Consciousness & Emotion

Agency, conscious choice, and selective perception

Consciousness & Emotion Book Series
Consciousness & Emotion Book Series publishes original works on this topic, in philosophy,
psychology and the neurosciences. The series emphasizes thoughtful analysis of the implications
of both empirical and experiential (e.g., clinical psychological) approaches to emotion. It will
include topical works by scientists who are interested in the implications of their empirical
findings for an understanding of emotion and consciousness and their interrelations.

Editors
Ralph D. Ellis Natika Newton
Clark Atlanta University Nassau County Community College, NY

Editorial Board
Carl M. Anderson Maxim I. Stamenov
McLean Hospital, Harvard University School of Medi- Bulgarian Academy of Sciences
cine, Cambridge, MA
Douglas F. Watt
Bill Faw Quincy Hospital, Boston, MA
Brewton Parker College, Mt. Vernon, GA
Peter Zachar
Eugene T. Gendlin Auburn University, Montgomery, AL
University of Chicago
Jaak Panksepp
Bowling Green State University, OH

Advisory Editors
Bernard J. Baars Alfred R. Mele
Wright Institute, Berkeley, CA Florida State University, Talahassee, FL
Thomas C. Dalton Martin Peper
California Polytechnic Institute, San Luis Obispo, CA University of Freiburg, Freiburg, Germany
Nicholas Georgalis Edward Ragsdale
East Carolina Univeristy, Greenville, NC New York, NY
George Graham Howard Shevrin
Wake Forest University, Wake Forest, North Carolina University of Michigan, Ann Arbor, MI
Valerie Gray Hardcastle Lynn Stephens
Virginia Polytechnic Institute, Blacksburg, VA University of Alabama, Birmingham, AL
Alfred W. Kaszniak Kathleen Wider
University of Arizona, Tucson, AZ University of Michigan, Dearborn, MI

Volume 1
Consciousness & Emotion: Agency, conscious choice, and selective perception
Edited by Ralph D. Ellis and Natika Newton
Consciousness & Emotion
Agency, conscious choice,
and selective perception

Edited by

Ralph D. Ellis
Clark Atlanta University

Natika Newton
Nassau County Community College, NY

John Benjamins Publishing Company

Amsterdam / Philadelphia
 TM The paper used in this publication meets the minimum requirements of
8

American National Standard for Information Sciences – Permanence of

Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data

Consciousness & Emotion : agency, conscious choice, and selective perception /

edited by Ralph D. Ellis and Natika Newton
p. cm. (Consciousness & Emotion Book Series, v. 1)
Includes bibliographical references and index.
1. Emotions and cognition. 2. Intentionalism.
BF311.C64455 2005
152.4-dc22 2004062311
isbn 90 272 3228 8 (Eur.) / 1 58811 596 8 (US) (Hb; alk. paper)

© 2005 – John Benjamins B.V.

No part of this book may be reproduced in any form, by print, photoprint, microfilm, or
any other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
JB[v.20020404] Prn:23/11/2004; 9:12 F: Z131CO.tex / p.1 (42-105)

Table of contents

Introduction ix
Ralph D. Ellis and Natika Newton

I. Emotional influences on perception and thought

Subjective prerequisites for the construction of an objective world 3
Gudmund J. W. Smith and Ingegerd Carlsson
Energetic effects of emotions on cognitions: Complementary
psychobiological and psychosocial findings 23
Luc Ciompi and Jaak Panksepp
Negative affective states’ effects on perception of affective pictures 57
F. Pahlavan and T. Lubart
Neural development: Affective and immune system influences 81
George F. R. Ellis and Judith A. Toronchuk
Consciousness, emotion and face: An event-related potentials
(ERP) study 121
Michela Balconi and Claudio Lucchiari
Phenomenal consciousness, sense impressions, and the logic
of ‘What it’s like’ 137
David Beisecker

II. Agency and choice

Exposing the covert agent 157
Anton Lethin
Doing it and meaning it: And the relationship between the two 181
Marek McGann and Steve Torrance
Anticipatory consciousness, Libet’s veto and a close-enough theory
of free will 197
Azim F. Shariff and Jordan B. Peterson
Freud’s phenomenology of the emotions 217
Thomas Natsoulas
 JB[v.20020404] Prn:23/11/2004; 9:12 F: Z131CO.tex / p.2 (105-119)

 Table of contents

Verbal expressions of self and emotions: A taxonomy with

implications for Alexithymia and related disorders 243
Louise Sundararajan and Lenhart K. Schubert

III. Agency and moral value

Apt affect: Moral concept mastery and the phenomenology
of emotions 287
Elisa A. Hurley
The Varieties of Religious Experience considered from the perspective
of James’s account of the stream of consciousness 303
Thomas Natsoulas
Index 327
JB[v.20020404] Prn:23/11/2004; 9:32 F: Z131AD.tex / p.1 (43-222)

Author addresses

Michela Balconi Ralph D. Ellis

Catholic University of Milan
Department of Psychology
Largo Gemelli 1
I-20123 MILAN
Italy
michela.balconi@unicatt.it
David Beisecker
Department of Philosophy
University of Nevada, Las Vegas
4505 Maryland Pkwy, Box 455028
Las Vegas, NV 89154
USA
Clark Atlanta University
Campus P.O. Box 81
ATLANTA, GA 30314
USA
ralphellis@mindspring.com
Glenn Geher
Department of Psychology JFT 314
State University of New York
75 South Manheim Boulevard
New Paltz, NY 12561
USA
Geherg@newpaltz.edu

Elisa Hurley
Ingegerd Carlsson
Philosophy Department
Department of Psychology
New North Building, 2nd Floor
Lund University, Box 213
Georgetown University
SE-221 00 LUND
37th and O Streets, NW
Sweden
Washington, D.C. 20009
USA
Luc Ciompi
Sozialpsychiatrische Anton Lethin
Universitätsklinik 300 Moncada Way
Murtenstrasse 21 San Francisco, CA 94127
CH-3010 BERN USA
Switzerland a.lethin@att.net

George F. R. Ellis Todd Lubart

Mathematics Department Université René Descartes
University of Cape Town Institut de Psychologie
Rondebosch 7701 Laboratoire de Psychologie sociale
Cape Town 71 Av. Edouard Vaillant
South Africa F-92774 BOULOGNE CEDEX
ellis@maths.uct.ac.za France
 JB[v.20020404] Prn:23/11/2004; 9:32 F: Z131AD.tex / p.2 (222-395)

 Author addresses

Marek McGann Lenhart K. Schubert

Dept. of Psychology University of Rochester
DBS School of Arts Computer Science
Balfe St., Dublin 2 Department Rochester
Ireland NY 14627-0226
USA
Thomas Natsoulas schubert@cs.rochester.edu
Department of Psychology
University of California, Davis Azim F. Shariff
Davis, CA 95616
Psychology Department
USA University of Toronto
Tnatsoulas@ucdavis.edu
100 St. George St.
Toronto, Ontario
Natika Newton
Canada
Nassau County Comm. College
Philosophy Dept.
Garden City, NY 11530 Gudmund J.W. Smith
USA Department of Psychology
nnewton@suffolk.lib.ny.us Lund University, Box 213
SE-221 00 LUND
Farzaneh Pahlavan Sweden
Université René Descartes
Institut de Psychologie Dr. Louise Sundararajan
Laboratoire de Psychologie Sociale 691 French Road
71, Av. Edouard Vaillant Rochester, NY 14618
92774 Boulogne cedex USA
France louiselu@pop3.frontiernet.net

Jaak Panksepp
Judith A. Toronchuk
Bowling Green State University
Psychology Department
JP Scott Center for Neuroscience,
Trinity Western University
Mind and Behavior Department
7600 Glover Road
of Psychology
Langley, B.C. V2Y 1Y1
Bowling Green, OH 43403-0228
Canada
USA
jpankse@bgnet.bgsu.edu
Steve Torrance
Jordan Peterson School of Health &
Psychology Department Social Sciences
University of Toronto Middlesex University
100 St. George St., Toronto, Ontario Enfield, EN3 4SF
Canada United Kingdom
JB[v.20020404] Prn:13/01/2005; 13:50 F: Z131IN.tex / p.1 (41-116)

Introduction

Ralph D. Ellis and Natika Newton

The papers in this volume are organized around the central theme of “Enac-
tivism”, a topic that is very dear to both editors and that has previously been
explored in several articles in the journal Consciousness & Emotion. The term
“enaction” was coined by Varela et al. (1991), and is related to the “embod-
iment” (Clark 1997) movement in cognitive psychology and philosophy of
mind. The term “embodiment” is often misused to mean little more than the
thesis that the subject has a body, or that mental processes have physiologi-
cal correlates. “Enactive” is more resistant to this misuse. Enactive processes,
as the name implies, are enacted rather than merely undergone by their sub-
jects. In the case of emotion, we are not merely the recipients of information
or the passive victims of input and learning. The organism first is engaged in
an ongoing, complex pattern of self-organizational activity, for the purpose of
maintaining a dynamical continuity of pattern across changes of subserving
micro-constituents and environmental conditions. This dynamical structure,
which allows organisms to seek out, appropriate, replace and reshape their
own micro-constituents as well as environmental affordances (within limits)
makes use of multiple shunt mechanisms, feedback loops, and other features
of complex dynamical systems. This self-organizational structure can be used
to distinguish between action (in the most primitive, minimal sense) and
mere reaction.
Not all of the authors in this volume speak the same theoretical language,
but they all contribute to understanding the active rather than passive basis of
emotional processes, and the way those in turn are needed to ground other con-
scious processes, including attention and perception. The papers are focused
on three aspects of the emotion-grounded enaction of conscious processes.
 JB[v.20020404] Prn:13/01/2005; 13:50 F: Z131IN.tex / p.2 (116-146)

 Ralph D. Ellis and Natika Newton

Part I. Emotional influences on perception and thought

The notion that emotion influences perception and thought in major ways –
not just “coloring” them in an additive way, but actually grounding them and
at least partly determining their content and perhaps even their possibility –
is an idea that has been gaining favor very recently. It is demonstrated empiri-
cally in several of the papers in the current volume. At a theoretical level, Smith
and Carlsson explore in a multi-disciplinary way the influence of emotionally-
determined subjective categories on the way we construct the objective world
in general. This paper will be especially interesting to phenomenologically-
oriented readers. Ciompi and Panksepp show from two different approaches
in what way emotion influences cognition. In an empirical study, Pahlavan
and Lubart, focusing on behavioral correlates and using an emotional face-to-
face recognition paradigm, conclude that perceptual objects selectively enter
our emotional awareness depending on their emotional valence. Balconi and
Lucchiari, taking a more neurophysiological approach, use event related po-
tentials to reach a conclusion that overlaps significantly with that of Pahlavan
and Lubart.
The David Beisecker paper is located squarely in the analytic philosophy
of mind tradition, and attempts to dispel the notion that the “what it’s like”
dimension of consciousness is unexplainable within a physicalist framework;
here too, consciousness emerges as something the subject does rather than in-
formation that is received by the subject. While it may be mysterious that one
person cannot receive the same information that another can, it is not so mys-
terious that one person cannot literally execute numerically the same actions
as another – though the actions of two different people may be similar or even
interactive.
To complete this section on emotional influences on perception and
thought, the paper by Ellis (no relation) and Toronchuk explores two theoret-
ical hypotheses that have already achieved a good bit of empirical grounding,
and both of which reflect in different ways the self-organizing processes under-
lying motivated organisms: Panksepp’s idea that endogenous emotional sys-
tems in the brain already motivate highly complex emotional processes that are
not based on learning and conditioning by association with simpler emotions;
and Edelman’s “neural Darwinism” hypothesis. What both of these theories
have in common is that they explain highly complex motivational influences
on behavior in terms of endogenous features of the way systems are organized,
rather than emphasizing classical and instrumental conditioning as ways for
complex valence-directed attitudes (emotions, motivations, feelings, and be-
JB[v.20020404] Prn:13/01/2005; 13:50 F: Z131IN.tex / p.3 (146-196)

Introduction 

havioral dispositions) to develop. The authors conclude that both the Panksepp
and the Edelman dimensions are needed, in a coherent synthesis.

Part II. Agency and choice

Part II explicitly addresses enactive approaches to the as-yet poorly understood

phenomenon of agency. In everyday experience, our own goal-directed agency
is one of the most commonplace features, yet it is so difficult to find a place for
it in the sciences of modernity that many have simply denied the very existence
of the phenomenon, concluding that our feeling that we direct our own actions
is a mere illusion – as recently argued by Daniel Wegner (2002). The authors of
this section all reject that solution, and attempt to develop an understanding
of agency that can also meet Wegner’s challenge by being grounded in good
science. The two papers by Lethin and by McGann and Torrance are review
papers outlining the self-organizational distinction between action and reac-
tion, or between activity and passivity, in specifically enactive terms. Shariff
and Peterson directly take up the central argument of Wegner, which is rooted
in the Libet (1999) finding that a “readiness potential” precedes the conscious
awareness of deciding to act. Shariff and Peterson offer an alternative account
of why Libet’s readiness potential precedes conscious awareness yet does not
eliminate the phenomenon of agency, based on original empirical findings as
well as analysis of other studies.
Thomas Natsoulas looks at the phenomenology of unconscious emo-
tional influences on consciousness, comparing the arguments of Freud with
the competing claims of James. At issue is the phenomenology of our access
to emotional states in general. Does the consciousness of an object also con-
tain consciousness of our own consciousness, or must we, as James argued, go
outside and “look back in” to “introspect” our consciousness, thus making it
into an object? But paradoxically, Freud, who believed like Husserl that self-
consciousness is built into all of our consciousness, and needs only to be sep-
arated out from it, also famously argued that there are unconscious emotional
processes that are every bit as complex and sophisticated as our conscious ones.
Natsoulas presents an integrated picture of how this is possible.
In the last paper of this section, Sundararajan and Schubert approach
the problem of unconscious emotional influences on our behavioral and con-
scious processes from the reverse direction, developing an entire methodology
for studying unconscious emotional contents – a psycholinguistic approach,
grounded in psychotherapy research. Their focus is the phenomenon of alex-
 JB[v.20020404] Prn:13/01/2005; 13:50 F: Z131IN.tex / p.4 (196-234)

 Ralph D. Ellis and Natika Newton

ithymia, the inability to access one’s own emotional contents and feeling pro-
cesses. Using psychotherapy data based on years of research with alexithymics
as well as normal subjects, they develop a way to correlate linguistic patterns
with unconscious emotional contents. This methodology should be highly
useful to those whose work involves either studying unconscious emotional
processes, or treating them therapeutically.

Part III. Agency and moral value

Finally, we should not overlook the fact that agency, the fact that we act rather
than just react to inputs, is an important dimension in all kinds of philosophi-
cal problems, including ethical ones. The two papers of this section explore this
moral dimension. Elisa Hurley uses a combination of phenomenological and
analytic methods to address the problem of moral concept mastery from the
standpoint of agency. And Thomas Natsoulas here again explores the thinking
of James, this time in The Varieties of Religious Experience, considered in terms
of James’s overall thinking about the nature of consciousness, from a critical
perspective.

References

Clark, A. (1997). Being there: Putting brain, body and world together again. Cambridge, MA:
MIT Press.
Libet, B. (1999). Do we have free will? Journal of Consciousness Studies, 6, 47–58.
Varela, F., Thompson, E., & Rosch, E. (1991/1993). The embodied mind. Cambridge, MA:
MIT Press.
Wegner, D. (2002). The illusion of conscious will. Cambridge, MA.: MIT Press.
JB[v.20020404] Prn:14/09/2004; 15:57 F: Z131P1.tex / p.1 (42-68)

P I

Emotional influences on perception

and thought
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.1 (41-102)

Subjective prerequisites for the construction

of an objective world

Gudmund J. W. Smith and Ingegerd Carlsson

Lund University, Sweden

Starting with the assumption that perception is not merely a reaction to

outside stimulation but a process of creation, although mostly brief and
inaccessible to conscious retrieval, the present paper attempts to determine
the prerequisites of a stable and secure account of reality. The authors’
leading hypothesis is that a dependable end product of adaptive and
constructive processes is contingent on open communication with its
subjective/emotional origins. A number of experiments offer empirical
support for such an assumption. These experiments also address problems
associated with creative functioning, anxiety, and defensive strategies. The
general conclusion of this survey is that how we conceive of the outside world
is determined, not only by the terminal stages of adaptive and constructive
processes but by the entire course of these events, including its
subjective/emotional origins.

Keywords: perceptgenesis, anxiety, psychological defenses, perception,

subjectivity, consciousness, emotional adaptivity

This paper argues that the stability of an objective conception of reality de-
pends on the accessibility of its subjective roots. Such a statement presupposes
that perception is not just conceived as a sensorial depiction of what happens
to exist out there, an on-the-spot imprint of the stimulus context. Instead,
perception is envisaged as the outcome of constructive processes, or recur-
rent pulses, originating in personal experience. This perceptgenetic position,
in many respects at odds with more fashionable assumptions in contempo-
rary psychology, will be spelled out more fully in what follows. We will also
use material from two previous studies to illustrate that lack of subjective un-
derpinnings threatens the firmness of our perceptual world. In a similar spirit
Tomkins (1991) held that without the “cement of affects” the individual’s sense
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.2 (102-149)

 Gudmund J. W. Smith and Ingegerd Carlsson

of coherence and sameness over time would be jeopardized. Since an open

communication between the consensual present and previous non-adapted
stages of the perceptgeneses is a requirement for creative functioning, as we
understand it, creativity will form a substantial part of the present survey.

Perceptgeneses

The notion that percepts are not just reflections of stimulation impinging on
sensory surfaces but rather products of constructive processes, however brief
and blocked from introspection, was first formulated in the nineteen-twenties.
These early phases of what was then termed Aktualgenese (in Germany) and
microgenesis (in the US) were revived half a century ago by Werner (1956).
Brown (2002) used the same term in a neurophysiological context. The
term perceptgenesis was coined in Sweden to account for the fact that most
empirical research in the area concerned perception (for a recent account, see
Smith 2001).
It is important to realize, however, that perception in early continental Ak-
tualgenese was kept separate from other functions, such as memorizing, and
particularly the area termed personality with its aura of emotions and irrational
dynamics. Indeed, the notion that perception could be an approach to person-
ality did not surface in the US until the middle of last century (Blake & Ramsey
1951). But if we assume that perceptgeneses represent series of subsequent
actualizations of the personal world, which are only gradually, in the course
of the process, constructed by the actual stimulus situation, the early process
stages would necessarily be more stamped by subjectivism than later stages. As
noted by Glicksohn (1998), referring to Smith and Westerlundh (1980), when
normally preattentive stages are forced, in a microgenetic experiment (see be-
low), to construct percepts as well as feasibly possible, these percepts have a
dreamlike quality to them.
What happens during the course of a perceptgenesis is thus seen as a release
step by step of the objective percept from the embrace of subjectivism. One im-
portant question in the present context would thus be how important for the
stability and dependability of the final percept a complete release would be, or
an abrupt release compared to a gradual one. Since these processes could not
very well be studied by way of introspection, special techniques have been in-
vented to extend them in time. One such technique involved fractionization of
the stimulus presentations, starting with very brief exposures in a tachistoscope
which were subsequently prolonged in a systematic fashion. The subject’s series
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.3 (149-201)

Subjective prerequisites for the construction of an objective world 

of reports of impressions was assumed to mirror a real perceptgenesis, elevated

from its preconscious hiding place to public scrutiny.

Adaptive serials

Since perceptgeneses reflect how we construct the world around us within

the sensorial givens, they cannot very easily be distinguished from processes
reflecting how we adapt to these givens. Consequently, the study of perceptge-
neses has been accompanied by studies of adaptation, i.e., processes with more
easily observable extension in time. Let the subjects be confronted again and
again with an unfamiliar, complex or contradictory situation and record how
far on the road to full accomodation they proceed during a fixed time segment,
or give them sequential chunks of an adaptive assignment and record the times
taken to complete them. When the series of measurements starts to level out,
the adaptive process is supposed to have reached a stage comparable to the final
stage (C-phase) in a perceptgenesis.
While perceptgeneses have proved to reflect important individual charac-
teristics, particularly when disquieting picture material has been administered,
adaptive curves have likewise been revealing. What has been of special impor-
tance is that the stimulus material be meaningful and unknown to the test
person. Threatening pictures can be used to unleash defensive reactions in
the viewer (Hentschel et al. 2004). In experiments with adaptive styles, such
contradictory situations as the Stroop Color-Word test have proved profitable
in diagnostic/clinical work (Smith 2001). This does not mean, however, that
perceptgeneses or series of adaptive actions lose their individual distinction un-
der more ordinary circumstances. The examples presented below will support
this point.

Self-nonself integration

A first illustration is taken from Hansson and Rydén (1987). The overall aim
of the study was to map the relation between differentiation and integration
of self and nonself, on the one hand, and modes of perceptual adaptation, on
the other.
The interplay between self and nonself was defined by means of the Spiral
Aftereffect Technique (SAT) as developed by Andersson (1972) and described
in more detail below. The SAT is constructed to reflect the balance between
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.4 (201-250)

 Gudmund J. W. Smith and Ingegerd Carlsson

self factors (long aftereffects) and nonself factors (inhibited aftereffects). Of

particular importance was that the initial attempts to come to terms with self-
and nonself-aspects of perception were kept separate and registered, as were the
ensuing, successive stages up to a stabilized final performance. In other words,
an adaptive process was studied. The development of the self-nonself balance
was supposed to predict how efficiently the individual could differentiate self
from nonself in another situation, i.e., the classical Rod-and-Frame Test (RFT).
Of the two complementary aspects of perceptual development, differentiation
and integration, the latter was particularly emphasized in the study.

The tests

The Spiral Aftereffect Technique (SAT) utilizes a rotating spiral, 2½ turns,

drawn in black ink on a white background. The aftereffect is projected on a
stationary circle. There are two introductory trials performed 20–30 min be-
fore the main test, which consists of 10 massed trials. Each induction lasts for 45
sec. The participant is instructed to tell when the effect, i.e. the circle seemingly
expanding or coming closer, ceases.
The Rod-and-Frame Test (RFT) was performed in a completely darkened
room. At a viewing distance of 1.3 m the participant faces a luminous frame
with 59.4 cm long sides, with a breadth of 6.5 mm, tilted 20 degrees to the
right. A luminous rod within this frame, 50.2 cm long and 16 mm wide, can
be moved by means of an electric engine. In two pretrials when only the rod
is illuminated the participant is asked to set the rod upright, starting from a
position of 20 degrees to the left. This is followed by 20 trials in the main ex-
periment with both rod and frame illuminated, the rod set 20 degrees to the
left and the frame 20 degrees to the right. The experimenter switches off the
illumination between trials. Deviation from subjective verticality, as obtained
in the pretrials, is calculated. Main deviation scores refer to subsequent groups
of 4 trials, i.e., 5 groups in all.

Procedure and subjects

The participant started with the SAT pretrials, continued with the RFT, and
finally took the main SAT trials. Among the 129 subjects participating in the
study, 68 were female university students and 21 student nurses and, finally, 40
obese patients (31 women and 9 men), most of them 18 to 30 years old.
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.5 (250-300)

Subjective prerequisites for the construction of an objective world 

Results

SAT curves were classified in terms of polynomials of 1 to 4 degrees (For details

of the statistics, cf. Hansson & Rydén 1987). These were the main categories:
flat curves, predominantly rising sequences, decreasing, and discontinuous
ones. Very low SAT durations in any part of the test correlate with substan-
tial deviations from verticality in the RFT. Decreasing SAT trends go together
with low RFT errors and improvement in the last sections of the test. Within
flat and rising SAT curves, brief and long durations compared to intermediate
ones are associated with considerable RFT deviations. Discontinuous trends
also predict either generally large errors and/or continuous deteriorations.

Conclusions

Generally “...very low durations of aftereffects, indicating self-aspects to be

poorly represented, make integration of self- and nonself-aspects difficult
or impossible and differentiation superficial” (Hansson & Rydén 1987: 536).
But even rapidly waning SAT durations correspond to abandonment of self-
aspects, allowing them to be only cursorily integrated. In a more gradual pro-
cess of adaptation, self- and nonself-aspects are more actively integrated. The
trend has to be continuous or else the RFT deviations will be considerable.
As the authors see it, self-nonself relationships with optimal capacity
for self-nonself differentiation, in terms of field-dependence/-independence,
would be found in SAT trends that decrease gradually and level out at any point
that is not extremely low. Still greater capacity for self-nonself differentiation
seems to be indicated by more elaborate strategies, where self- and nonself-
aspects are more intricately interwoven, as indicated by frequent changes of
direction within an overall decreasing trend.
This seems to support our general hypothesis that without a close and mu-
tual interaction between subjective and objective factors in the adaptive pro-
cess, the end-product will be unstable and inconclusive. The best way to make
sure of a dependable account of a situation is thus not to abandon subjectivity
as quickly and definitely as possible but to gradually disengage the subjective
roots of a perceptual process in an ongoing interplay with outside reality.
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.6 (300-332)

 Gudmund J. W. Smith and Ingegerd Carlsson

P-phase to C-phase interaction

Another empirical example derives from a study of creativity in children aged 4

years and upwards to 16 (Smith & Carlsson 1990). The children were subjected
to the Creative Functioning Test (CFT). This consists of two parts. In the first
part a still-life motif is projected perceptgenetic fashion in a tachistoscope, to
begin with at exposure times just below the visual threshold. These are then
systematically prolonged until the viewer (if possible) has delivered three cor-
rect reports in a row. The experimenter particularly notes (1) the number of
mutually different subjective interpretations of the stimulus theme (P-phases);
(2) the stability of the final C-phase. In a second part of the test the procedure
is turned upside-down with continuously diminishing exposure values. If the
subject is willing to retrieve subjective themes, often those already mentioned
in the first part of the test, this is seen as a sign of creativity. To begin with,
however, we will only deal with the first part.

Participants

The first subgroup to concern us here were 47 4–6-year-olds. In order to

group them according to cognitive maturity they were tested with Piaget’s
three-mountain test. Here three differently shaded mountains are placed on
a foursided plate of wood. Four pictures are taken of this doll’s landscape, each
from one of the four different perspectives. The experimenter places a doll at
the child’s position and subsequently at the three remaining positions and asks
the child to pick a photo corresponding to what the doll could be supposed to
see. A totally random series of answers (class 1a) corresponds to the lowest level
of cognitive maturity. At the next level (1b) the child has comprehended that
a problem is involved, without really mastering it, and at 2a it makes at least
some correct choice. From 2b over 3a and 3b the child gradually approaches
full cognitive maturity with respect to the problem posed by the Piaget test.

Results

Among the cognitively least mature children (subgroup 1a) 50% had only 0-1
P-phase themes as against 18% among the remaining children. In subgroups
1a + 1b no child attained a stable C-phase, in the other subgroups 33%. If we
divide the entire sample in two halves we get a lower half (1a, 1b, 2a) of 24
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.7 (332-383)

Subjective prerequisites for the construction of an objective world 

children and an upper half (2b, 3a, 3b) of 23. In the lower half only 17% could
stabilize a C-phase; in the upper half 39%. Only 3% of the former reported 7 or
more P-phase themes compared to 42% of the latter. Stabilization of a C-phase
at the end of the perceptgenesis obviously goes together with an increasing
investment in subjective interpretations of the stimulus.
In the same study there was another group of 86 children aged 7–8 years
and 10–11 years. Most of these children could stabilize a C-phase, all of them
among the oldest ones. Simultaneously the number of P-phase themes in-
creased. Among 55 7–8-year-olds 45% had 3.5 or more themes; among 31
10–11-year-olds, 78%. Increasing age and maturity thus does not necessarily
imply disqualification of subjective interpretations, but rather the opposite.
This subjective expansion goes together with an increased inclination to re-
cover P-phase themes in part two of the test, i.e., the so-called reversed or
descending perceptgenesis. In the group of 10–11-year-olds, 29% recovered
complete P-phase themes; in the group of 7–8-year-olds, only 9%.
As we move up the age ladder to the 12-year-olds, there is a significant
(p < .05) reduction of the number of subjects recovering complete P-phase
themes in the descending genesis. The number of such themes reported in
the first section of the test, however, remains approximately the same. What
made the 12-year-olds reluctant to return to the subjective beginnings of their
perceptgenesis appeared to be a lowered tolerance of anxiety, evidenced by an
increasing dominance of compulsive defenses as spotted in the Meta-Contrast
Technique (MCT, to be dealt with in more detail later).

Conclusion

In both studies the accessibility of subjective roots seems necessary to guaran-

tee the stability of the final command of the stimulus situation. This conclusion
could be turned upside-down, however, at least when it comes to the percept-
genetic experiment. Thus, only a stable C-phase, interpreted as a secure rela-
tion to outside reality, appears to guarantee the reconstruction of preliminary,
subjective phases. In order to grapple with this problem we have to consider
in more depth what happens in the course of an adaptive sequence, or in a
perceptgenesis, and also try to unravel the enigma of creativity.
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.8 (383-429)

 Gudmund J. W. Smith and Ingegerd Carlsson

Perceptgenesis, a closer look

One way to describe a perceptgenesis could be to state, as above, that its begin-
nings are dreamlike but feel gradually closer to reality as the process continues.
In other words, a perceptgenesis could be regarded as a continuity of states of
consciousness. Switching to adaptive serials we can, in an analogous way, see
them as an ongoing interplay between self-factors and nonself-factors. In or-
der for the self-factors to be finally overcome they have to be treated in close
interaction with nonself-factors. Looking at proper perceptgeneses once again
we find the early stages more emotional, closer to me or more ego-involved,
but also more chaotic. An abrupt discharge of these stages would probably
jeopardize the feeling of coherence between emotion and perception.
Considering the perceptgenetic methodology of continually prolonging
exposure times it could seem natural for a traditional psychologist to envision
a perceptgenesis as a series of reports differing only in clarity and detail. The
quantitative change from one exposure to the next could perhaps look like this.

1. A vague contour.
2. The contour is more conspicuous, perhaps a nose.
3. More of the face is coming into focus, an eye too.
4. There is some sort of background that I didn’t notice before, etc.

The theme remains the same and appears gradually like a photo developing in
a processing laboratory.
But a perceptgenesis can just as well proceed in leaps of qualitative change.
Let us start with the nose once again.
2. A nose.
3. No, a whole person standing on a side-walk.
4. Now it looks rather like a landscape, a promontory in a stormy sea.
5. A person again, in a big overcoat.
etc.

Since reports in a perceptgenetic experiment are reconstructions, likely to be

adapted to a level of intelligible communication, or in other words rationalized,
many phases in the genesis, particularly early ones, may seem more ordered,
less dreamlike and chaotic, than their non-reclaimed origins. Moreover, when
further transcribed and recast by a traditional experimenter, such a genesis is
likely to lose even more of its original character. This is wholly in line with a
common assumption, hard to eradicate, that preattentive reactions to sublim-
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.9 (429-476)

Subjective prerequisites for the construction of an objective world 

inal stimuli are nothing but weak copies of supraliminal reactions to the same
stimuli. In the same way as a traditional cognitivist may disregard exciting qual-
ities in a perceptgenesis, the participant him/herself may act as a censor and
keep back all too open indications of self-involvement in the perceptgenesis
(cf. below).
But why not advocate a seemingly more reasonable view: that the sooner
you cut off the subjective roots the more stable and objective the end-product
would be? Such a view is consistent with the belief that subjectivity, disclosed
by its emotional coloring, is nothing but a remnant of our primordial past. In
order for our adaptation to be as orderly and efficient as possible this remnant
should be abandoned quickly and definitely. Admittedly, inability to accept re-
ality, or remaining stuck in one’s own subjectivity, would be counteradaptive.
On the other hand, there are many indications that rejections of subjectivity
are rather signs of defensive denial than of adaptive efficiency. This can be elu-
cidated by means of various perceptgenetic tests of defensive activity. Here we
will choose the Meta-Contrast Technique (MCT, Smith 2001) as an illustration,
partly because it was also applied in the study of children above.

