philosophy and in the Cognitive Science Programme atUniversity College Dublin. Various sections of the book were tried outon a diverse group of students, many of whom took the time to provideme with valuable feedback. I am grateful to all of them. My philosophicallife in Dublin would have been greatly impoverished without the friend-ship and philosophical stimulation provided by my colleagues in theAporo research network. My thanks to them are especially heartfelt.

In the summer of 2009 I spent a very productive month in the

almost perfect research conditions afforded by the Australian NationalUniversitys Research School of Social Sciences. I am hugely gratefulfor that opportunity and for the many helpful conversations I enjoyedthere.

The final stage of the books preparation was completed at theUniversity of British Columbia, where the support of my new colleagueshas been faultless.

I have also learned a lot from audiences at the various venues, in con-

ferences, job talks, and colloquia, where various prototypes of this booksarguments have been presented. Thanks to all of them too.

This book is a rather impersonal, imperfect, and academic work. It seemsa little inappropriate to give it a personal dedication, but I do dedicate it:to my parentswith love and thanks.


11/199

This page intentionally left blank


12/199

xi

Contents

1 Highlights of a Difficult History 32 The Underlying Metaphysical Issue 243 Rejecting the Process-First View 354 Cognitive Unison 47

5 The Causal Life of Attention 886 Consequences for Cognitive Psychology 1177 Philosophical Work for the Theory of Attention 136

Notes 171Bibliography 174Index 183


13/199



14/199

Attention Is Cognitive Unison


15/199



16/199

3

1

Highlights of a Difficult History

1.1 THE PRELIMINARY IDENTIFICATION OF OUR TOPIC

There is one remark from William Jamess The Principles of Psychologythat all those who study attention can quote by heart. It is the remarkthat everyone knows what attention is (James 1890: 381).

This remark does not, by itself, tell us very much. The quoting ofit nonetheless serves two purposes, thanks to which it continues to bequoted with a frequency that nothing else in Jamess masterwork canmatch. The first purpose is to serve as evidence of attentions importance.

Everyone knows what attention is because the role played by attention isso central to cognition that nobody could possibly overlook it. The secondpurpose is to stand in lieu of a definition. If everyone knows what atten-tion is, then there is no need for psychologists interested in attention tostart out by identifying which phenomenon is under investigation, and nocause for philosophical complaint when they fail to do so.

James, of course, had more to say. Having told us that everyone knowswhat attention is, he goes on to tell us that what everyone knows is that

attention is the taking possession by the mind, in clear and vivid form, ofone out of what seem several simultaneously possible trains of thought(James 1890: 381). Philosophical controversies begin when we try to sayanything more than this. The best tactic yet discovered for postponingthose controversies, and so for allowing psychologists to get on with thekind of work that their methods are equipped for, is to quote Jamessbold remark and to leave things at that.

In this book we shall be facing up to the much postponed philosophi-

cal controversies about the nature of attention. But, as a preliminary,pretheoretical identification of our subject matter, it is still Jamessremark that provides us with our best starting place. We can expandupon that remark like this: For minds like ours, in environments likeours, more than one sequence of mental states is possible. We end upwith the train of thought that we actually have partly by chance, butpartly because certain things catch our attention, and partly becausewe direct our attention onto certain things. A theory of attention is an

attempt to give an account of this nonrandomness in our coming to have


17/199

4 Attention Is Cognitive Unison

the train of thoughts that we in fact end up with. It is an attempt toexplain theselectivity of our mental engagement with the world. That iswhat psychologists working on attention seek to explain. It is what this

book is about.Jamess remark that everyone knows what attention is has an air of

unanswerability about it, and an air of finality. It suggests that a morephilosophically ambitious analysis of attention, such as we shall beattempting in chapter 4, would be superfluous. But this air of unan-swerability should be looked on with suspicion. It is a signature traitof the bluffer. Jamess remark was, I think, a piece of bluff. When ThePrinciples of Psychology was written it was far from clear that anybody

knew what attention was. Karl Groos admitted as much, not very longafter The Principles appeared: To the question, What is Attention? , hewrote, there is not only no generally recognized answer, but the differ-ent attempts at a solution even diverge in the most disturbing manner(Groos 1896, cited by Tsotsos, Itti, and Rees 2005: xxiv).

James himself provides evidence that there was some confusion amonghis peers as to which phenomena were, and which were not, attention-involving. His chapter on attention includes, under the title To How

Many Things Can We Attend at Once?, a discussion of experimentsinto the ability that we now call subitizing (that is, the ability to seestraightaway whether one, two, or three stimuli are present, withouthaving to go through any procedure of counting them). But James goeson to say that it is obvious that such observations decide nothing at allabout attention, properly so called (384).

Subitizing was by no means the only phenomenon whose relationto attention was controversial at the time when James was writing.

Disagreements about which phenomena should be taken as the paradig-matically attention-involving ones were widespread at that time, as theywere in the decades before and after. In this chapter I shall be arguingthat, underpinning these disagreements about attentions explanatoryremit, there was a metaphysical issue. This issue is of the first impor-tance for the attempt to provide an explanatory account of what atten-tion really is. Various historical factors have conspired to make it an issuethat has rarely been explicitly discussed.

One reason for revisiting the history of this issue is to show that meta-physical matters need not be arcane ones. They are woven into our foun-dational assumptions about the way in which mental phenomena shouldbe explained, and woven, also, into our assumptions about the methodsby which those phenomena can be examined. The research program ofcognitive psychology is so fruitful and so fast-moving that it is easy forexperimental psychologists to be tempted by the idea that a priori philo-sophical inquiry into the metaphysical underpinnings of the phenomena

that they investigate is a futile pursuit, or, at least, that it is a pursuitlong made obsolete by the advent of their empirical methods. The his-tory that we shall be revisiting here shows this attitude to be mistaken.


18/199

Highlights of a Difficult History 5

The metaphysical question that this book attempts to articulate, and toanswer, is a question that has to be settled at the first stage in the construc-tion of any explanatory theory of attention. It is a question that psychol-

ogys founding fathers never quite squared up to. The current methodsof empirical psychology do not enable us to dodge that question. On thecontrary, it is our attempt to understand the explanatory importance ofapplying those methods that obliges us to take sides on it.

1.2 THREE APPROACHES

Although it was not the first work of psychology to discuss the topic ofattention at length, Jamess Principles came at a time that was crucial topsychologys development as a science, and it was unusually influential.At the time when James was writing there were at least three differentways of approaching the topic of attention, each of which had its ownadvocates.

There were those who took the most fundamental thing about atten-tion to be its involvement in willed action, and in the coordination of

bodily movements. There were those who took the most fundamentalthing about attention to be its involvement inperception, and the achieve-ment of sensory clearness. And there were those who took the mostfundamental thing about attention to be its involvement in the direc-tion ofthinking, and in the coordination of ones internal monologue. Wecan produce examples that show each of these three approaches to havesome intuitive appeal.

