Table Of ContentAGENT CAUSALITY
SYNTHESE LIBRARY
STUDIES IN EPISTEMOLOGY,
LOGIC, METHODOLOGY, AND PHILOSOPHY OF SCIENCE
Managing Editor:
JAAKKO HINTIKKA, Boston University
Editors:
DIRK VAN DALEN, University of Utrecht, The Netherlands
DONALD DAVIDSON, University of California, Berkeley
THEO A.F. KUIPERS, University of Groningen, The Netherlands
PATRICK SUPPES, Stanford University, California
JAN WOLENSKI, Jagiellonian University, Krakow, Poland
VOLUME 283
FRED VOLLMER
Department of Psychology,
University of Bergen, Norway
AGENT CAUSALITY
'~ ·
SPRINGER-SCIENCE+BUSINESS MEDIA B.V.
A C.I.P. Catalogue record for this book is available from the Library of Congress.
ISBN 978-90-481-5272-8 ISBN 978-94-015-9225-3 (eBook)
DOI 10.1007/978-94-015-9225-3
Printed an acid-free paper
AII Rights Reserved
© 1999 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 1999
Softcover reprlnt of the hardcover 1s t edition 1999
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
incIuding photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
CONTENTS
Preface
Vll
CHAPTER I.
PROBLEMS WITH THE COMMON SENSE THEORY OF ACTION
CHAPTER2.
FRANKFURT'S SOLUTION 23
CHAPTER3.
THEORIES OF AGENT CAUSALITY 25
CHAPTER4.
PERSONS 42
CHAPTERS.
REFLECTIVE CONSCIOUSNESS 72
CHAPTER6.
EMPIRICAL RESEARCH 93
CHAPTER 7.
THE BPI-PHENOMENALIST PROBLEM 103
CHAPTERS.
IS THE SELF A SOCIAL CONSTRUCTION? 115
CHAPTER9.
DARKNESS, MADNESS AND CHILDHOOD 134
CHAPTER 10.
OVERALL SUMMARY AND CONCLUSIONS 154
REFERENCES 157
NAME INDEX 165
SUBJECT INDEX 167
PREFACE
We act for reasons. But, it is sometimes claimed, the mental states and
events that make up reasons, are not sufficient conditions of actions.
Reasons never make actions happen. We- as agents (persons, selves,
subjects) - make our actions happen. Actions are done by us, not
elicited by reasons. The present essay is an attempt to understand this
concept of agent causality. Who -~ or what - is an agent ? And how
- in virtue of what - does an agent do things, or refrain from doing
them?
The first chapter deals with problems in the theory of action that
seem to require the assumption that actions are controlled by agents.
Chapters two and three then review and discuss theories of agent cau
sality. Chapters four and five make up the central parts of the essay in
which my own solution is put forth, and chapter six presents some data
that seem to support this view. Chapter seven discusses how the theory
can be reconciled with neuro-physiological facts. And in the last two
chapters the theory is confronted with conflicting viewpoints and phe
nomena.
Daniel Robinson and Richard Swinburne took time to read parts of
the manuscript in draft form. Though they disagree with my main
viewpoints on the nature of the self, their conunents were very helpful.
I hereby thank them both.
Parts of chapters five and six use material previously presented in
an article published in Journal for the Theory of Social Behaviour,
1995, 25, 175-190. I am grateful to the Publishers for permission to use
this material.
Vll
1. PROBLEMS WITH THE COMMON SENSE
THEORY OF ACTION
1.1.
The idea that actions are caused or controlled by agents has its
background in some problems with the common sense theory of
intentional action. I shall start, therefore, by giving a sketch of the
theory in its standard form (see e.g. Davidson, 1980; and Fodor, 1987,
1989), and then go on to discuss the problems.
1.2.
For anyone to perform an intentional action, a number of conditions
must be fulfilled. Normally - disregarding purely mental acts - a
person can't perform an intentional action unless he moves his body in
some way, or keeps it still, e.g. stands up or stays seated.
When a person moves his body and performs an intentional action,
he also has to have some understanding of what he is doing, enabling
him to describe his behavior, e.g. "I stood up", "stretched my legs",
"voted for the proposal". Actions, according to Anscombe (1957) and
Davidson (1980), are only intentional under certain descriptions,
namely those in terms of which the agent himself would describe what
he has done. Thus "standing up" might be a description under which
p' s behavior was intentional, whereas "voting for a proposal" might
not be. Perhaps p had been asleep, not knowing that a proposal had
been put to the vote, and just stood up because everyone else suddenly
did so. Stated another way, behavior that is intentional must have a
meaning for the agent, this meaning determining what the person can
be said to have done intentionally.
Moving one's body is by most authors considered to be the basic
component in any action (but see section 3.5)-basic in the sense that
2 CHAPTER 1
it is something one does directly, and is the means by which one does
other things. By pulling the trigger e.g., I fire the gun, shoot someone,
and kill him. Moving one's body has effects (and conventional
meanings). Consequently, in describing what a person has done, one
can focus on the bodily movements he has performed (he pulled the
trigger), yielding a basic description - or on one or several of the
effects (he killed someone), yielding non-basic descriptions. Since
moving one's body is normally something a person is aware of, basic
descriptions will usually describe things the person did intentionally.