Adaptation and defense

The MCT is built on the operation of pairs of stimuli. One stimulus in a pair,
presented in perceptgenetic fashion at gradually increasing exposure values,
is intended to let the viewer construct a stable perceptual frame of reference.
Another stimulus induces the viewer to start a perceptual process inconsistent
with or even hostile to the frame percept. We shall here dwell on the latter type
of stimulation where attempts at warding off the hostile appearance are most
easily detectable.
The stimulus intended to produce the frame percept (called B hereafter)
depicts a young person sitting at a table with a window close by. When the
viewer has stabilized his/her perception of B, exposure times are cut back to a
standard value where the picture theme can still be comprehended. The other
stimulus (called A) is then exposed immediately before B, first at subthreshold
values which are gradually prolonged.
The viewer sits before a TV screen, size 25×25 cm, where the stimulus
pictures appear and is told to report everything he/she has seen. The percept-
genesis leading to the B percept is usually unproblematic. But since stimulus A
depicts a hostile, grimacing face projected in the window behind the young per-
son in B (with whom the viewer is supposed to identify) the viewer may react
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.10 (476-525)

 Gudmund J. W. Smith and Ingegerd Carlsson

defensively. While the hostile A gradually penetrates the window in most nor-
mal cases, to be sure in various guises but soon revealed as something disquiet-
ing, defensive reactions imply that the threat is either denied or misrepresented
as something innocuous.
We shall dwell on a number of defensive strategies particularly relevant
for the theme of the present topic. Behind the delineation of various strategies
and their relations to a broader symptomatology stands a long clinical research
tradition, reaching back to the nineteen-sixties and accounted for in a recent
manual (Smith, Johnson, Almgren, & Johanson 2002).
In classical psychoanalysis defense is regarded as a reaction to anxiety sig-
nals portending danger. What has to be warded off by various defensive strate-
gies accessible to the individual is not outside perils but threats coming from
within. Among these are unacceptable wishes, clashing with the individual’s
self-image. While this may be a true picture of how defense is often activated it
is certainly not the whole picture. In the present context we would particularly
like to give prominence to dread of chaos or, even, lack of discernible structure
as occasion for defensive reactions. That would be sufficient reason for hold-
ing back early phases in a perceptgeneses, preventing them from being nakedly
reconstructed in a perceptgenetic test situation. In everyday life such defenses
would operate on any associations to and reminiscences of these early roots.
But how does anxiety manifest itself in the MCT? There is a broad spec-
trum of signs, from open fear manifested in the experimental situation to
reports of broken structures like cracks in the walls of B, reports of so-called
zero-phases where nothing meaningful can be distinguished, or of inadequate
defenses admitting blackness to leak through, undue emphasis on darkness
in B, and mild signs that the structure of B is intermittently becoming more
vague or hazy.
Emphasis on blackness and other signs of moderate anxiety are common
in normal people whereas such signs as total loss of meaning are shown to be
pathological. One important type of defense has been termed isolation because
of its frequent appearance in patients with obsessive-compulsive disturbances.
In some patients isolation manifested itself as covering the window in B (where
the threatening A-face is projected) with a screen, often white coated (white
being the color of innocence). In others the viewer simply denied that the
new object penetrating the window was in any way threatening. The most
rigid form of isolation was found among compulsive character neurotics where
nothing but the last, correct phase of the A-process was permitted to be rec-
ognized. A successful defense apparently implied that emotions were denied
access to consciousness, emotions as they are reflected in early, subjective per-
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.11 (525-576)

Subjective prerequisites for the construction of an objective world 

ceptgenetic phases. It is illuminating to note that when the isolating defense

cracked and blackness showed in the crevices the patient was usually ridden by
severe anxiety.
Depressive patients were often characterized by stereotypy. In these cases A
manifested itself as an uninterpreted “something” in the window of B. At this
point the perceptual development was halted over a number of phases (at least
five according to the manual) and the report of that “something” repeated in a
stereotyped manner. The perceptgenesis was obviously narrowed to a kind of
tunnel preventing any free mobility of associations to the stimulus theme. The
final result was even more obstruction of communication with early stages of
the genesis.
When the threat is changed to a less threatening object the defense is
termed repression, partly because it is associated with histrionic (hysteroid)
types of personality. The prototypical sign of repression is the report of a
stiffened face, e.g., a statue or a mask, made harmless by its immobility. The
category was early on extended to include other objects, like a stone, a house,
a tree, a landscape. Geneses with such avoidance reactions are often very lively
and colorful, reflecting what appears characteristic of Thespian characters. The
ontogenetic origins of repression could, moreover, be traced to the behavioral
level in cognitively immature children, who staved off the threat by shutting
their eyes or turning away from the projection screen.
Repression obviously serves its defensive purpose to avoid direct con-
frontation with the appalling A-stimulus. At the same time the richness in
content of geneses with signs of repression announces that their early, emo-
tional origins are not definitely cut off but can be reached via circuituous
routes. This defensive style is reminiscent of the use of metaphor in poetry
in order to illuminate but at the same time to keep at a safe distance..
Projection is another interesting defensive strategy. In the MCT it implies
transferring subjective influences from the A-geneses to the B-percept. Thus
B is reported as changed before A has penetrated the surface of conscious-
ness. Projection serves as a protection against direct confrontation with the
meaning of A. But the A-process is not arrested or strangled. The flexibility
characterizing the B-percept in these cases indicates susceptibility to precon-
scious influence. Mild variants of projection, implying change in perspective,
distance, etc., are termed sensitivity. More pronounced signs of projection need
not be restricted to B-changes but often also concern A. Unconscious identifi-
cation with the threat may thus lead to descriptions of A as benign, a person
of good repute, etc. However, in the more psychotic forms of projection –
which can also be exposed in the MCT as total and abrupt transformations
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.12 (576-628)

 Gudmund J. W. Smith and Ingegerd Carlsson

of the B-percept – the obstacles to open communication with early phases of

the perceptgenesis seem to be more formidable.
As indicated by these descriptions, some types of defense (isolation and
depressive stereotypy) appear to cut off the projection of early perceptgenetic
stages in an abrupt and definitive manner while other types (repression and
projection) still allow a certain early-to-late communication, even if via de-
tours. As will be elucidated in more detail later, this distinction between groups
of defenses is important for our understanding of how some defenses may be
unequivocal obstacles to creative functioning and how others may be more per-
missive. This is also important in view of the fact that defensive operations are
more common in everyday life than we have hitherto assumed, and are not just
instances of psychopathology.

Creativity

The Creative Functioning Test (CFT) was described in connection with the
study of creativity in childhood and adolescence above. As said, it consists
of two parts, an ascending series of stimulus presentations and a descending
one, using the same stimulus in both series. The ascending series where expo-
sure times are systematically prolonged produces an ordinary perceptgenesis.
In this series the number and character of P-phase themes constitute the most
important variable. The descending series with its systematically abbreviated
exposure times starts when the ascending series has attained a stable C-phase.
Here the main interest is whether stimulus deviant themes are reported toward
the end of the series, or whether the participant sticks to what he/she knows to
be correct.
In our view, creativity is a generative way of shaping our conception of
reality, i.e., to see reality unfettered by contemporary conventions of thinking.
In order to do so, we argue, it is necessary to keep the lines of communication
with one’s uncultured self open. Or to use Ernt Kris’s (1952) terminology, to
make a partial regression in the service of the ego. In reference to the CFT
this implies a willingness to go beyond the limits of correctness established by
the C-phase of the ascending series and accept deviant interpretations of the
stimulus in the descending series. These interpretations are often reminiscent
of or even identical with the P-phase reports in the ascending series.
The stimuli used in the two existing versions of the CFT are not primarily
threatening as they are in the MCT. Thus the constructors have tried to min-
imize the risk that defensive strategies will completely distort or strangle the
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.13 (628-684)

Subjective prerequisites for the construction of an objective world 

series. Still, any unknown stimulus can be experienced as a threat until it has
been correctly recognized as innocent. We therefore expect that habitual defen-
sive inclinations as revealed in the MCT will also influence the CFT results. In
order to give an indication of the possible variety of CFT protocols, we want to
present a few excerpts from three protocols given as examples in the test man-
ual. We also refer to the manual for proofs of the validity and reliability of the
test (Smith & Carlsson 2001).
A CFT protocol judged as highly creative (numbers referring to exposure designations).
The protocol contained 3.5 P-phase themes in the ascending series, including
human themes.
The correct interpretation of the stimulus was abandoned at the end of the de-
scending series and replaced by interpretations reminiscent of the P-phase themes.

Ascending series
8b. Could be illuminated figures in a cave. Odd figures, dressed in something white.
9a. Yes, two figures, one light and one dark. Felt it was something threatening there. The
dark one attacking the light one.
9b. Forms building up. Lighter and darker. Perhaps people close together...in some joint
action...if they embraced or fought.
11b. A naked person, like when you are being X-rayed. And then a dark shadow that
someone tried to help to an upright position.
12a. Seems more gruesome. There was a dead body, a victim of violence. The claire-
obscure is frightening in the dark form.
18. Obviously some kind of container in the background. An oval or a circle in perspec-
tive. A figure standing, a human figure.
20a. A relatively vague person and a vessel with a surface. Don’t know if there is water in
it.
(A bowl and a bottle suggested by the test leader).
20b. In that case the bottle has a very human outline. Because I saw a shoulder. But it could
also be a bottle painted that way.

Descending series
19. I accept the bottle.
18. Yes, it looks like a bottle and a basket. Cezanne.
13. I’m back with the human being again, the dejected, vulnerable. Looks like one who
has been X-rayed, treated in a hospital.
11b. The same person, more simple. Defenseless. And a threat lurking in the darkness
closing in on the bright part.
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.14 (684-749)

 Gudmund J. W. Smith and Ingegerd Carlsson

8a. If I hadn’t kept the person on my retina I wouldn’t have seen her this time. Recognized
the contours.
5. Indistinct, just a flash.
A CFT protocol judged as median creative.
Numerous reports of slight change in the descending series. No P-phases re-
claimed.
11. The same bottle and bowl.
10a. I had the impression that the bottle was somewhat taller.
8a. Now it’s moving towards the edge of the screen, one cannot see the whole bottle any
more.
7a. It covers practically half the screen. More towards the middle. Rounded in my direc-
tion, convex.
5b. It’s beginning to disappear away from me, to diminish.
A CFT protocol judged as low creative.
1.0 P-phase themes in the ascending series, human. No themes reclaimed.

Ascending series
6b. Looks like the shadow of a man, walking in a landscape. Fateful.
11a. The form more like a standing bottle. More straight now. The background more
coherent now with the big shapes.
14a. Yes more distinct, a pot and a bottle.

Descending series
14. No change.
11a. The whole motif is visible now.
9a. The light changes. Some on the bottle and spots on the pot.
5a. A gleam of light and a faint silhouette of a bottle and in the bend some light.
4a. Not much. The screen annoys me.

Creativity, defense, and anxiety

It is time to return to the study of creativity in children and youngsters. As men-

tioned, there was a significant reduction among the 12-year-olds of the number
of subjects who recovered complete P-phase themes in their descending gene-
ses, like the participant in the first CFT example presented above. At the same
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.15 (749-800)

Subjective prerequisites for the construction of an objective world 

time, signs of anxiety in the MCT were strikingly thinned out and signs of
isolation defense strengthened. While compulsive strategies obviously checked
the retrieval of subjective themes in the descending part of the CFT they also,
at the same time, subdued the echoes of anxiety in the MCT protocols.
Test results from 142 children affirm that isolation and kindred strategies
have a restraining influence on creative ideation. It is of particular interest to
note the negative association between these strategies and signs of anxiety. If we
take anxiety as an indication of open contact between the conscious endstage
of a perceptgenesis and its preconscious antecedents, it seems evident that iso-
lation functions as a protective shield against emotional subjectivity, and hence
as an obstacle to non-rational, deviant influences on perception and thought.
Depressive stereotypy should have a similar effect.
These findings from the Smith and Carlsson (1990) study are partly sup-
ported by a later investigation (Carlsson 2002). Two extreme groups scoring
either high or low on the CFT were selected from a larger cohort of undergrad-
uate students. There were 12 subjects in each of these subgroups who also took
the MCT and completed two anxiety inventories.
The following MCT categories were tried against the CFT results: anxiety,
repression, isolation, projection/sensitivity, regression and depression. Since
depressive signs were rare in this group of people that category could not be
tested. A special index, the MCT sum of different categories was also calculated.
The higher this index the more variable was the battery of defensive strate-
gies. Finally, the number of various P-phase themes in the ascending series
was counted.
As expected, the high creative subgroup had more P-phase themes in the
ascending genesis than the low creative one (t = 5.63, p < .001). If grave anxiety
was excluded and counted on a par with no anxiety, the high creative subjects
got more markings of (moderate) anxiety than the low creative counterparts
(chi square = 8.40, p = .004). There was, however, only one significant correla-
tion for single defensive categories. Projection, including weak signs, was more
common among high creatives. If sensitivity was separated as a category of its
own no significant association with creativity appeared. What we are here deal-
ing with is most probably signs of moderate projection, grave signs being most
unlikely in the present group.
There was a significant difference (t = 3.7, p = .035) with respect to defen-
sive variability. The more creative the more blended the defensive arsenal. This
can be interpreted as meaning that when, e.g., isolation is mixed with other
defensive counter-measures its negative impact on creative ideation is checked.
In the same way strategies like repression (in the wide meaning used here) do
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.16 (800-852)

 Gudmund J. W. Smith and Ingegerd Carlsson

not influence creative functioning positively unless they allow other strategies
to operate at the same time. Only then do they attain a permeable nature.

An experiment with flight phobia

As an additional support for the main thesis of the present paper we would
like to briefly refer to a study of flight phobia using the Spiral Aftereffect Tech-
nique (SAT) described previously (Amnér 1997). It was known before that long
aftereffects are indicative of manifest anxiety. Surprisingly, however, the flight
phobics did not only cluster at the upper end of the SAT scale. They were also
found at the lower end, among subjects with very brief effects.
Interviewing the participants the experimenter found two distinctive kinds
of flight phobia. Subjects with long aftereffects were generally anxious, non-
self influences coloring their private universe. Subjects with brief effects did
not refer to a general state of anxiety but pinpointed specific causes of con-
cern in the situation at hand. They heard ominous changes in the sound of the
engines; they saw the cabin crew exchange meaningful glances. Even if they su-
pressed the subjective end of their experience of the flight, their world revealed
its frailty. Only those with medium aftereffects did not care, i.e., the controls.

Discussion

In spite of the heterogeneity of experimental situations and test subjects used

to highlight the theme of the present study, it would not be too farfetched to
conclude that the accessibility of subjective themes in the perceptgenesis is as-
sociated with the stability of the perceptual end-product, the C-phase. But as
already intimated, this reasoning could also be turned upside-down. Thus only
a stable contact with reality makes the subject willing to face the arcane early
stages of his/her perceptgeneses. A secure anchorage may increase the tolerance
of anxiety. As assumed, anxiety is often a natural reaction to exposure to the
chaotic and elusive subjectivity characterizing early process stages. But to turn
the reasoning upside-down once again: a stable conception of reality is not the
same as a mere reflection of outside stimulation.
A convincing support for this way of reasoning was found in the experi-
ment with adaptive serials conducted by Hansson and Rydén (1987) and re-
viewed in some detail previously in this text. In participants who mastered the
differentiation of subjective and objective perspectives in the rod-and-frame
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.17 (852-897)

Subjective prerequisites for the construction of an objective world 

situation, the serials reflecting performance in the spiral aftereffect test were
not characterized by a swift restriction of self influences but by gradual release
of the nonself perspective, preferably after a period of intimate integration. We
also referred to a study of flight phobia by Amnér (1997) where the spiral af-
tereffect technique was a central instrument. Subjects who almost completely
denied self factors in their adaptive serials did not achieve a safe reality contact.
Instead, their experience of the flight situation was filled with ominous signs of
impending danger.
As regards the genuine perceptgeneses it is more difficult to directly prove
that C-phases without antecedent P-phases are less stable than those with sub-
jective beginnings. However, when signs of anxiety decrease in the MCT and
signs of defense become more prominent and inflexible, with inhibited recon-
struction of P-phases as a consequence, the most straightforward conclusion
would be that the person suffers from impaired tolerance of anxiety and hence
a less solidly rooted sense of security. This could be easily demonstrated in a
group of 4–6-year-olds taking the CFT. The apparent stability of the C-phase
in CFT among older children tended to obscure this association. But by us-
ing the more menacing picture material of the MCT the brittleness of reality
contact in, e.g., compulsive people could be unveiled.
These results have implications for how creativity should be understood.
As we see it, creative functioning depends on an open interchange between the
subjective and objective poles of perceptgenesis or, to employ the vocabulary
of the Spiral Aftereffect Technique, between self and nonself factors. Creativ-
ity is not synonymous with subjective flooding of consciousness. That would
impede the formation to intelligible material of the new, deviant ideas sup-
posedly brewing outside rational control. But it is also difficult to envisage a
creative person where contact with the subjective roots of perception has been
cut off. Even in the fields of science, the generation of new ideas depends on the
availability of nonadapted levels of experiencing. In a similar way, perception
is usually taken for granted until a disturbance, perhaps a brain injury, reveals
the delicate underlying contrivance.
To summarize, a stable and dependable conception of reality depends on
an unbroken connection with the preparatory stages of the process of construc-
tion, in this context represented by reconstructed perceptgeneses and adaptive
serials. This implies that how we conceive of the outside world is determined,
not only by the terminal stages of the process but by the entire course of adap-
tive events. Openness towards the subjective, or pre-perceptual, origins of our
depiction of reality implies a double bind: not only is the final stage dependent
on the preceding stages but our contact with the early stages is conditioned
 JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.18 (897-964)

 Gudmund J. W. Smith and Ingegerd Carlsson

by the stability of the end-stage. Given these contingencies the process of real-
ity construction can continue unhindered by serious defensive distortions, i.e.,
with a sense of safety – one important prerequisite for creative functioning (cf.
also Sandler 1960; Edelman 1989; Migone & Liotti 1999).
Trying to be onesidedly objective is to be prisoner of an idée fixe. If you
want to break loose you must first of all reestablish contact with your subjective
self, with the source of those processes through which your experiential exis-
tence is shaped. To thus make contact with your emotional self gives you a sense
of consistency, even in the midst of change (Emde 1999). Present adaptive en-
deavours, particularly those of a pioneering character, are deeply rooted in past
experiences as reflected by early sections of a perceptgenesis. True objectivity
is not the mirror image of outside stimulation, but is attained by way of pro-
cesses originating in the self. The former is a soon withering cutting flower; the
latter could be a flower rooted in fertile soil and with growth potential. In an
analogous way, discussing the psychoanalytic theory of consciousness Brown
(2000: 62) contends that “if Cs is caused by or rests on non-conscious activ-
ity, the better part of the state of Cs is ignored by excluding the non-conscious
portion”.
Starting from processes originating in the self, our picture of the world
is constantly being transformed by way of the creation of new symbols. An
objectified “truth” is not the mark of creation, but instead ongoing efforts to
consider both subjective and objective perspectives.

References

Amnér, G. (1997). Fear of flying in civil airline passengers: A manifold problem with various
motivational roots. Lund: Studentlitteratur.
Andersson, A. L. (1972). Personality as reflected in adaptive regulation of visual aftereffect
perception: A review of concepts and empirical findings. In A. L. Andersson, A. Nilsson,
E. Ruuth, & G. Smith (Eds.), Visual aftereffects and the individual as an adaptive system
(pp. 159–171). Lund: Gleerup.
Blake, R. R. & Ramsey, G. V. (Eds.). (1951). Perception: an approach to personality. New York:
Ronald.
Brown, J. W. (2000). Mind and nature. Essays on time and subjectivity. London: Whurr
Publishers.
Brown, J. W. (2002). The self-embodying mind. New York: Barrytown.
Carlsson, I. (2002). Anxiety and flexibility of defense related to high or low creativity.
Creativity Research Journal, 14, 341–349.
Edelman, G. (1989). The remembered present: A biological theory of consciousness. New York:
Basic Books.
JB[v.20020404] Prn:23/11/2004; 9:46 F: Z13101.tex / p.19 (964-1027)

Subjective prerequisites for the construction of an objective world 

Emde, R. (1999). Moving ahead: Integrating influences of affective processes for devel-
opment and for psychoanalysis. International Journal of Psychoanalysis, 80, 317–339.
Glicksohn, J. (1998). States of consciousness and symbolic cognition. The Journal of Mind
and Behavior, 19, 105–118.
Hansson, S. B. & Rydén, O. O. (1987). Relationship between differentiation and integration
of self and nonself: An investigation in terms of modes of adaptation. Perceptual and
Motor Skills, 64, 523–538.
Hentschel, U., Smith, G. J. W., Draguns, J. G., & Ehlers, W. (Eds). (2004). Defense mech-
anisms: Theoretical, research, and clinical perspectives. Amsterdam: Elsevier.
Kris, E. (1952). Psychoanalytic explorations in art. New York: International Universities Press.
Migone, P. & Liotti, L. (1999). Psychoanalysis and cognitive-evolutionary psychology: An
attempt at integration. International Journal of Psychoanalysis, 80, 1–27.
Sandler, J. (1960). The background of safety. International Journal of Psychoanalysis, 41, 191–
198.
Smith, G. J. W. (2001). The process approach to personality. New York: Plenum.
Smith, G. J. W. & Carlsson, I. (1990). The creative process. Psychological Issues, Monograph
57. Madison, C: International Universities Press.
Smith, G. J. W. & Carlsson, I. (2001). CFT – the Creative Functioning Test. Lund: Department
of Psychology.
Smith, G. J. W., Johnson, G., Almgren, P.-E., & Johanson, A. (2002). MCT – the Meta-
Contrast Technique. Lund: Department of Psychology.
Smith, G. J. W. & Westerlundh, B. (1980). Perceptgenesis: A process perspective on per-
ception-personality. Review of Personality and Social Psychology, 1, 94–124.
Tomkins, S. (1991). Affect, imagery, consciousness, Vol. III, The negative affects: Anger and
fear. New York: Springer-Verlag.
Werner, H. (1956). Microgenesis in aphasia. Journal of abnormal and social Psychology, 52,
347–353.
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.1 (41-102)

Energetic effects of emotions on cognitions

Complementary psychobiological
and psychosocial findings

Luc Ciompi* and Jaak Panksepp**

Socio-Psychiatric University Clinic, Switzerland / Bowling Green State
University, USA

This paper seeks complementary themes between psychosocial and

neuroscientific approaches to understanding interactions between emotions
and cognitions, using the “affect-logic” approach of Ciompi and the
“affective neuroscience” approach of Panksepp. Both views envision
meaningful distinctions between emotional and cognitive processes, with the
former being large-scale energetic states of brain and body that reflect
evolutionarily adaptive action systems of the internal world, while the latter
are more informationally encapsulated perception based processes that
distinguish differences in the external world. In the intact organism, they are
fully interactive; each can control the other. Emotions establish global,
non-linear dynamic control over perceptual processes, memory and learning,
and cognitions can trigger and regulate emotional processes. A variety of
research strategies and predictions, ranging from the neurobiological to the
psychosocial, are entertained.

Keywords: emotion, affect,cognition, brain, logic, neurodynamics, chaos

theory

During the last two decades, it has become increasingly evident both in neu-
roscience (Damasio 2000; Davidson 2000; Derryberry & Tucker 1992; Lane &
Nadel 2000; LeDoux 1996; Panksepp, 1998) and in psychology (Dalgleish &
Power 1999; Lazarus 1999), psychopathology (Ciompi 1988, 1997a, c, 1999;
Flack & Laird 1998), social psychology (Forgas 2000a, b; Kowalsky 1999; Lawler
& Thye 1999; Parrott 2001) and affect science (Davidson, Scherer & Goldsmith
2003; Manstead, Frijda & Fischer 2004) that emotions and cognitions are con-
tinually interacting in almost all mental activities. This, of course, does not
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.2 (102-155)

 Luc Ciompi and Jaak Panksepp

mean that they cannot be meaningfully distinguished at either psychological

and/or neural levels; they most certainly can (Panksepp 2003a). However, the
exact nature of these interactions remains far from clear, despite decade-long
debates between leading psychologists on the relative primacy of affect over
cognition or vice-versa (cf. Clore & Ortony 2000; Lazarus 1999; Zajonc 2000).
Certain authors argue that emotions and cognitions are so intimately con-
nected that it makes little sense to hold the classic distinction between passions
and reason (e.g., Lane & Nadel 2000). Others maintain that most data speak
for bidirectional causality with emotions often activated by cognitive stimuli,
and cognitive processes that can be selected and regulated by emotions (Dama-
sio 1994; Derryberry & Tucker 1992; Forgas 2000a, b; Izard 1993; Lazarus
1999; LeDoux 2000; Panksepp 1998a, 2004). The study of affective-cognitive
interactions is further hampered by the fact that there exist no generally rec-
ognized definitions of the key-concepts of emotion and cognition, nor any
consensus on their precise delimitations. The interdisciplinary exchange and
comparison of data is therefore difficult, and new insights that might result
from multidisciplinary approaches are not fostered.
In this paper, we seek to overcome some of these difficulties by comparing
selected findings on affective-cognitive interactions from two different per-
spectives, between which remarkable convergences are gradually emerging: A
low-level neuroscientific approach on the one side, and a high-level psychoso-
cial1 approach on the other. In line with our previously developed concepts
of “affect-logic” (Ciompi 1988 1997a, c, 2003) and “affective neuroscience”
(Panksepp 1982, 1998, 2001, 2003), we will defend the view that the distinction
between emotion and cognition is not only possible and meaningful, but essen-
tial for understanding how they actually interact. While adhering to the notion
of circular causality, we will mainly focus on the largely neglected problem of
regulating and energizing influences that emotions, in our view, continually ex-
ert on almost all cognitive functions. The overall aim of this interdisciplinary
dialog is to identify new working hypotheses for research that might be more
evident in a truly bifocal view. In addition, an effort is made to clarify some as-
pects of the above-mentioned definitional problems on the basis of differential
evolutionary roots of emotions and cognitions, and to bring certain nonlin-
ear dynamic perspectives to bear on the cognition-emotion interaction debate
(also see Lewis & Granic 2000).
Our coverage of the literature must be highly selective, since both fields are
expanding remarkably rapidly during the current “affect revolution”. More-
over, our attention will be focused on a few so-called basic emotions like
interest/curiosity, fear/panic, anger, joy or sadness (cf. e.g. Ekman 1984; Izard
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.3 (155-221)

Energetic effects of emotions on cognitions 

1971, 1992) for which the neurobiological substrata have been provisionally
identified (Panksepp 1982, 1998a). Concerning countless additional nuances,
our position is that many of the more complex emotions reflect epigenetic
emergents of cognitive and socio-cultural abilities interacting with basic emo-
tional systems (Ciompi 1997c; Panksepp & Panksepp 2000). Terms like affects,
emotions or feelings are provisionally used here according to their largely over-
lapping meanings in different related fields of research. In the final discussion,
we propose, however, a possible partial solution of the definitional issues.

. Neurobiological Perspectives

Distinctions between affective and cognitive structures

From the perspective of affective neuroscience, the core affective/emotional

systems of the brain (very largely sub-neocortical, and limbic/paleocortical)
can be distinguished from the more recently evolved higher cognitive struc-
tures (thalamic-neocortical axis) on the basis of various criteria outlined pre-
viously (Panksepp 2003a) and summarized in Table 1.
These distinctions force us to consider emotional brain functions in
quite different ways than has been traditional (e.g., as just another form of
information-processing), and recognize that core emotional systems operate in
holistic dynamic ways (i.e., networks of neurons yield energetic brain and bod-
ily field-dynamics) as opposed to discrete information processing algorithms.
Such distinctions provide novel ways of looking at those pervasive emotion-
cognition interactions that are currently of great interest to psychologists and
other brain/mind scientists. Let us first briefly discuss the meaning of these
six distinctions (derived from Panksepp 2003a), for they are essential for the
types of emotion-cognition interaction perspectives we plan to share. At the

Table 1. Distinct characteristics of emotional/affective and cognitive/informational

systems of the brain

Affective processes (Values) Cognitive processes (Information)

– more subcortical more neocortical
– less computational (analog) more computational (digital)
– intentions in action intentions to act
– action to perception perception to action
– neuromodulator codes neurotransmitter codes
(e.g., more neuropeptidergic) (e.g., more glutamatergic, etc.).
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.4 (221-269)

 Luc Ciompi and Jaak Panksepp

most fundamental level, we view the core of affective/emotional processes to

be more deeply organic (i.e., their status cannot be understood without a “net-
work doctrine” and molecular coding approaches to global neurodynamics)
while cognitions are more deeply informational (i.e., their status may be dra-
matically clarified by information-only approaches). Thus, the governance of
mind and behavior is a multi-tiered, hierarchical process with several distinct
types of mechanisms – some more instinctual and built into the system by
evolution, and the other more flexible and deliberative, relaying heavily on
learning mechanisms.
State functions vs. channel functions. Marcel Mesulam (2000) highlighted that
some aspects of the brain operate via discrete information channels (e.g.,
sensory-perceptual processes) while others operate more globally to control
wide swaths of brain activity (e.g., obvious examples are the biogenic amine
transmitter, brain-wide “spritzers” such as norepinephrine, dopamine and
serotonin, that regulate neuronal arousability in global ways, but many neu-
ropeptides also operate similarly to organize discrete emotions and motiva-
tions). In our estimation, those systems represent the best understood global
state controls, but they are rather non-specific in terms of discrete emotions. In
line with such a distinction, we envision emotional operating systems to consist
of large ensembles of neurons working together, under the “symphonic” con-
trol of neuropeptides to produce coherent organic pressures for action, thought
and feelings (from such a conception, an “energy” dimension for the core
emotional state control systems is apt, and is supported by various affective
neuroscience research strategies, especially neuropeptide regulation of such
global processes – Panksepp 1998a; Panksepp & Harro 2004). This distinction
can help us conceptually differentiate brain processes that produce highly re-
solved perceptual qualia from those aspects that are less sensorially distinct,
holistic-energetic, and in the category of raw feels. The global state-patterns
elaborated by such brain networks may generate an essential psychoneural con-
text for perceptual consciousness – establishing a solid organic grounding for
specific cognitive mental activities linked to discrete information-processing
channels.
Computational vs. non-computational forms of consciousness. This view claims
that channel-functions, since they are dependent on the coding of neuronal
firing patterns in anatomically delimited channels, may be instantiated using
symbol-manipulating computational models. On the other hand, the more
organically instantiated forms of affective consciousness, although also depen-
dent on neuronal systems, are not computational in the same sense. These
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.5 (269-329)

Energetic effects of emotions on cognitions 

systems depend on extensive networks in which the patterns of neuronal firings

do not convey discreet information, but rather ensembles of neurons develop
analog pressures within the brain/mind, creating certain types of holistic feel-
ings and actions. We believe that these global state envelopes, most evident in
the instinctual emotional urges of animals, are critically dependent on not only
a variety of generalized multi-functional chemistries (e.g., amino acids and
biogenic amines) but also emotion and motivation specific neurochemistries
that are emerging as promising targets for new psychiatric drug development
(Panksepp & Harro 2004).

Intentions-in-action vs. intentions-to-act. During mind-brain evolution, the

state-control systems of the brain help to establish embodied instinctual be-
havioral patterns along with internally experienced affective states (Panksepp
1998a, b). These instinctual arousals constitute ancient psycho-behavioral con-
trols allowing fundamental forms of cognitive activity and intentionality to
emerge as an intrinsic part of the action apparatus. This, we believe, is what
John Searle (1983) was referring to in his classic distinction between inten-
tion in action and intentions to act. Only with a more differentiated sensory-
perceptual cognitive apparatus, such as that which emerged with higher cor-
tical encephalization, can certain organisms operate in a virtual reality of
cognitive-type activities, and thereby select and generate more deliberative
behavioral choices (i.e., intentions to act) based upon the nuances of their
perceptual fields.

Action-to-perception processes vs. perception-to-action processes. This distinction

overlaps with the previous one. It assumes that the affective/emotional state-
control systems help direct and focus specific attentional-perceptual fields.
Only with the emergence of cognitive neural mechanisms capable of resolv-
ing highly detailed perceptual fields did the possibility emerge of buffering
decision-making by higher executive processes (largely through the working-
memory capacities of frontal lobes). Thus, sensorial awareness was trans-
formed into perceptual guidance devices, permitting higher-order cognitive
actions by organisms (yielding eventually the widely accepted perception-to-
action processes, which are closely related to intentions-to-act processes). If
ancient action-to-perception processes are fundamental for affective experi-
ence and primary-process intentionality (i.e., intentions-in-action), we may be
better able to generate some general principles whereby emotion-cognition in-
teractions transpire in the brain (i.e., certain perceptions and cognitions may
be strongly linked to the field dynamics of emotional systems).
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.6 (329-358)

 Luc Ciompi and Jaak Panksepp

Neurochemical codes vs. general glutamatergic computations. Neuroscientists

have long recognized that a distinction needs to be made between the rapidly
acting neurotransmitters that directly generate action potentials (with gluta-
mate being the prime example of an excitatory transmitter), and those neu-
romodulatory influences which bias how effectively the rapidly acting trans-
mitters operate (with the abundant neuropeptides being prime examples of
neuromodulators that may regulate emotionally and motivationally specific
state-variables in widely-ramifying neural networks). In other words, there
are distinct neurochemical codes for the core states of the nervous system,
and a better understanding of these codes will allow us to better conceptual-
ize how state functions become dynamic governors of channel functions, with
the former weighing more heavily in generating distinct forms of affective con-
sciousness and the latter being more important in resolving cognitive details.
For instance, there will be neuropeptides that are able to conduct a neurotrans-
mitter orchestra into new global field-dynamics (Panksepp, Panksepp & Harro
2004). In this view, one can conceptualize how affective and cognitive processes
are remarkably interpenetrant, while at the same time recognizing that the
conceptual distinction can help guide both neuroscientific and psychosocial
inquiries.