On behalf of the first approach we can produce examples in which

the coordination and execution of action seem to be a paradigm case ofattention. One obvious example is the attention required by gymnasts.A less obvious example is the attention that one pays when trying tomove a partially numb limb. Both cases involve attention, but in neithercase is there a perceived object the details of which we are trying tobring into view. The attention given to such cases therefore seems to beattentive action.1

On behalf of the second approach we can produce examples in which

sensory focalization seems to be a paradigm case of attention. The atten-tion paid by the piano tuner before making some fine adjustment to thepitch of a string gives us an example along these lines. Before making hisadjustment the piano tuner is not performing any action. His task is thepurely perceptual one of listening to the pitch. The attention given tosuch a task therefore seems to be attentiveperception.

On behalf of the third approach we can produce examples in whichpursuing a train of thought, without perceptual input or behavioral out-

put, seems to be a paradigm case of attention. An example along theselines is the attention that we pay when performing mental arithmetic.Mental arithmetic certainly requires attention, but it may involve no


19/199


attempt toperceive anything, nor any attempt to execute an action. Theattention given in such a case therefore seems to be attentive thinking.

Such examples show that, among the paradigm cases that a theory

of attention needs to account for, there are cases of attention in action,cases of attention in perception, and cases of attention in cogitation.

The three approaches that these examples make plausible were eachdeveloped in some detail around the time when James was writing.The perception-centered approach was endorsed by several psycholo-gists in Germany and in the United States (for a contemporary review,see Mnsterberg and Kozaki 1894). It was influentially advocated byE. B. Titchener in his 1908 Lectures on the Elementary Psychology of

Feeling and Attention, and again in his 1910 article Attention as SensoryClearness. The action-centered approach was taken in the 1888 edi-tion of Alexander Bains The Emotions and the Will(although in earliereditions Bain had been less sure of it). And the third approachtheapproach centered on the role of attention in the coordination of reason-ingcan be found in work from this period by G. F. Stout. Stout pres-ents the view in 1891, in the introduction to an article on Apperceptionand the Movement of Attention, where he writes that

intellectual ends are attained by an appropriate combination of move-ments of attention, just as practical ends are attained by an appropriatecombination of movements of the body. If, therefore, we are to explainthe process of thinking, we must clearly determine the nature of activeAttention. (25)

It would be misleading to suggest that the plausibility of these threeapproaches led to a period of sustained normal scientific debate between

the advocates of three well-articulated theories. What actually happenedwas that nobody had much of an idea about how the question that dividedthese three approaches might be settled. By the beginning of the twentiethcentury, the unsatisfactoriness of this theoretical situation had become acause for occasional complaint. W. B. Pillsbury began his 1908 book onattention by remarking that the contemporary attention theories werein a chaotic condition (ix). And F. H. Bradley, in a paper from 1902,complained that some of the theories that were being offered as attempts

to explain attention had lost touch with their purported subject matterentirely. He writes that In the case of attention the abuse [of words] haseven been carried to such a point that attention has been used to includeand cover what everyone does and must call a state of inattention (1).

1.3 BRADLEYS PROTEST

Bradley did not tell us whose abuse of words he was objecting to whenhe complained that the word attention had been used to cover whateveryone does and must call a state of inattention, but it is a safe bet that


20/199


James Ward was among those whom he had in mind. In an article writtenin 1885, surveying the entire field of psychology for the ninth edition ofthe Encyclopedia Britannica, Ward claimed that, for scientific purposes,

attention should be construed very broadly indeed. This claim was onethat Ward continued to advance in his 1918 bookPsychological Principles.In chapter 2 of that book Ward recommends extending the denotationof this term [attention] so as to include even what we ordinarily callinattention (49).

On the face of it Wards suggestion looks to be recipe for confusion,and for exactly the abuse of words that Bradley was complaining about.If we allow attention to be used in a way that includes what we ordinar-

ily call inattention then it seems we allow the term to become uselesslybroad. That may be a little unfair to Ward. He was not recommend-ing that attention be used in an entirely unconstrained way. (He was, Ithink, trying to advance the view that the word attention, thanks to itsverbal form, attend, is better suited than the word consciousness toplay the role that the latter word has had assigned to it.) But, whether ornot Wards view was ultimately a sensible one, we can see why Bradleythought that such liberal use of the word attention was worth com-

plaining about. If we use the word attention in such a way that all con-scious states involve attention to some degree, even those idle, tuned-outstates that we naturally describe as instances of inattention, then we loseour grip on the selectivity and directedness that our preliminary charac-terization of attention took to be of its essence.

Bradleys reluctance, in 1902, to name the theorists whose uncon-strained use of the word attention he objected to may have been partlyowing to the fact that Bradley himself, in 1886, had advocated a view

of attention that was notable for the slightness of the constraints that itimposed. The view found in Bradleys earlier work is that

any function whatever of the body or the mind will be active atten-tion if it is prompted by an interest and brings about the result of ourengrossment with its product. (316)

That is clearly not a view that puts much constraint on the things thatcan be called attention. It may be no surprise, therefore, that the later

Bradley, when complaining that the use of attention had become toounconstrained, was happy for his earlier view to be neglected. A footnoteto his 1902 paper mentions that the view defended there departs fromhis 1886 view but tells us that it seems not worth while to ask in detailhow much (2).

Bradleys earlier view was something that he never developed in anydetail. The chief object that he claimed for the paper in which thatview is found was not to elucidate his positive theory. It was to record

a kind of protest (Bradley 1886: 21). The remark quoted above, aboutinterest and engrossment, is more or less all the elucidation that the posi-tive theory gets.


21/199


It will be one of my claims in what follows that the approach taken byBradleys early positive theory is, in fact, well worth pursuing, but sinceBradleys chief object was not the development of that theory but the

recording of a kind of protest, we need to know, if we are to understandhim, what it was that he was protesting against. Several aspects of thecontemporary treatments of attention seem to have been objectionableto him:

I observe a tendency to break up the life of the soul, to divide it intoactive and passive factors, or to suppose a passive beginning with asupervening activity, the latter by some identified with an irreducibleact of attention. I believe this tendency to be a serious obstacle to psy-chology, and there is another tendency not less injurious. Attentionmay be given such a position that the reader cannot tell if it is pri-mary or derivative, or, if primary, whether it is an original elementor something that supervenes; or, again, whether it is one of a class ofactivities, or one function exerted on different objects. (Bradley 1886:305306)

The first part of this remarkthe bit about a tendency to break upthe life of the soulis certainly evocative of what might have beenobjectionable in the conceptual schematizing of Bradleys contempo-raries. The second part of the remark is rather too jargonized to be illu-minating. What does Bradley mean by being primary, and how is theproperty of being primary meant to contrast with the property of beingan original element? What is the notion of supervenience that is inplay here? (It clearly is not the same as the notion that metaphysicianscurrently denote with that word.) One suspects that, in writing this pas-sage, Bradley was rather more concerned with illustrating the reigning

confusion than with showing us a way out of it.We can bring Bradleys positive theory more clearly into focus, together

with his point of protest, by contrasting his explanatory approach notonly with the approach of Ward, for whom attention and conscious-ness were virtually synonymous, but also with the approach taken byJames in The Principles of Psychology. The approach that James advocatedwas one version of the approach that Bradley objected to, and Jameshimself devotes a good part of his chapter on attention to what he takes

to be his point of dispute with Bradley. It will be helpful, therefore,to look at Jamess approach before returning to get a clearer view ofBradleys protest, and of the theoretical position that motivated it.