But, since an agent is not aware of, and cannot anticipate, all the things
that happen as effects of his bodily movements, non-basic descriptions
do not always describe things the person did intentionally. Descriptions
of non-intended effects are, according to Davidson (1980), still
descriptions of actions. Once a man has moved a finger intentionally,
"each consequence presents us with a deed (p.53)".
A person who performs an intentional action normally also has a
reason for doing what he thinks he is doing. My reason for standing up
and stretching my legs e.g. might be that I feel a cramp in my leg that I
want to get rid of, and that I believe the cramp may go away if I get up
and stretch my legs. The desire and belief, in tum, may lead me to form
the intention (or decide) to stand up and stretch my legs. Desires,
beliefs and intentions that make up reasons, are mental states that are
about something (intentional). They have contents that are expressed in
propositions. And in virtue of their propositional contents, as
understood and described (non-transparently or subjectively) by the
persons they belong to, reasons justifY or rationalize actions as
described by their agents. There is a logical connection (some like to
say) between actions (under certain descriptions) and reasons with
certain contents. Wanting to be rid of the pain, and believing that
getting up will make it go away, imply that it would be good to get up,
that getting up is the right thing to do. (Strictly speaking, reasons imply
judgments about actions, not actions themselves).
Though ordinarily a person can both say what he is doing and give a
separate reason for his action, this is not always the case. Sometimes
PROBLEMS WITH THE COMMON SENSE THEORY OF ACTION 3
the agent has no reason beyond what is mentioned in the description of
the action. "Why did I stand up?"-"I just wanted to".
A last criterion of intentional action is that the reason an agent has
for performing such an action (the reason that justifies performing it),
also causes him to perform it, and so explains why he performed it.
Assuming that reasons cause actions has some implications. According
to Fodor (1989), if wanting to stretch my legs caused me to stand up,
then wanting to stretch my legs must, under certain conditions, the
ones that were satisfied when I stood up, be a sufficient condition for
my standing up. If wanting to stretch my legs (under certain
conditions) is a sufficient condition for standing up, then whenever I
want to stretch my legs, and conditions ... are satisfied, I stand up. This,
in tum, means that there must be laws connecting reasons and actions,
though these laws will not be strict laws, but laws "hedged" by ceteris
paribus clauses. Fodor (1989) believes that such laws can be found,
and that it is possible to know when their ceteris paribus conditions are
satisfied. Davidson (1980), on the other hand, though not doubting that
reasons cause actions, suspects that the ceteris paribus conditions
surrounding reasons and actions are so numerous and complex that it
will be impossible to specify them and know when they are satisfied.
For this reason Davidson believes that intentional (or psychological)
laws can never be more than rough correlations. According to
Davidson, the precise (scientific) laws covering reasons and actions are
likely to be laws couched in neuro-physiological terms.
Now to the problems.
1.3.
The first challenge to the common sense theory is the problem of
deviant causal chains. This challenge is typically presented in the form
of (more or less contrived) examples in which an agent has reasons for
doing X, forms the intention to do X, and then, because of the intention
to do X, does X. It is then claimed that, according to the common sense
theory, X should be an intentional action, whereas in reality it is not.
4 CHAPTER 1
There are, according to Brand ( 1984 ), two kinds of deviant causal
chains. In the one type, called consequential waywardness, "the agents
begin the activities in the appropriate ways; there is no problem about
the connection between the antecedent mental events and the initiation
of the behavior. The activities, however, are not completed as expected
(p.18)". In the other type, called antecedential waywardness, "there is
interference between the mental antecedent and the resultant bodily
behavior (p.18)".
The most well-known example of consequential waywardness is
perhaps Chisholm's (1966) nephew who wants to inherit his uncle's
fortune, believes that he will only inherit this fortune if he kills his
uncle, and then forms the intention to kill him and thereby get the
money. He believes that his uncle is at home and drives toward his
house. But his desire to kill his uncle agitates "him so severely that he
drives excessively fast, with the result that he accidently runs over and
kills a pedestrian who, unknown to the nephew, was none other than
the uncle (p.30)". That is, he intends to kill his uncle, and as a result of
this intention does something that leads to his death. Yet it is obvious
that killing his uncle was not something the nephew did intentionally.
But this poses no threat to the common sense theory, in the form we
have presented it. For according to that theory, if n killed his uncle
intentionally, he must have had a reason for doing so. Because of that
reason he must have done something causing his uncle's death. And he
must have understood that what he did was killing his uncle, or was
something that might cause his uncle's death. But, in Chisholm's
example, what the nephew did, in his own mind, had nothing
whatsoever to do with killing his uncle. And so, though n had a reason
for killing his uncle, which caused him to do something that resulted in
his uncle's death, according to the common sense theory, n could not
be said to have killed his uncle intentionally.
Moya (1990) has constructed "a modified version of Chisholm's
case. We have again a man who wants to inherit a fortune and believes
that by killing his uncle he will be able to inherit a fortune.
Accordingly, he forms the intention to kill his uncle by running him
over at about 7.30 in the evening. He knows that his uncle crosses