Neurobiological aspects of affective-cognitive interactions

We will first consider three general ways in which affective-cognitive interac-

tions may emerge: i) How “low energy” cognitive attributions provoke “high
energy” emotional responses; ii) How aroused emotional responses may lead
to global shifts in attention, perceptual and cognitive encodings; and ii) How
aroused emotional systems activate not only emotion-dependent actions but
also a rapid retrieval of state-dependent cognitive information, which can be-
come chronically changed because of the use-dependent plasticity of all basic
emotional systems.
Attributional triggers. The fact that very small cognitive changes are able to pre-
cipitate enormous emotional outbursts is one key reason we need cognition-
emotion distinctions. An interpretation of a glance, a shift in the tone of the
voice, or an ambiguous comment that can be interpreted as an insult, can all
provoke high energetic states that we call emotions. The fact that fairly mod-
est changes in the informational content of an incoming cognitive signal can
have such large effects provides striking support for the conclusion that there
are evolutionarily prepared emotional states of the nervous system. How these
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.7 (358-407)

Energetic effects of emotions on cognitions 

“trigger spots” emerge as a function of emotional learning has been studied ex-
tensively through the use of classical conditioning paradigms in animals, such
as the influential fear conditioning studies of LeDoux (1996), although such
basic learning studies provide little insight into the evolutionary nature of the
core emotional systems.
These types of trigger interactions are embedded in another global
cognition-emotion interaction: Namely, that high cortical activity generally
tends to inhibit subcortical emotional systems, and when individuals become
intensely emotionally aroused (i.e., the felt affective intensity of an emotion
is high), those states are generally accompanied by decreased cortical process-
ing (for an overview of such brain imaging work, see Liotti & Panksepp 2004).
Clearly cortico-cognitive processes can inhibit emotional arousal, as has been
demonstrated in fMRI studies where the subcortical arousals that accompany
erotic feelings can be actively inhibited by recruiting higher frontal cortical ex-
ecutive functions (Beauregard et al. 2001). It is only during mild emotional
states that the cognitive and affective processes may be operating more syn-
ergistically, which may suggest that there are quite different forms of affect
logic for low and high emotional arousals, following perhaps a classic Yerkes-
Dodson inverted-U related functions between arousal and optimal behavioural
control.
Emotionally induced shifts in attention, perceptual focusing and cognitive encod-
ings. Although it is widely accepted that emotional arousal has such cognitive
consequences, including distinct effects on attention, decision-making, judg-
ments and memory (Burke & Mathews 1992; Lowenstein et al. 2001), there
is, in fact, only modest neuroscience data highlighting how these interactions
operate, for instance in cognition-emotion interactions such as temporal lobe
inputs to amygdala (McGaugh & Cahill 2003). The neurobiological mecha-
nisms by which such effects are achieved in humans remain largely unknown.
However, in animal research there is quite a large literature on pharmacolog-
ically and physiologically induced state-dependent learning (Cameron 2002;
Colpaert & Balster 1988), and many of the molecules that have been used are
mood modifiers in humans (Panksepp 2004a).
Use-dependent plasticities of emotional systems and retrieval of mood-congruent
information. In addition to emotion-specific learning capacities, the animal
work has indicated that every emotional system can exhibit semi-permanent
changes in the vigor of the emotional system itself. This has been most clearly
demonstrated for fearfulness and irritability, where the stimulation of sub-
cortical fear and anger systems can chronically shift an animal’s long-term
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.8 (407-456)

 Luc Ciompi and Jaak Panksepp

temperamental bias (Adamec & Young 2000). These effects should have long-
term consequences on cognitive tendencies, as has been most clearly exhibited
in the flash-backs and ruminative tendencies of individuals suffering of post-
traumatic stress disorders (PTSD) (van der Kolk 2004). One of the ongoing
challenges is to determine how such long-term neural changes can be reversed.
There is also only modest neuroscience data on how mood-congruent in-
formation is actually retrieved in the brain. Hence we will simply highlight
some of the ways in which such issues can be studied in humans. In gen-
eral, it is harder to do good cognitive work in animals, just as it is easier to
do good basic emotion research on the animal models. An example of the
type of work that could be achieved is highlighted by a recent paper by Zack
& Poulos (2004), indicating that arousal of brain dopamine systems (which
has been conceptualized as a SEEKING-Wanting-Expectancy system) with am-
phetamine in problem gamblers leads them to have urges to gamble and their
semantic networks are primed to generate thoughts related to gambling. As ad-
ditional emotion-specific chemistries are discovered, it will be most important
to carefully study how the dynamics of mental contents shift as a function of
affect biasing molecules (Panksepp 1999; Panksepp & Harro 2004).

Basic emotional systems

Several basic emotional systems in sub-neocortical regions of the brain have

recently been identified. Cognitive interactions with them need to be worked
out on a system by system basis, since each one has its own characteristics, as in
sadness promoting pessimistic cognition and happiness promoting more op-
timistic cognitions (e.g., MacLeod & Salaminiou 2001). Here we will briefly
summarize the types of cognitive changes that we should expect from three
major negative affect systems (FEAR, RAGE and PANIC, or sadness) and three
positive affect systems (CARE, PLAY and SEEKING). Although each of these
emotional systems remains sparsely studied in cognitive and psychosocial con-
texts, they all presumably follow the same patterns as the basic motivations
such as thirst, hunger, and lust. When one is very thirsty or hungry, it is al-
most impossible to keep the cognitive apparatus from dwelling on the intensity
of the feelings and how one might be able to alleviate the sustained distress.
The same goes for various feelings of pain as well the urgent feelings of sex-
ual arousal. It is an especially well demonstrated fact how frequently sexual
thoughts emerge in the mind of young males, and presumably this is because
there neurochemistries of sexuality can sustain certain types of pressures and
biases on the cognitive apparatus (Panksepp 1998a).
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.9 (456-513)

Energetic effects of emotions on cognitions 

Again, we would emphasize that during intense emotional arousal, much

of the cortico-cognitive apparatus becomes less active (Liotti & Panksepp
2003), and probably more obsessive in terms of ideational flexibility, at least
for the negative emotions. On the other hand, the positive emotions tend to
broaden decision making processes, allowing one to recruit new ideas into
an ever widening network of associations (Fredrickson 2002) and increasingly
productive social interactions (Isen 2004).
Let us briefly consider each of the emotional systems mentioned above.
Since it is difficult to access the cognitions and thoughts of other species in
which the details of the core emotional circuits have been extensively stud-
ied, we must focus here on classically conditioned learning issues and use-
dependent plasticities. In our estimation the semi-permanent changes in the
reactivity of these systems that can emerge through experience (i.e., “use-
dependent plasticity”) may link up especially well to chronic psychiatric con-
ditions, and we will provide one prominent exemplar of that type of change for
each of the following six basic emotional systems.
FEAR: Classical conditioning of fear (i.e., the development of unconscious trig-
ger points to the FEAR circuitry) has been extensively studied (Davis 1999;
LeDoux 1996), and attempts have been made to envision this type of re-
search within the context of basic emotion theory (for a recent summary, see
Panksepp 2004b). These trigger spots may help explain the development of spe-
cific phobias, but such classical conditioning models are not as important for
understanding the chronic changes that can occur in FEAR systems to provoke
chronic anxiety. For this, the kindling models are very useful, whereby one elec-
trically stimulates the FEAR system briefly each day for about a week until an
epileptogenic sensitisation has emerged.. Once the system has been thus sensi-
tized, animals (e.g., cats) exhibit chronic changes in fearful reactivity, cowering
and being fearful of stimuli (e.g., rats) to which they would normally respond
with curiosity and predatory intent (Adamec & Young 2000). Although, no
one has found ways to reduce this type of sensitization once it has developed, a
few neurochemical manipulations (e.g. choleocystokinin receptor antagonism)
have been effective in aborting the emergence of the sensitization (Adamec,
Shallow & Budgell 1997). Presumably if such maneuvers were implemented in
humans during traumatic experiences, one might be able to abort the emer-
gence of chronic anxieties such as those that characterize post-traumatic stress
disorders (PTSD).
RAGE: The study of the details of anger in the brain, and the description of
how this system learns has diminished to a trickle during the past 30 years.
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.10 (513-569)

 Luc Ciompi and Jaak Panksepp

However, we know much about the anatomy and neurochemistry of the sys-
tem (Siegel, Roeling, Gregg, & Kruk 1999), and it is well established that one of
the major environmental triggers of anger is frustrative non-reward. The cog-
nitive consequences of anger are obsessive thoughts of retribution and revenge
for perceived slights, offenses or other abuses of one’s freedom and dignity.
Presumably there are cortical zones that are especially likely to dwell on such
issues (Murphy, Nimmo-Smith, & Lawrence 2003). However, more important
for psychiatrically significant irritability are the changes in aggressive tempera-
ment that can be achieved by stimulating the RAGE system in a manner similar
to that just described for the FEAR system (Adamec & Young 2000). Animals
with kindling of such systems exhibit a very “short fuse” and are likely to attack
to minimal provocations/attributions.
PANIC (separation distress): During grief, the mind obsessionally returns to
mental images of the lost loved one, to dwell on how the loss of the attachment
bonds might have been averted. The natural mental landscape needs to be bet-
ter described for this and all the other basic emotional responses. There has
been no work on the use-dependent plasticity of this system, but it is likely to
be related to panic attacks and variants of PTSD and as pervasive and profound
for mental life as the changes in the better studied FEAR system of the brain,
which may operate synergistically. Related to this issue, it is reasonably well
established that early social loss has long-term consequences which facilitate
a depressive outlook on life, and social loss long been considered to be a ma-
jor factor that contributes to susceptibility to clinically significant depression
(Heim & Nemeroff 1999)
CARE: When one is sad, the effects of caring social contacts are especially pow-
erful mood facilitators that can lead to rebound from depressive ruminations.
Nurturant maternal urges – tender loving care, in the vernacular – can be sen-
sitized by exposure of animals to infants (Panksepp 1998). Rat mothers exhibt
a species-typical type of care – ano-genital and other bodily licking – which
has recently been shown to have life-long positive effects on the offspring.
The positive effects include widespread benefits for the nervous system that
can be encapsulated in a phrase – the offspring are more inquisitive, coura-
geous and less likely to be severely influenced by stress (Meaney 2001). These
benefits become permanent habits, the benefits of which are then passed on
trans-generationally in non-genetic ways.
PLAY: The cognitive relationships to rough-and-tumble PLAY systems have
been even less well mapped than the other emotions, but it is likely that this
type of emotional engagement helps solidify social habits that can promote
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.11 (569-605)

Energetic effects of emotions on cognitions 

a better understanding of what one can and cannot do in social situations.

While captivated by playful energies, one’s mind tends to go quite naturally
to new associations that have the potential for sustaining and amplifying the
fun. Perhaps these short-term changes can help widen behavioural options
that the organisms has in all future types of emotional encounters. Some longer
term social-emotional consequences of play have been documented (Panksepp,
Burgdorf, Turner & Gordon 2003), and the possibility that play can have long
term therapeutic effects for various childhood disorders, especially attention-
hyperactivity problems has been addressed (Panksepp, Burgdorf, Gordon &
Turner 2002). At present, the overall concept is that play may not only fa-
cilitate the programming of social circuits, but also provide tonic effects on
neuronal growth factors (Gordon et al. 2003) which may facilitate frontal lobe
maturation, and thereby executive psychological functions for promote a more
thoughtful engagement with the world and other lives (Panksepp 2001a).
SEEKING-Wanting-Expectancy: The pursuit of every resource – from food to
romatic love–requires the energetization of this general purpose systems for
our desires, or appetitive engagements with the various fruits of the environ-
ment. A great deal of work has been conducted on how this system participates
in learning (Berridge & Robinson 2003; Ikemoto & Panksepp 1999; Schulz
2002) leading to a variety of terminologies that is often detracting from com-
municative clarity. However, most ideas are converging on the assumption
that the system is essential for appetitive eagerness, regardless of the reward
that is being pursued (Panksepp & Moskal 2004). The remarkable fact is that
this system readily sensitises when animals are exposed to psychostimulants
like amphetamines and cocaine that activate this system, and the sensitised
animals then exhibit a general elevation of pursuit strength for a variety of
incentives (Nocjar & Panksepp 2002). It is to be expected that a psychostimu-
lant sensitised brain would tend to exhibit different types of thought patterns
than “normal,” but there is very little psycho-ethological work on that issue
(Panksepp 2003b, c; Panksepp, Nocjar, Burdgorf, Panksepp & Huber 2004).
The scientific analysis of how the “mental apparatus” is modified by
chronic changes in the arousability of brain emotional systems remains in its
infancy. To do this properly, new psychological approaches such as “psycho-
ethology” need to be implemented that permit the analysis of fluctuating
psychological contents of the mental apparatus under various conditions. In
general, we favor the view that a much greater amount of the overall inter-
active equation will be solved by focusing on how basic emotional arousals
modify the cognitive apparatus than in the way cognitive attributions provoke
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.12 (605-662)

 Luc Ciompi and Jaak Panksepp

emotions. In other words, once an emotional episode is triggered, then the re-
sulting ruminations are pulled into self-organizing orbits that are substantially
broader than the initial attributional instigators of the emotional episodes (see
Parkinson 1995 for a fuller development of this interactive cascade).
In addition, the three following general neuroscience concepts are of great
interest from psychosocial perspectives:
1) Emotion-related and cognition-related circuits are anatomically and function-
ally very closely linked and intertwined. Emotional, cognitive and behavioural
components implied in the above mentioned systems form inborn functional
entities that are further differentiated by learning and other mechanisms of
neural plasticity. Although the subcortical emotional operating systems are
concentrated in primitive brain areas that have not been typically thought of
in cognitive terms, they do interact with many higher brain areas, that may
be “centers of gravity” for emotion-cognition interactions (Panksepp 1988)
many now confirmed with modern brain imaging (Murphy et al. 2003; Phan
et al. 2002). The main areas are anterior cingulate, insula, temporal and var-
ious frontal cortical zones. Exactly what type of processing occurs in those
regions is by no means clear, but from a neurodynamic perspective, we en-
vision that basic emotional systems serve as attractors for various perceptual
and cortico-cognitive activities, yielding individual and personality specific
attractor landscapes for how different people cope with emotional situations.
2) All incoming cognitive stimuli are linked with a situation-dependent emo-
tional value or “color”. This phenomenon has important consequences for all
further information processing, as will be discussed in the psychosocial sec-
tion. Typically the one way linear-causal approach to this issue is based on
systematic studies of classical conditioning, where neutral conditional stimuli
become capable of generating specific emotional responses (e.g., conditioned
fear, such as freezing) when paired with unconditionally aversive events such a
foot shock (LeDoux 1996). There are, however, enormous limits to such simple
one-way causal analyzes to model what might actually be happening in the hu-
man mind. The dynamic systems perspective provides a compelling metaphor
of how non-linear causal processes may operate in higher brain areas such as
described above to yield attractor basins, in which ruminating cognitive ac-
tivities may be trapped, as in fear or rage thoughts that prevail in chronic
conflict-situations (Panksepp 2000; Ciompi 1997c).
3) Subcortical circuits for quick emergency reactions have recently been detected
that provide a neural substratum for largely sub-conscious emotional regula-
tions of cognition (Davis 1999; LeDoux 1996). These important findings, so far
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.13 (662-709)

Energetic effects of emotions on cognitions 

mainly restricted to fear-related circuits at the neurobiological level, provide

compelling neurobiological models for unconscious cognitive activities, simi-
lar to those for long postulated by psychoanalysts. Although these phenomena
may appear to be unconscious from the perspective of cognitive conscious-
ness, they may, however, be the very core of the affective energies that should
be recognized as a distinct form of affective consciousness (Panksepp 2003a).

. Psychosocial perspectives

Much of the reported neuroscientific evidence corresponds to well known

phenomena on the mental and social level. However, because of linguistically-
mediated psychological access, numerous additional emotion-cognition in-
teractions can be observed on this level that might, potentially, lead to new
questions and hypotheses. In the following, we firstly point to a number of
striking convergences between psychosocial and neurobiological findings, and
move on then to psychosocial phenomena for which neurobiological evidence
remains scanty.

Converging neurobiological and psychosocial findings

First, affective and cognitive emotional phenomena are also clearly distin-
guishable on the psychosocial level. To reiterate, cognitive phenomena might
essentially be understood as sensory distinctions or, on a more differentiated
level, as distinctions of distinctions of distinctions, and their further compu-
tation (Ciompi 1997c; Spencer-Brown 1979). Emotions are, in contrast, much
more comprehensive phenomena with multiple mental and somatic aspects.
For instance, at the conclusion of their comprehensive overview on definitional
issues, Kleinginna & Kleinginna (1981) describe emotions as “. . . a complex
set of interactions among subjective and objective factors, mediated by neu-
ral/hormonal systems, which can (a) give rise to affective experiences such as
feelings of arousal, pleasure/displeasure; (b) generate cognitive processes such
as emotionally relevant perceptual effects, appraisals, labelling processes; (c)
activate widespread physiological adjustments to the arousing conditions; and
(d) lead to behaviour that is often, but not always, expressive, goal-directed,
and adaptive.” Both on the neurobiological and on the psychosocial level,
cognitions thus appear as much more specific, discrete, “digital” and “com-
putational” phenomena than emotions which have a definitely more global,
holistic, and generally also a longer lasting character. In addition, emotions
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.14 (709-759)

 Luc Ciompi and Jaak Panksepp

include important energetic and dynamic aspects that are lacking in purely
cognitive phenomena (see below).
The detection of a few distinct emotional brain systems, too, closely cor-
responds to a long established psychological view: Namely the notion of a
limited number of evolutionary rooted basic emotions whose exact nature and
number remains, however, controversial. Thus, Ekman (1984) postulated 6 ba-
sic emotions (anger, disgust, fear, joy, sadness, surprise) on the basis of his
extended transcultural studies, while other authors also include feelings like
shame, guilt and several other emotions (for details, see Ciompi 1997c; Izard
1971, Ortony & Turner 1990). Hopefully, the ongoing neurobiological research
will be able to clarify which emotions should truly be considered as “basic” and
evolutionary rooted, and which ones rather represent epigenetic-learning and
culture-dependent cognitive modulations.
Another essential neurobiological finding – namely the fact that emotion-
related and cognition-related cerebral circuits are closely intertwined and contin-
ually interacting in almost all mental activities – corresponds to a great number
of psychological observations that all speak for practically ubiquitous affective-
cognitive interactions. Already Freud emphasized the omnipresence of emo-
tional influences on thinking, and vice- versa. Piaget’s monumental research on
the genesis of cognitive structures in humans lead him to the same conclusion,
in spite of the predominantly cognitive focus of his “genetic epistemology”
(Piaget 1981). The same is true for the extended psychosocial studies on condi-
tioned reflexes, learning and operant conditioning (e.g. through the systematic
application of rewards or punishments), and hence for almost all learning the-
ories from Pavlov to Skinner and Hull, even though their emotional aspects
were routinely neglected.
Striking convergences exist, furthermore, between the reported obligatory
neuronal linking of incoming cognitive stimuli with simultaneously experienced
emotions and a wealth of analogue psychosocial phenomena, among them
again, especially, conditioned reflexes and learning experiences, and also the
afore mentioned posttraumatic stress disorders. It is very likely that the above
reported possibilities of neuronal sensitisation of emotional systems (e.g. by
the so-called kindling mechanisms) play an important role in the pathogenesis
of PTSD and of various other psychiatric conditions (e.g. phobic or obsessional
disorders) that are characterized by overanxious or overaggressive reactions at
minimal stimulations. Sensitisation by traumatic events may also be implied in
the (individual or collective) emergence of a so-called “fear-logic”, “hate-logic”,
or “logics of war” whereby predominating affects like fear or rage literally
“enslave” all thinking and behaving (Ciompi 1997c). In all these phenom-
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.15 (759-814)

Energetic effects of emotions on cognitions 

ena, the afore-mentioned attractor-like properties of basic emotions become

particularly obvious.

General and specific operator-effects of emotions on cognition

We turn now to emotion-cognition interactions that have mainly been ob-

served on the psychosocial-phenomenological level, with few notions concern-
ing the underlying neurobiological mechanisms. Among them, a number of
so-called operator-effects of emotions on attention, memory and combinatory
thought (Ciompi 1997c) are of particular interest. An operator is a variable that
acts on another variable and modifies it. A striking example of the immediate
modification of form and content of cognition under the influence of emotions
like surprise, interest or fear is already provided by the evolutionary rooted at-
tention reaction that can be observed both in newborn and adult humans, as
well as in most animals. Attention-shifting effects of emotions, for which some
neurobiological evidence is just beginning to appear (Anderson et al. 2004;
LeDoux 2000; Matthews & Wells 1999), are since ancient times intuitively used
in all kinds of everyday activities like selling, advertising, political or religious
rhetoric. They are well known also in clinical psychology, psychopathology and
psychoanalysis. Recent systematic psychosocial research confirms a variety of
focusing or defocusing, stimulating or inhibiting, hierarchy-creating and or-
ganising effects of emotions on attention and perception (Mathews & Wells
1999), on memory (Eich & Macauley 2000; Ellis & Moore 1999), and on com-
prehensive thought (Berkowitz et al. 2000; Niedenthal & Halberstadt 2000).
There are two types of operator-effects of emotions on cognitions: General
ones that are basically similar in all kind of emotions, and specific ones that
differ from one emotion to another (Ciompi 1997c).
General operator-effects include the fact that emotions
– stimulate or, on the contrary, inhibit cognitive activities, that is, that they
act on them as energy-regulating “motors” or “brakes”
– focus the attention on emotion-congruent cognitive objects, thus tending
to establish an emotion-dependent hierarchy of perceiving and thinking
– preferentially store and mobilize emotion-congruent cognitions in mem-
ory, and
– tend to link emotion-congruent cognitive elements and to combine them
in larger cognitive entities. A good example is the emergence of global
affective-cognitive judgements of the type “a nice person”, “an ugly town”,
“a wonderfull country” etc. that are the result of an agglomeration of
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.16 (814-898)

 Luc Ciompi and Jaak Panksepp

a great number of single affective-cognitive elements of corresponding

affective value.

Specific operator-effects include the fact that

– interest and curiosity arouse emotional energies and direct them toward
specific cognitive objects (Ciompi 1997c; Panksepp 1998a)
– fear and anxiety increase the distance or induce flight from anxiogenic
objects (Le Doux 1996).
– aggression or rage defend or extend ones own borders. Additional evo-
lutionary rooted functions are defending the offspring, catching preys,
concurring against rivals, and reacting at frustrations of all kind (Lorenz
1966; Panksepp 1998a).
– pleasure, joy and other “positive feelings” decrease the distance and create
emotional bonds with cognitive objects such as specific persons, places,
situations, theories and ideologies (Bowlby 1973; Ciompi 1997c).
– sadness, mourning and grief have the vital function of loosening dysfunc-
tional bonds by detaching emotional energies from lost objects and thus
permitting to re-establish contact with significant others after a severe loss.

Far from being just “irrational”, both general and specific operator effects of
emotions on cognition have highly survival-relevant functions. Their com-
mon function is the experience-based simplification of the infinite complexity
of the cognitive world, by reducing it to a few behaviour-relevant categories
like interesting/indifferent, harmless/dangerous, usefull/useless, etc. through
corresponding adaptations of attention, memory and combinatory thought.
Such effects are by no means only present in clearly manifest emotional
states, but also in numerous apparently unemotional every-day phenomena
such as culture-dependent prejudices and mentalities, fashions, value systems,
political or religious ideologies, etc., where regulating and integrating effects
of initially intense feelings gradually become automatic and almost uncon-
scious by habituation. Emotion-regulated patterns are progressively consol-
idated by the fact that once established, pleasant and easy going ways of
thinking and behaving are continually repeated, whilst unpleasant (conflicting,
tension-creating) ways are as much as possible avoided. Similar mechanisms
are covertly at work, even in seemingly neutral scientific and other rational
thinking, given that abstract contradictions and conflicts, too, are emotion-
ally unpleasant and uneconomic, whereas good solutions are pleasant and
economic, and therefore attractive (Ciompi 1988, 1997b).
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.17 (898-918)

Energetic effects of emotions on cognitions 

The energetic dimension of emotions

Another particularly important aspect of emotions in our view, namely their

energetic dimension remains astonishingly neglected in both neurobiological
and psychosocial research, in spite of having for a long time been integrated
in several comprehensive models of mental functioning: Already the psycho-
analytic approach located the energy for all thinking and behaving in innate
drives and corresponding basic emotions, and Piaget’s systematic studies on
the genesis of mind and thought in children also identified emotional energies
as the essential “motors” or energizers of all cognitive activity (Piaget 1981). In
the theory of affect-logic, too, basic affects are understood as directed energetic
states selected by evolution, each basic affect being characterized by a specific
pattern of energy consumption (Ciompi 1988, 1997b). It must be emphasised
that the notion of energy is not used just as a metaphor in this context, but
in the sense of measurable patterns of goal-directed conventional biological
energy consumption, as already well-known in stressful sympathetic arousal
states that are related to fear, rage and tension (especially in flight or fight), in
contrast to the relaxed parasympathetic states that are related to “positive feel-
ings” like pleasure, joy or lust which characterise activities like food-intake, rest,
socialisation, sexuality, etc. (Selye 1946). To our knowledge, only one study has,
so far, attempted a precise quantification of a diversity of emotion-energetic
patterns as reflected in expressive movements (Clynes 1987). However, differ-
ent emotion-specific patterns of energy distribution both in brain and body
could nowadays be localized and quantified (e.g. by current imaging and/or
bioelectric techniques).
That emotions have an energetic dimension has also been repeatedly re-
vealed in the preceding section on biological aspects, so for instance, when
emotional states exert an “organic pressure” on cognition, when they ener-
gize the context-relevant neuronal networks, or act as dynamic governors for
different modes of thinking and behaving. This dimension appears, however,
with particular clarity on various psychosocial levels. Activating or inhibit-
ing energetic effects of emotions on speed, form and content of cognition
are already frequently observed in normal everyday mood fluctuations. Un-
der exceptional circumstances such as sudden joys or griefs, traumatic events
etc., they become, however, much more obvious, especially when individual
emotions spread out on various collective levels, thanks to the mechanisms
of “emotional contagion”, as described by Hatfield et al. (1994). The poten-
tially devastating effects of mass-phenomena like mass-panic, mass-aggression
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.18 (918-982)

 Luc Ciompi and Jaak Panksepp

or mass-enthusiasm, too, are primarily related to the huge polarized emotional

energies that collective emotions may liberate.
Spectacular effects of the activating or, on the contrary inhibiting power
of emotions are also observed in psychopathological conditions like mania or
melancholia: In elated manic states, the acceleration of all cognitive functions
may reach the degree of a so-called logorrhea (relentless high-speed thinking
and speaking activity, with continually changing associations and progressive
destructuring of syntax, approaching incoherence). Simultaneously, all psy-
chomotor functions are accelerated, while the need for sleep and food intake
is dramatically diminished. In pathological depression or melancholia, on the
contrary, ideation, speech and all other cognition-guided activities are severely
inhibited and impoverished, with a general slowing down that may reach the
degree of complete mutism and immobility (cf. Ciompi 1997a, c; DSM III-R
1987; Flack & Laird 1998). Simultaneously, attention, memory and thought are
narrowly focused on sad contents, thanks to the described operator-effects of
melancholic feelings. Facial expression, voice, bodily posture and practically
all other psychomotor and vegetative functions are equally inhibited. In states
of fear, panic or rage, too, speed and energetic pressure of cognitive functions
are generally increased, with, however, the possibility of a sudden shift to the
evolutionary-based phenomenon of freezing at a critical point of emotional
tension. In the normally relaxed or mildly emotional state, cognitive activity is,
in contrast, characterized by a modally flexible, quick and economic mode of
thinking and feeling, without major fluctuations to affective extremes.
Fully acknowledging the energetic dimension of emotions has important
theoretical and practical implications: It clearly identifies the relevant dynamic
forces at work both behind individual and collective affective-cognitive dy-
namics, it explains their direction and the organising effects of emotion on
cognition, and it also opens new ways of understanding the dynamics of sudden
non-linear shifts of the prevailing affective-cognitive patterns of functioning
that are addressed in the next paragraph. In psychopathology, psychotherapy
and pharmacotherapy, too, a full conceptual integration of the energetic as-
pects of emotions leads to numerous practical and theoretical consequences
(Ciompi 1997c).

Non-linear aspects of emotion-cognition interactions

Mental and social systems are certainly among the most complex systems we
know. It is therefore not at all surprising that they show non-linear dynam-
ics under certain conditions. Preliminary neurobiological studies have already
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.19 (982-1019)

Energetic effects of emotions on cognitions 

revealed that so-called ”chaotic” brain-dynamics can appear in the dopamine-

metabolism (King & Barchas 1983), and in the EEG of epileptics (Babloyantz
& Destenxhe 1986; Kovalik & Elbers 1994) or of psychotics (Koukkou et al.
1993). Here again, more extended information is, however, only available in
the domain of psychosocial phenomena. Of particular interest is the obser-
vation that increasing emotional tensions are capable of provoking, at a critical
breaking point, a sudden global reorganization of the prevailing feeling-thinking-
behaving patterns (Ciompi 1997b, c, 1999; Freeman 2004; Kelso 1995; Mech-
sner et al. 2001; Paulus 2003; Gottschalk et al. 1995; Tschacher & Dauwalder
1999). Such nonlinear bifurcations occur, for instance, when a conflict sud-
denly escalates from verbal to physical violence, or when a generally prevailing
“logic of peace” turns into a “logic of war”. Similar mechanisms are at work
in religious conversions or political revolutions (Ciompi 1988; Sargant 1957).
Critically increasing emotional tensions often also play a crucial role in the out-
break of schizophrenic psychosis, as confirmed by numerous empirical studies
on the impact of so-called “expressed emotions” (see Ciompi 1997c; Kavanagh
1992; Leff & Vaugh 1985).
In addition, so-called butterfly-effects (small causes having enormous ef-
fects) can occur in unstable mental and social systems, such as, e.g., sudden
outbursts of disproportioned rage and aggression after minimal provocation.
If we assume that emotions correspond to goal-directed biologic energies, and
that emotional energies play an analogue role in psychosocial systems as con-
ventional energy plays in physical or chemical systems, all these observations
are consistent with a dynamical systems approach, and in particular with the
observation that in all kind of open dynamic systems, the input of energy op-
erates as a critical control parameter capable of provoking, at a critical level, a
sudden nonlinear bifurcation from one overall mode of functioning to another
(Haken 1990; Prigogine & Stengers 1983; Schuster 1989).
Another interesting aspect of non-linear psychosocial dynamics is the al-
ready noticed emergence of attractor-like effects of emotions on both individual
and collective thought and behaviour processes. In chronic family conflicts, for
instance, or in obsessive, depressive or manic states, all thinking and feeling
may endlessly circle along the same inescapable trajectories, under the influ-
ence of the described operator-effects of emotions on attention, memory and
thought. Typically attractor-like affective-cognitive dynamics (e.g. in the sense
of the mentioned “logic of fear” or “logic of hate”), that may be propagated by
emotional contagion, can also be observed on various macro- or microsocial
levels – the former for instance in the endless Israel-Palestine conflict, where for
decades huge parts of both individual and collective feeling and thinking has
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.20 (1019-1066)

 Luc Ciompi and Jaak Panksepp

been trapped in escalating circles of mutual hate and violence. All these obser-
vations suggest that affective-cognitive processes may have a so-called fractal
structure (Mandelbrot 1983), that is that they show self-similar dynamics on
any mental or social scale, in terms of dynamic systems theory. The main mech-
anism that leads to these scale-independent self-similarities is very probably the
omnipresence of the described operator-effects of emotions on cognitions on
all psychosocial (and possibly also neurobiological) levels (Ciompi 1997b, d;
Ciompi & Baatz, in press).