1.4 JAMESS DISJUNCTIVE THEORY

At the time when he wrote The Principles of Psychology James was sym-

pathetic to something rather like the perception-centered approachto attention. Early in chapter 11 ofPrinciples, shortly after the remarkabout everyone knowing what attention is, James writes that sensory


22/199


focalization is of [attentions] essence (381382). Later in that chap-ter he attempts to provide an account of the intimate nature of atten-tion by identifying the particular processes that constitute attentions

instances. Two processes are identified:

The accommodation or adjustment of the sensory organs; andThe anticipatory preparation from within of the ideational centres con-

cerned with the object to which attention is paid. (411)

The first of these processes is a familiar one. By the accommodation oradjustment of the sensory organs James means such things as pointingones ears in the right direction, bringing ones eyes into focus, taking

a sniff, and so on. The second of the processes is described by Jamesin somewhat less familiar terms, but his point is again quite straight-forward. Anticipatory preparation of ideational centres is simplyJamess way of characterizing imagination. His claim here is that someinstances of attention consist in imagining the things one is attendingto, or looking for. Taking both of these processes together, we find thatJames is advocating a disjunctive theory according to which attention isconstituted either by the processes underpinning imagination, or else by

the processes of sensory orienting.James presents this theory of attention as if it were a theory that enjoys

a good deal of intuitive appeal. These two physiological processes, hewrites, immediately suggest themselves as possibly forming in combi-nation a complete reply to the question of The Intimate Nature of theAttentive Process (1890: 411). In suggesting that this disjunctive analy-sis of the attentive process immediately suggests itself as complete Jamesgoes too far since there are instances of attention that seem to involve

neither imagination nor sense organ orientation. But if Jamess claim hadbeen only that imagination and sense organ orientation can each playa central role insome instances of attention, then his claim to intuitiveappeal would ring true. We can find clear examples that illustrate whatJames must have had in mind here, just as we could find the examples ofthe attentive gymnast, attentive piano tuner, and attentive mental arith-metic solver to illustrate the action-centered, perception-centered, andreasoning-centered views from the works of Bain, Titchener, and Stout.

Examples illustrating the plausibility of the first disjunct of Jamessclaimthe claim that sense-organ adjustment can be a way ofattendingare relatively commonplace: If, sitting in the audience at thesymphony, I want now to attend to the strings, having previously beenattending to the brass, then it seems that an appropriate reorientation ofmy sense organs is one way to go about doing it. This overt attention, asmore recent discussions have dubbed it, is the easy case. For the seconddisjunct of Jamess claimthe claim that preparatory imagination can

be attention-constitutingthe clearest examples are not quite as famil-iar. But once we have seen them, we can see that they are also strong intheir intuitive appeal. The example mentioned by James himself is the


23/199


following: When listening to a note played on the piano one can hear anumber of harmonic overtones in addition to the fundamental frequen-cies of the note played. Suppose you want to listen for the third overtone

of a particular note. (If you are a piano tuner then this will be the sortof thing that you have to do routinely.) One way to go about attendingto the note so as to hear the third overtone is by first listening to thepitch at which that overtone sounds, and then imaginatively keepingthat pitch in mind while listening as the note is struck. In this case theimagining of the pitch and the attention that one pays in listening for theovertone seem to be the same thing.

James took this example from Hermann von Helmholtzs 1870 trea-

tise on sound perception. For those of us who are not piano tuners it maynot be an everyday example of attention, but such shifts in what one islistening for in a single auditory stimulus are familiar enough. They arecertainly possible, and they are certainly attention-involving. We needour theory of attention to allow for such shifts. The part of Jamess the-ory that is concerned with imagination is an attempt to do so.

Some readers will be sure to have a qualm at this point, since I havejust made a psychological claim on the basis of introspection and infor-

mal observation. Many psychologists disapprove of introspection andof informal observation. They have perfectly legitimate reasons fordoing so. Ideally, one would like to puts Jamess and Helmholtzs ideasabout attention and preparatory imaging to a more rigorous empiricaltest. I know of no studies that do exactly that, but Harold Pashler, inhis 1997 textbook on attention, reports a study in which subjects wererequired to form a visual mental image of an item and then, forgettingabout that image, to perform a task requiring them to detect the occur-

rence of a digit in a rapidly presented sequence of pictures (249). Theresults that Pashler observed comport nicely with what James suggested.When presented with a sequence of pictures it seems that, irrespectiveof the demands of their task, subjects do give attention to the picturesof recently imagined items, just as James would have predicted. Theevidence that attention is paid to these items comes from the fact thatthey elicit an attentional blink: If the digit that the subject has beenlooking out for comes shortly after the imagined image, then subjects are

liable to overlook it. In Pashlers experiments, then, prior imagination ofa thing does seem to bring it about that the thing is attended, much asJames thought.

Whatever their empirical credentials, the examples of switching atten-tion to a different instrument in the symphony and switching attentionto a different aspect of a note played on the piano give us some insightinto Jamess disjunctive theory, and into his reasons for finding it plau-sible. With that theory before us, as an example of the sort of thing that

F. H. Bradley was protesting against, we can now begin to see the sourceof Bradleys dissatisfaction.


24/199


1.5 THE SOURCE OF BRADLEYS DISSATISFACTION

We have seen that Jamess claim about attentions links to orienting and

to imagining is plausible enough. What was it, then, that Bradley foundin approaches such as Jamess that he wanted to protest about? The prob-lem that Bradley had with the theorists of attention who were writing atthe end of the nineteenth century was not that they had failed to identifyany psychological phenomena in which attention was clearly displayed.That was very far from being the problem. James, we have seen, identi-fied phenomena in which attention seems clearly to be displayed. So didHelmholtz, Bain, Titchener, and Stout. All of their proposals seemed

plausible (and they continue to seem plausible, even in the face of morerecently gathered evidence). The source of Bradleys dissatisfaction withhis contemporaries was not that they had identified no clear examplesof attention-involving phenomena, but that there were too many suchexamples, and the various examples of attention-involving phenomenathat these different theorists had identified seem not to have much incommon as far as their constituent processes go. Nor did any one ofthese examples seem to have explanatory priority over the others. It

therefore seemed that the contemporary accounts of the processes thatare involved in particular instances of attention did little to help us informulating a generalizable explanatory theory of what attention is. Andthe project of identifying any particular process as the special activity ofattention therefore seemed wrongheaded. When Bradley titled his 1886essay Is There Any Special Activity of Attention? it was in order to urgethat the question be answered in the negative.