. Discussion and resulting working hypotheses

A comprehensive picture of affective-cognitive interactions in humans and

higher mammalians is gradually emerging from the reported puzzle of neu-
robiological and psychosocial findings. It can be summarized as follows: It
is very likely that emotions and cognitions are circularly interacting in most
mental activities. On the one hand, specific cognitions activate specific emo-
tions on the basis of a limited number of affective-cognitive-behavioral sys-
tems selected by evolution that are further differentiated by experience. On
the other hand, specific emotions mobilize and regulate specific cognitions by
their inborn operator-effects on perception, attention, memory and thought.
All incoming cognitions get an experience-based emotional value or “colour”
that greatly influences all subsequent information processing. Cognitions with
similar emotional values tend to become agglomerated. Thanks to its operator
effects and attractor-like properties, each basic emotion is related to its particu-
lar mode of regulating perception, its distinct memory banks, and its own way
of combining both in the complex emotion-cognition activating and/or in-
hibiting processes that we call “thinking”. Affects thus function as the essential
“switchers” from one type of thinking to another. Rapidly increasing emotional
tensions are, in addition, capable of provoking sudden non-linear modifica-
tions of the prevailing thinking and behaving patterns. All these interactions
appear to be self-similar (or “fractal”) on any individual and social level.
All this speaks for our central assumption that basic emotions correspond
to evolution-selected patterns of energy flow and consumption both on the
brain/mind and on the brain/body level. Directed emotional energies orient all
thinking and behavior by preferentially focusing conscious attention on new,
important, potentially dangerous, difficult or otherwise particularly “exciting”
cognitions. By habituation and differentiation, such initially high energetic
modes of processing are, however, gradually transformed into the more eco-
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.21 (1066-1115)

Energetic effects of emotions on cognitions 

nomic modes of semi-automatic everyday functioning in which low emotional

energies operate on largely subconscious levels, without, however, loosing their
regulating operator-effects on both thought and behavior. In sum, affects ap-
pear as elementary forces with Janus-like properties: Comparable to natural
physical forces like fire, water or wind, on the one hand their energies are
capable, when adequately channelled and modulated, of promoting subtle in-
dividual or socio-cultural performances. But when out of control, the same
energies can, on the other hand, have disruptive effects on all possible levels of
cognitive information processing and behaviour.
This integrative view of emotions expands on the traditional understand-
ing of emotions as rather short, conscious and typically reactive events ac-
companied by a cluster of expressive and neuro-vegetative processes. This is,
of course, essentially a question of definition. Without going into a lengthy
discussion of definitional issues that are tangled, as mentioned, by widely vari-
able meanings given to terms like feeling, affect/affectivity, emotion and mood,
we wish, however, to emphasize our conviction that a superordinate concept
reflecting the underlying deep functional and phenomenological unity of all
these emotion-like phenomena (subjective and objective, mental and somatic,
energetic and phenomenological, conscious and unconscious, short-term and
long-term) is desperately needed. On such a base, which also recognizes the
energetic dimension and the self-similar effects of emotions on various hierar-
chical levels, substantial progress in the study of affective-cognitive interactions
might be possible.
It is precisely on such an interdisciplinary basis that we propose to include
all phenomena commonly called feelings, emotions, affects or moods under
the general definition of specifically directed psycho-somatic energetic states of
variable quality, intensity, duration and degree of consciousness. Cognitions, in
contrast, may be defined as the capacity of perceiving and further processing
sensory differences (Ciompi 1988, 1997c, 2004). Both definitions are deeply
evolution-rooted: In fact, already unicellular organisms function by sensory
distinctions (e.g. between basic or acid, warm or cold, bright or dark milieus)
that interact with global organismic states (e.g. relaxed or tense, or parasym-
pathetic and sympathetic states) in which we might see the first glimmers
of systemic changes that are eventually experienced and expressed as con-
scious “emotions” or “affects”. Simultaneously, the proposed definitions have
the advantage of being mutually exclusive. The clear separation of two differ-
ent but constantly interacting classes of phenomena should greatly facilitate
the study of their interactions. They also integrate all known aspects of both
emotions and cognitions without major contradictions to existing semantic
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.22 (1115-1151)

 Luc Ciompi and Jaak Panksepp

traditions, and should thus promote interdisciplinary communication and col-

laboration. Furthermore, the proposed definitions put emotion-like phenom-
ena appropriately close to psychosomatic states like hunger, thirst, lust or pain,
whereas cognitive phenomena remain close to digital information-processing
and computer-like cybernetics. Last but not least, the proposed inclusion of the
energetic aspect of emotions brings a powerful dynamic element, commonly
neglected, to the forefront in the currently prevailing cognition-based under-
standings of both mental and social processes. This is not to deny that in most
“civilized” social interactions, a common aim is to keep the emotional ener-
gies at a sufficiently low lever that they not impair the flow of well-monitored
cognitive information flow.
A major problem is that there probably exists no generally acceptable term
for such an enlarged notion of emotion-like phenomena, given that the two
best candidates – namely “emotion” or “affect” – are already used, but in
opposite senses, as supraordinate notions by different authors who have re-
cently been reflecting on definitional problems: Ciompi (1988, 1997c), Clore
& Ortony (2000), Damasio (2000), Ellis & Moore (1999) and Forgas (2000a),
for instance, propose the term affect (or affectivity) as an “umbrella notion”,
whereas Kleinginna & Kleinginna (1981), Panksepp (1998a, 2000a, b, 2001a,
b) and others understand emotion as the supraordinate term, tending to re-
strict the notion of affect to the conscious subjective experience of an emo-
tion (see Solms & Nersessian 1999 for discussion) and to distinguish, on this
basis, between the supra-ordinate category of emotional processes, of which
“affective feelings” are a subcomponent. Standard sources like the “Oxford
English Dictionary” or the “Webster” are of no help, because they truly mir-
ror the mentioned scientific ambiguities. A third possible solution, namely
“thymic phenomena” (from the greek thymos that means both “feeling” and
“the forces of life”) that is already used in this sense in “thymoleptics” (=
emotion-normalizing antidepressive drugs) is probably too remote from cur-
rent semantic practice for having a chance of being accepted.
In the hope that some terminological consensus will eventually emerge by
sheer necessity, we leave this question open and move on to finally proposing
eight tentative working hypotheses as result of our reflections and possible guide-
lines for further research. Some of them are already well supported, but others
are not. Again, we focus essentially on emotions for which the neurobiological
bases are already partly known. Methodological issues will not be addressed in
any detail.
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.23 (1151-1206)

Energetic effects of emotions on cognitions 

1. Different basic emotional states correspond to different pattern of energy con-

sumption in both body and brain.

As discussed above, this hypothesis and its numerous practical and theoretical
implications is the central – and probably most innovative – aspect of our un-
derstanding of emotions. It could primarily be tested by modern measures of
local and global bio-energetic flows, and secondarily by additional psychoso-
cial and neuro-vegetative measures. If adequately confirmed, it should put the
whole field of emotion-research on more precise quantitative grounds. Last
but not least, it would eventually permit investigators to use both basic and
differentiated energetic patterns as indicators for different emotions that, so
far, are mostly assessed by indirect conventional research methods. This could
be advanced by development of approaches such as those pioneered by Clynes
(1987), as well as metabolic and neurophysiological brain imaging procedures
such as PET and fMRI (Liotti & Panksepp 2004).
2. Mild to modest emotional arousal generally increases the overall amount of
cognitive processing, while intense arousal tends to inhibit it.

This hypothesis appears as a consequence of the preceding one. Both neuro-

biological and psychosocial evidence suggest, however, that it must be further
differentiated, because different emotions and different emotional intensities
apparently activate (or sometimes, as in panic, rage or depressive states, slow
down or even block) cognition in quite different ways (Liotti & Panksepp
2004). An additional difficulty is the fact that adequate methods for quantify-
ing cognitions are still widely lacking on the neurobiological level, and methods
for following the flow of thought in humans, such as the study of psychoana-
lytic transcripts under different emotional states, remains poorly developed,
but should be capable of being implemented with automated voice recording
and data analysis approaches (Panksepp 1999).
3. Incoming cognitive stimuli associated with emotional arousal are automati-
cally linked and stored with the accompanying emotional quality or value.

This hypothesis, too, is, as reported, already partly confirmed both in neuro-
biology and on the mental and social level, especially by studies on classical
conditioning. We further believe that once established affective-cognitive con-
nections have wide-ranging consequences for all further information process-
ing (see hypotheses 4–6). More research on all levels is consequently needed
in order to identify how affective-cognitive connections are neurobiologically
established, and how they are further processed.
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.24 (1206-1262)

 Luc Ciompi and Jaak Panksepp

4. In different basic emotional states, different sets of cognitions are preferentially

stored and/or reactivated in memory.

This hypothesis summarizes an essential aspect of the described operator-

effects of emotion on cognition. Although closely related to the empirically
well known phenomenon of state-dependent learning, it needs more psychoso-
cial and especially neurobiological empirical consolidation, too. As we begin to
understand some of the unique chemistries of specific emotional systems (e.g.,
neuropeptides, as discussed by Panksepp & Harro 2004), it will be probably
possible to determine the extent to which various emotion-related memories
and cognitive patterns can be reactivated by specific neurochemical brain con-
ditions. The number of neuropeptides that have emotion specific effects should
allow robust evaluation of such connections.
5. Cognitions with a similar emotional value are preferentially assembled into
higher-order affective-cognitive-behavioral units.

This assumption represents another important aspect of the reported operator-

effects of emotion on cognition. It is at the basis of the attractor-like emergence
of comprehensive affect-specific logics (e.g. “fear-logic”, “hate-logic” etc.) and
of specifically “coloured” personal or cultural value-systems, belief-systems,
prejudices and ideologies. This phenomenon is also sustained by the activity of
one or more of the above described emotion-dominated brain systems that in-
tegrate extended cognitive, affective, hormonal-neurovegetative and behavioral
patterns into higher-order functional units. Over time, such patterns might
be stabilized and become chronic by various neuronal plasticity and kindling
mechanisms.
6. Sudden non-linear shifts from one prevailing global affective-cognitive pattern
to another can occur, when the degree of emotional tension reaches a critical
level.

This hypothesis is derived, as explained above, from a dynamical systems inter-

pretation of a great variety of unexpected psychological and psychopathologi-
cal phenomena (e.g. the shift from a “logic of peace” to a “logic of war”, or from
normal to psychotic modes of thinking and behaving). Already Pavlov’s expe-
riences with dogs, which were confirmed by numerous empirical studies both
on the animal and the human level, have revealed sudden overall behavioural
changes provoked by increasingly difficult and stressful cognitive tasks. Given
the frequency of non-linear mental and social processes, this approach seems
quite promising, especially when combined with the proposed energetic under-
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.25 (1262-1312)

Energetic effects of emotions on cognitions 

standing of emotions. Here, too, much additional research is, however, needed.
One additional experimental possibility would be to study mental reactions in
response to artistic stimuli (music, film, dance, etc.) where there are sudden
shifts of prevailing emotional moods and dynamics (Panksepp & Bernatzky
2002). Likewise, in studying humour, one could try to capture such global shifts
before and after people “get” a joke. Increasing social tensions could, for in-
stance, be quantified by classifying, counting and comparing emotion-loaded
titles in newspapers and other media.
7. At least for the basic emotions, the described affective-cognitive interactions
are self-similar in both small-scale and large-scale psychosocial processes.

This assumption, too, stems from dynamic systems theory. It is suggested

by numerous psychosocial observations which speak for analogue operator-
effects of emotions on cognition – and consequently for analogue reactions,
for instance, to frightening or, on the contrary, to pleasant stimuli – on any
individual and collective level. This issue has not yet been explored by for-
mal research, to our knowledge. Several methods could be used for assessing
self-similar individual and collective emotion-cognition interactions, among
them the reported measurement of expressive movements (Clynes 1987) and
other emotion-specific features (e.g. vocalization or body-posture) in real life
emotional situations such as sports, artistic, or mob movement patterns.
8. A similar specific “affective imprint” is necessary not only for the generation
and maturation, but also for the eventual reactivation of functional affective-
cognitive circuits.

This “imprint-hypothesis”, first formulated by Ciompi (1988) on the base of

mainly psychosocial observations, is now also supported by the above men-
tioned neurobiological findings in animals that speak for beneficial effects of
body-care and play on the maturation of certain crucially important limbo-
prefrontal circuits. Numerous empirical data on early parent-infant relations
and pro-social behaviour in humans point in the same direction (Bowlby 1973;
Fivaz-Depeursinge 1999; Stern 1995). The assumption that the same affective
tuning that initially facilitates the neuronal maturation of specific affective-
cognitive circuits could eventually stimulate the reactivation (and possible fur-
ther differentiation) of these very circuits in similar situations is providing
a very parsimonious explanation for most of the previously reported “op-
erator” effects of emotions: Coherent electrodynamic/neurochemical mech-
anisms, such as certain brain activity rhythms, may be responsible not only
for the establishment of specific affective-cognitive-behavioural links, but also
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.26 (1312-1366)

 Luc Ciompi and Jaak Panksepp

for the preferential focusing of attention on affect-congruent cognitions, for

the preferential storage and remobilization of affect-congruent cognition from
memory, and for the tendency of assembling cognitions with similar affective
colour into larger cognitive entities (cf. hypotheses 3–5). This last integra-
tive hypothesis, albeit conceptually reasonable, presently has no clear neuro-
biological evidence.

. Conclusions

Comparing recent neurobiological and psychosocial findings on interactions

between emotion and cognition from the perspective of two largely com-
plementary models – namely “affective neuroscience” on the one hand, and
“affect-logic” on the other – seems to lead, as hoped, to some innovative ideas
and predictions capable of stimulating further research. However, some inher-
ent difficulties of such interdisciplinary efforts also became evident, even if they
are not fully discussed here. They range from “soft” and often subtle differences
in the use of concepts, theoretical orientations and value systems, to hard-core
questions of specific methodologies on how to best pursue research efforts. We
hope, however, to have at least partly succeeded in overcoming such difficulties,
thanks to a common commitment to empirical research as base for all scientific
progress, especially in the complex field of emotion and cognition.
Obviously, the neurobiological basis of a great number of conceptually
well recognized mental and social phenomena remain to be empirically elu-
cidated. Psychosocial phenomenology is, therefore, an ongoing challenge for
neuroscience – and vice-versa. One of the more important outcomes of this
interdisciplinary comparison is, in fact, the recognition that each approach
has different strengths to add to an eventual synthesis. For instance, despite
advances in imaging brain correlates (e.g., Liotti & Panksepp 2004; Mayberg
2004), it will be hard to specify how the detailed flow of cognitive activity (e.g.,
the progression of thoughts) could ever be quantified or causally manipulated
at the basic neuroscience level, and it will clearly continue to be easier and
more germane from a “mental apparatus” perspective to evaluate these issues
at the human psychological level, especially with causal manipulations such as
Transcranial Magnetic Stimulation (Schutter, Van Honk, & Panksepp 2004).
Phenomena like subjectivity and consciousness can hardly be explored on
other but the mental level in humans, even though credible approaches to
the study of affective feelings are emerging in the neuroscience community
(Knutson, Burgdorf & Panksepp 2002; Panksepp 1998a). This is important
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.27 (1366-1419)

Energetic effects of emotions on cognitions 

because basic emotional systems and their evolutionary roots and effects can
only be studied in great neuronal detail in animal models. Neuroscience also
provides powerful new tools, such as neurochemical ones to modify specific
brain systems, some of which can be employed in humans to see what conse-
quences there are for the intensity and flow of psychological output. Because
of the abundance of new causal manipulations, such as neuropharmaceuticals
that can modulate specific neuropeptides systems, there is a real possibility of
bridging such seemingly distant approaches (Panksepp & Harro 2004). Such
hard-core methods for testing new hypotheses should be interfaced with the
uniquely rich field of psychosocial observation. Thus, in our estimation, there
is much to be gained scientifically by allowing such very different paradigms to
work synergistically toward a mutually desired and desirable synthesis.

Notes

* Former medical director of the Socio-Psychiatric University Clinic, Faculty of Medicine,

Berne/Switzerland
** Department of Psychology, Bowling Green State University, Bowling Green, OH 43403,
USA
. We use the term “psychosocial” as a supraordinate notion including several overlapping
fields like psychology, social psychology, psychopathology and sociology

References

Adamec, R., & Young, B. (2000). Neuroplasticity in specific limbic system circuits
may mediate specific kindling induced changes in animal affect – implications
for understanding anxiety associated with epilepsy. Neuroscience and Biobehavioral
Reviews, 24, 705–723.
Adamec, R. E., Shallow, T., & Budgell, J. (1997). Blockade of CCK(B) but not CCK(A)
receptors before and after the stress of predator exposure prevents lasting increases
in anxiety-like behavior: Implications of anxiety associated with posttraumatic stress
disorder. Behavioral Neuroscience, 111, 435–449.
Anderson, M. C., Ochsner, K. N., Kuhl, B., Cooper, J., Robertson, E., Gabrieli, S. W.,
Glover, G. H., & Gabrieli, J. D. E. (2004). Neural systems underlying the suppression
of unwanted memories. Science, 303, 232–235.
Babloyantz, A., Destexhe, A. (1986). Low-dimentional chaos in an instance of epilepsy. Proc.
Natl. Acad. Sci., 83, 3513–3517.
Beauregard, M., Levesque, J., & Bourgouin, P. (2001). Neural correlates of conscious self-
regulation of emotion. Journal of Neuroscience, 21, RC165.
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.28 (1419-1515)

 Luc Ciompi and Jaak Panksepp

Berkowitz, L., Jaffee, S., Eunkyung, J., & Troccoli. T. (2000). On the correction of feeling-
induced judgemental bias. In Forgas, P. (Ed), Feeling and thinking. The role of affect in
social cognition. Cambridge, UK: Cambridge University Press.
Berridge, K. C., & Robinson, T. E. (2003). Parsing reward. Trends in Neurosciences, 9, 507–
513.
Bowlby J. (1973). Attachment and loss. Basic Books, New York.
Burke, M., & Mathews, A. M. (1992). Autobiographical memory and clinical anxiety.
Cognition and Emotion, 6, 23–35.
Cameron, O. G. (2002). Visceral sensory neuroscience: Interoception. Oxford: Oxford
University Press.
Ciompi, L. (1988). The psyche and schizophrenia. The bond between affect and logic. Harvard
University Press, Cambridge/Mass. and London/UK.
Ciompi, L. (1997a). The concept of affect logic. An integrative psycho-socio-biological
approach to understanding and treatment of schizophrenia. Psychiatry, 60, 158–170.
Ciompi, L. (1997b). Non-linear dynamics of complex systems. The chaos-theoretical
approach to schizophrenia. In Brenner, H.-D., Böker, W., Genner, R. (Eds.), Toward a
comprehensive therapy of schizophrenia (pp. 18–31). Hogrefe & Huber, Seattle-Toronto-
Bern-Göttingen.
Ciompi, L. (1997c). [The emotional bases of thinking. Outline of a fractal affect-logic].
Die emotionalen Grundlagen des Denkens. Entwurf einer fraktalen Affektlogik.
Vandenhoeck & Ruprecht, Göttingen.
Ciompi L. (1999). An affect-centered model of the psyche and its consequences for a
new understanding of nonlinear psychodynamics. In Tschacher, W., Dauwalder, J. P.
(Eds.), Dynamics, synergetics, autonomous agents. Nonlinear system approach to cognitive
psychology and cognitive science (pp. 123–131). World Scientific, Singapure-New Jersey-
London-Hong Kong.
Ciompi, L. (2003). Reflections on the role of emotions in consciousness and subjectivity,
from the perspective of affect logic. Consciousness and Emotion, 4, 181–196.
Ciompi, L., & Baatz, M. (2004). Do mental and social processes have a fractal structure? The
hypothesis of affect-logic. In Losa, G. (Ed.), Fractals in biology and medicine. Birkhäuser,
Basel (in press).
Clore, D. L.,& Ortony, A. (2000). Cognition in emotion: Always, sometimes, or never? In
Lane, R. D., & Nadel, L. (Eds.), Cognitive neuroscience of emotion (pp. 24–61). Oxford
Univ. Press, New York.
Clynes, M. (1987). Sentics: The touch of emotions. New York: Doubleday.
Colpaert, F. C., & Balster, R. L. (Eds.). (1988). Transduction mechanisms of drug stimuli. New
York: Springer-Verlag.
Dalgeish, T., & Power, M. J. (Eds.). (1999 ). Handbook of cognition and emotion. John Wiles
& Sons, Chichester-New York-Weinheim-Brisbane-Singapore-Toronto.
Damasio, A. R. (1994). Descartes’ error. Emotion, reason and the human brain.
Grosset/Putnam, New York.
Damasio, A. R. (2000). A second chance for emotion. In Lane, R. D., & Nadel, L. (Eds.),
Cognitive neuroscience of emotion (pp. 12–23). Oxford Univ. Press, New York-Oxford.
Davidson, R. J., Jackson D. C., & Kalin N. H. (2000). Emotion, plasticity, context, and
regulation: Perspectives from affective neuroscience. Psychol. Bull., 126, 890–909.
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.29 (1515-1613)

Energetic effects of emotions on cognitions 

Davidson, R. J., Scherer, K. R., & Goldsmith, H. H. (Eds.). (2003). Handbook of affective
sciences. Oxford University Press, New York.
Davis M. (1999). Functional neuroanatomy of anxiety and fear: A focus on the amygdala. In
D. S. Charney, E. J. Nestler, & B. S. Bunney (Eds.), Neurobiology of Mental Illness (pp.
463–474). Oxford University Press, New York.
Derryberry, D., & Tucker, D. M. (1992). Neural mechanismes of emotion. J. of Consulting
and Clinical Psychology, 60, 329–338.
DSM III-R Diagnostic and Statistical Manual of Mental Disorders. Third Edition revised.
American Psychiatric Association, Washington 1987.
Eich, E., & Macauley, D. (2000). Fundamental factors in mood-dependent memory. In
Forgas, P. (Ed.), Feeling and thinking. The role of affect in social cognition (pp. 109–130).
Cambridge University Press, New York.
Ekman, P. (1984). Approaches to emotion. Hillsdale, NJ: Erlbaum.
Ellis, H. C., & Moore, B. A. (1999). Mood and memory. In Dalgeish, T., & Power, M. J.
(Eds.), Handbook of cognition and emotion (pp. 193–227). Wiley, Chichester-New York-
Weinheim-Brisbane-Singapore-Toronto.
Fivaz-Depeursinge, E., & Corboz-Warnerey, A. (1999). The primary triangle. D devel-
opmental systems view of mothers, fathers, and infants. Basic Books, New York.
Flack W. F., & Laird, J. D. (Eds.). (1998). Emotions in psychopathology: theory and research.
Oxford University Press, New York Oxford.
Forgas, P. (Ed). (2000a). Feeling and thinking. The role of affect in social cognition. Cambridge
University Press, New York.
Forgas, P. (2000b). Feeling and thinking. Summary and integration. In Forgas, P. (Ed),
Feeling and thinking. The role of affect in social cognition (pp. 387–406). Cambridge
University Press, New York.
Fredrickson, B. L. (2002). Positive emotions. In C. R. Snyder & S. J. Lopez (Eds.), Handbook
of positive psychology. Oxford University Press, London.
Freeman, J. W. (to be published, 2004). A biological theory of brain function and its
relevance to psychoanalysis. In John Muller, Joseph Brent, & Craig Piers (Eds.), The
Self-Organizing Subject of Psychoanalysis. Volume 16 in the series, Psychiatry and the
Humanities. Johns Hopkins University Press.
Gordon, N. S., Burke, S., Akil, H., Watson, J., & Panksepp, J. (2003). Socially induced brain
fertilization: Play promotes brain derived neurotrophic factor expression. Neuroscience
Letters, 341, 17–20.
Gottschalk, A., Bauer, M. S., & Whybrow P. C. (1995). Evidence of chaotic mood variation
in bipolar disorder. Arch. Gen. Psychiatr., 52, 947–959.
Haken, H. (1990). Synergetics. An introduction. Springer, Berlin.
Hatfield, E., Cacioppo J. T., & Rapson, R. L. (1994). Emotional contagion. Cambridge
University Press, Paris.
Heim, C., & Nemeroff, C. B. (1999). The impact of early adverse experiences on brain
systems involved in the pathophysiology of anxiety and affective disorders. Biological
Psychiatry, 46, 1509–1522.
Ikemoto S, & Panksepp, J. (1999). The role of nucleus accumbens dopamine in motivated
behavior: A unifying interpretation with special reference to reward-seeking. Brain
Research Reviews, 31, 6–41.
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.30 (1613-1699)

 Luc Ciompi and Jaak Panksepp

Isen, A. M. (2004). Some perspectives on positive feelings and emotions. In Manstead,

A. S. R., Frijda, N., & Fischer, A. (Eds.), Feelings and emotions: The Amsterdam Sym-
posium (pp. 263–281). Cambridge, UK: Cambridge University Press.
Izard, C. E. (1971). The face of emotions. New York: Appleton-Century-Crofts.
Izard, C. E. (1992). Basic emotions, relations among emotions, and emotion-cognition
relations. Psychol. Rev., 99, 561–565.
Izard, C. E. (1993). Organizational and motivational functions of discrete emotions. In
Lewis, M., & Haviland, J. (Eds.), Handbook of emotion (pp. 631–641). Gauilford Press,
New York, London.
Lewis, M. D., & Granic, I. (Eds.). (2000). Emotion, development, and self-organization:
Dynamic systems approaches to emotional development. Cambridge Unviersity Press:
Cambridge, UK.
Kavanagh, D. J. (1992). Recent developments in expressed emotion and schizophrenia. Brit.
J. Psychiat., 160, 601–620.
Kelso, J. A. S. (1995). Dynamic patterns: The self-organisation of brain and behaviour. MIT
Press, Cambridge.
King, R., & Barchas, J. D. (1983). An application of dynamic systems theory to human
behaviour. In Basar, E., Flohr, H., Haken, H., & Handell, A. I. (Eds.), Synergetics of
the brain. Springer, Berlin.
Kleinginna, P. R., & Kleinginna, A. (1981). A categorized list of emotion definitions, with
suggestions for a consensual definition. Motivation and Emotion, 5, 345–359.
Koukkou, M., Lehmann, D., Wackermann, J., Dvorak, I., & Henggeler, B. (1993).
Dimensional complexity of EEG brain mechanims in untreated schizophrenia. Biol.
Psychiat., 33, 397–407.
Kovalik, Z. J., & Elbert, T. (1994). Changes of chaoticness in spontaneous EEG/EMG.
Integrative Physiological and Behavioral Science, 29, 270–282.
Kowalski, R., & Leary, M. R. (Eds.). (1999). The social psychology of emotional and
behavioural problems: Interfaces of social and clinical psychology. Amer. Psychol. Assoc.
Washington, DC.
Lane, R. D., & Nadel, L. (Eds.). (2000). Cognitive neuroscience of emotion. New York: Oxford
University press.
Lawler, E. J., & Thye, S. R. (1999). Bringing emotions in the social exchange theory. Annual
review of sociology, 25.
Lazarus, R. S. (1999). The cognition-emotion debate: A bit of history. In Dalgeish, T., &
Power, M. J. (Eds.), Handbook of cognition and emotion (pp. 3–19). John Wiles & Sons,
Chichester-New York-Weinheim-Brisbane-Singapore-Toronto.
LeDoux J. (1996). The emotional brain. The mysterious underpinnings of emotional life. Simon
and Schuster, New York.
LeDoux, J. (2000). Cognitive-emotional interactions: listen to the brain. In Lane, R. D., &
Nadel, L. (Eds.), Cognitive neuroscience of emotion (pp. 128–155). Oxford Univ. Press,
New York-Oxford.
Leff, J., & Vaughn, C. (1985). Expressed emotions in families. Its significance for mental illness.
Guilford Press, New York-London.
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.31 (1699-1797)

Energetic effects of emotions on cognitions 

Liotti, M., & Panksepp, J. (2004). On the neural nature of human emotions and implications
for biological psychiatry. In J. Panksepp (Ed.), Textbook of Biological Psychiatry (pp. 33–
74). Wiley: Hoboken, NJ.
Lorenz, K. (1966). On aggression. Harcourt, Brace, and World, New York.
Lowenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings.
Psychological Bulletin, 127, 267–286.
MacLeod, A. K., & Salaminiou, E. (2001). Reduced positive future-thinking in depression:
Cognitive and affective factors. Cognition and Emotion, 15, 99–107.
Mandelbrot, B. B. (1983). The fractal geometry of nature. Freeman, New York.
Manstead, A. S. R., Frijda, N., & Fischer, A. (2004). Feelings and emotions: The Amsterdam
Symposium. Cambridge University Press, Cambridge, UK.
Matthews, G., & Wells, A. (1999). The cognitive science of attention and emotion. In
Dalgeish, T., & Power, M. J. (Eds.), Handbook of cognition and emotion (pp. 171–192).
John Wiles & Sons, Chichester-New York-Weinheim-Brisbane-Singapore-Toronto.
Mayberg, H. (2004). Depression: A neuropsychiatric perspective. In J. Panksepp (Ed.),
Textbook of Biological Psychiatry (pp. 197–230). Wiley: Hoboken, NJ.
McGaugh, J. L., & Cahill, L. (2003). Emotion and memory: Central and peripheral
contributions. In R. J. Davidson, K. R. Scherer, & H. H. Goldsmith (Eds.), Handbook of
affective sciences (pp. 93–116). Oxford University Press, New York.
Meaney, M. J. (2001). Maternal care, gene expression, and the transmission of individual
differences in stress reactivity across generations. Annual Review of Neuroscience, 24,
1161–1192.
Mechsner F., Kerzel, D., Knoblich, G., & Prinz, W. (2001). Perceptual basis of bimanual
coordination. Nature, 144, 69–73.
Mesulam, M. (Ed.). (2000). Principles of behavior and cognitive neurology. 2nd ed. New York:
Oxford University Press.
Murphy, F. C., Nimmo-Smith, I, & Lawrence, D. (2003). Functional neuroanatomy of
emotions: A meta-analysis. Cognitive, Affective, & Behavioral Neuroscience, 3, 207–233.
Niedenthal, P. M., & Halberstadt, J. H. (2000). Grounding categories in emotional response.
In Forgas, P. (Ed.), Feeling and thinking. The role of affect in social cognition (pp. 357–
386). Cambridge University Press, New York.
Nocjar, C., & Panksepp, J. (2002). Chronic intermittent amphetamine pretreatment
enhances future appetitive behavior for drug- and natural-reward: Interaction with
environmental variables. Behavioural Brain Research, 128, 189–203.
Ortony, A., & Turner, T. K. (1990). What’s basic about basic emotions? Psychol. Rev., 97,
315–331.
Panksepp, J. (1982). Toward a general psychobiological theory of emotions. Behavioral and
brain sciences, 5, 407–467.
Panksepp, J. (1988). Brain emotional circuits and psychopathologies. In M. Clynes & J.
Panksepp (Eds.), Emotions and psychopathology (pp. 37–76). New York: Plenum Press.
Panksepp, J. (1998). Affective neuroscience. The foundation of human and animal emotions.
Oxford University Press.
Panksepp, J. (1999). Emotions as viewed by psychoanalysis and neuroscience: an exercise in
consilience, and accompanying commentaries. NeuroPsychoanalysis, 1, 15–89.
 JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.32 (1797-1893)

 Luc Ciompi and Jaak Panksepp

Panksepp, J. (2001a). The long-term psychobiological consequences of infant emotions:

Prescriptions for the 21st century. Infant Mental Health Journal., 22, 132–173.
Panksepp, J. (2001b). The neuroevolutionary cusp between emotions and cognitions.
Implications for understanding consciousness and the emergence of a unified mind
science. Evolution and Cognition, 7, 141–163.
Panksepp, J. (2003a). At the interface of affective, behavioral and cognitive neurosciences.
Decoding the emotional feelings of the brain. Brain and Cognition, 52, 4–14.
Panksepp, J. (2003b). Can anthropomorphic analyses of “separation cries” in other animals
inform us about the emotional nature of social loss in humans? Psychological Review,
110, 376–388.
Panksepp, J. (2003c). Commentary on “Understanding addicitive vulnerability”. Neuro-
Psychoanalysis, 5, 21–29.
Panksepp, J. (Ed.). (2004a). Textbook of Biological Psychiatry. Wiley: Hoboken, NJ.
Panksepp, J. (2004b). Basic affects and the instinctual emotional systems of the brain: The
primordial sources of sadness, joy, and seeking. In Manstead, A. S. R., Frijda, N., &
Fischer, A. (Eds.), Feelings and emotions: The Amsterdam Symposium (pp.174–193).
Cambridge University Press, Cambridge, UK.
Panksepp, J., & Bernatzky, G. (2002). Emotional sounds and the brain: The neuro-affective
foundations of musical appreciation. Behavioural Processes, 60, 133–155.
Panksepp, J., Burgdorf, J., Gordon, N., & Turner, C. (2002). Treatment of ADHD with
methylphenidate may sensitize brain substrates of desire. Consciousness & Emotion, 3,
7–19.
Panksepp, J., Burgdorf, J., Turner, C., & Gordon, N. (2003). Modeling ADHD-type arousal
with unilateral frontal cortex damage in rats and beneficial effects of play therapy. Brain
and Cognition, 52, 97–105.
Panksepp, J., & Harro, J. (2004). Future prospects in psychopharmacology. In J. Panksepp
(Ed.), Textbook of Biological Psychiatry (pp. 627–660). Wiley: Hoboken, NJ.
Panksepp, J., & Panksepp, J. B. (2000). The seven sins of evolutionary psychology. Evolution
and Cognition, 6, 108–131.
Panksepp, J., & Moskal, J. (2004). Dopamine, pleasure and appetitive eagerness. An
emotional systems overview of the trans-hypothalamic “reward” system in the genesis
of addictive urges. In Susan Barsch (Ed.), The Cognitive, Behavioral and Affective
Neursociences in Psychiatric Disorders. In press.
Panksepp, J., Nocjar, C., Burgdorf, J., Panksepp, J. B., & Huber, R. (2004). The role of
emotional systems in addiction: A neuroethological perspective. In R. A. Bevins (Ed.),
50th Nebraska Symposium on Motivation: Motivational Factors in the Etiology of Drug
Abuse (pp. 85–126). Lincoln: Nebraska.
Parkinson, B. (1995). Ideas and realities of emotion. London: Routledge.
Paulus, M. P., & Braff, D. L. (2003). Chaos and schizophrenia. Does the method fir the
madness? Biological Psychiatry, 53, 3–11.
Parrott, W. G. (2001). Emotions in social psychology: Essential readings. Psychology Press,
Taylor & Francis, Philadelphia, PA.
Phan, K. L., Wager, T., Taylor, S. F., & Liberzon, I. (2002). Functional neuroanatomy of
emotion, A meta-analysis of emotion activation studies in PET and fMRI. Neuroimage,
16, 331–348.
JB[v.20020404] Prn:23/11/2004; 9:59 F: Z13102.tex / p.33 (1893-1959)

Energetic effects of emotions on cognitions 

Piaget, J. (1953/1954). Intelligence and affectivity. Their relationship during child devel-
opment In Brown, T. A., & Kaegi, C.E. (Eds.), Annual Review Monograph. University of
California Press, Palo Alto, 1981.
Prigogine, I., & Stengers, I. (1983). Order our of chaos. Heinemann, London.
Sargant, W. (1957). Battle for the mind. Doubleday, New York.
Schultz, W. (2002). Getting formal with DA reward. Neuron, 36, 241–263.
Schuster, H. G. (1989). Deterministic chaos. VCH Verlagsgesellschaft GmbH, Weinheim.
Schutter, D. J. L. G., Van Honk, J., & Panksepp, J. (2004). Introducing repetitive transcranial
magnetic stimulation (rTMS) and its property of causal inference in investigating the
brain-function relationship. Synthese, 141, 155–173.
Searle, J. R. (1983). Intentionality: An essay in the philosophy of mind. New York: Cambridge
Univ. Press.
Selye, H. (1946). The general adaptation syndrome and the diseases of adaptation. J. Clin.
Endocrinology, 6, 117.
Siegel, A., Roeling, T. A., Gregg, T. R., & Kruk, M. R. (1999). Neuropharmacology of brain-
stimulation-evoked aggression. Neuroscience and Biobehavioral Reviews, 23, 359–389.
Stern, D. N. (1995).The motherhood constellation. Basic Books, New York.
Spencer-Brown, G. (1979). Laws of form. Durron, New York.
Tschacher, W., Dauwalder, & J. P. (Eds.). (1999). Dynamics, synergetics, autonomous agents.
Non-linear system approach to cognitive psychology and cognitive science (pp. 123–131).
World Scientific, Singapure-New Jersey-London-Hong Kong.
Van der Kolk, B. A. (2004). Psychobiology of postraumatic stress disorder. In J. Panksepp
(Ed.), Textbook of Biological Psychiatry (pp. 319–344). Wiley: Hoboken, NJ.
Zack, M., & Poulos, C. X. (2004). Amphetamine primes motivation to gamble and
gambling-related semantic networks in problem gamblers. Neuropsychopharmacology,
29, 195–207.
Zajonc, R. B. (2000). Feeling and thinking. Closing the debate over the independence of
affect. In Forgas, P. (Ed.), Feeling and thinking. The role of affect in social cognition.
Cambridge University Press, New York.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.1 (41-109)

Negative affective states’ effects

on perception of affective pictures

F. Pahlavan and T. Lubart

Université Paris V

A preliminary study of the International Affective Photo System (IAPS) with

a French population was conducted. Ten photos from three different
valence-based categories (pleasant, neutral, unpleasant) were selected (total =
30 photos). It was hypothesized that participants’ emotional judgments could
be modulated by immediate subjective experience of specific affective states
(Congratulation, Insult) in an affectively congruent manner. Specifically, it
was predicted that individuals under emotional states should show greater
affective valence judgments for affective stimuli of the same valence and
faster reaction times for affect-congruent stimuli.
Each photo was evaluated, under negative, neutral and positive feedback
conditions, on valence, arousal, and dominance dimensions using the
Self-Assessment Manikin scale (SAM; Lang 1980). Participants were 36
university-level psychology students.
The results showed, first, an affect congruency effect in terms of more
processing time when the valence of the emotional state and the information
presented in the stimulus were the same. Second, men compared with
women were more affected by the information presented in the stimuli under
emotional conditions, particularly under the insult condition, whereas
women were more influenced by this information under the neutral
condition. Third, for the memory-related measures for life events, the highest
scores were found for pleasant pictures.
These results are discussed in relation to modulation of the effects of
specific affective states by socialization style and self-regulatory motivations.