To readers familiar with the recent empirical literature on atten-

tion the diversity among the processes that Bradleys various contem-poraries took to be involved in paradigm cases of attention may soundlike a familiar fact, and the metaphysical lesson that Bradley took fromitthat there is no special activity of attentionmay be thought to besomething that we now find obvious. Recent discussions of attention fre-quently note the plurality and heterogeneity among the paradigm casesof attention, and urge us to accept that our theories of attention mustbe corresponding heterogeneous (for an influential example of such rea-

soning, see Parasuraman 1998). Those who attempt summaries of thecurrent debates typically preface their discussions with a remark to theeffect that there may be numerous component processes behind ourattentional abilities (Driver 1998: 298), or else they claim that atten-tion is an umbrella term for a variety of psychological phenomena(Styles 2006: 1) and take this as showing that to try to define attentionas a unitary concept is not possible and to do so would be misleading(Styles 2006: 9). We shall return to these ideas from the current litera-

ture in chapter 6. The point to note presently is just that their famil-iarity should not be allowed to mislead us as to the point that Bradley


25/199


26/199


constitute paradigm cases of employment. Nor would they simply bemistaken about how many such processes there are. The mistake is amuch more fundamental one. It is a mistake about the way in which an

attempt to explain employment should proceed. In taking it that theexplanation should proceed by identifying the processes that constituteemployment these imagined scientists would be making a metaphysicalmistake: a mistake about the sort of thing that employment is.

It bears emphasis that the thing we find ourselves wanting to say aboutthe imagined project of identifying the process of employment is notthething that the current textbooks say about attempts to identify the pro-cess of attention. We do not give a correct diagnosis of what has gone

wrong in the imagined debate about employment by saying that employ-ment is an umbrella term, or that it is a term for which no definitioncan be given, or that analogue-James and analogue-Titchener each had apart of the truth. We do not say that the theories of analogue-James andanalogue-Titchener should be reconstrued as contributions to a hetero-geneous, family-resemblancebased theory of employment. What we sayabout analogue-James and analogue-Titchener is what Bradley says of hiscontemporaries attempts to understand attention: It is all wrongheaded.

For the purposes of explanation there is no point in starting a catalog ofthe processes that constitute employment. Any process whatever, of thebody or the mind, will be a case of employment, as long as it is done inthe right way and for the right reasons. A theory of employment shouldspecify the way and the reasons. It should not catalog the constituentprocesses.

Bradleys complaint against a project like Jamess was not that thewrongprocesses had been identified as the attention constitutors, or that

too few processes had been identified. Bradley did not deny Jamess claimthat imagination and sense-organ accommodation had important roles toplay in many instances of attention (Bradley 1886: 307). Nor, when dis-cussing Bains claims about motor coordination, did Bradley take issuewith the idea that motor coordination can be attention constituting. Nor,one imagines, would he have objected to Stout and Titchener when theyclaimed that there are instances of attention in which cogitation andsensory clearness are crucial. Bradleys protest was a metaphysical one.

He was objecting to the idea that the identification of these processescontributed even a part of the explanation of what attention is. His ownview was that there is no explanatory value in starting a catalog of theattention-constituting processes because any process could constituteattention, as long as certain conditions were met (conditions that hisremarks about engrossment and interest gestured at but did not ade-quately specify). This is what he means when he writes:

We have found nothing in attention that is not derivative, nothingwhich could justify our placing it among the primary elements ofmind. In attention there is either no activity at all beyond the common


27/199


processes of redintegration and blending, or, if the activity exists, itselfis not attention. Any function whatever of the body or the mind willbe active attention if it is prompted by an interest and brings about the

result of our engrossment with its product. There is no primary act ofattention, there is no specific act of attention, there is no one kind ofact of attention at all. (Bradley 1886: 316)

Jamess question about the intimate nature of the attentive process was,in Bradleys view, a question with a false metaphysical presupposition,not because it supposed that there was just one or two such process, butbecause it supposed that attention is a matter of executing particularprocesses at all.

The metaphysical point of Bradleys protest is worth laboring, notonly because it is a point that is easily mistaken for the less radical claimsthat we find in the current literature, but also because it is a point thatgets lost almost as soon as Bradley makes it. Bradley himself, although hewas no slouch when it came to making metaphysical points elsewhere,never spelled the point out very clearly, partly because, as we have seen,he was more concerned with registering a kind of protest than witharticulating his positive view, and partly because, as we have also seen,

he changed his mind before that positive view was ever brought clearlyinto focus. Nor did anybody else manage to give a very clear articulationof the metaphysical issue that separated Bradley from James, and fromhis other contemporaries.

The failure to articulate this issue was partly owing to the fact that thephilosopher/psychologists of this period lacked the necessary metaphysi-cal vocabulary. Works from this time frequently display a proliferation ofjargon, corresponding to the points where different authors take differ-

ing metaphysical positions, and this leads to a lack of clarity about whatneeded to be explained and how. An example is William Cypless 1880book,An Inquiry into the Process of Human Experience, in which Cyplesattempts to define organization as the interhappening of structuralstatics with related dynamical activities (10). Such definitions earnedCyples the scorn of his reviewers (Sully 1880), but they were a reflectionof a widespread uncertainty about which of the concepts employed inthe course of stating psychological theories were in need of clarification

and about how such clarifications were to be given. Psychology did notyet have an established policy about what a psychological explanationshould look like. It lacked an established vocabulary for articulating themetaphysical issues that such explanatory questions depend on.

This lack of an established vocabulary in which to frame psychol-ogys underlying metaphysical issues prevented Bradley and James fromgiving a clear account of the metaphysical point about which they weredisagreeing. Bradley, as we have seen, frames the issue in terms of spe-

cial activities, original elements, supervenient activities, and activeand passive factors (Bradley 1886: 306). James, in contrast, framesthe issue as a dispute between a cause theory and an effect theory


28/199


29/199


The dominance of behaviorism in this period resulted in attentionbeing demoted from its central place in psychology. As a result, themetaphysical/explanatory issue on which Bradley and James had been

divided became all but invisible.This was not because the behaviorists had nothing at all to say about

attention. John Dashiells 1927 Fundamentals of Objective Psychologyoffered a behaviorist account of attention as a form of posturing (284).And, some decades before behaviorism came to be established as aschool of thought,Thodule Ribots 1889 book, Psychologie de lattention,had offered a strikingly behaviorist approach to attention, which Ribotasserts with characteristic rhetorical flamboyance:

Are the movements of the face, the body, and the limbs and the respi-ratory modifications that accompany attention, simple effects, out-ward marks, as is usually supposed? Or are they, on the contrary, thenecessary conditions, the constituent elements, the indispensable factors ofattention? Without hesitation we accept the second thesis. (25)

Despite the boldness of Ribots proto-behaviorist treatment of atten-tion, the attempts, once behaviorism got under way, to identify a set of

movements, postures, or respiratory modifications as essential to atten-tion was generally recognized to be hopeless. As Gilbert Ryle notes, itis not only attention that resists simple behaviorist analysis but also theclass of heed concepts more generally:

When a man is described as driving carefully, whistling with concentra-tion or eating absent-mindedly the special character of his activity seemsto elude the observer, the camera and the Dictaphone. Perhaps knittedbrows, taciturnity and fixity of gaze may be evidence of intentness; but

these can be simulated, or they can be purely habitual. (1949: 133)

The problem for the behaviorist is that attention may be given to allsorts of stimuli, and all sorts of responses may be performed attentively.For that reason attention is exactly the sort of thing that a behavioristtheory, given in the form of rules for stimulus/response mapping, willalways struggle to accommodate. Most of the behaviorists who set theagenda for psychology in the first decades of the twentieth century were

therefore happy for attention to drop out of view.