Keywords: Insult, positive approbation, feedback, emotional context,

evaluation, perception, affective stimuli.

Numerous studies have demonstrated an impact of moods on evaluative judg-

ments. We seek to determine whether specific negative emotional states could
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.2 (109-177)

 F. Pahlavan and T. Lubart

influence differently evaluative judgments about affective stimuli. After a brief

review of recent explanations of affect-congruent effects on cognitive contents
and cognitive processing, we expose positions concerning affect-incongruent
effects. Then the results of a study are reported.

Affect-congruent effects
Early affect-cognition research was concerned with the influence of affective
states on the content of cognition, especially for explaining affect-congruent
effects in an automatic, bottom-up manner (Bower 1981). Taking two major
orientations, one based on the memory approach (Forgas, Bower, & Krantz
1984), and the other inferential approach (Schwarz & Clore 1983), recent the-
oretical conceptions are rather integrative and propose implications for both
processes.
The Affect Infusion Model (AIM; Forgas 1995), as an integrative affect-
cognition theory, claims that incorporation of affective information into think-
ing occurs whenever individuals engage in heuristic (affect-as-information,
Schwarz & Clore 1983), or substantive strategies (affect-priming, Bower 1981)
as opposed to direct access or motivational strategies (Forgas 1999, 2000). Ac-
cording to Forgas (2000), affect infusion under heuristic processing occurs be-
cause of the affect-as-information mechanism (Schwarz & Clore 1983). Thus,
affect infusion due to current mood could provide a shortcut to infer how
the individuals feel about a particular issue or result from a priming effect for
affective congruent thoughts (Forgas 2000: 357).

Affect-incongruent effects
Affect-congruent effects on memory, evaluation and judgment are not al-
ways seen, in particular with negative affect (Berkowitz & Troccoli 1990; Bless,
Bohner, Schwarz, & Strack 1990; Martin 1999; McFarland & Buehler 1997; Par-
rot & Sabini 1990). Some researchers explain this lack of affect-congruency in
terms of individuals’ consciousness about implications of their feelings (Mar-
tin et al. 1993) and/or awareness that their feelings might bias their judgment,
resulting in an over-correction of their judgmental process and their perfor-
mance (see Berkowitz &Troccoli 1990; Da Gloria, Pahlavan, Duda, & Bonnet
1994; Pahlavan, Duda, & Bonnet 2000b). Berkowitz (1999) suggests that this
over-correction can occur when people are (a) aware of their affective state, (b)
want to arrive at an accurate evaluation, and (c) are mentally active. Another
possible explanation of diverse affect-incongruent effects is that specific emo-
tional states that are typically grouped together may have very different effects,
especially in motivational terms.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.3 (177-215)

Affective perception 

Affect congruent effects and motivational implications

Because there are motivational implications of affective states, some affect-
cognition research focused on the influence of affects with a specific valence
on information processing strategies. There is evidence for important and
asymmetrical effects of positive and negative affective states on cognition (Isen
2000). It has been found that people in positive moods tend to think more
rapidly, more superficially, reach a decision more quickly, use less information,
avoid demanding systematic processing, and are more confident about their
decisions; whereas negative mood has been shown to generate more system-
atic, analytic, and vigilant processing strategies (Forgas 1989, 1991, 1998; Isen
1984, 1987; Schwarz 1990; Schwarz & Bless 1992; Sinclair & Mark 1992). These
differential information processes are also assumed to have different motiva-
tional implications. Accordingly, positive affects should be related to approach
or appetitive processes, whereas negative affects are rather supposed to imply
avoidance or aversive processes. Although research has demonstrated differen-
tial implications of the affective processes in behavioral activation, the exact
nature of this involvement is unclear. Some researchers assumed that these dif-
ferential implications are due to the valence of the affective states, but others
suggest implication of motivational direction of the affect.

Affect- or motivation-congruent effects?

Past research is unable to address whether affective behavioral activation is
associated with emotional valence or motivational orientation, because they
examined the effect of mood states rather than specific emotional states con-
founding affective valence and motivational direction. Additionally, the va-
lence of affective states and the direction of the motivation are supposed to
be directly related. It means that all positive affects are supposed to be related
to approach and all negative affects to avoidance strategies. Nevertheless, these
qualities could be differentiable.
Higgins (1987) distinguished between “agitated” and “dejected” negative
emotions. Accordingly a threat of negative outcomes results in an agitated neg-
ative state (fear, threat, anger, or edginess) and motivation to avoid negative
outcomes. But a lack of positive outcomes induces a dejected negative state
(sadness or disappointment) inducing motivation to approach positive out-
comes. Thus, resulting strategies and tactical decisions for resolving life tasks
are strongly influenced by individuals’ expectancies.
For Schwarz (1990), approach and avoidance situations are characterized
by a basic asymmetry in the amount of analytic reasoning that they require.
To attain a positive goal, it is usually sufficient to pursue one of the many pos-
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.4 (215-262)

 F. Pahlavan and T. Lubart

sible ways of obtaining the positive outcome. But to avoid a certain outcome
all possible causal links that may produce this outcome must be taken into
account. Thus, according to Schwarz (1990), agitated negative affect states as-
sociated with avoidance motivation should trigger a more elaborate analytic
processing style than dejected negative affective states which are associated
with approach motivations. Moreover, agitated negative states could focus in-
dividuals’ attention on information relevant to avoidance motivation, whereas
dejected negative states focus attention on information that is relevant to an
approach motivation.
Therefore, perception of causality to explain discrepancies between perfor-
mance and outcomes could induce, in individuals high in achievement motiva-
tion, negative emotional states (agitated/dejected) and orient their judgments
and behaviors (approach/avoidance tendencies). However, as mentioned by
Weiner (1982), although individuals high in achievement motivation attribute
failure to a lack of effort and success to personal control (Weiner & Kukla 1970),
when experiencing a discrepancy between their performance and outcomes
due to the others’ behavior they could feel anger.
Generally, anger is considered to be a negatively-valenced emotion because
the situations in which anger is evoked are viewed as unfavorable or incon-
gruent with one’s goals (Lazarus 1991). Considered as an agitated negative
state, anger is not necessary related to avoidance tendencies. In fact, anger is
an emotion that is supposed to evoke approach behavioral tendencies (Ek-
man & Friesen 1975; Levenson 1994), because it is often associated with at-
tack in the absence of a defensive motivation (Berkowitz 1993). Research on
learned helplessness phenomena showed increased performance on an unre-
lated second task in individuals who responded with anger to failure at a first
task (Worthman & Brehm 1975), and decreased performance whenever indi-
viduals responded with depression to failure (Mikulincer 1988). In the same
vein, some studies found that infants expressing anger to the extinction of a
reinforcement displayed the highest levels of pulling movements (associated
with approach motivation; Lewis, Alessandri, & Sullivan 1990). Other stud-
ies showed gender-based modulation of these behavioral tendencies due to the
failure or other kinds of aversive stimulation in adults (Da Gloria et al. 1994;
Pahlavan et al. 2000b).
Additionally, some researchers found that the angry people displayed the
same pattern of judgments as those of happy people, showing some preference
for heuristic strategies. Sad people, in contrast, tended to be detail-oriented
and more thorough in their processing (Bodenhausen, Kramer, & Süsser 1994;
Bodenhausen, Sheppard, & Kramer 1994).
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.5 (262-306)

Affective perception 

These suggestions indicate an influence of affective states on informa-

tion processing style, and inferential processes in motivational terms. Affective
implication induces a self-evaluation or self-appraisal process, which in turn
could have motivational implications dependent on individuals’ inferences
about the purpose of the situation. Consequently, individuals’ strategies and
tactical decisions to cope with affective situations could also be mediated by
their motivational tendencies.
The present study was designed to examine how a positive or negative
feedback (insult vs positive approbation) could influence information process-
ing in individuals high in achievement motivation. Past research showed that
arbitrary frustration and provocation influenced affective state and induced
anger (see Berkowitz 1993). Verbal provocation has frequently served as an
important elicitor of negative affects and aggressive actions. Insulting state-
ments, even mild ones, could lead to initiation of aggressive behaviors and
retaliation. The physical discomfort experienced by a person may be subor-
dinate to the symbolic elements that are incorporated in the attack (Greenwell
& Dengernick, 1973). And what makes it possible for intelligent systems to be-
have correctly as a function of what they represent (beliefs, goals, . . .) is that
the representations are encoded in a system of physically instantiated symbolic
codes (Pylyshyn 1999: 5). Therefore, cutting comments and sarcastic remarks
often hurt as much as, and sometimes more than, physical attacks, particularly
when they cause one to “lose face” in front of another person (Richardson,
Leonard, Taylor, & Hammock 1985). Thus, retaliation against an insulting of-
fender could be motivated by an individual’s desire to restore his or her public
image (Tedeschi & Felson 1994).
Being motivated to have a positive rather than negative self-image, individ-
uals should seek to reduce any discrepancy from a valued self-image produced
by their or others’ behavior. Thus, their attention could be selectively oriented
toward the positive rather than negative aspect of the situation. These pref-
erentially tuned processes could produce a better retrieval of positive rather
than negative events and biased self-knowledge (Higgins & Krulanski 2000).
Therefore, their strategic decision to maintain or restore public image could be
influenced by these tuned processes.
It is possible to maintain or restore one’s public image in several ways, such
as by ignoring the fact that one has been hurt. If a person cannot deny be-
ing hurt, he or she may try to restore a positive image by retaliation to justify
his/her positive image, but the victim’s response is affected by the attacker’s
awareness of success or failure of the intended attack (Ohbuchi &Kambara
1985). If a victim thinks that the attacker is aware of the success of the in-
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.6 (306-365)

 F. Pahlavan and T. Lubart

tended attack, the victim should show a higher level of retaliation in order not
to lose face.
The present studies examine how specific emotional states induced by
experimental manipulation (Congratulation versus Insult) could influence af-
fective evaluations made by participants high in achievement motivation (uni-
versity students).
We propose that emotions can affect evaluation in a relatively passive and
automatic manner (see Berkowitz 2000). These primed affects could modify
judgment through connections between the affective system and the motor re-
sponse system (Winkielman, Zajonc, & Schwarz 1997: 437). Our past research
has shown these affective priming effects on motor decisions (Da Gloria, Duda,
Pahlavan, & Bonnet 1989; Da Gloria et al. 1994; Pahlavan, Bonnet, & Duda
2000a; Pahlavan et al. 2000b). Accessibility of specific cognitive procedures
(related to avoidance or approach strategies) increased through affective ma-
nipulation (Congratulation vs. insult) should facilitate their application and
their transfer to a subsequent task of evaluation (Smith & Branscombe 1987;
Zillmann 1983). Focusing on conscious affective evaluation of stimuli, the
emotions induced by an experimental emotional manipulation can influence
the judgments of to-be-evaluated stimuli.
In this optic, we used a picture perception methodology designed by Lang
and his colleagues. We took advantage of the properties of photos of real-
life scenes from the International Affective Picture System (IAPS) and the
established reliability of the Self Assessment Manikin (SAM) scales (Valence,
Arousal, Dominance) to obtain and measure affective judgments and judg-
mental behaviors (see Bradley et al. 1992; Lang, Bradley, & Cuthbert 1997: 106;
Lang, Greenwald, Bradley, & Hamm, 1993). Specific configuration of latency
and judgments of perceived feeling should reveal specific motivational inclina-
tions induced by experimental manipulations.
Thus, in the case of the current studies, measures of affective valence or
arousal are considered to be indicators of action tendencies. Ratings of plea-
sure could be considered as reflecting one’s tendency to approach a stimu-
lus, whereas displeasure reflects a tendency to escape. Similarly, judgments
of arousal could represent the amount of vigor associated with a given be-
havior. Dominance ratings may reflect individuals’ feeling of control in the
situation and index the interactive relationship that exists between individuals
and stimuli to be judged (with high dominance associated with the one having
maximum control in the situation).
The aim of the present experiment was, by means of the picture perception
task, to examine the organization and modification of emotional perception
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.7 (365-405)

Affective perception 

and its motivational consequences when individuals have experienced a spe-

cific emotional state. In the case of the present study, valence and arousal are
considered to represent primitive motivational parameters (strategic frame,
Lang, Bradley, & Cuthbert, 1990; Lang, Bradley, & Cuthbert 1997) that define
both a general disposition to approach (appetitive disposition) or avoidance
(aversive disposition) stimulation and the vigor of the directional tendency.
Measures on these dimensions could also be helpful for understanding the
processes involved in retrieval of past life events (Higgins & Krulanski 2000;
Leventhal 1993).

Specific Hypotheses: A relationship between subjects’ judgments and their im-

mediate experiences of emotion was postulated. It was hypothesized that the
latency and judgments of perceived level of pleasure, arousal, and dominance
of the stimuli materials as well as recall of relevant life events could be modu-
lated by the emotional context of their presentation in an affectively congruent
manner. Specifically, we predicted that individuals under the insult condition
should show greater negative valence judgments for negative stimuli and faster
reaction times for these affect-congruent stimuli.

Method

Overview
Male and female volunteers participated in a study of emotion and its rela-
tionship to perception. During an initial phase of study, subjects completed
a questionnaire measure of personal experiences of traumatic life events. The
participants were then instructed that a series of pictures would be displayed
and that they should attend to each picture the entire time it was exposed on the
screen because they would have to rate their feelings about the picture on three
scales. After a practice phase, a female experimenter introduced the experi-
mental treatment (either insults, neutral remarks, or encouragement). Then
the participants rated 30 pictures extracted from International Affective Photo
System (Lang, Bradley, & Cuthbert 1995) using the Self-Assessment Manikin
scales (Lang 1980 as cited in Bradley & Lang 1994).

Participants
Thirty-six right-handed University of René Descartes introductory psychol-
ogy students (18 females, and 18 males) aged between 18 and 29 years (M =
22.44 years, SD = 2.95) were recruited by a poster in the university’s building.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.8 (405-464)

 F. Pahlavan and T. Lubart

Left-handed subjects and those who had experienced a recent traumatic event
were excluded.

Stimuli
Thirty color images, extracted from the International Affective Photo System
(IAPS; Lang, Bradley & Cuthbert 1995) were used. There were three cate-
gories (pleasant, neutral, unpleasant) of 10 photos selected based on intra-and
inter-categorical matching. The intra-categorical matching was based on va-
lence ratings obtained from a previous group validation study reported by Lang
and his colleagues (Lang, Bradley, & Cuthbert 1995, average of mean valence
and standard deviation (see Annex 1) for each category were: for pleasant:
M = 7.18; SD = 1.88, for neutral: M = 4.44; SD = 1.52; and for unpleas-
ant : M = 2.02; SD = 1.36). The inter-categorical matching was based on the
type of the central elements of the pictures (object, animal, human beings),
and their number. The human photos were matched on the age and gender of
the person in the picture.

Apparatus and response measurement

The presentation of a digitized version of the photographs, image duration
(6 s), collection of ratings, and latency of the ratings were controlled and
recorded by a 486 IBM-compatible computer. Valence, arousal, and domi-
nance ratings for each photo were obtained using a digitized version of the
Self-Assessment Manikins (SAM; Lang 1980 as cited in Bradley et al. 1994).
The time between removal of the picture and the rating response by subjects as
well as rating time for each SAM dimension were recorded.

Experimental conditions
There were three conditions: Positive Approbation, Insult, and Neutral treat-
ment. In each condition, participants completed a block of three trials designed
to aid habituation to the materials and the experimental procedure. The com-
puter was programmed to break down after the practice trials, to let one of the
experimenters introduce the experimental treatment, which was either indirect
congratulating (“It’s really excellent, you understood everything! We will con-
tinue.”), indirect insulting (“It’s really bad, you understood nothing! We will
continue.”), or being neutral (“I have the results, I’m going to register it. We
will continue.).
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.9 (464-508)

Affective perception 

Experimental task
After the three practice trials and the introduction of experimental conditions,
all the participants received 30 trials. A fixed random order of presentation of
the photos was used (see annex 1). For each experimental trial, the participants
had to evaluate their affective reactions (valence, arousal, and dominance di-
mensions) while looking at a picture. For the life events question, participants
replied by using one of three response keys corresponding to the following
propositions: (A) the photo leads you to recall a specific personal life event
that you can date, (B) the photo leads you to recall some non-specific personal
life event that you cannot date, and (C) the photo does not lead you to recall
any personal life event. Then there was a final presentation of 10 trials with
only pleasant pictures, which was hypothesized to induce, in all participants, a
positive mood.
Each trial consisted of four specific events: (a) a 5-second presentation of a
screen bearing the caption “Attention, next picture”, (b) the to-be-rated picture
was displayed by the computer, for 6 seconds, (c) the three SAM scales were
presented to participants, one at a time in a fixed order randomized across
photos (see Annex 1), (d) the computer displayed the question about memory
of one’s life events. The participants used the computer’s mouse to respond
to the SAM scale to rate the displayed picture on each scale. For the life events
question, participants replied by using one of three response keys (A, B, and C).

Procedure
Participants in all conditions were told that the purpose of the experiment
was to assess their opinions and reactions to different pictures. After entering
a sound-attenuated, dimly-lit room, the participants were invited, by one of
the two female experimenters, to sit on a chair facing a computer monitor on
which a uniformly gray image was presented and to complete a questionnaire
measuring life events. The other female experimenter evaluated the question-
naire for traumatic events. If no recent traumatic events were reported, then
the participant was invited to fill out a consent form and read instructions dis-
played on the computer. The first experimenter told participants that a series
of pictures would be presented by the computer and they should pay atten-
tion to each picture the entire time it was shown on the screen because they
would have to rate their feelings about them on a series of three scales, and to
answer some questions about memory for the pictures. Participants were in-
structed to view each picture for the entire time that it was on the screen and
to refrain from making their rating until presentation of the scales. They were
then instructed how to use the computerized SAM display to make ratings of
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.10 (508-558)

 F. Pahlavan and T. Lubart

valence, arousal, and dominance. Participants were also instructed that an ini-
tial block of three trials was designed to aid habituation to the experimental
task. The computer was programmed to stop after the practice trials, and to
let the first experimenter introduce the experimental treatment. Then the ex-
perimenter restarted the program and the experiment continued. Immediately
after the presentation and ratings of all pictures, participants completed a ques-
tionnaire evaluating their impression of the experimenter, and a questionnaire
evaluating their irritability as a personality trait (Caprara & Renzi 1981). Fi-
nally, they were thoroughly debriefed. Probed for suspicion and skepticism, no
subjects reported knowledge of the study’s purpose.

Data analyses
A 2 × 3 × 3 × 3 (Sex of the subjects × Failure condition × Scale type × Pic-
ture valence) factorial ANOVA was used in which the latter two factors were
treated as repeated measures. Analysis of participants’ scores on the irritability
scale showed a significant main effect for feedback conditions, (F (2, 30) = 3,95,
p < .03), corresponding to a significant difference between positive and nega-
tive feedback (F (1, 20)= 10,89, p < .004; M Congratulation = 93,67 vs M Neutral =
107,25 vs M Insult = 110,92). Therefore, for all analyses, irritability scores were
integrated as a co-variable.

Results

Manipulation check for affect-induction

In line with Berkowitz and Troccoli (1990), the participants evaluated the
experimenter as significantly more quick-tempered, F(2, 29) = 4.22, p <
.03; M = 1.73 vs M = 1.58 vs M = 2.58), more offhanded, F(2, 29) = 5,52,
p < .00; M = 3.30 vs M = 1.75 vs M = 2.50) and more obstinate,
F(2, 29) = 2.58, p < .09; M = 1.45 vs M = 1.33 vs M = 2.25) under the
insult condition compared to the other conditions.

Reliability of affective ratings

Correlations were computed between the normative American-based pleasure,
arousal, and dominance mean ratings of the photos (Lang et al. 1995) and
the mean ratings obtained in the current study with subjects under a neutral
experimental condition. Affective ratings across the two samples were highly
related, which validates in part the selection of photos based on American
norms (Valence, r = .97; Arousal, r = .87; Dominance, r = .89).
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.11 (558-640)

Affective perception 

Affective ratings
Table 1, shows mean valence, arousal and dominance ratings for each of the
three series of pictures in function of experimental conditions and sex of the
participants.1 Analysis for statistical power of these data yielded a medium
effect size (fvalence = 0.15, fArousal = 0.12, fDominance = 0.21).
Significant main effects of picture valence (F (2, 60) = 61,79, p < .000) and
scale type (F (2, 60) = 2,11, p < .005) were found. The participants’ global af-
fective evaluations were lowest when they were exposed to unpleasant pictures
and highest in the case of the pleasant ones (MPleasant = 5,81 vs MNeutral = 4,94
vs MUnpleasant = 4,28). Indeed, they felt rather aroused and dominant but

Table 1. Affective Rating as a function of Sex of the subjects, Experimental condition,

Picture Valence, and Scale type

Sex of the subjects/ Picture valence/

Experimental Scale type
Conditions
Unpleasant Pleasant Neutral
Valence Arousal Dominance Valence Arousal Dominance Valence Arousal Dominance
Male subjects
Positive approbation
M 2.90a 6.25 a 4.42 b 6.77 a 5.00 b 5.85 a 5.22 a 4.07 a 5.57 a
SD 1.74 1.90 2.01 1.84 2.31 1.87 1.88 2.24 1.90
Neutral
M 2.22 a 7.05 a 3.17 a 6.92 a 5.08 a 5.97 a 5.02 a 3.82 a 5.17 a
SD 1.06 1.55 1.67 1.46 2.68 1.64 1.43 1.74 1.42
Insult
M 2.85 a 6.75 a 4.17 a 6.70 a 5.02 a 6.45 a 4.77 a 4.45 a 5.78 a
SD 1.28 1.87 2.04 1.59 2.10 1.99 0.97 2.39 1.53
Female subjects
Positive approbation
M 1.92 a 7.23 a 3.25 b 5.27 a 4.77 a 5.37 a 4.95 a 3.90 a 5.40 a
SD 0.90 1.20 1.63 1.83 2.00 1.34 1.57 1.73 1.46
Neutral
M 2.08 a 7.20 a 3.28 c 6.40 a 5.43 a 6.33 a 5.32 a 4.92 a 6.35 a
SD 1.27 2.04 1.75 2.10 2.55 2.36 1.87 2.48 1.74
Insult
M 2.28 a 6.50 a 3.43 a 6.12 a 3.75 c 6.42 a 4.67 a 3.95 a 5.63 a
SD 2.29 1.85 1.47 1.83 2.00 1.85 1.43 1.99 1.82
Note: Means in the same column that do not share subscripts differ at p < .05 in the LSD
tests post-hoc.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.12 (640-777)

 F. Pahlavan and T. Lubart

less pleased independently of the valence of the pictures (MValence = 4,63 vs

MArousal = 5,28 vs MDominance = 5,11).
Analyses showed a significant interaction involving pictures’ valence and
scale type, which supported the reliability of the selection of the picture based
on valence and arousal evaluations, F (4, 120) = 109,08, p < .000. Judgments
of unpleasant photographs indicated the most negative (M Unpleasant = 2,38 vs
M Pleasant = 6,83 vs M Neutral = 3,62), and the most aroused feelings compared
with the other categories of the pictures (M Unpleasant = 6,53 vs M Pleasant = 4,84
vs M Neutral = 6,06), but their dominance evaluation was the lower (M Unpleasant =
4,99 vs M Pleasant = 4,18 vs M Neutral = 5,65).
A marginal significant interaction effect involving experimental condi-
tions was found in association with sex of the subjects, F(2, 30) = 2.83, p <
.08; male subjects compared to female subjects rated their feelings as more
intense under emotional conditions, especially under the insult condition
(Men: M Congratulation = 5,12; M Insult = 5,21 vs Women: M Congratulation = 4,78;
M Insult = 4,75). The opposite trend was found under the neutral condition
(Men: M Neutral = 4,93 vs Women: M Neutral = 5,26). A similar significant in-
teraction effect was found when comparing the insult condition to the neutral
condition, F (1, 19) = 4,85, p < .04, indicating that men and women did not
evaluate their feelings for affective pictures in the same manner after being in-
sulted. Men seem to be more emotionally influenced by emotional experiences,
particularly by an insult, than women for whom more affective reactions were
rather found in the absence of emotional experience.

Decision Time
Table 2 shows Rating Times in each condition for male and female subjects for
each of the 3 sets of pictures rated on the three dimensions of valence, arousal
and dominance.
The results concerning the response time for making affective judgments
about the photos on the different scales showed significant main effects cor-
responding to picture valence (F(2, 60) = 5,35, p < .007) and scale type
(F(2, 60) = 5,92, p < .005). These significant main effects indicated faster
decision times for neutral compared with negative or positive photographs
(M Neutral = 3189 msec vs M Unpleasant = 3448 msec vs M Pleasant = 3377 msec), and
for valence compared with arousal or dominance ratings (M Valence = 3161 msec
vs M Arousal = 3290 msec vs M Dominance = 3564 msec).
An interaction effect revealed modulation of the effect of the scale type by
the valence of the pictures for affective rating times (F (4, 120) = 7,55, p < .000);
judgments of feelings to pleasant photographs indicated the slowest decision
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.13 (777-793)

Affective perception 

Table 2. Rating Time as a function of Sex of the subjects. Experimental condition.

Picture valence. and Scale type

Sex of the subjects/ Picture valence/

Experimental Scale type
Conditions
Unpleasant Pleasant Neutral
Valence Arousal Dominance Valence Arousal Dominance Valence Arousal Dominance
Male subjects
Positive approbation
M 3530b 3482b 4051b 3156b 3665b 3844b 3448b 3517b 3904b
SD 2003 446 1772 1708 1653 1551 2122 2147 1988
Neutral
M 3512a 3738a 3964a 3258a 3817a 3607a 3332a 3161a 3940a
SD 2006 1556 1928 1820 1908 1880 2127 2190 2022
Insult
M 3424a 3105a 4120a 2997a 3781a 2981c 2976a 2580c 2855c
SD 1835 1889 1718 1560 2228 1461 2032 1776 1777
Female subjects
Positive approbation
M 3043a 3490a 4068c 3815c 4086a 3376a 3355a 3479a 3661a
SD 1660 2045 1737 2149 2072 2051 1994 1878 2290
Neutral
M 2954a 2995a 3980a 3130a 3297a 3368a 2809c 3182a 3541a
SD 1515 1847 1498 1654 2039 1507 1987 2069 1875
Insult
M 2580c 2685c 3348d 2963a 2915c 2737d 2618d 2243d 2802d
SD 1157 1439 1329 1445 1282 1013 1482 1749 1191
Note: Rating Times are expressed in milliseconds.
Means in the same column that do not share subscripts differ at p < .05 in the LSD tests
post-hoc.

times on valence (M Unpleasant = 3174 msec vs M Pleasant = 3220 msec vs M Neutral =

33090 msec), and arousal (M Unpleasant = 3249 msec vs M Pleasant = 3594 msec
vs M Neutral = 3027 msec), dimensions. However, the evaluations of domi-
nant feelings to unpleasant compared to other pictures were particularly slower
(M Unpleasant = 3922 msec vs M Pleasant = 3319 msec vs M Neutral = 3450 msec).
Analyses of the effect of success or failure feedback compared to neutral
feedback showed the same significant main and interaction effects. Addition-
ally, the analyses showed that the effect of Congratulation compared to in-
sult feedback on affective rating times depends on both emotional conditions
and the valence of the pictures, F(2,40) = 3,40; p < .05. In spite of gener-
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.14 (793-928)

 F. Pahlavan and T. Lubart

ally increased rating times under the Congratulation condition, compared to

insult condition, the participants under the Congratulation condition took
less time to process unpleasant pictures, but more to process pleasant pic-
tures (M Unpleasant = 3610 msec vs M Pleasant = 3657 msec), whereas the reverse
tendencies occurred under the insult condition (M Unpleasant = 3210 msec vs
M Pleasant = 3063 msec). The fastest rating times were found for neutral pictures
under the insult condition (M Insult = 3561 msec VS M Neutral = 2679 msec).
With regard to these results, it seems that positive approbation increased,
whereas an experience of insult decreased processing of the pictorial informa-
tion, particularly when the valence of pictorial information was the same as the
valence of the emotional experience. In the neutral condition, the affective rat-
ing times reflected general tendencies and increased affective rating times for
emotional information, especially negative ones.

Retrieval of life events

Table 3 presents, by condition, gender, and valence of the pictures, the mean
Retrieval of life events. The retrieval responses were coded as follows: 3 points
for a specific life event, 2 points for a non-specific life event, and 1 point for
a non-recalled life event. A 2 × 3 × 3 factorial plan (Gender × Emotional
conditions × Picture valence) with repeated measures on the last factor was
used to analyze the means of these scores.
Analyses of the data concerning the retrieval of life events evoked by the
photos revealed a significant main effect for photo valence. The configuration
of the data indicated the highest retrieval scores in response to positive photos,
whereas exposure to negative pictures tended to induce lowest retrieval scores
(F(2, 60) = 16.82, p < .00; M Unpleasant = 10,17 vs M Pleasant = 20,47 vs M Neutral =
18,33). No effects relating to the experimental manipulations were found.
These results provide evidence in favor of the notion that people are moti-
vated to have positive rather than negative thoughts, and orient selectively their
attention to positive versus negative acts during performance. Consequently, a
better retrieval of positive versus negative acts after performance results from
these preferentially tuned processes (Isen 1984; Higgins et al. 2000).

Analyses of the affective and time ratings by valence and content of the picture
Analyses of the affective and time ratings by valence and content of the picture
confirmed principal effects found in the case of the global analyses. Addi-
tionally, these analyses revealed diverse effects of processing of the affective
information related to social life by social agent under specific emotional states.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.15 (928-946)

Affective perception 

Table 3. Retrieval of Life Events as a function of Sex of the subjects. Experimental

condition. Picture valence
Sex of the subjects/ Picture valence/
Experimental Scale type
Conditions
Unpleasant Pleasant Neutral

Male subjects
Positive approbation
M 12.00 19.83 18.17
SD 7.3 4.53 4.6
Neutral
M 11.17 21.00 17.67
SD 9.39 3.79 2.80
Insult
M 9.50 21.50 18.33
SD 4.46 5.32 4.55
Female subjects
Positive approbation
M 7.67 20.00 17.33
SD 2.73 3.29 1.89
Neutral
M 11.17 21.33 19.67
SD 6.71 3.20 3.88
Insult
M 9.50 19.17 18.83
SD 2.88 3.71 2.64
Note: Values represent rate mean of score of retrieval.
Means in the same column that do not share subscripts differ at p < .05 in the LSD tests
post-hoc.

In fact, when analyses by pictures were conducted, we observed an effect

of the experimental manipulation for some specific pictures. For example, an
interaction associating failure and sex of the subjects, in the case of the overall
exploratory analyses, was found to be also significant for certain of categories
of pictures. Therefore, it may be important to systematically analyze the pres-
ence of effects across the different stimuli used in emotion-centered research
because the presence or absence of trends may be due more to the specific
selection of stimuli used than to a truly generalized phenomenon.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.16 (946-985)

 F. Pahlavan and T. Lubart

Discussion

The major results of this experiment show an effect of affect manipulation on

judgments and decision times conditioned by sex of subjects, dimensions of
the judgments, and the specific content of the pictures. Typically, research has
used a large set of pictures (at least 60 pictures) for resolving the problem of
heterogeneity of picture content, averaging across individual pictures. How-
ever, because our primary objective was to study effects of a specific emotion
on perception and memory, the use of a large set of pictures would dissipate
the effect of the experimental manipulation of emotion. For this reason, a more
limited set of the pictures (30 pictures) was chosen as a function of their valence
ratings, and precautions were taken to balance other information concerning
these pictures. Nevertheless, in the case of this study, analyses by pictures2
confirmed the major effects observed for overall exploratory analyses for the
experimental manipulation.
Overall, our results reveal that men and women process affective infor-
mation differently whenever they experience a specific emotional state. More
precisely, it seems that men are more affected than women are by emotional
experiences, particularly when an offender insulted them in presence of a third
person. Women expressed more intense affective reactions in the absence of
emotional experience. Considering the rating times for these effects, we can
notice that the rating times to express the observed differential affective rat-
ings were, for men and women, the fastest under the insult condition and
the slowest under the congratulation condition (Men: M Insult = 3202msec
vs M Congratulation = 3622 msec vs M Neutral = 3592 msec; Women: M Insult =
2766 msec vs M Congratulation = 3597 msec vs M Neutral = 3251 msec). Therefore,
in spite of faster global processing times under the insult condition, affective
reactions differed. Faster processing for men results in the most intense reac-
tions, whereas for women it results in the least intense ones. More interesting,
to process information present in the stimuli under the neutral condition, men
took more time than women to express the least emotional reactions. Under
the neutral condition women expressed their most affective reactions.
The faster rating times under emotional experiences support our hypoth-
esis about the relation between anger and approach tendencies. Experience of
an insult induced, even in individuals high in achievement motivation, a state
of anger and approach tendencies. However, this acceleration of processing af-
fected affective reactions differently in men and women. For men, recorded
affective ratings reflect accumulation of the effects of emotional states and af-
fective reactions induced by information presented in the stimuli. For women,
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.17 (985-1036)

Affective perception 

affective ratings seem to be a function of a relative comparison between their

existing emotional state and affective reactions induced by the information
present in the stimuli. Under the emotional conditions, they felt less affected
by information present in the stimuli. By a process of self-awareness or over-
correction, and following the instructions of the experimenter, women seemed
to evaluate their feelings to the pictures without taking into account their
existing emotional states.
Research on men’s and women’s socialization and their differential devel-
opment of mental models of the social world explain, in part, these differences
(Eagly & Stefen 1986; Österman, Bjorkqvist, Lagerspetz, Kaukianen, Landau,
Fraczek, & Caprara 1998). A growing body of research suggests that, because
parents discuss emotions more with their daughters than with their sons, girls
are socialized to be more attuned to emotions, particularly in interpersonal
terms. For Higgins (1991), one consequence of this differential socialization
would be the development of stronger self-other contingencies. It suggests that
women tend to have an interdependent view of the self, and men construct the
self in terms of independent self-guides (Moretti & Higgins 1999). According
to Roberts (1991), women attach more informational value to the evaluative
feedback. Thus, men may rather tend to use objective, external criteria to eval-
uate themselves, whereas women could try to accomplish their assignment
and seek congruency with others’ standards. Consequently, their expression
of emotions may differ because different self-construals imply the pursuit of
diverse social goals.
Concerning affective rating times, we found that the effect of congratula-
tion feedback compared to insult feedback on affective rating times depends on
the valence of the pictures. In spite of general increased rating times under the
congratulation condition, processing times for unpleasant pictures decreased
more in the congratulation condition than the insult condition; whereas to
process pleasant pictures under the same condition, participants required to
more time. The fastest rating times were found for neutral pictures under the
insult condition.
As a matter of fact, processing of the neutral pictures under emotional con-
ditions could be considered as corresponding to the effect of the experimental
conditions, whereas processing of the affective pictures under the neutral con-
dition reflect rather processing times due to the affective reactions induced
by the pictures. Thus, with regard to our results, it seems that an experience
of positive approbation increases, whereas an experience of insult decreases
processing of the pictorial information, except when the valence of pictorial
information is the same as the valence of the emotional experience.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.18 (1036-1080)

 F. Pahlavan and T. Lubart

For the neutral condition, the affective rating times reflect general tenden-
cies, with increased affective rating times for emotional information, especially
negative ones. This slower processing could correspond to systematic and an-
alytic processing due to activation of the negative knowledge by negatively va-
lenced emotional information presented in the stimuli. Affective ratings in this
case reflect rather standard evaluations (M Unpleasant = 4,17 vs M Pleasant = 6,02 vs
M Neutral = 5,10).
Therefore, affect congruency in the case of this experience is observed
in terms of an increase in processing times. In fact, this slower processing
corresponds to the most intense affective reactions to pleasant pictures un-
der the congratulation conditions, and the least under the insult condition
(Congratulation: M Unpleasant = 4,33 vs M Pleasant = 5,67 vs M Neutral = 4,85;
Insult: M Unpleasant = 4,33 vs M Pleasant = 5,75 vs M Neutral = 4,88). For non-
emotional information presented in the stimuli, experience of insult decreased
processing times.
For retrieval of life events, we found classical evidence in favor of the no-
tion that people are motivated to have positive rather than negative thoughts,
and selectively orient their attention to positive versus negative acts during
performance. Consequently, a better retrieval of positive versus negative acts
after performance results from these preferentially tuned processes (Isen 1984;
Higgins et al. 2000).
The general conclusion about this experiment is that the affect congruency
effects in the case of the specific emotional states vary considerably and de-
pend on affective context, individual developmental history, self-implication,
and the nature of emotional states. The multidimensional imagery approach
used in the current research shows the complexity of affect-congruent effects,
and suggests the necessity of a theoretical model in which all the levels of be-
havioral regulations are integrated for explaining these effects. Further research
on this topic needs to focus on a better comprehension of these self-regulatory
motivation effects, especially when under the specific emotional states where
individuals are exposed to ambiguous targets.

Acknowledgements

We thank Pierre Taranne for his assistance in the development of the computer
version of materials used in the experiment.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.19 (1080-1153)

Affective perception 

The experiments reported in this paper were performed by Nathalie CU-

PIC, Sophie GIFFARD, and Elise VERNET during preparation of their Master’s
degree. We thank Leonard BERKOWITZ for helpful comments.

Notes

. Given the small size of the samples, the scores were tested for heterogeneity of variance.
The values of analyses of variance tests of homogeneity for affective ratings were inferior to
the significant values with n = 6 and k = 6, except for Levene F in the case of the valence
rating for negative photos. So, in spite of the small size of the samples, we can conclude, with
some reserve concerning the valence rating of negative photos, that the data do not provide
evidence that the assumption of homogeneity of variance has been violated.
. Additional analyses by valence and content of the pictures involving emotional condi-
tions are available upon request.

References

Atkinson, J. W., & Feather, N. T. (Eds.). (1966). A theory of achievement. New York: Wiley.
Atkinson, J. W. (1957). Motivational determinants of risk-taking behavior. Psychological
Review, 64, 359–372.
Bandura, A., Cervone, D. (1983). Self-evaluative and self-efficacy mechanisms governing
the motivational effects of goal systems. Journal of Personality and Social Psychology, 45,
1017–1028.
Berkowitz L. (1999). On the correlation of feeling-induced judgmental biases. In J. P.
Forgas (Ed.), Feeling and Thinking: The role of affect in social cognition (p. 131–152).
Cambridge: Cambridge University Press.
Berkowitz L. (Ed.). (2000). Causes and consequences of feelings. Paris, Cambridge: Cambridge
University Press.
Berkowitz, L. & Troccoli, B. T. (1990). Feelings, direction of attention, and expressed
evaluations of others. Cognition and Emotion, 4, 305–325.
Bless, H. & Fielder, K. (1995). Affective states and the influence of activated general
knowledge structures. Personality and Social Psychology Bulletin, 21, 766–778.
Bless, H. (1999). The interplay of affect and cognition: The mediating role of general
knowledge structures. In J. P. Forgas (Ed.), Feeling and Thinking: The role of affect in
social cognition (p. 1–28). Cambridge: Cambridge University Press.
Bless, H., Bohner, G., Schwarz, N., & Strack, F. (1990). Mood and persuasion: A cognitive
response analysis. Personality and Social Psychology Bulletin, 16, 331–345.
Bodenhausen, G. V., Kramer, G. P., & Süsser, K. (1994). Happiness and stereotypic thinking
in social judgment. Journal of Personality and Social Psychology, 66, 621–632.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.20 (1153-1246)

 F. Pahlavan and T. Lubart

Bodenhausen, G. V., Sheppard, L. A., & Kramer, G. P. (1994). Negative affect and social
judgment: The differential impact of anger and sadness. European Journal of Social
Psychology, 24, 45–62.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129–148.
Bradley, M. M., Greenwald, M. K., Petry, M., & Lang, P. J. (1992). Remembering pictures:
Pleasure and arousal in memory. Journal of Experimental Psychology: Learning, Memory,
and Cognition, 18, 379–390.
Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: The self-assessment manikin and
the semantic differential. Journal of Behavior Therapy and Experimental Psychiatry, 25,
45–59.
Caprara, G. V., & Renzi, P. (1981). The frustration-aggression hypothesis vs. Irritability.
Recherches de Psychologie Sociale, 3, 75–80.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2rd ed.). New Jersey:
Lawrence Erlbaum Associates.
Cross, S. E., & Madson, L. (1997). Models of the self: Self construals and gender. Psychological
Bulletin, 122, 5–37.
Da Gloria J., Duda, D., Pahlavan F., & Bonnet P. (1989). “Weapons effects” revisited: Motor
effects of the reception of aversive stimulation and exposure to pictures of firearms.
Aggressive Behavior, 15, 265–271.
Da Gloria J., Pahlavan F., Duda D., & Bonnet P. (1994). Evidence for a motor mechanism of
pain-induced aggressive instigation in humans. Aggressive Behavior, 20, 1–7.
Eagly, A. H., & Steffen, V. J. (1986). Gender and aggressive behavior: A meta-analytic review
of the social psychological literature. Psychological Bulletin, 100, 309–330.
Ekman, P., & Friesen, W. V. (1975). Unmasking the face: A guide to recognizing emotions from
facial clues. London: Oxford, England: Prentice-Hall.
Fiedler, K. (1999). Toward an integrative account of affect and cognition phenomenon using
the BIAS computer algorithm. In J. P. Forgas (Ed.), Feeling and Thinking: The role of
affect in social cognition (pp. 223–252). Cambridge: Cambridge University Press.
Forgas (1995). Mood and judgment: Affect infusion model (AIM). Psychological Bulletin,
21, 747–765.
Forgas, J. P. (1989). Mood effects on decision – making strategies. Australian Journal of
Psychology, 41, 197–214.
Forgas, J. P. (1991). Mood effects on partner choice: Role of affect in social decisions. Journal
of Personality and Social Psychology, 61, 708–720.
Forgas, J. P. (1998). On being happy and mistaken: Mood effects on the fundamental
attribution. Journal of Personality and Social Psychology, 75, 318–331.
Forgas, J. P., & Bower, G. H. (1987). Mood effects on person perception judgments. Journal
of Personality and Social Psychology, 53, 53–60.
Forgas, J. P. & Bower, G. H. , & Krantz, S. (1984). The influence of mood perceptions of
social interactions. Journal of Personality and Social Psychology, 20, 497–513.
Forgas, J. P. (1999). The interplay of affect and cognition: The mediating role of general
knowledge structures. In J. P. Forgas (Ed.), Feeling and Thinking: The role of affect in
social cognition (pp. 1–28). Cambridge: Cambridge University Press.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.21 (1246-1327)

Affective perception 

Higgins, E. T. & Tykocinski, O. (1992). Self-discrepancies and biographical memory:

Personality and cognition at the level of psychological situation. Personality and
Psychology Bulletin, 18, 527–535.
Higgins, E. T. (1991). Development of self-regulatory and self-evaluative processes: Costs,
benefits, and tradeoffs. In M. R. Gunner & L. A. Sroufe (Eds.), Self processes and
development: The Minnesota Symposia on child development, Vol. 23 (pp. 125–165).
Hillsdale, NJ: Erlbaum.
Higgins, E.T., and Krulanski, H. (2000). Motivational Science: The nature and functions of
wanting. In E. T. Higgins & H. Krulanski, H. (Eds), Motivational Science: Social and
personality perspectives (pp. 1–20). Philadelphia: Psychology Press.
Higgins, E. T. (1987). Self-discrepancy theory: A theory relating self and affect. Psychological
Review, 94, 319–340.
Higgins, E. T. (1996). Ideal, ought, and regulatory focus: Effect and motivation from distinct
pains and pleasures. In P. M. Gollwitzer & J. A. Bargh (Eds.), The Psychology of Action:
Linking cognition and motivation to behavior (pp. 91–114). New York: The Guilford
Press.
Higgins, E. T., Strauman, T., & Klein, R. (1986). Standards and the process of self-evaluation:
Multiple affects from multiple stage. In R. M. Sorrentino & E. T. Higgins (Eds.),
Handbook of motivation and cognition: Foundations of social behavior, Vol. 1 (pp. 23–63).
New York: Guilford Press.
Isen, A. M. (1984). Towards understanding the role of affect in cognition. In R. S. Wyer
& T. K. Srull (Eds.), Handbook of social cognition, Vol. 3 (pp. 179–236). Hillsdale, NJ:
Erlbaum.
Isen, A. M. (1987). Positive affect, cognitive processes, and social behavior. In L. Berkowitz
(Ed.), Advances in Experimental Social Psychology, Vol. 20 (pp. 203–253). San Diego,
CA: Academic Press.
Isen, A. M. (2000). Positive affect and decision making. In M. Lewis & J. M. Hoviland-Jones
(Eds.), A handbook of emotions (2nd ed.) (pp. 417–4351). New York: The Guilford Press.
Isen, A. M., Means, B., Patrick, R. & Nowichi, G. (1982). Some factors influencing decision-
making strategy and risk-taking. In M. S. Clark & S. T. Fiske (Eds.), Affect and
cognition: The 17th annual Carnegie Symposium on Cognition (pp. 243–261). Hillsdale,
NJ: Erlbaum.
Lang, J. P., Bradley, M. M., & Cuthbert, B. N. (1990). Emotion, attention, and the startle
reflex. Psychological Review, 97, 377–395.
Lang, J. P., Bradley, M. M., & Cuthbert, B. N. (1995). International affective picture systems
(IAPS): Instruction manual and affective ratings. Technical Report A-4, The center for
Research in Psychophysiology, University of Florida.
Lang, J. P., Bradley, M. M., & Cuthbert, B. N. (1997). Motivated attention: Affect,
activation, and action. In J. P. Lang, R. F. Simons, M. T. Balaban (Eds.), Attention
and orientating sensory and motivational processes (pp. 97–135). New Jersey: Lawrence
Erlbaum Associates.
Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm (1993). Looking at the pictures:
Affective, facial, visceral and behavioral reactions. Psychophysiology, 30, 261–273.
Lazarus, R. S. (1991). Emotion and adaptation. London: Oxford University Press.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.22 (1327-1415)

 F. Pahlavan and T. Lubart

Levenson, R. L. (1994). The search for autonomic specificity. In P. Ekman & R. J. Davidson
(Eds.), The nature of emotion: Fundamental questions (pp. 252–257). New York: Oxford
University Press.
Lewis, M., Alessandri, S. M., & Sullivan, M. W. (1990). Violation of expectancy, loss of
control, and anger expression in young enfants. Development Psychology, 26, 745–751.
Leventhal, H. (1993). A componential, self-regulative systems view of Berkowitz’s cognitive
neoassociationistic model of anger. In R. S. Wyer & T. K. Srull (Eds.), Advances in Social
Cognition, Vol. VI (pp. 135–146).
Martin, L. L. (1999). Moods do not convey information: Moods in context do. In J. P.
Forgas (Ed.), Feeling and Thinking: The role of affect in social cognition (pp. 153–177).
Cambridge: Cambridge University Press.
Martin, L. L., Seta, J. J., & Crelia, R. A. (1990). Assimilation and contrast as a function of
people’s willingness and ability to expend effort in forming an impression. Journal of
Personality and Social Psychology, 59, 27–37.
Martin, L. L., Achee, J. W., Ward, D. W., & Harlow, T. F. (1993). The role of cognition and
effort in the use of emotions to guide behavior. In R. S. Wyer & T. K. Srull (Eds.),
Advances in Social Cognition, Vol. VI (pp. 147–157).
McFarland, C. & Buehler, R. (1997). Negative affective states and the motivated retrieval
of positive life events: The role of affect acknowledgement. Journal of Personality and
Social Psychology, 73, 200–214.
Mikulincer, M. (1988). Reactance and helplessness following exposure to unsolvable
problems: The effect of attributional style. Journal of Personality and Social Psychology,
54, 679–686.
Moretti, M. M. & Higgins, E. T. (1999). Internal representations of others in self-regulation:
A new look at a classic issue. Social Cognition, 17, 186–208.
Österman, Bjorkqvist, Lagerspetz, Kaukianen, Landau, Fraçzek, & Caprara (1998). Cross-
Cultural evidence of female indirect aggression. Aggressive Behavior, 24, 1–8;
Pahlavan, F., Bonnet, P. & Duda, D. (2000). Human motor responses to simultaneous
aversive stimulation and failure on a valued task. Psychological Reports, Perceptual and
Motor Skills, 86, 232–242.
Pahlavan, F., Duda, D., & Bonnet, P. (2000). Direction of human motor responses by men
and women to aversive stimulation. Psychological Reports, Perceptual and Motor Skills,
90, 415–422.
Parrot, W. G. & Sabini, J. (1990). Mood and memory under natural conditions: Evidence
for mood incongruent recall. Journal of Personality and Social Psychology, 59, 321–336.
Pylyshyn, Z. (1999). What’s in your mind? In E. Lepore and Z. Pylyshyn (Eds.), What is
cognitive Science? Malden, Massachusetts: Blackwell Publishers Ltd.
Roberts, T. (1991). Gender and influence of evaluations on self-assessments in achievement
settings. Psychological Bulletin, 109, 297–308.
Schwarz, N., & Bless, B. (1992). Constructing reality and its alternatives: An inclu-
sion/exclusion model of assimilation and contrast effects in social judgment. In L. L.
Martin & A. Tesser (Eds.), The construction of social judgments (pp. 165–193). Hillsdale,
NJ: Erlbaum.
JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.23 (1415-1486)

Affective perception 

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution and judgments of well-being:
Informative and directive functions of affective states. Journal of Personality and Social
Psychology, 45, 513–523.
Schwarz, N. (1990). Feelings as information: Information and motivational functions of
affective states. In E. T. Higgins & R. M. Sorrentiono (Eds.), Handbook of motivation
and cognition: Foundations of social behavior, Vol. 2 (pp. 527–561). New York: Guilford
Press.
Sinclair, R. C., & Mark, M. M. (1992). The influence of mood state on judgment and action:
Effects on persuasion, categorization, social justice, person perception, and judgmental
accuracy. In Positive affect, cognitive processes, and social behavior. In L. L. Martin
& A. Tesser (Eds.), The construction of social judgments (pp. 165–193). Hillsdale, NJ:
Erlbaum.
Smith, E. R., & Branscomb, N. R. (1987). Procedural mediated social inferences: the case of
category accessibility effects. Journal of Experimental Social Psychology, 23, 361–382.
Weiner, B., & Kukla, A. (???). An attributional analysis of achievement motivation. Journal
of Personality and Social Psychology, 15, 1–20.
Weiner, B. (1982). The emotional consequences of causal attribution. In Clark M. C. & Fiske
S. T. (Eds.), Affect and cognition. Hillsdale, NJ: Erblaum.
Wilson, T. D., & Brekke, N. (1994). Mental contamination and mental correction: Unwanted
influences on judgments and evaluations. Psychological Bulletin, 116, 117–142.
Winkielman, P., Zajonc, R. B., & Schwarz, N. (1997). Subliminal affective priming resists
attributional intervention. Cognition and Emotion, 11, 433–465.
Wortman, C. B., & Brehm, J. W. (1975). Responses to uncontrollable outcomes: An
integration of reactance theory and the learned helplessness model. In L. Berkowitz
(Ed.), Advances in Experimenyal Social Psychology, Vol. 8 (pp. 278–336). New York:
Academic Press.
Zillmann, D. (1983). Arousal and aggression. In R. G. Geen & E. I. Donnerstein (Eds.),
Aggression: Theoretical and empirical review, Vol. 1 (pp. 75–102). New York: Academic
Press.
 JB[v.20020404] Prn:4/11/2004; 16:36 F: Z13103.tex / p.24 (1486-1582)

 F. Pahlavan and T. Lubart

Annex 1

Order of presentation of the photos. IAPS code number, their valence, order of pre-
sentation of affective dimensions, and means and standard deviations according
to American norms:
Order. Description. Code Valence Dim Order Valence. Arousal. and Dominance Means and SD
1-Revolver (6230) U DAV : 2.37 (1.57). 7.35 (2.01). 2.15 (2.09)
2-Erotic woman (4180) P AVD: 6.21 (2.57), 5.54 (2.89), 5.44 (2.89)
3-Baby with tumor (3170) U DVA: 1.46 (1.01), 7.21 (1.99), 2.70 (1.89)
on face
4-Erotic man (4490) P VAD: 6.27 (1.95), 6.06 (2.42), 5.03 (1.91)
5-A silhouette under (9210) N DVA: 4.53 (1.82), 3.08 (2.13), 4.55 (1.90)
the rain
6-Mutilated woman’s face (3030) U AVD: 1.91 (1.56), 6.76 (2.10), 3.69 (2.10)
7-Baby’s face (2050) P VDA: 8.20 (1.31), 4.57 (2.53), 7.71 (2.53)
8-Old man in pajamas (2520) N DVA: 4.13 (1.90), 4.22 (1.69), 4.44 (2.33)
9-Mutilated woman’s body (3140) U DVA: 1.83 (1.17), 6.36 (1.97), 3.20 (2.17)
10-An empty basket (7010) N DVA: 4.94 (1.07), 1.76 (1.48), 6.70 (1.48)
11-Snicker bar (7430) P AVD: 7.11 (1.78), 4.72 (2.29), 5.86 (2.02)
12-Mutilated man’s face (3080) U VAD: 1.48 (0.95), 7.22 (1.97), 2.85 (2.10)
13-Erotic scene (4800) P ADV: 6.44 (2.22), 7.07 (1.78), 5.51 (2.11)
14-A pit bull (1300) U DAV: 3.55 (1.78), 6.79 (1.84), 3.49 (2.10)
15-Man with burns. in bed (3100) U ADV: 1.60 (1.07), 6.49 (2.23), 3.00 (2.16)
16-An umbrella (7150) N DAV: 4.72 (1.00), 2.61 (1.76), 5.55 (2.01)
17-French fries (7460) P DAV: 6.81 (2.08), 5.12 (2.49), 5.78 (2.26)
18-Women at the exit
of a church (2700) N ADV: 3.19 (1.56), 4.77 (1.97), 4.44 (2.04)
19-Young man’s face (2200) N AVD: 4.79 (1.38), 3.18 (2.17), 5.44 (2.17)
20-Bearded man’s face (2210) N DAV: 4.70 (0.93), 3.08 (1.76), 5.23 (1.78)
21-Elderly woman (2590) N DVA: 3.26 (1.92), 3.93 (1.94), 4.31 (2.14)
22-Male erotic (4520) P VDA: 7.07 (1.64), 5.48 (2.27), 5.48 (1.58)
23-Woman’s face (2130) N VDA: 4.08 (1.33), 5.02 (2.00), 5.10 (2.00)
24-Naked couple (4690) P AVD: 6.83 (1.94), 6.06 (2.21), 6.12 (2.18)
25-Gnnet (neutral
background) (1450) N VAD: 6.37 (1.62), 2.83 (1.87) 6.75 (1.87)
26-Man with blood (3010) U DAV: 1.71 (1.19), 7.16 (2.24), 2.88 (2.41)
27-Couple in a cemetery (9220) U DVA: 2.06 (1.54), 4.00 (2.09), 3.13 (1.97)
28-Female Erotic (4220) P ADV: 8.02 (1.93), 7.17 (2.69), 5.33 (2.12)
29-Dirty toilettes (9300) U ADV: 2.26 (1.76), 6.00 (2.41), 4.12 (2.59)
30-A rabbit (1610) P ADV: 7.82 (1.34), 3.08 (2.19), 6.77 (2.19)
Note: V = valence; A = arousal; D = dominance; U = unpleasant; N = neutral; P = pleasant
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.1 (41-107)

Neural development
Affective and immune system influences

George F. R. Ellis* and Judith A. Toronchuk**

University of Cape Town, South Africa / Trinity Western University,
Canada

This paper proposes that the developmental processes of Edelman’s neural

Darwinism fit together in a very coherent way with the present increasing
understanding of the importance of the affective dimension in neuroscience.
A synthesis of these two features, with the evolutionarily determined primary
affective systems together with the immune system providing the value
system required by neural Darwinism, provides an integrative viewpoint
relating psychological issues at the macro level to neurobiological processes
structuring neuronal connections at the micro level. We look at the various
implications of such an integrative viewpoint relating genetically determined
affective systems to higher cortical functions, considering successively
developmental and functional issues, primary and secondary emotions,
psychological issues, evolutionary issues, language, genetic issues,
neurological issues, and potential outcomes of the proposal. We suggest that
the “wet-wiring” nature of neurotransmitter mediated synaptic connections
may be related to this integration. We then consider the implications of
molecularly based links between the brain and the immune system, showing
this too might play a significant role in the processes of neural Darwinism.
Indeed this could possibly relate to the evolutionary origin of
affective systems.

. Introduction

Two recent contributions have advanced understanding of brain function: Ger-

ald Edelman’s “Neural Darwinism” (Edelman 1989, 1992; Edelman & Tononi
2001), dealing with how brain development and function can be well under-
stood in terms of a process of natural selection applied to neural connections,
and Jaak Panksepp’s formulation of “Affective Neuroscience” (Panksepp 1998,
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.2 (107-145)

 George F. R. Ellis and Judith A. Toronchuk

2001), addressing how neurobiological systems mediate the basic emotions.1

We point out here that these theories can complete each other in a very satisfac-
tory way, providing a synthesis which strengthens support for each of the the-
ories individually, and also provides an extended understanding of important
interactions in the brain. In brief: on the one hand, major features of the basic
value system crucial to Edelmann’s neural Darwinism but not fully elucidated
by him can be provided by the affective neuroscience of Panksepp. On the other
hand, important aspects of the mechanism implementing Panksepp’s proposal
that “valenced affective feeling states provide fundamental values for the guid-
ance of behavior” can be explicated by emphasizing neural Darwinism in a way
that more fully accentuates current understanding of developmental biology.
This proposed synthesis (which might perhaps be called “Affective Neural
Group Selection”) then gives a useful standpoint from which to investigate the
relations between affective neuroscience and neural Darwinism, and to con-
sider aspects of developmental and evolutionary psychology. In this synthesis
the emotional neural systems in the brain play an important role in selection
of higher cortical connections. We present our proposal in terms of a central
hypothesis which then leads to a series of conjectures that flesh it out, these
in turn leading to a series of detailed questions that are susceptible to devel-
opment and investigation. We indicate what some of the latter are, but do not
have time or space to develop them in detail here. However this overall struc-
ture does indeed show that our central hypothesis is a fruitful one which can
lead to interesting further investigations.
The second theme of this paper addresses the role of immune system-brain
interactions in such neuronal selection, suggesting these links too can be inter-
preted as playing a significant role in neural group selection processes. It is well
understood that not only are there causal links from the brain to the immune
system (Sternberg 2000; Webster et al. 2002; Kiecolt-Glaser et al. 2002) but
also from the immune system to the central nervous system (Mulla & Buck-
ingham 1999; Maier & Watkins, 2000; Larson 2002; Maier 2003). In addition,
it is suggested here that immune mechanisms might have been significant in
the evolutionary emergence of the hierarchy of signals that result in neural
Darwinism. This part of the paper takes our initial theme and extends it in
a consistent way to further areas.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.3 (145-209)

Neural development 

. Neural Darwinism

Edelman argues that generalised principles of Darwinian natural selection

(“Neural Darwinism”) must apply in the developmental process controlling
detailed neural connections in each individual’s brain (Edelman 1989, 1992;
Edelman & Tononi 2001).2 The theory3 has three main elements:
1. Developmental selection,
2. Experiential selection,
3. Re-entry
(Edelman 1989: 4–8; Edelman 1992: 81–98; Edelman & Tononi, 2001: 79–92).
The chemical gradients of molecules, which in some cases are the same as those
used by the immune system, steer neurons in the generally right direction and
these molecules may also prevent pre-programmed apoptosis. Competition for
synapses then weeds out those which receive less use. In other words gen-
eral tuning occurs by chemical gradients, but epigenetic experience provides
specific tuning of those pathways.
The key feature that concerns us here is that, after developmental processes
establish a great variety of connection patterns between neurons, a process of
synaptic selection occurs within neuronal groups as a result of behavioural
experiences that modify affective states. These changes occur because certain
synapses are strengthened and others weakened without macroscopic changes
in the anatomy, although there might be microscopic changes such as per-
forated synapses or changes in dendritic spines. “This selectional process is
constrained by brain signals that arise as a result of the activity of diffusely
projecting value systems, a constraint that is continually modified by success-
ful output” (Edelman & Tononi 2001: 84; see also Deacon 1997: 202; Schore
1994: 162, 253, 257).4 An example of a value system is the noradrenergic sys-
tem, originating in the locus coeruleus and projecting diffusely to the en-
tire brain, releasing norepinephrine. The unit of selection is neuronal groups
(Edelman 1989: 43–69; Edelman 1992: 95–99). The “groups” that are selected
experientially are those neurons which participate in ongoing, reciprocal sig-
nalling between widely separated regions of the brain (forming “re-entrant
networks”). This parallel interchange allows for coordination and synchronous
firing in the individual components of the neural group enabling coherent
coordinated output. Edelman suggests that this re-entrant firing provides the
basis for both learning and consciousness.
This argument extends the Darwinian type of understanding from the evo-
lutionary processes that historically led to the existence of the brain and shaped
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.4 (209-250)

 George F. R. Ellis and Judith A. Toronchuk

the genotype to also underpinning both brain developmental processes and

brain functioning, thus directly affecting the phenotype. This is in accord with
the way that such processes are now understood to underlie the functioning of
the immune system through clonal selection5 (Burnet 1959; Edelman 1992: 77–
78). Thus such principles are already known to occur in human physiological
functioning in the immune system, giving the same benefits as discussed here:
putting in place a mechanism that can deal efficiently with conditions already
encountered, but that can also deal adequately with situations that have never
before been encountered by the organism. Through this mechanism, “In a very
literal sense, each developing brain region adapts to the body in which it finds
itself ” (Deacon 1997: 205). Thus this provides a neural mechanism underlying
traditional learning theory, also imbedding it in the broader understanding of
the adaptive power of Darwinian selection processes, seen as one of the most
important mechanisms in biology.