Behaviorism ceased to be the dominant paradigm in psychology over thecourse of the 1950s. Attention returned to the psychologists explana-tory agenda over the same period. These two tendencies were intimatelylinked. The Second World War had lent a new urgency to questionsabout such things as a persons capacity to pay attention to multipleradar screens for prolonged periods. The research that such questions

prompted revealed interesting effects. These cried out for experimentalstudy, and for a theory to explain them. The rehabilitation of attention asa topic for scientific study was called for, but that rehabilitation required


30/199


31/199


Broadbent employed the intellectual resources that are used by commu-nications engineers to negotiate those practicalities.

Information technology was progressing dramatically at the time

when Broadbent and Chomsky were writing. It was progressing rapidly,with the invention in September 1958 of the integrated circuit. And itwas progressing conspicuously, with the launch of Sputnik, and withthe introduction, in December of that year, of subscriber trunk dial-ing to telephone exchanges in the United Kingdom. Broadbent saw thetheoretical resources that were used in developing such communicationstechnologies as a tool that could be used throughout psychology. In theconclusion ofPerception and Communication, he explicitly presented this

as an attack on, and as an alternative to, positivism-influenced behavior-ism. But at a much early stage in the discussion, before these broadermethodological themes have been broached, he writes:

Perhaps the point of permanent value which will remain in psychologyif the fashion for communication theory wanes, will be the emphasison problems of capacity. [. . .] The fact that any given channel has alimit is a matter of central importance to communication engineers,and it is correspondingly forced on the attention of psychologists whouse their terms. (1958: 5)

This introduction of the notion of capacity limitations into discussionsof perception was, as Broadbent predicted, hugely and permanentlyinfluential. It was from claims about capacity, thought of in informationprocessing terms, that all theories of attention in the decades followingBroadbent would be built.

According to Broadbents own capacity-based theory of attention

there is a single attentional bottleneck at which capacity limitationsare especially pertinent, a bottleneck that arises because the informa-tion coming in from the senses is processed by two systems operatingin series. The first system, Broadbent thought, has a large capacity forinformation processing. The second has a much smaller capacity. Thebottleneck produced by the connection of the two is the locus of atten-tion in the sense that, although all stimuli are automatically subjectedto processing by the first, large-capacity system, the stimuli that make it

through the bottleneck into the small-capacity system count, ipso facto,as stimuli to which attention is paid.

This two-systems-and-a-bottleneck picture was intended as acommunication-theoretic rendering of the everyday idea that simple fea-tures of ones environment, such as the fact that there are people talkingin the next room, come to ones awareness involuntarily, whereas thedetails of these things, such as the contentof the conversation that istaking place, can be detected only for one or two of the things that are

going on, with the question ofwhich things have their details detecteddepending on the focus of ones attention.

By introducing the information-theoretic notion of capacity to dis-cussions of perception, Broadbent allowed various theories of attention


32/199


33/199


for empirical inquiry rather than for a priori philosophizing. Prior toBroadbent the philosophical neglect of attention was owing to the popu-larity of positivistic and of Wittgensteinian qualms about the explana-

tory credentials of all inner acts, with attendings being just one example.In 1957, the year before Broadbents Perception and Communicationwas published, these Wittgensteinian qualms were in evidence in PeterGeachsMental Acts. Geach mentions attention in connection with theexplanation of the way in which perception puts one in a position toform a judgment about the content perceived, but he mentions attentiononly in order to dismiss its explanatory credentials. So far as I can see,writes Geach, it is quite useless to say the relevant sense-perceptions

must be attended to, either this does not give a sufficient condition, orelse attended to is a mere word for the very relation of judgement tosense perception that requires analysis (64).

In the context of this Wittgenstein-inspired hostility to internal acts,the metaphysical issue that divided Bradley and James remained absentfrom the philosophical agenda. There were, however, occasional breaksin this silence on topics related to attention. They can be found in theliterature prompted by the discussion of heed concepts in Ryles 1949

The Concept of Mind. The most sustained treatment of attention in thisperiod is Alan Whites 1964 monographAttention.

White shared Ryles sense that a straightforward behaviorist treatmentof heed concepts was unsatisfactory, and he thoughtas Ryle himselfmay have done in later yearsthat the more sophisticated behaviouristtreatment offered in The Concept of Mindfailed to address these worries(see Ryle 1971, viii). White also shared Ryles enthusiasm for the projectof extracting philosophical lessons from the norms governing discourse

in ordinary language. This can sometimes disguise the depth of Whitesinsight. When we find him taking pains over points like the following,his philosophy can sometimes sound like a rather uninformative gram-mar lesson:

When we say of someone that he noticed what he or another was doing,the what is an interrogative pronoun because noticing is here aninstance of noticing that. By contrast, the what of attending to whatoneself or someone else is doing is a relative pronoun meaning that

which he is doing. (1964: 18)

Current philosophers tend to regard this fixation on natural language asa phase that philosophy did well to get out of, and understandably so,but the bad reputation of ordinary language philosophy is in some waysundeserved. White knew perfectly well that our philosophical theory ofattention should tell us about attention itself, and not just about the rulesgoverning our talk about it. His discussion of the grammar of attention

talk was not intended as an end in itself. It was intended to enable thediagnosis of fallacies in the arguments about the phenomenons underly-ing nature.


34/199


Among the lessons that White wanted to draw from his examinationof the behavior of heed concepts was a lesson about the way in whichRyles treatment of those concepts needed to be modified. Ryle had

treated heed concepts as belonging in the class that he dubbed mongrelcategoricals (1949: 135138). By this he meant that statements employ-ing heed concepts not only tell us whatis taking place, but also situatethat event relative to a set of dispositions and hypotheticals. Ryle pro-vides some examples that make the point clear. That bird is migratingis one of his paradigm cases:

The description of a bird as migrating has a greater complexity than

the description of it as flying in the direction of Africa, but this greatercomplexity does not consist in its narrating a larger number of inci-dents. Only one thing need be going on, namely, that the bird be at aparticular moment flying south. It is migrating tells not more stories,but a more pregnant story. (136)

Ryles theory is that he is attending to what he is doing, like it ismigrating, tells a pregnant story in the sense that it gives us hypotheticalinformation, about what might have been the case, along with descrip-

tive information, about what actually is the case. The hypothetical infor-mation that it gives is information about what the attentive man doestypically, or about what he would have done in similar but slightly dif-ferent situations.