. The affective connection

The key issue then is what provides the fitness characterisation determining
whether particular connections are strengthened or not. In Edelman’s terms,
this is the value system guiding the neural Darwinism, which he relates phys-
ically to a fan of connections spreading out from a relatively small number
of monoaminergic, cholinergic, and histaminergic neurons located in various
brainstem and hypothalamic nuclei (see Edelman & Tononi 2000: 46). It is pro-
posed in this paper that the signals provided by the set of primitive emotional
functions described by Panksepp (1998, 2001) are the key signals in the value sys-
tem guiding neural selection. This would tie brain functioning to vital capacities
developed by evolutionary processes, strongly related to survival, and giving a
specific set of mechanisms to implement Panksepp’s hypothesis that “affect is a
central organizing process for sentience” (Watt 1999), namely diversity of the
primary repertoire, selection based on epigenetic modification of synaptic con-
nections resulting in a secondary repertoire, and re-entrant signalling based on
reciprocally connected neural maps (see Edelman 1989).
Panksepp presents in his work a careful neurologically based taxonomy
of basic emotional processes, each related to specific neurotransmitters and
associated with activity in specific subcortical brain areas. These are the evo-
lutionary heritage we share with many members of the animal kingdom. They
play a fundamental role in human behaviour: “the basic emotional states pro-
vide efficient ways to mediate categorical types of learned behavioural changes.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.5 (250-369)

Neural development 

... emotional feelings not only sustain certain unconditioned behavioural ten-
dencies but also help guide new behaviours by providing simple value coding
mechanisms that provide self-referential salience, thereby allowing organisms
to categorize world events efficiently so as to control future behaviours ... [they]
may provide efficient ways to guide and sustain behaviour patterns, as well
as to mediate certain types of learning” (Panksepp 1998: 14–15). That seems
just what is required to explicate in detail the value system needed by neural
Darwinism (Edelman & Tononi 2001: 87–90). Emotions, in this context, are
then the core of pre-organized mechanisms which “help the organism classify
things or events as ‘good’ or ‘bad’ because of their possible impact on survival”
(Damasio 1995: 117).
The basic emotional systems identified by Panksepp (1998) are the follow-
ing:
E1: The SEEKING system: general motivation, seeking, expectancy
(pp. 52–54, 144–163).
E2: The RAGE system: rage/anger (p. 54 and pp. 187–205).
E3: The FEAR system: fear/anxiety (p. 54 and pp. 206–222).
E4: The LUST systems: lust/sexuality in the male and female (p. 54,
pp. 225–245).
E5: The CARE system: providing parental care/nurturance (p. 54,
pp. 246–260).
E6: The PANIC system: panic/separation, need of care (p. 54, pp. 261–
279).
E7: The PLAY system: rough-housing play/joy (pp. 280–299).
Panksepp gives a detailed characterisation in each case, including associated
key brain areas and neurotransmitters (for a summary, see Panksepp 2001: 147;
and Table 1 below). These systems – processing sensations such as heart rate,
palpitations, breathing changes, vasoconstriction etc, and leading to activity in
the cortical regions – embody the basic preferences or biases of the organism
that guide its further behavioural development.
These basic emotional systems underlie the higher-level systems that de-
velop in the brain (Panksepp 1998: 300–323). Various inputs to the SEEKING
system to do with thermal balance, hunger, thirst, sexual arousal, etc. enable
it to provide the basis of maintaining homeostasis. This system also drives the
basic impulse to search, investigate, and make sense of the environment. The
foundation of learning is then provided by satisfaction or dissatisfaction as-
sociated with the success or failure of one’s endeavours as motivated by the
SEEKING system. Presumably as brain development takes place this underlies
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.6 (369-378)

 George F. R. Ellis and Judith A. Toronchuk

Table 1. Some of the basic emotional systems together with their associated brain
areas and key neuromodulators. Serotonin and norepinephrine, which have more non-
specific effects, are omitted, as are higher cortical areas. Derived from Panksepp (2004)
and Watt (1999)

System Neural Networks Neurotransmitter Aspects

Seeking ventral tegmental area Dopamine (+), Appetitive states,
(VTA), lateral glutamate (+), anticipatory
hypothalamus, nucleus neurotensin (+), eagerness,
accumbens, PAG, diffuse Opioids (+), other exploration,
mesolimbic and neuropeptides (seeks non-
mesocortical outputs, specifically)

Rage medial amygdala, BNST, substance P(+), Ach Anger,

medial and perifornical (+), glutamate(+) frustration,
hypothalamus, dorsal PAG hot aggression

Fear lateral and central Glutamate(+), DBI, escaping danger

amygdala, medial CRH, CCK,
hypothalamus, dorsal PAG alpha-MSH, NPY

Lust Amygdala, BNST, preoptic Steroids (+), pleasure,

and ventromedial vasopressin, oxytocin, gratification
hypothalamus, basal LH-RH, CCK
forebrain, ventral PAG

Panic Dorso-medial Opioids(–), oxytocin separation

diencephalon, PAG (–), prolactin (–), CRH distress, social
bonding

Caring Anterior cingulate, BNST, oxytocin (+), prolactin Nurturance,

preoptic hypothalamus, (+), dopamine, opioids maternal care
VTA, ventral PAG

Play Dorso-medial Opioids (+ in small Play, joy, social

diencephalon, PAG amounts, –in larger affection
amounts)
Key: PAG = periaqueductal gray, BNST = bed nucleus of the stria terminalis, VTA = ventral tegmental area; CCK =
cholecystokinin, CRH = corticotrophin releasing hormone, DBI = diazepam binding inhibitor, LH-RH = leutenizing
hormone release hormone MSH = melanocyte stimulating hormone, NPY = neuropeptide Y.

the development of systems that carry out specific tasks to aid these functions,
in particular systems that anticipate what may happen by means of some kind
of modelling of the external world, including anticipation of the behaviours
of others.
The explicit hypothesis we make is
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.7 (378-440)

Neural development 

Hypothesis: The basic emotional systems E1–E7 identified by Panksepp, to-

gether with inputs from the endocrine and immune systems, are necessary
and sufficient to provide the value system of neural Darwinism identified by
Edelman and Tononi.

This makes explicit the way in which emotions underlie rationality, extending
the understanding of the significance of emotions beyond that recognised by
Damasio, namely
(i) the production of a specific reaction to an inducing situation,
(ii) the regulation of the internal state of the organism so that it can be pre-
pared for a specific reaction (Damasio 2000: 53–56),

to also including
(iii)shaping of brain functionality, including the aspects classified as rational.

This proposal agrees with his statement that “emotions are curious adapta-
tions that are part and parcel of the machinery with which organisms regulate
survival” (Damasio 2000: 54). They do so both in the short term through facil-
itating homeostasis, and in the long term through facilitating the development
of intellect (Damasio 2003). In this way “all mammals, indeed all organisms,
come into the world with a variety of abilities that do not require previ-
ous learning but which provide immediate opportunities to learn” (Panksepp
1998: 25).
Thus what is proposed here can provide a neural mechanism underlying
behavioural development and learning, and is therefore supported by all the
body of evidence that emotion is developmentally and functionally important.
In genetic terms, “the genome helps set the precise or nearly precise struc-
ture of a number of important systems and circuits in the evolutionarily old
sectors of the human brain” (Damasio 1995: 109); these innate circuits then
“intervene not just in bodily regulation but also in the development and adult
activity of the evolutionarily modern structures of the brain .... whose precise
arrangement comes about under the influence of environmental circumstances
complemented and constrained by the influences of the innately and precisely
set circuits concerned with biological regulation” (Damasio 1995: 109; Dama-
sio 2003: 30–59). It is through this process that the visual, auditory, sensory,
and motor systems all contribute to the development of the central nervous
system as a whole.
The present proposal is that this happens through the mechanism of affec-
tive neural Darwinism outlined above. Furthermore, it is suggested that apart
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.8 (440-485)

 George F. R. Ellis and Judith A. Toronchuk

from immune and endocrine system inputs discussed below, the seven basic
emotional systems E1–E7 together with sensory input and life-experience are
sufficient to sculpt the developing brain; there are no other unidentified sectors
of the value system. This means that when taken together with physiological
needs, they are sufficient to underlie any system of “human universals” such as
that proposed by Brown (1991).6 Of course such broad concepts are hiding a
vast number of essential details, and it is mainly in the explication of the details
that scientific progress must be measured. It is this assumption of sufficiency
that gives this proposal bite and enables it to be tested in a variety of ways dis-
cussed in the next section, and allows it to help guide the required explication
of the details.

. Implications of affective neural Darwinism

What are the benefits of this view? From the viewpoint of neural Darwin-
ism, it provides a crucial link between macro-behavior and the formation of
synaptic connections by identifying the macro-features of the value system in
psychological terms. This system is characterised in functional terms in Edel-
man & Tononi 2001: 87–90, but not related there to specific macro-features
of behaviour. This identification clarifies the causal nature of that system in
relation to individual development and function. From the viewpoint of the
affective neurosciences, it is largely unsurprising because it is just a slightly
more specific version of what some already believe (e.g. Panksepp 2001b); it
emphasizes explicitly that the effects of primary affective states include adap-
tive selection of neuronal groups as well as their immediate effect on neuronal
functioning. However although the affective viewpoint has been substantiated
by recent studies (e.g. Panksepp 2003c), it is nevertheless still contested and not
gaining as much ground as one might have expected (Panksepp 2002, 2003b).7
The main benefit here may well be in supporting that view by indicating how
naturally it fits with a neural Darwinism viewpoint.
What features might be expected and so could serve as a test of this pro-
posal? It serves as a unifying theme that will ultimately be tested by how well
it links neurological processes on the one hand and psychological issues on the
other, tying them in to evolutionary and developmental processes. That is, the
test is how effective it is as an integrating theme suggesting how neurological
processes may be tied in to macroscopic aspects of consciousness.
On this view, the primary emotions E1 to E7 characterised above become
the lynch-pin linking neurophysiology to experience and the social and phys-
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.9 (485-547)

Neural development 

ical environment. They link macro-events to neural micro-structure by top-

down action from the macro to the micro scale. Consequently they are a key
both to brain physiological development and to evolutionary development of
secondary emotions and higher cognitive functions. This should play itself out
in terms of both evolutionary psychology and developmental theory, and hence
in terms of psychiatry (relating both to physical and developmental disorders)
and learning theory. In particular it should manifest in terms of behavioural al-
terations associated with physiological changes in the primary emotional areas,
and in terms of commonalities with our animal relatives. This section outlines
how that might work out.

. Developmental and functional issues

The key factor underlying brain development is the fact that the stored infor-
mation in the human genome is far too little to control brain development by
itself. The Human Genome Project has revealed (Baltimore 2001; Wolfsberg
et al. 2001) that there are of the order of 45,000 genes in the human genome;
but there are about 1013 cells in the human body and 1011 neurons in a human
brain. Consequently – remembering that this genetic information has to cover
development of all other bodily structure as well as the brain – there is not a
fraction of the information required to structure in detail any significant brain
modules, let alone the human brain as a whole (Damasio 1995: 108), for there
is only one gene per 107 neurons and per 109 neural connections.
Human development has to take place in the face of these genetic limita-
tions, and this is a key reason why neural Darwinism is necessary as an effective
mechanism structuring the brain and adapting it to both the local molecu-
lar and encompassing social environments. Neural Darwinism also provides
individual organisms the advantage of flexible neural development toned to
individual somatic conditions such as differences in the size of individual mus-
cles. We can make this specific through the following set of conjectures that
flow out of our basic hypothesis:
Conjecture 1: the more hard–wired parts of the CNS are the spinal cord,
brain stem and diencephalon together with its sensory inputs while the cere-
bellum and cortical structures are genetically determined only in terms of
basic neuronal structure and broad connections between different brain ar-
eas. The detailed connectivity in the cerebellum and telencephalon are deter-
mined by neuronal group selection.8
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.10 (547-584)

 George F. R. Ellis and Judith A. Toronchuk

By hard-wired, we mean neuronal connections are determined largely by genes

read in the local molecular context, and therefore susceptible to immune and
endocrine influences during early development; however, these connections
are thereafter fixed except that their strengths may vary according to use in
a standard Hebbian way. The chemical milieu in which neurons find them-
selves early in development can play a role in organization of the brainstem
emotional systems, but thereafter these connections are, relative to those in
telencephalic structures, more stable and fixed. The reason for this is that it
is important that the brainstem structures should not be subject to constant
change, as they themselves provide the value system for later emotional devel-
opment; there is no other value system, by hypothesis. Many of these structures
also exhibit use-dependent plasticity, as mentioned above, but by itself this
does not guide development in a way dependent on the interplay between
a value system and experience; rather it plays a supporting role in such de-
velopments. The situation would become dynamically unstable, going into
unbounded oscillations or exponential instability, if neural Darwinism applied
to the connections determining the value system through the values implicit
in that system itself. These structures must under most circumstances be fixed
and unchanging in order to provide long-term stability; thus they must be rela-
tively free of perturbation. Nevertheless extreme environmental circumstances
during development may provoke dysfunctional changes in the value system.
Conjecture 2: Consequently, the only genetically specified brain functions
are the unconscious basic regulatory capacities plus instinctive drives to-
gether with the basic emotional systems and basic sensory-motor pathways.
No higher cognitive capacities are hard-wired: they are determined by de-
velopmental interaction with the internal and external environments in
conjunction with the internal (self-referential) effects of the brain on itself.
Conjecture 3: The secondary emotions and higher cognitive capacities de-
velop under the guidance of the value system provided by the basic emotional
systems together with broad information from the immune and endocrine
systems. Consequently these higher capacities are shaped by the dimensions of
the basic emotional systems, which are the filter through which environmen-
tal effects and experience are evaluated and used to guide brain development.

On the view put here, there is a clear distinction between primary and sec-
ondary emotions. The primary emotions are affective states arising primarily
from subcortical brain structures (the amygdala, basal forebrain, hypothala-
mus, and nuclei of brainstem tegmentum) that we share with other mammals.
On the other hand secondary emotions arise through the effects of the primary
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.11 (584-626)

Neural development 

emotions on cortical regions (ventromedial prefrontal, cingulate and insular

cortex) in the course of daily life and social interaction. In brief: primary
emotions are hard-wired during early development, secondary emotions are
soft-wired products of experience. Both may give rise to conscious awareness.
In the main, secondary emotions correspond to social emotions developing
out of the primary emotions with a social character (CARE and PANIC, also
to some extent LUST and PLAY); this issue is discussed below. Note that some
secondary emotions may also occur in other animals, developing through their
social interactions.
The key question underlying these conjectures is what use is made of the
very limited number of genes available for brain development. The view here
is that this genetic information is used to hard-wire the brainstem and dien-
cephalic structures, as outlined above, and to soft-wire the rest. Higher level
brain development will be only generically set genetically, and then will be de-
termined in detail by neural Darwinism in response to life experience, with
the value system guiding that process provided by the less malleable primary
emotional systems E1–E7 as listed above.
Intrinsic personality differences will then depend on the relative strengths
of these seven affective systems, which are sufficient to determine development
of higher-level aspects of the brain.9 One can in cartoon fashion imagine a
genetic genie together with the early developmental environment setting the
values of the relative strengths of the systems E1–E7 for each individual on
the dials of an instrument panel; these relative strengths are then fixed for life,
determining that individual’s basic emotional type. This is the initial mate-
rial which then interacts with the physical and social environment to develop
the adult personality (Donald 2001); hence the systems E1–E7 provide the
emotional palette that guides brain development.
How this works out will depend on the manner in which the social envi-
ronment responds positively or negatively to these basic emotional capacities.
Different personality development, in terms of secondary emotion and cog-
nitive development, will occur according to the relative strengths of the basic
systems E1–E7 on the one hand and the responsiveness and emotional tenor
of the environment in each of these dimensions on the other. An analogy is
helpful here: consider colour vision. Because of the genetically determined
hard-wiring of the long, medium, and short wavelength photoreceptor sys-
tems, colour vision is determined by the relative sensitivities of these systems.
The possible nature of colour vision disorders follows from this: there are seven
possible types of colour blindness depending on which of these primary colour
systems is impaired (with disorders related to non-functionality of a single
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.12 (626-683)

 George F. R. Ellis and Judith A. Toronchuk

colour system being more likely than those with two or three of them being
dysfunctional).10 These can be tested for and related to genetic disorders.
If the view proposed here is correct, the pallet of primary emotions char-
acterised above similarly determines the different kinds of emotional disorders
that may occur for genetic reasons. Further, vision is determined developmen-
tally as well as genetically: if a person or animal were kept in an environment
where some colours never occurred (for example, they were kept continuously
in a room illuminated only by very long wavelength light so that it avoided
stimulating one set of receptors) the associated visual ability would presumably
decay, and they would at later stages be unable to see the colours that had been
omitted in their developmental environment. This suggests that analogous, al-
though much more complex, emotional disorders could arise for individuals
brought up in emotionally stunted environments. The possible natures of these
disorders would again be determined by the pallet of primary emotions.
In functional terms, this whole setup may be regarded as a feedback control
system, using genetically determined inherited goals to guide the development
of the organism in accord with information implicit in those goals about sur-
vival needs in the world in which the organism lives. This kind of use of
information in feedback control systems is characteristic of the way that higher
level order emerges in hierarchically structured complex systems (Ellis 2003).
The information is mainly used to attain equilibrium of the organism in its
environment, that is, the process is one of homeostasis. According to Milsum
(1966), all human physiological systems can be regarded as means to attain-
ing homeostasis. This theme is presented in a different way by Damasio, who
emphasizes homeostatic regulation carried out by a multibranched affective
tree (Damasio 2003: 30–59). The bottom level consists of metabolism, basic re-
flexes, and immune responses. The next level is pain and pleasure behaviours
associated with reward and punishment - the first stage of the “value” system.
The next level is drives and motivations including play, exploration, sex: also
part of the value system. Then he has “emotions-proper” such as fear, anger,
shame, guilt, and at the top “feelings”, a term which he reserves for qualia-like
conscious awareness of emotional states, but he emphasizes that all of these
levels (including emotions and feelings) are regulators of homeostasis. This is
basically in accord with our view; we address the relation of these higher levels
to the value system in our discussion of secondary emotions below.
From our viewpoint, we have a causal hierarchy characterised in an ex-
tremely simplified form in Figure 1. The emotional system sends information
on the affective state of the organism to the cortical areas, thereby influencing
the way neurons fire in the cortex and helping shape the outcome of current
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.13 (683-736)

Neural development 

Cognition, feeling
secondary emotions

Primary emotional
neural systems

Immune
system

Figure 1a. The basic set of interactions between the higher level cognitive functions,
the emotional system, and the immune system, go both ways. They involve immediate
(physiological) effects, longer term (developmental) effects, and very long term (evo-
lutionary) effects. The suggestion here is that the latter two proceed by basically the
same evolutionary mechanism, albeit operating on very different timescales. Thus in
broad brush terms the immune system underlies the emotions, which in turn underlie
intellect.

neuronal operations there; thus they inform the cognitive functions and mod-
ify their outputs. This is bottom-up action from the emotional system to the
cortex. Similarly the higher cortical areas send information on the way the
higher brain views the situation to the emotional system, thus informing it and
modifying its outputs; this is top-down action from the cortex to the emotional
system. However additionally, through the value system, bottom-up action by
the emotional system helps shape the network of neuronal connections in the
cortex; it is crucial that there is no converse downwards effect from the cortex
to the emotional system.
We may summarise by saying that emotion and intellect interact with each
other operationally, but emotion shapes intellect developmentally. In short:
emotion underlies intellect. This hypothesis can be tested overall by extending
the range of investigations of factors underlying human cognitive abilities (Car-
roll 1993) to include the relative strengths of the primary emotional systems
E1–E7 as outlined above. This may help illuminate the key unresolved issue of
why some people are cleverer than others (Deary 2001, Chapter 3).
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.14 (736-745)

 George F. R. Ellis and Judith A. Toronchuk

Neocortex

amygdala hippocampus

hypothalamus

pituitary

Autonomic
HPA axis nervous system

Adrenal gland

Immune system

Figure 1b. More detailed view of the basic set of interactions between the higher level
cognitive functions, the emotional system, and the immune system.

. Primary and secondary emotions

A key point is the relation between primary and secondary emotions (Ekman
& Davidson 1994: 5–48; Damasio 2003). On the view put forward here, sec-
ondary emotions would arise through the effects of the primary emotions on
the cortex in the course of social interaction, the primary emotions based in
subcortical limbic structures being our genetic heritage from our animal for-
bears. The secondary emotions are not part of the value system, as they don’t
originate in the same loci as the primary emotions. They are however indirectly
able to be effective in determining neural connections through the mechanisms
of neural Darwinism, because of their powerful ability to affect the state of pri-
mary emotions, which then act as the value system. In this way the secondary
emotions can modulate both the immediate state of the primary emotions, and
their longer-term effects through sculpting neuronal connections.
It is clearly crucial to clarify which are secondary and which are pri-
mary emotions. As an example, Damasio (2003: 44) suggests universal primary
emotions are,
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.15 (745-804)

Neural development 

P1. happiness, P2. sadness, P3. fear,

P4. anger, P5. surprise, P6. disgust,
and (pp. 45 and 156) characterises developmentally emergent secondary emo-
tions as,
S1: embarrassment, shame, guilt;
S2: contempt, indignation;
S3: sympathy, compassion;
S4: awe/wonder/elevation, gratitude, pride;
S5: jealousy, envy.

Because of animal behaviour patterns (Manning & Dawkins 1998: 395–401;

Slater 1999: 200–204), there is a good case to add to these,
S6: social rank/dominance,
S7: belonging/exclusion.

Now these proposals may be compatible with our functional distinction be-
tween primary and secondary emotions, but this needs testing in each case;
where there is a disagreement, we suggest our functional distinction should
take precedence because of its grounding in neurostructure. Furthermore the
classification of secondary emotions is not settled, and one can suggest that
they will be universal in humanity, because of the universal nature of the devel-
opmental experiences leading to their existence. To what degree they are shared
by higher primates and other animals is a fascinating subject for further study.
The point is that it isn’t 100% clear which ones should be considered pri-
mary and which are derivative. Others would have different lists than those
of Damasio given above.11 The need is to come up with a coherent classi-
fication that makes sense in the light of the mechanisms proposed here and
also relates sensibly to Panksepp’s list E1–E7, which is not the same as the list
P1–P6 above. Those listed as primary must be sufficient to underlie develop-
ment of all present day intellectual and emotional capabilities, including the
secondary emotions, as outlined in Conjectures 1 to 3 above, and should also
be related to activity in specific subcortical structures and neurotransmitters
utilized in their activation.12 Thus a revised categorisation of primary and sec-
ondary emotions, and a characterisation of which primary emotions are most
important for developing which secondary ones,13 is an important task.
The outcome of all this should be either a confirmation that Panksepp’s
list E1–E7 is indeed adequate to serve as the basis of higher emotional and
cognitive development, or a proposal for further additions to that list in order
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.16 (804-849)

 George F. R. Ellis and Judith A. Toronchuk

that it is indeed sufficient. There could for example be a case for adding a basic
underlying emotional system
E0: PLEASURE/PAIN, an overall “state-of -the-system” assessment, re-
lated to overall happiness/welfare/unhappiness and corresponding to
Damasio’s emphasis of the importance of these overall aspects of emotion
(Damasio 2003: 32–34).

To make this good one would have to identify a corresponding physical lo-
cus and associated neurotransmitters. A possibility is that this dimension is
related to the noradrenergic system, originating in the locus coeruleus and
projecting diffusely to the entire brain (Kingsley 2000: 131; Edelman & Tononi
2001: 84), and closely related to the SEEKING system (Panksepp 1998). Further
elucidation is needed. Pleasure/pain may just be part of the seeking system, al-
though the way this affective dimension is characterised by Damasio differs
from Panksepp’s description of E1.
Reconsideration might decide that some of the primary emotions listed
by Damasio (see above) should be considered as additions to Panksepp’s list.
Alternatively the conclusions might be that some of them, for example P6: Dis-
gust in which the insular cortex (Wicker 2003) plays a major role, should rather
be classified as secondary emotions. Furthermore P5: Surprise seems to be a re-
flex rather then an emotion (it does not convey specific values suggesting any
specific actions are appropriate), and so should either be taken off the list of
primary emotions, or incorporated with E0.
Additionally it is plausible that in view of our animal forebears’ social
structures, some of the social emotions may in fact be primary (i.e. hardwired
through the evolutionary process and related to specific genes) because of their
high survival value, suggesting a re-classification of some of the secondary
emotions identified above as primary. This might specifically apply to S6: so-
cial rank/ dominance, which would then be re-classified as primary rather than
secondary, and possibly also to S1: embarrassment, shame, guilt. Alternatively
one would like a definite proposal as to how these can reliably emerge in each
individual from the primary emotions through the processes of social interac-
tion, for example based on E5: The CARE system and E6: The PANIC system,
which are basic emotions related to sociality and so could possibly provide the
basis needed for these to develop as secondary emotions.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.17 (849-908)

Neural development 

. Psychological issues

The developmental environment might be characterised in a similar way in

terms of its responsiveness and depth in these emotional dimensions; the
ultimately resulting personality will depend on the interaction between the
environment and the intrinsic capacities in each dimension.
Two theoretically possible kinds of experiments indicate how it would in
principle be possible to test the above hypotheses. They are,
Experiment 1 (structural effects): surgically, genetically or chemically in-
activate the various basic emotional systems E1–E7 in young children or
animals, and see what the developmental results are;
Experiment 2 (environmental effects): raise children or other mammals in
a milieu where they receive only cognitive rather than emotional stimula-
tion, and see what the developmental effects are.

In the latter case, one could have a set of four comparison groups:
E+ I+ , an emotionally and cognitively stimulating environment;
E+ I– , an emotionally stimulating but cognitively deprived environment;
E– I+ , an emotionally deprived but cognitively stimulating environment;
and
E– I– , an emotionally and cognitively deprived environment.

It is well known that in the latter case children will lag developmentally, will
lose weight, and may even die from lack of caring attention. For example Coe
and Lubach (2003) have recently reviewed the effects on the health of infant
monkeys and children of the interaction of immune and psychological factors
during separation. This matrix of comparison will determine which effects are
more significant, and the hypothesis will be that it is the emotional aspect that is
developmentally more important. One could further separate this out according
to emotional category by depriving the subjects separately of stimuli/responses
in each category E1–E7. Of course it might be impossible to deprive a child
in just one emotional category, or even to completely separate emotional and
intellectual stimuli, but still one could find environments where there were
sufficient differences in these dimensions to look for such effects.
Although these variables cannot be experimentally manipulated in hu-
mans, there are situations in which this has been effectively what has happened,
through either genetic defects or physiological trauma in the first case, and
through impoverished home or environmental circumstances in the second.
However this is not an all–or–nothing affair: one can create a causal matrix to
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.18 (908-967)

 George F. R. Ellis and Judith A. Toronchuk

test for degrees of such genetic and environmental effects, in particular testing
for various relative strengths of the basic emotional systems on the one hand,
and to what degree emotional/cognitive aspects are encouraged and responded
to in particular family/school/life situations, on the other. Thus a meta-analysis
could be carried out using studies already in the existing literature. (There is
a great deal of relevant modern animal information on such issues, includ-
ing neuroscience varieties. For an excellent review of the animal literature see
Panksepp 1998.)
The causal interactions here are of course immensely complex; neverthe-
less if the basic thesis proposed here is true, it should in principle be possible to
characterise the relative strengths of the basic systems E1–E7 by psychological
tests, and similarly to characterise the developmental environment along the
same dimensions, and then to relate these two causal features to psychological
tests of resulting secondary emotions and cognitive capacities, so determining a
causal matrix for these primary relations. The ultimate way to test the proposed
integration would be characterisation of developmental and functional disor-
ders resulting when the various brain regions corresponding to the systems
E1–E7 either suffer damage at an early age, or are developmentally defective ab
initio. These would be contrasted with physiological damage leading to impair-
ment of operations of the secondary emotions S1–S7, which are known to lead
to a series of functional disorders (Damasio 1995, 2003). This all clearly ties
in well to the concept of emotional intelligence (Goleman 1996) as well as the
importance ascribed to emotion in design issues by Donald Norman (2003).
This leads on naturally to studies of the relation to child developmental
issues (Schore 1994): what kinds of interactions in terms of the primary emo-
tions E1–E7 are conducive to sound child development, and what to patholog-
ical development? What social interventions may be useful as a result? A vast
amount has of course been written on this topic, and in particular Panksepp
(2001b) has already developed this theme to a considerable degree in the con-
text of affective neuroscience. Relating the sufficiency proposals made here to
that literature would be a useful exercise.
It might even be possible to develop a relation of adult psychological per-
sonality categorisation schemes to the basic factors E1–E7, perhaps suggesting
related psychological disorders and interventions (Panksepp, 2002). Clearly
that would be a very complex and demanding project; nevertheless the hypoth-
esis of the key developmental role of the basic emotions E1–E7 might be useful
in this context too. The Five Factor Model (FFM) is viewed by most contempo-
rary personality theorists as representing the best current model of personality
structure. Developed through factor analysis, it suggests that personality can
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.19 (967-1005)

Neural development 

be best described in terms of five basic dimensions: extraversion (E), emo-

tional instability (or neuroticism N), openness to experience (O), agreeableness
(A) and conscientiousness (C). Data collected in several distinct cultures from
both genders and various ages all yield five similar factors (e.g. McCrae & Costa
1997; Costa et al. 2001; McCrae 2001) suggesting that personality trait struc-
ture is universal. Other theorists have suggested that E and A form the axes of
an interpersonal circumplex characterising interpersonal behavior (Trapnell &
Wiggins 1990) on which they can be alternatively labeled affiliation and dom-
inance. If these five orthogonal factors are in fact universal, they would have
arisen either through genetic influence or experience common to all humans,
and it would be expected that they bear some relationship to primary emotions.
It might be that SEEKING provides a physiological basis for E for example, and
PLAY for O. PANIC and CARE together might contribute to A while RAGE and
FEAR provide aspects of N.
The FFM may provide a mechanism to characterize different basic person-
ality types according to the relative strengths of the basic emotional systems:
the ‘settings on the dial’ corresponding to each aspect E1–E7 of the basic emo-
tions mentioned above.14 In extreme cases, some of the settings might be either
zero or exceptionally high, with the corresponding basic emotion either miss-
ing, with corresponding personality disorders resulting (for example E3– is a
lack of emotion type E3), or pathologically dominant, with the opposite emo-
tional disorder emerging (for example E3+ is over-dominance of emotion type
E3). Injury might cause destruction of one or more of the basic emotional ca-
pacities, leading to similar emotional-lack disorders, the difference from the
previous ones being that these problems will in this case start at some specific
event in the individual’s history, rather than being there ab initio. The FFM has
already been shown in numerous studies to differentiate various personality
disorders (Lynam & Widiger 2001) and has indeed already been incorporated
in the Affective Neuroscience view (Davis et al. 2003; Reuter & Hennig 2003).
There is a clear relation of the above to learning theory in terms of explor-
ing emotional relations underlying learning success (in contrast to approaches
based on rational aspects alone). One might suggest here for example that it
is E1: The SEEKING system that is the primary driver in the learning of skills
and abstract understanding, but with the E7: The PLAY system also playing a
significant role, generating alternative possibilities that express creativity; and
consequently that these are the ones that would be most significantly correlated
with educational success; and indeed there is a large literature on the impor-
tance of play in the development of the mind (Bruner et al. 1976; Hughes 1998;
Frost et al. 2000). This proposal helps to explain why the PLAY system is im-
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.20 (1005-1057)

 George F. R. Ellis and Judith A. Toronchuk

portant enough to be genetically determined through the evolutionary process

(being a primary rather than secondary emotion).
Many schooling systems rely on E3: The FEAR system as a main moti-
vational driver, which may well work in the short-term but at the expense of
attaching such negative emotional tones to the relevant topics as to make prob-
able failure of educational achievement in the long term. This would agree with
what Skinner and other behaviourists pointed out: that organisms learn more
readily through reward than punishment. However the implications of the
present view are more than that: they are that we should pay considerable atten-
tion to the affective climate in our educational systems as well as to the intellectual
quality of activity (c.f. Salovey and Sluyter 1997). Of course high quality intel-
lectual activity can inspire pupils and so be a driver in the affective domain, but
that cannot be taken for granted, and it is possible that good educational out-
comes may be at least as dependent on the emotional atmosphere in the classroom
as on the intellectual quality of material engaged with.