White rejected Ryles idea that in playing the piano, he is attendingwas a mongrel categorical of this sort. His reason, which we consider inmore detail in chapter 4, was that it is migrating can serve as an expla-nation of it is flying south, whereas he is attending cannot explain the

fact that he is playing the piano. At most, it explains a different fact: thathis piano playing is successful. Describing the birds flight as an instanceof migration provides us with information that zooms out and locates theflight in a broader pattern of dispositions that the bird, and birds like it,displays. Describing the pianists playing as an instance of attention, bycontrast, zooms in and tells us more about the way in which the play-ing is taking place. As White puts it: Whereas He is practising on thepiano gives a more complex narration of the same incident as that nar-

rated by He is playing the piano. He is attending to what he is playingon the piano narrates a more complex incident (1964:14).Ryles and Whites preference for talking in formal mode, about con-

cepts andsentences rather than things in themselves, is, on this occasion,merely an expository technique. When translated into the material modeWhites claim is recognizable as a metaphysical one. It is the claim thatthe metaphysical relationship between attending and the activities thatare performed attentively must be different in kind from the metaphysical

relationship between migrating and the flights that are instances of it.The alternative picture, with which White wants to replace Rylesmongrel categorical account, is a picture that is close kin to the picture


35/199


found in Bradley. Bradley, as we have seen, rejected the idea that identi-fying the attention-constituting processes gave us what we need in orderto know what attention is. He thought that attention was constituted

by the processes of redintegration and blending, but he thought thatthis fact was of no explanatory consequence, since a catalog of atten-tion-constituting processes, like the catalog of employment-constitutingactivities, is the wrong form for an explanation of attention to take. InWhites work, the idea that the explanation of attention cannot proceedby identifying which processes take place becomes the claim that atten-tion does not name any specific activity; it indicates the circumstancesin which the activity occurs and thus signifies what, on this occasion, it

amounts to or is a form of (1964: 6).White and Bradley are, then, in broad agreement about the negative

claimthat attention is not explicable by the identification of its con-stituent processes. They also offer something broadly similar by way of apositive alternative. White spells it out as follows:

Simply to say that someone is attending, or paying attention, gives usno more clue as to what activities he is engaged in than simply to saythat he is practising. What attending tells us is that his activities andenergies, whatever they are, are directed to and focused on somethingwhich occupies him. (1964: 7)

For Bradley and White, and in a slightly different way for Ryle, claimsabout attention are not made true simply by the facts aboutwhich pro-cesses are taking place in the attentive thinker. In Bradleys treatment thisis because claims about attention tell us, not about which processes aretaking place, but about their immediate causal context: The attention-

constituting processes are those that are prompted by interest and thatbring about engrossment (1886: 316). In Ryles treatment it is becauseclaims about attention tell us something about processes and also tellus something hypothetical: Intellectual processes constitute attentionwhen they occur in a context such that similar processes would havetaken place in a range of counterfactual situations (1949: 135136). InWhites treatment, the idea is that claims about attention do not tell uswhich processes are taking place. Instead they tell us how those pro-

cesses relate to the agents other activities and to his goals.A theory like Bradleys or Whites (or Ryles) has different explanatorygoals and operates at a different level from a theory like Broadbents,but this should not be taken as an indication that the theories are not inconflict. Although theories at different levels of explanation need notcompete, this does not mean that they never exclude one another. Thereis a genuine disagreement between a theory like Broadbents and a the-ory like Whites. It is a disagreement over the fundamental metaphysi-

cal/explanatory question that divided Bradley and James. White, likeBradley, takes it that the identification of processes at the subpersonallevel is the wrong sort of thing to provide an explanation of attention.


36/199


37/199

24

2

The Underlying Metaphysical Issue

2.1 EXPLANATORY TACTICS

Our theory of attention should put us in a position from which we cananswer both of the following questions:

What is attention?1.What is it for something to be done attentively?2.

There are two questions to be answered here, but there are not two bitsof philosophical work needing to be done. Once we have an answer to

either one of these questions we are in a position to derive our answer tothe other: If we have an independent answer to question 1 then we mayderive our answer to question 2 by saying, To do something attentivelyis to do the thing in the manner that typifies instances of the process ofattention (a process that we identified in our answer to question 1). If wehave an independent answer to question 2 then we can derive our answerto question 1 by saying, Attention is the doing of something in the atten-tive manner (a manner that we identified in our answer to question 2).

We need to start with an answer to one question or the other, but it isnot obvious which it should be.1

In chapter 1 we saw William James advocating a theory that took thefirst approach. Jamess theory started by telling us what attention is. Hisclaim was that it is either sensory orienting or preparatory imagining.From here James could, if he had wanted to, derive an account of atten-tiveness simply be saying that it is the manner that typifies instances ofthese processes.

We contrasted Jamess approach, as did James himself, with theapproach advocated by F. H. Bradley. Bradley starts with question 2.He claims that an activity is performed attentively if and only if theagent is engrossed and prompted by interest. As long as this mannerwas instantiated, Bradley suggested, we have an instance of attention. Ifwe want an answer to the question What is attention?, then the cor-rect way to go about getting it is by deriving that answer directly fromthe answer that we give to the question about performing attentively.

Attention is just the doing of something in that manner.


38/199

The Underlying Metaphysical Issue 25

Rather than being, as James took it to be, a disagreement about causesand effects, or about spiritual and mechanical forces, the disagreementthat divides these two explanatory strategies can be traced to their differ-

ent commitments as to the metaphysical category into which they assigntheir explananda. And these different metaphysical commitments, in turn,have different methodological ramifications: Starting with the wrong viewof attentions metaphysics will lead us to look for the wrong sort of expla-nation, and it will lead to our using the wrong sorts of methods with whichto look. In chapter 3 I present an argument that gives us a reason (althoughnot a conclusive reason) to suspect that most current theories of attentionhave indeed started with the wrong metaphysics. To see that argument we

shall need an account of what the underlying metaphysical issue is. Thischapter is concerned with giving such an account. It will make precise theidea that if we side with James we are committed to the idea that a pro-cess, or a set of processes, can be identified as being responsible for atten-tion, whereas if we side with Bradley we are committed to the idea thatit is something other than their constitutive processessomething aboutmanner or contextthat instances of attention have in common.

2.2 THE BASIC DISTINCTION

The lesson to be learned from James and Bradley, if put in maximallygeneral terms, is that something fundamental about the metaphysicalstatus of any phenomenon, x, is revealed by the order of explanatorypriority between our answers to the questions What isx? and What isit for something to be donex-ishly? We can begin to make this clear by

considering the order in which these questions must be taken in a pairof cases wherexstands for perfectly unmysterious phenomena of intui-tively contrasting metaphysical sorts.

Suppose that one were faced with an intelligent but ignorant inter-locutor who wanted to know whatcombustion is. Suppose also that ourinterlocutor had had some exposure to discourse about combustion andwanted answers to the two questions analogous to those with which thischapter started: (1) What is combustion? and (2) What is it for some-

thing to occur combustively? We would need to start our analysis with ananswer to the first of these. We explain the phenomenon of combustionby identifying it with a particular processthe process of burningandby giving a description of what that process is. Once we have identifieda process of combustion (and thereby answered the first question), wecan, with no further empirical or philosophical work, derive an accountof what it is for something to happen combustively. We need do no morethan say that, in order for something to happen combustively, it must

happen in the manner of a typical instance of combustion (that is, of aninstance of the process of burning). For combustion, then, our answer


39/199


40/199


of their instances map onto different taxonomies of events. The pointsthat we need, in order to clarify the metaphysical distinction that we areconcerned with, are points about taxonomies ofevents, but these points

can be seen more clearly when they are applied to taxonomies ofobjects.It will help to consider these first.

There are various different taxonomic principles that might be usedto carve up the set of all objects into categories. Some of the resultingtaxonomies will be useful for various purposes. Some will be completelyuseless. Among the useful taxonomies are those that group objects onthe basis of the substance from which they are made. Since there is morethan one way to individuate substances, there is more than one way to

give a substance-based taxonomy of objects. One substance-based tax-onomy groups all plastic things into one set, all wooden things intoanother, and so on. A different taxonomy might divide up the plasticthings more finely, grouping all the PVC into one set, all the acrylicthings into another, and so on. There are lots of ways that a taxonomy ofobjects might be given.

Suppose now that we are given the set of objects and asked to find away of specifying the set of all paperweights. If we are allowed to use

only the classificatory resources of a substance-based taxonomy then wewill not be able to do it. The problem we face is notthat paperweightsare multiply realizable, so that there are many substances from which apaperweight could be made. It is true that paperweights can be realizedin multiple substances, but it is not this that stops us from delineat-ing the paperweights with the resources of a substance taxonomy: Theproblem of multiple realizability could be solved simply by specifyinga union of several sets. What stops us from being able to delineate the

paperweights using the distinctions provided by a substance taxonomy isthat being made of the same substance as a paperweight does not sufficefor being a paperweight. That makes it impossible to rule in all of thepaperweights without also ruling in some of the nonpaperweights.

There is a cheating way to avoid this problem, which is to define amaximally fine-grained taxonomy of substancesa taxonomy on which,for example, plastic-that-was-made-in-one-a-certain-list-of-factories-on-a-Tuesday-morning is a different substance from plastic-that-was-made-

in-one-of-a-certain-list-of-factories-on-a-Tuesday-afternoon. A tax onomylike that might enable us to draw a line around all and only the paper-weights. But any such taxonomy would be so fine-grained that it wouldcease to be useful as a taxonomy of substances. Any taxonomy that is recog-nizable as a taxonomy by substances is cut by the set of paperweights.

When we apply it to events, this notion of a set that cuts a taxon-omy enables us to clarify the metaphysical distinction that underpinsthe difference between those phenomena that are amenable to process-

first explanation and those that are amenable to adverb-first modes ofexplanation. As with the set of objects, the set ofevents can be carvedup in different ways, and the different taxonomies that result will beuseful for different purposes. One sort of taxonomy, useful for several


41/199


purposes, is the taxonomy that groups events according to the processesthat they instantiate. Suppose now that we are given the task of specify-ing the set of instances of haste. If we are allowed to use only the classi-

ficatory resources of a process-based taxonomy then we will be unableto do it. Again the problem has nothing to do with multiple realizability.The problem is not that there are too many sorts of things that can bedone hastily. The problem is that instantiating the same processes as aninstance of haste is not sufficient for being an instance of haste. The setof instances of haste cuts any taxonomy by process.

It would be a mistake to take these points as showing (although itmight, in fact, be true) that paperweights and haste are ontologically less

basic than lumps of metal or instances of combustion: No point aboutorders of ontological basicness can be made via a taxonomy-cutting argu-ment since, had we chosen different taxonomies, the situation wouldhave been the other way around. No substance taxonomy is sufficientfor delineating the set of paperweights, but if we were using a taxon-omy that classifies objects byfunction, then we could delineate the set ofpaperweights quite easily, and there would now be no way to delineatethe set of lumps of metal. Again there is an exact parallel in the case of

events. No taxonomy that classifies events by process is adequate to thetask of delineating the set of instances of haste, but if we were using ataxonomy that classifies events by their manner of occurrence then wecould delineate the instances of haste quite easily, and there would nowbe no way to delineate the set of instances of combustion. Taxonomycutting, then, is a way of articulating the distinctions between meta-physical categories. It is not a tool for the ordering of hierarchies accord-ing to ontological basicness.

The distinction between process-first phenomena and adverbial phe-nomena can now be analyzed as follows: The set of instances of an adver-bial phenomenon cuts any taxonomy by process (as reflected by the factthat the things that can be done hastily can also be done nonhastily).The set of instances of a process-first phenomenon cuts any taxonomyby manner of performance (as reflected by the fact that imitations of themanner in which combustion occurs need not themselves be instances ofcombustion). The distinction that we are interested in can therefore be

made precise if we can find a way to distinguish between taxonomies byprocess and taxonomies by manner of performance. And now our phil-osophical task is relatively straightforward: A distinction between thesedifferent sorts of taxonomic principles, at least in the case of taxonomiesof events, is easy to draw.

The distinction that we need in order to complete our account of themetaphysical difference between process-first phenomena and adverbialphenomena is a distinction between taxonomies that classify events on

the basis of process and taxonomies that classify events on the basis ofmanner of occurrence. The natural way to characterize the contrast hereis this: The question of which processes happen is settled by the facts


42/199


about what things happen to what. The question of whatmannerthoseprocesses have is settled by the facts about how the things that happenhappen. What happens to what is a matter of the gaining and losing of

properties by objects. How the things that happen happen is a matterof the gaining and losing of properties by events. To put the same pointmore precisely: A taxonomy is a taxonomy on the basis of process if thetaxonomy classifies events on the basis of the having or gaining of a prop-erty by an object. A taxonomy is a taxonomy on the basis of manner ifthe taxonomy classifies events on the basis of the having or gaining of aproperty by an event.

Putting all this together gives us the result that, if a phenomenon

belongs in the process-first category, then the set of instances of thatphenomenon can be delineated using the resources of a taxonomy thatgroups events on the basis of the having or gaining of a property by anobject. And if a phenomenon belongs in the adverbial category, thenthe set of instances of that phenomenon can be delineated using theresources of a taxonomy that groups events on the basis of the having orgaining of a property by an event.

The contrapositive versions of these claims are particularly pertinent

for the argument that follows and are also worth noting here: A phenom-enon does notbelong in the category of adverbial phenomena if the set ofinstances of that phenomenon cuts all taxonomies that group events onthe basis of the having or gaining of a property by an event. A phenome-non does notbelong in the category of process-first phenomena if the setof instances of that phenomenon cuts all taxonomies that group eventson the basis of the having or gaining of a property by an object.

What I have done here is to introduce and to make precise a distinctionbetween two metaphysical categories, corresponding to a difference intwo explanatory approaches. We have seen that there is a metaphysicaldifference between those phenomena that are explained by first answer-ing a question like the first of those we started with and those phe-nomena that are explained by first answering a question like the secondof those we started with. I have put this distinction on a more precisefooting by showing how, via the notion of taxonomy cutting, it can be

explained in terms of the metaphysical distinction between properties ofobjects and properties of events. All of this is intended to make preciseour intuitive notion that there are some thingssuch as combustion,decay, digestion, and evolutionthat are to be explained in one way,and other thingssuch as haste, stealth, wit, style, and employmentthat are to be explained differently. The reader who is skeptical of themetaphysical distinction between properties of objects and properties ofevents (on which my analysis rests) will be skeptical about whether my

discussion moves in a way that constitutes philosophical progress. I amsure some readers will be skeptical of precisely that, but the analysis hasto stop somewhere, and I propose to stop it here, with the distinction


43/199


between properties of objects and properties of events as our unanalyzedanalyzer.

This distinction between adverbial and process-first phenomena, not

a distinction between causes and effects, nor a distinction between spiri-tual and mechanical forces, is what James and Bradley were really dis-agreeing about.

2.4 ADVERBIALISM, MULTIPLE REALIZABILITY,

AND NATURAL KINDS

Having said what it is to be an adverbial or a process-first phenomenon,we can now clarify some important points about this distinction andabout the claim, which is what we are working toward, that attentionmight belong on the adverbial side of it. We can, in particular, see howthis claim differs from the various things that might be meant when psy-chologists say that attention is an umbrella term (see chapter 1, 1.5).

We have already seen that the adverbialists claim is not, in the firstplace, a claim about multiple realizability. Philosophers of mind have

often focused on issues arising from the multiple realizability of mentalphenomena. They have had good reasons for doing so, but multiple real-izability is not the heart of the metaphysical matter that is in contentionhere. A phenomenon may be realized by multiple processes because it isa disjunctive kind of process; or because it is a process-genus with multi-ple processes as its component species; or because it is a process-family,with multiple processes as it members. Such phenomena are realized bymultiple processes, but they are not ruled out of the category of process-

first phenomena. The sets of their instances can be delineated using onlyconjunctions or disjunctions of the distinctions given to us by a processtaxonomy on the set of events. They dont cut such taxonomies.

The claim that attention is an adverbial phenomenon should not bemistaken for the claim that attention can be realized by multiple pro-cesses. Nor should it be confused with the claim that the instances ofattention do not form a natural kindof process. The analogy betweentaxonomies of objects and taxonomies of events makes this clear: The

instances of jade do not form a natural kind of substancesome instancesof jade are nephrite, some are jadeite. But the set that is the union of jade-ite instances and nephrite instances contains all and only the instances ofjade, and so there is a taxonomy by substance that the set of instances ofjade does not cut. This is exactly as it should be. The question of whethersomething is an instance of jade is a question that is settled by the factsabout what substance the thing is made of. The metaphysics of jade isthat of a substance, even though jade is not a natural kind of substance.

The claim that jade is not a substance term would be stronger than theclaim that the instances of jade do not form a natural kind of substance(and it would be false, even though the second claim is true).


44/199


This point about taxonomies of objects transfers directly to taxono-mies of events. The claim that attention does not have the metaphysicsof a process is stronger than the claim that the instances of attention do

not form a natural kind of process. The claim that attention does nothave the metaphysics of a process, unlike the claim that attention is nota natural kind of process, rules out the possibility that attention can beidentified with a disjunction or family of several processes, any one ofwhich is sufficient for the instantiation of attention.

The methodological significance of this difference between the adver-bialists claim and the not-a-natural-kind claim can be seen if we considerthe standard view of the psychology ofmemory. The processes respon-

sible for short-term memory are probably very different, cognitivelyspeaking, from the processes responsible for long-term memory; and theprocesses responsible for episodic memory are probably very different,cognitively speaking, from those responsible for semantic memory, and soon (Baddeley 1990, chapter 3; Schacter, Wagner, and Buckner 2000). Itfollows from this that memory is not a natural kind of cognitive process.But that fact does not present a problem for the attempt to explain mem-ory via the identification of cognitive processes. We can explain memory

perfectly well by the piecemeal identification of various processes cor-responding to its various instances. Adverbial phenomena, in contrast,cannot be explained by the piecemeal identification of their componentprocesses. It is impossible to explain what haste is by looking closely atthe component processes of various instances of haste, just as it is impos-sible to explain what employment is by looking at the processes that peo-ple execute in the course of being employed. However closely you studythe component processes of the instance of haste that took place in my

kitchen this morning, you will only find out about the processes involvedin pouring water over a tea bag and adding milk. An examination of theseprocesses will tell you nothing about haste in general and nothing abouthaste per se. If it turns out that attention belongs in the adverbial categorythen that, unlike the not-a-natural-kind claim, will rule out the possibil-ity of explaining attention via a piecemeal process-by-process approach,of the sort that we use when explaining memory.

We shall return to these methodological and explanatory conse-

quences of adverbialism in chapter 6, once we have got a worked outversion of an adverbialist theory of attention on the table. The point tonote here is just that the denial of the process-first approach to attentionis distinct from the denial that the instances of attention form a naturalkind of process, and that the first claim, unlike the second, is incompati-ble with the idea that an account of what attention is can be given via theexamination of attentions constituent process. A piecemeal explanatoryapproach (such as some psychologists advocate when they say that atten-

tion is an umbrella term, and such as we use when explaining memory)is committed to the process-first view and incompatible with the adver-bialist alternative.


45/199


46/199


particular adverbial phenomenon occurs are facts about how the thingsthat happen happen. The metaphysical difference between adverbialphenomena and process-first phenomena can be thought of as a differ-

ence as to the sort of facts that make it the case that the phenomenonin question occurs.

The sense of making it the case that is at work when we think aboutthe adverbial/process-first distinction in this way is clearly not a causalsense. The facts about what happens to certain oxygen and carbon atomswhen I light the gas are the facts that make it the case that there is aninstance of burning, but not because these facts cause the gas to burn.They are facts about the burning itself. This noncausal metaphysical rela-

tion of making it the case is one for which we would like to have a thor-ough philosophical account. For present purposes it is enough to notethat this variety of making it the case is a determination relation, in thesense that claims with the form p makes it the case thata is X corre-spond to supervenience claims in the following way:

The fact that an item has propertiesp1. . .p

nmakes it the case that that

item is an instance of X only if it is metaphysically necessary that anytwo items that are the same in respect of propertiesp

1

. . .pn

are also thesame in respect of whether they are instances of X.

What this says is that if any two items that are the same in respect ofp

1. . .p

ncan differ in that one is an X and the other is not, then something

other than the propertiesp1. . .p

nmakes it the case that the one of them

is an X. Or, to put it another way, it says that for some arbitrary thatis an instance of X, the facts aboutp

1. . .p

nmake it the case thatis an

X only if the question of whether is an X is settled by the facts about

whether hasp1 . . .pn. The tea making that took place in my kitchen thismorning was hasty, but the tea making that took place this afternoonwas not hasty. The two tea makings were the same in respect of theprocesses that they involved (boiling water, putting a tea bag in the cup,and so on), but one was an instance of haste and the other was not. Itmust therefore have been something other than the facts aboutwhichprocesses were involved that made it the case that the tea making thismorning was an instance of haste.

This link between

attention et cognition

Documents