. Evolutionary issues

The proposal put here has significant implications for evolutionary psychol-
ogy (Barkow et al. 1992). Specific evolutionary psychology proposals should
only claim to produce brain modules that have in fact been genetically imple-
mented, and the suggestion here is that these are only the primary emotions
E1–E7 plus automatic reactions. Claims of any further brain modules need
solid substantiation in terms of showing they are indeed (a) physiologically ex-
istent and (b) genetically determined, taking into account the limitations on
genetic information mentioned above.
Thus this proposal should enable a re-visiting of the specific claims of evo-
lutionary psychology, seeing if one can relate them to the primary emotions
E1–E7 and subsequent reliable development of secondary emotions S1–S7,
without supposing any other genetically determined systems. Here, as indi-
cated above, we define basic emotions as those that are genetically determined
in the context of the developmental environment and secondary emotions as
those that then are developmentally determined.
One task would be to show why each of the genetically determined pri-
mary emotions E1–E7 plays such an important role in individual functioning in
the face of threats to survival, that it became hard-wired into the species. This
re-visitation might result in postulation of specific further basic genetically de-
termined modules needed for explanatory completeness, that could then be
searched for physiologically and psychologically, for example (as noted above)
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.21 (1057-1107)

Neural development 

checking if some of the social emotions are hard wired. In terms of evolution-
ary game theory, the idea would be to determine what might be an evolutionary
stable set of emotions that could underlie evolutionary stable strategies. There
is a clear relation of this proposal to studies of primates and our other close
animal relations who share a large part of our genetic heritage, so ethology is
also very relevant (Panksepp 2003a).
Apart from looking at our close relatives such as apes for the corresponding
emotional systems and their developmental effects, for example looking at the
role of the amygdala in such animals (Aggleton & Young 2002), relevant tests
could be carried out on animals such as mice, lab rats, and naked mole rats, the
latter being of particular interest in this regard because of their eusociality. By
doing so one might be able to establish the kind of link envisaged here between
emotional systems and behavior in those animals, and so establish relevant as-
pects of the path of its evolutionary development (Panksepp 2002). One could
also look for its precursor traces in much simpler animals such as C. elegans
and Drosophila, thereby tracing even earlier parts of this evolutionary history.

. Language

The key issue of language and symbolism separates humans from all other an-
imals (Deacon 1997; Hauser et al. 2002). There may be a significant difference
in the way the seeking system operates in humans as opposed to in all other
animals in order to allow language development in conjunction with the vocal
apparatus allowing speech. This mechanism must provide the basis for brain-
language-culture co-evolution (Deacon 1997; Donald 1991). As argued above
there is not sufficient genetic information available to specifically determine
construction of language modules (Pinker 1994), but rather the mechanisms
to develop such modules must evolve; see Edelman (1992), Deacon (1997),
Panksepp (1998). Similar issues arise in relation to mathematics and numeracy
(Butterworth 1999; Devlin 2000).
It may well be that language develops in response to strong emotional pres-
sures related to development of culture and social interaction between ever
more conscious beings in a social context (Bonner 1980; Donald 1991, 2001;
Tomasello 2003). Thus apart from E1: The SEEKING system, one might expect
that together with E7: The PLAY system (important in all cognitive develop-
ment as discussed above, and known to be important in language development,
see e.g. Bruner 1985; Rinaldi et al. 2000), it is E5: The CARE system that is cru-
cial in language development in the individual, because it is the mother-child
dyad that is crucial in early mental development (Schore 1994) and communi-
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.22 (1107-1152)

 George F. R. Ellis and Judith A. Toronchuk

cation between them must be the essential foundation of individual language

development; this will be strongly driven by parent-child bonding. Hence one
can look at these basic emotional capacities in those with language disorders
both in terms of psychological tests of those basic capacities, and checking for
structural or chemical variations in areas important to these emotional sys-
tems. One might also include the proposed emotions S6: social rank/dominance
and S7: belonging/exclusion as drivers of language ability, whether these are in
fact primary or secondary emotions.
Furthermore in evolutionary terms one would look for some substantial
change in these systems at the time that language arose, subsequently seen as
a difference between us and our closest primate relatives. One could compare
such possibilities of a change in the SEEKING or CARE system with the claim
that the sudden expansion of the cortex, based on the number of times the
neural precursor cells divide, might have been enough to lead to language,
given that other apes have already well developed emotional responses includ-
ing curiosity and nurturance. However that scenario fails to explain why the
expansion of the cortex took place. It makes sense to ask if some other change
took place first, that then led to the need for greatly increased memory and/or
computational capacity.
The further challenge here will be to explain from this perspective the kinds
of developmental issues that have led Pinker and others to postulate a specific
intrinsic language module (Pinker 1994). These apparent in-built grammatical
propensities should arise out of the way the generic pattern-seeking apparatus
of the cortex interacts with the seeking system in response to environmental
stimuli. Showing how this happens is an intriguing and substantial challenge;
Tomasello (2003) may have already done what is needed here, but his proposal,
based on intention-reading and pattern-finding on the one hand, and on usage
based linguistics (emphasizing that language structure emerges from language
use) on the other, could usefully be revisited in the light of the affective con-
nection (which he does not emphasize). The CARE system will plausibly play
a major role in intention reading; while pattern-finding (and associated world
modelling) would seem to be central to the SEEKING system’s functioning.

. Genetic issues

Given the claim that the basic emotional faculties are genetically determined,
clearly one could try to locate the specific genes related to each such capac-
ity E1–E7 (functionally based in the limbic system). One can in principle
determine which sets of genes or quantitative-trait loci (QTL) are related to
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.23 (1152-1202)

Neural development 

which basic emotional system by looking for correlations between disorders

in development in the corresponding brain areas and the characteristics of
the individual’s genome. These disorders should lead to developmental conse-
quences as indicated above, so those can be looked for also, realizing however
that many disorders are complicated by complexity of interactions in several
pathways. For example, Fullerton et al. (2003) have recently reported a genetic
linkage scan identifying QTLs that influence variation in the personality trait
of neuroticism.
In particular, genetic studies of families showing inherited language dis-
abilities should be particularly informative in relation to the development of
language, as briefly mentioned above. Lai et al. (2001) have recently reported
finding a mutation on chromosome 7 which severely disrupts language but not
other abilities. One could presumably try to trace the developmental pathways
associated with the relevant genes, and compare the human genome with ani-
mal genomes in this context in order to search for differences associated with
the rise of language. We mention this possibility of what would be immensely
complex studies, without attempting to give any details at this point, because
they would be an important part of the overall project and our proposal would
be incomplete without indicating this link.

. Neurological issues

We have not attempted here any detailed characterisation of the brain areas
and neural connections underlying the above proposals, and of course a vast
amount is known about these connections (see e.g. Schore 1994 and references
therein) and in particular about the regions and neurotransmitters associated
with specific primary emotional systems (see Panksepp 1998; and Chapter 4
of Solms and Turnbull 2001, partially summarised in Table 1 above). The final
requirement to pull the present proposal together is to pursue those neural
connections in detail and see how the connectivity between the various parts
of the limbic system and the various areas in the cortex might implement the
suggested neural Darwinism mechanisms based on the primary emotions.
Thus we need to relate this proposal to a much more detailed characteri-
sation of how the different affective systems connect to different brain regions.
This will not be attempted in this brief paper, for it is a large task whose elucida-
tion will require much further endeavour. However the overall point is that the
different primary emotions E1–E7 are of primary significance for the different sec-
ondary emotions and cognitive powers, and the major neural connectivities should
reflect this set of functional relations. Two further comments are as follows:
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.24 (1202-1251)

 George F. R. Ellis and Judith A. Toronchuk

First, the emotions listed above are valenced: many are negative emotions
suggesting either avoidance or confrontation (‘fight or flee’), but some
are positive emotions suggesting reinforcement. These two valences (to
some degree depending on the context) must presumably be handled in
a different way through the effects of their associated neurotransmitters,
which may either strengthen or weaken neural connections (but not in
simple overall correlation to the valence of the emotion). Explicating this
would seem a worthwhile exercise.

Second, we note that it is possible for neurons to be directly wired to each other,
rather than so often using the rather odd mechanism of sending neurotrans-
mitters across a synaptic gap to convey information from an axon to a dendrite.
Why has this strange soup-based mechanism evolved, instead of the simpler
use of direct wiring?15 One possible answer is that this mechanism is needed in
order that neural Darwinism can take place. By using neurotransmitters to con-
vey information across synapses, one enables non-local modulation of synaptic
effects by neurotransmitters from the value system that are conveyed to all the
neurons in a cortical region, the catecholamine neurotransmitters sometimes
diffusing from their source to fill a volume rather than being transmitted to
precisely chosen synaptic connections. Those synapses in this region that are
currently active may then be affected by these diffuse neurotransmitters16 but
those that are inactive will not, and this may be presumed to be a key aspect of
neural Darwinism.
Thus the strange “wet-wiring” across synapses may be needed to enable
neuronal group selection to function. If this is correct, one might make the
following tentative hypothesis:
Conjecture 4: Those neuronal connections that are genetically hard-wired
(for example in the eye, the synapses within the retina itself i.e. between bipo-
lar cells, horizontal, amacrine and ganglion cells) will often be directly wired
electrical synapses; but all those whose detail is determined by neural Dar-
winism will be neurotransmitter-based chemical synapses. This difference
would then be reflected in some appropriate way in the genes determining
the neural structuring.

This would be supported by noting the increased efficiency that direct wiring
will allow, but on the other hand evolution often does not attain maximal ef-
ficiency, and rather cobbles together solutions based on the material at hand;
in which case chemical synapses might be used in these cases too, even though
that is relatively inefficient.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.25 (1251-1300)

Neural development 

An interesting point is that chemical synapses may also be needed to enable

memory to function, apparently providing a different rationale for their use.
But the issue then is, Who chooses what to remember? Information is thrown
away by the billions of bits each day, with only the important remembered
out of the flood of information arriving in our brain; but determining what
is important requires a value system to provide a decision, so this may in fact
be related to the proposal here: wet connections are required both for neural
Darwinism and for memory.

. Potential outcomes

The above characterisation of relations that might follow from our overall hy-
pothesis provides a variety of ways of checking its validity. On pursuing them,
we should be able to check the adequacy of the main hypothesis by developing
implications of the relation of the value system identified (the basic emotional
systems E1–E7) to neuroscience on the one hand and the integrated package of
genetics, evolution, and development on the other. This is in principle doable
as outlined above. This serves as a test of the outlined proposal and may suggest
areas where it is deficient. In particular, we can consider
(a) Are the basic emotions plus the immune and endocrine systems sufficient to
characterise the value system, or should there be other kinds of components
added? If so what are they and how would one test for them?

The point here is that whatever such extra components might be, because of
their developmental importance (which would follow precisely because they
are part of the value system), in turn leading to evolutionary significance (else
they would not be genetically determined), they must have determinable psy-
chological effects which can be searched for. The suggestion here is that the
basic emotions E1–E7, with a possible addition of E0 as characterised above,
are sufficient as well as necessary to determine the value system. But that can
be tested as a hypothesis. If there are other factors needed they should be char-
acterisable; any such proposal will then lead to a revision of all the aspects
outlined above.
(b) Is the list of basic emotions E1–E7 complete or does it need supplementation?

Apart from seeking for omissions of basic emotions needed for causal com-
pleteness, such as E0, the main issue here is separating primary and secondary
emotions as discussed above, with the former genetically determined but the
latter developmentally determined. For example, should there be recognised
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.26 (1300-1361)

 George F. R. Ellis and Judith A. Toronchuk

a primary emotion corresponding to social status or rank (“pecking order”)?

Various studies of animal behavior might suggest some such system exists in
our animal relatives, and has strong survival value so that it is embodied in our
genetic structure. But then its physical correlates should be determinable. We
note here that play may be a substrate for the epigenetic emergence of adult
dominance, especially when combined with anger and fear systems. Alterna-
tively, as dominance is one axis of the interpersonal circumplex, perhaps it is
indeed a primary emotion.
Any such understandings will in principle have implications for child psy-
chology/psychiatry on the one hand and learning/teaching theory on the other,
as briefly indicated above. In essence such implications then form the platform
for further testing of the basic hypothesis, either confirming it, or showing its
inadequacy and the need for further development of the basic idea.

. Relation to the immune system

One might also speculate on conceivable implications regarding the interaction

between the immune system and the brain. It is known that these two systems
interact with each other at multiple levels and in a bi-directional manner (see
Sternberg 2000 and references therein). If the conjecture proposed here were
correct, similar Darwinian processes would underlie both neural and clonal se-
lection in the nervous system and the immune system (for the immune system
case, see Burnett 1959). One could therefore consider whether the immune sys-
tem might, through similar mechanisms, also be engaged in shaping the brain
at a sub-emotional level (see Figure 1).
In evolutionary terms, if one accepts the notion proposed by Edelman,
described above, that the emotional neurocircuitry of the brain, including no-
radrenergic systems activated during fear or stress, play a role in shaping higher
order brain functions as discussed in the first part of this paper, then it is
not unreasonable to propose that the immune system may have evolutionarily
played a similar role in setting emotional brain systems. There is much evi-
dence that cells and molecules of the immune system play an essential role in
shaping neuronal connections, synapse formation and even neuronal cell death
and survival in health, during development, throughout life, during learning,
and in the course of inflammatory and infectious diseases of the brain. Fur-
thermore, immune molecules also alter the emotional circuitry of the brain at
a systems level (e.g. Maier & Watkins 1998; Mulla & Buckingham 1999; Maier
& Watkins 2000; Larson 2002; Davidson et al. 2002; Maier 2003).
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.27 (1361-1404)

Neural development 

There are an immense variety of immune molecules, with many used both
in the immune and nervous systems. It is known that some affect neuronal
function at the cellular level and may also modify structural relationships be-
tween neurons. Thus, the immune molecule interleukin 1 (IL-1) plays a role
in memory consolidation, especially certain types of memory that depend
on place recognition in the hippocampus (Watkins & Maier 2000). Maier et
al., (2003) have shown that IL-1 treatment of mice during context depen-
dent conditioning interferes with the establishment of contextual memory for
the aversive stimulus. The same immune molecule, IL-1, affects an immedi-
ate electrical/cellular event important in establishment of memory, long-term
potentiation. Schneider et al. (1998) have shown that just as neurotransmit-
ters do, IL-1 itself can induce long-term potentiation. There are also structural
parallels between molecules expressed in the immune and central nervous sys-
tems, which play an important role in the establishment of connections and
signalling between immune cells on the one hand and neurons on the other.
Thus a molecule usually expressed on immune cells, MHC class 1, which in the
immune system is critical for lymphocyte-macrophage interactions with anti-
gen and subsequent lymphocyte maturation, is also expressed on neurons and
is important in synapse plasticity (Huh et al. 2002).
Immune molecules act as growth factors in the CNS (Benveniste 1998),
and immune molecules and cells also play an important role in neuronal cell
death and survival (Bajetto et al. 2002). It is now clear that interleukins are pro-
duced by some non-neuronal cells of the nervous system – including microglia
and astrocytes – some of which are derived from or closely related to the im-
mune cells, macrophages. Immune molecules released from these cells within
the nervous system act in many ways like growth factors, either enhancing neu-
ronal survival under some conditions, or promoting cell death, or apoptosis,
under other conditions. Immune molecules can affect neuronal cell death and
survival whether they are released from resident non-neuronal cells of the ner-
vous system, or from immune cells that enter the nervous system in the course
of inflammation (Schwartz & Cohen 2000; Rothwell et al. 1997; Rothwell &
Luheshi 2000). The role that these cells and molecules play in neuronal cell
death, survival and synapse formation can thus be thought of as shaping brain
development in much the same way that pruning may shape a tree by stim-
ulating some branches to grow and removing others. That is another way of
saying they participate in neuronal group selection, but in this case based on
immunological rather than affective states.
In addition to this effect of immune molecules on neurons and neuronal
function at a molecular and cellular level, immune molecules also have im-
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.28 (1404-1436)

 George F. R. Ellis and Judith A. Toronchuk

portant effects on brain function at a systems level. Thus, immune molecules

produced outside the nervous system, at sites of inflammation, in immune or-
gans, or during infection, can alter a variety of brain functions. Interleukins
in this sense act like hormones, molecules that are produced at one site in the
body and affect the function of an organ distal to their site of origin. Many im-
mune molecules, including interleukins and interferons, have been shown to
activate the brain’s hormonal stress response (reviewed in Webster et al. 2002),
alter mood and cognition, induce sleep and fever (Krueger & Majde 1994),
and induce a set of behaviours called sickness behavior (Dantzer et al. 1998:
Watkins & Maier 2000). The latter are conserved across species and consist
of decreased locomotion, loss of appetite, social isolation, loss of interest in
sex – i.e. a pattern of behaviours that causes the organism to turn in upon itself
rather than focusing on the outside world. At the same time, interleukins also
cause a profound change in mood, with induction of symptoms in many ways
overlapping with those of depression (Davidson et al. 2002; Dantzer et al. 1998;
Miller 1998).
Thus at molecular, cellular and functional levels, immune cells and
molecules have profound immediate and long-lasting effects on the nervous
system in general, and on emotional sub-systems within the nervous system,
in particular. Thus the immune system acts in a functional way on the CNS,
in particular activating affective states; we are suggesting that additionally, part
of this function can be seen as adding to the value system that guides neuronal
group selection, in this case conveying danger signals implied by homeostatic
disequilibrium.
Interestingly, while the amygdala-based fear system is amongst those pro-
posed to mediate selection of adaptive pathways of higher brain functions, it
is the parasympathetic vagal system, originating in the brainstem nucleus of
the tractus solitarius, which mediates incoming immune signals to the emo-
tional systems of the brain. One could question the evolutionary advantage of
these two opposing arms of the autonomic nervous system in selection of adap-
tive pathways at two sequential levels of this hierarchy. One possible answer
might be that the inhibitory effects of the parasympathetic nervous system on
the adrenergic sympathetic nervous system, could serve as a natural negative
feedback loop, protecting the overall system from spiralling out of control.17
There are two points that arise. The first is what specific effects might
arise through long-term programming of emotional systems by immune sys-
tem molecules during the life of an individual. It might extend to affecting the
way the primary systems are activated: attaching negative emotional valence to
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.29 (1436-1494)

Neural development 

specific kinds of conditions and hence developing avoidance reactions to such

situations, and perhaps playing a role in depression.
The second is, what role could this have played in evolution? We propose
Conjecture 5: The immune system could have played a selective and
survival role in setting emotional systems in the first place during evo-
lutionary history.

The issue here is not how the different emotional systems became differentiated
but rather how any emotional system at all came into being. This is a crucial
step on the way to full consciousness. Once the basic capacity was there it could
evolve to respond to the major environmental issues confronting the individ-
uals in a population, resulting in the basic affective reactions. That capacity
would then evolve to the present primary emotional systems that are indeed
genetically laid down and realised in response to the local environment during
embryonic development. Once they have come into being, the immune system
would still help guide neural development but not in as specific a way as the
primary emotions.
The proposal here is that in the causal hierarchy shown in Figure 1, im-
mune signals to the emotional subsystems of the nervous system, and in turn,
neurotransmitter signals from emotional systems to cortical areas of higher
brain function, could represent not only a current functional hierarchy but also
an evolutionary developmental sequence. Consider early animal development
where a rudimentary CNS is in operation but no affective system has yet been
formed; the animal is unable to consciously “feel” (it behaves as if it is emo-
tional though; it shows what might be called emotional behaviours, but on a
reflexive level). But later animals can feel; the issue is how the possibility of af-
fective states first arose. Now we have seen that the immune system can signal
to and affect the CNS, conveying the information of health-threatening states
to the CNS; this is a possible rudimentary form of value system as discussed
above. It is then natural to suggest an evolutionary sequence whereby
(i) a primitive innate immune system was present in the earliest metazoans,
then
(ii) the evolution of the innate immune system provided several basic mech-
anisms that were co-opted in evolutionary development by the nervous
system,
(iii) the immune system started functioning as a primary component of the
value system for the CNS, and
(iv) further co-opting of the immune mechanisms allowed a primitive form
of neural Darwinism to be utilized in the CNS of early vertebrates.
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.30 (1494-1548)

 George F. R. Ellis and Judith A. Toronchuk

The common molecules would have facilitated the development of synapses

that are reliant on neurotransmitters, because of the survival value of recog-
nising the threats to life from the events and environments associated with
infectious attack. As consciousness developed, in whatever way that happened,
it might be natural for this immune system interaction with the CNS to provide
the first affective states, essentially the feeling of “being ill”.18 The immune sys-
tem also provided the first differentiation of “self ” from “non-self ”, an ability
which reaches its highest level in human consciousness.
Once a first set of such affective states had come into existence, presumably
corresponding to the suggested basic affective dimension E0: PLEASURE/PAIN
suggested above, they could have been adapted and developed into the full set
E1–E7 identified by Panksepp and additionally become genetically determined,
because of the evolutionary advantage they provided through acting as a value
system. Neural Darwinism in the individual would have proved itself to be a
winning strategy and hence could have developed further affective capacities
in an efficient way, through the usual evolutionary process of natural selection
acting on the developing brain in a sequence of animal species.
Conjecture 6: It was through this process of immune system interaction with
the CNS that neural Darwinism came into being as a brain-structuring
mechanism in the course of evolutionary history.

This suggestion has the potential to explain why some molecules are both im-
mune system molecules and also neurotransmitters. It could conceivably even
help explain how chemical synapses came into being in the first place, or at
least why they are so common.

. Implications of the immune system link

The proposal here suggests an evolutionary sequence that can be searched for
through its survival in present day animals: first the basic immune system came
into being, then neurons connected together to form a CNS developing a sim-
ple response capacity, then the basic immune system developed into a clonal
selection system, and only then the major emotional neural systems coming
into being. Hence one can test the proposal made here by looking at the lev-
els of primary emotions and immune system development in simpler animals
(for example nematodes and insects, as well as more complex animals such as
birds and mammals), using this to trace the evolutionary sequence and hence
see if it is in accord with the idea proposed here or not. This investigation could
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.31 (1548-1605)

Neural development 

also clearly be related to the study of the genomes of these animals and their
developmental pathways.
Our proposal also suggests that infectious attacks may develop not only
immediate feelings of ill-health but also longer term emotional aversion to the
kinds of conditions that lead to those infections. Presumably one could test that
kind of effect in animals such as mice, finding corresponding behavioural ef-
fects. One could then try to trace how far down the scale of complexity such
effects existed: for example could snails be shown to behave in this way? This
would clearly relate to the evolutionary issue just mentioned.

. Conclusion

There is an immense and complex literature on neuronal development and

function. Our suggestion is that the overall causal linkages we have sketched
here might represent significant organising principles applying within that
complex reality. We would not be so bold as to make such a proposal ab initio;
rather we gain some confidence because our proposal is based on unifying two
theories that are already well developed and supported by an extensive liter-
ature, namely affective neuroscience on the one hand and neural Darwinism
on the other. By supporting the two theories independently, that extensive lit-
erature serves to indicate that they are each valid in their own right, and if
both are valid, this then provides some support for a proposal such as ours for
combining the two into a unified whole.
Several pieces of evidence have been proposed to bolster the hypothesis of
Neuronal Group Selection. These include the following:
– something like this is needed to account for the great variability in human
brain structure, contrary to any process of construction according to a preset
algorithm (Edelman 1992: 27, 82);
– this allows the brain to optimally adapt to the local physical and cultural
environment (Deacon 1997: 206; Siegel 2001), while also being able to face up
to new circumstances in a adaptive way;
– there is far too little genetic information available to structure the brain
in a more direct way, as discussed above;
– it links to the fact that neuronal connections are set up in a random way
on the micro level, and then pruned by micro processes.
On the other hand there is ever increasing recognition of developmental and
neuronal importance of the affective dimension. The first part of our paper
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.32 (1605-1674)

 George F. R. Ellis and Judith A. Toronchuk

suggests these two developments fit well together. If this proposed syntheses is
true, this would imply three key features:
– First, immune system, higher brain, and emotional system evolution,
development, and function would all be based on the same generalised Dar-
winian selectional principles (the way this happens being set out in the main
Hypothesis); a significant unification of understanding and extension of ap-
plication of the basic Darwinian insight in terms of allowing developmental
emergence of high level structures through processes of natural selection op-
erating on various time scales; this is widely regarded as the only viable way of
enabling teleonomic higher level structures to emerge through self-structuring
processes in evolutionary history, and extends that proposal to developmen-
tal processes. This unification has intriguing links to functional detail at the
neuronal level (see Conjecture 4).
– Second, it makes concrete proposals as regards the neurobiological func-
tioning underlying the role of emotions and the immune system in brain
evolution and development as well as in learning, namely with primary emo-
tions providing the essential part of the fitness function that shapes the result
of natural selection processes operating developmentally in the phenotype (see
Conjecture 3). This understanding provides sound links between evolutionary
theory, neurology, developmental biology, and aspects of psychology (partic-
ularly learning theory) and ethology as outlined above, by explicating a key
multilevel link of the kind envisioned as underlying social neuroscience (Ca-
cioppo et al 2002). It also could provide proposals for the function of signifi-
cant brain systems whose physiological role is otherwise unknown (e.g. some
of the monoamine systems, see Kingsley 2000: 131).
– Third, it makes specific proposals as regard the age-old nature-nurture
issue, see Conjectures 1 and 2; this issue has important implications for social
policy and politics, as so clearly shown by Stephen Pinker in his book The Blank
Slate (Pinker 2002).
In the second part of the paper, we suggest that investigation of a corresponding
link from the immune system to brain development and evolution could be
interesting. This proposal is based on two features: first, there is a substantial
literature on influences the other way, from the brain to the immune system.
But given this causal connection, it then makes sense to look also for effects of
the immune system on the brain (this may be thought of as an application of
the heuristic principle, where there is an action, there is often a reaction).
Second, there is evidence that immune molecules do indeed have the ca-
pacity to influence neurons in the CNS. The question then is what form that
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.33 (1674-1730)

Neural development 

influence might take when considered at the macro level, and a natural extension
of the ideas of the first part suggest this might also take the form of a contri-
bution to neural Darwinism, in this case with the immune system influencing
both evolution and function of the affective systems (Conjecture 5). Indeed this
interaction could possibly have played a significant role in the first evolutionary
development of affective systems at the macro level and of chemical synapses
at the micro level, and through this process have facilitated the emergence of
neural Darwinism (Conjecture 6). This proposal does not deny the immense
complexity of the immune system in its own right, but does indicate interest-
ing lines to follow in terms of understanding its possible interactions with the
CNS – another immensely complex system. This interaction would leave traces
in the historical record; particularly one might attempt to show that immune
system development preceded emergence of simple affective responses.
Overall, the point of this paper is that the macro-role of neural Darwin-
ism is not fully explicated until the nature of the value system has been made
clear. We have made such a proposal that unites neural Darwinism with a ma-
jor aspect of recent neuroscience, namely the appreciation of the importance
of affect. The resulting synthesis provides a unification with interesting im-
plications as outlined above. It has a plausibility arising from the fact that it
proposes a concrete mechanism to underlie the effectiveness of affective macro
processes, for which there is an ever-growing body of evidence.
Our final major recommendations flowing from all of this (relating to
Section 4.3 above) are as follows:
Recommendation 1: A series of psychological tests should be devised that
aim to determine the relative strengths A(EI ) of the basic emotional sys-
tem E1–E7 (see Section 4.1 above) as well as the strengths A(SI ) of a
suitably chosen set S1-S7 of secondary emotions;
Recommendation 2: These tests are then used to search for correlations of
the primary emotional factor strengths A(EI ) with those of the secondary
emotions A(SI );
Recommendation 3: These tests are also used in conjunction with tests
of the different aspects of intelligence (a general factor g, group factors
gi , and specific factors gi (s) within each group, see Deary, 2001 for a
summary of such factors) to search for correlations with the intelligence
factors g, gi , gi (s) firstly of the primary emotional factor strengths A(EI )
and secondly of the secondary emotions A(SI ).
These tests should enable firmly experimentally based investigation of the basic
hypothesis proposed here, in line with the strong experimental approach laid
out by Deary (2001).
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.34 (1730-1805)

 George F. R. Ellis and Judith A. Toronchuk

Acknowledgements

GE thanks the members of the University of Cape Town Consciousness Study

Group for helpful discussions, and particularly David Kibel for drawing my at-
tention to the writings of Jaak Panksepp. We thank Jaak Panksepp and Alan
St Clair Gibson for helpful communications, and Esther Sternberg for very
helpful contributions.

Notes

* Mathematics Department, University of Cape Town, Rondebosch 7701, Cape Town,

South Africa. email address: ellis@maths.uct.ac.za.
** Psychology and Biology Departments, Trinity Western University, 7600 Glover Road,
Langley, B.C. V2Y 1Y1 Canada. email address: toronchu@twu.ca.
. See also Damasio (1995, 2000, 2003); Schore (1994); Le Doux (1998); Lane and Nadel
(2000); Davidson et al (2002).
. See also Schore (1994); Deacon (1997); LeDoux (2002); and references therein.
. The name neural Darwinism can be criticised because there is no element of reproduc-
tion with genetic inheritance and variation; instead there is variation of connections created
by imprecise synaptic formation, followed by neuronal group selection. This is however just
an argument over semantics; it does not affect the nature of the theory, which centres on the
core Darwinian feature that it is survival of the fittest that structures emergent complexity.
It may be more accurate to call the proposal, “Neuronal Group Selection”.
. Schore (1994) relates it to the idea of “parcellation” (pp. 19, 250, 258).
. An antigen selects particular lymphocytes out of a very diverse population of pre-existing
lymphocytes for clonal expansion, enabling the immune system to respond to antigens that
it has never encountered before.
. Summarised in Pinker (2002), see the Appendix.
. See for example the concise survey of psychology by Butler and McManus (2000), where
the role of emotions in behaviour is consigned to a few remarks on pages 67–68, comment-
ing that for many they are regarded simply as an impediment to rational thinking.
. Many neocortical connections could also emerge in tissue culture, but these connections
would be broad and general and not represent detailed connectivity such as that provided by
experience. If there is no experiential input, then the connections cannot be representative
of such.
. In mathematical terms, they form an orthogonal basis for the space of primary emotions.
. It is usually claimed there are four types, because all of the conditions with one type of
cone or no cones leave one with monochrome vision only (there must be two or more cone
types to have different responses to different wavelengths). However these monochromatic
images will each be different, corresponding to what can be seen with different colour filters.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.35 (1805-1889)

Neural development 

. As another example, Evans (2001) lists hardwired basic emotions as Joy; Distress; Anger;
Fear; Surprise; Disgust, and universal higher cognitive emotions as Love; Guilt; Shame;
Embarrassment; Pride; Envy; Jealousy.
. There is not a unique relation here, as there is great overlap in the transmitter systems,
plus many transmitters have both excitatory and inhibitory effects.
. Preliminary assessments in this regard are given by Damasio (2003: 156).
. We assume from here on that by ‘E1–E7’ we mean whatever classification of primary
emotions emerges from the reconsideration just discussed.
. We thank Alan St Clair Gibson for raising this issue.
. This possibility is not taken into account by current models of synaptic plasticity rules
(see Dayan & Abbot, 2001, Section 8.2) and consequently the possibility of value based learn-
ing is not considered in present studies of the neuronal bases of learning (ibid., Sections 8.3
and 8.4). However the possibility of classical conditioning and reinforcement based learning
(and specifically the role of dopamine) is recognised at the macro-level (ibid. Chapter 9).
. We thank Esther Sternberg for this suggestion.
. This theme will be pursued further in an upcoming paper (Toronchuk & Ellis 2005).

References

Aggleton, John P., & Young, Andrew W. (2000). The engima of the amygdala: On its
contribution to human emotion. In Richard D. Lane & Lynn Nadel (Eds.), Cognitive
neuroscience of emotion (p. 106). Oxford: Oxford University Press.
Bajetto, Adriana, Bonavia, Rudy, Barbero, Simone, & Schettini, Gennaro (2002). Charac-
terization of chemokines and their receptors in the central nervous system:
physiopathological implications. Journal of Neurochemistry, 82, 1311–1329.
Baltimore, David (2001). Our genome unveiled. Nature, 409, 814–816.
Barkow, J. H., Cosmides, L., & Tooby, J. (1992). The adapted mind: Evolutionary psychology
and the peneration of culture. Oxford: Oxford University Press.
Benveniste, E. N. (1998). Cytokine actions in the central nervous system. Cytokine and
Growth Factor Reviews, 9, 259–275.
Bonner, J. T. (1980). The evolution of culture in animals. Princeton: Princeton University
Press.
Brown, Daniel E. (1991). Human universals. New York: McGraw Hill.
Bruner, J. S., Jolly, A., & Sylva, K. (Eds). (1976). Play: Its role in development and evolution.
Harmondsworth: Penguin Books.
Bruner, Jerome (1985). Child’s talk: Learning to use language. New York: W.W. Norton.
Burnet, F. Macfarlane (1959). The clonal selection theory of acquired immunity. Cambridge:
Cambridge University Press.
Butler, Gillian, & McManus, Freda (2000). Psychology: A very short introduction. Oxford:
Oxford University Press.
Butterworth, B. (1999). What counts: How every brain is hardwired for math. New York: The
Free Press.
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.36 (1889-1981)

 George F. R. Ellis and Judith A. Toronchuk

Cacioppo, John T., Berntson, Gary G., Adolphs, Ralph, Carter, C. Sue, Davidson, Richard
J., McClintock, Martha K. et al. (Eds). (2002). Foundations in social neuroscience.
Cambridge, Mass: MIT Press.
Carroll, B. J. (1993). Human cognitive abilities: A survey of factor analytic studies. Cambridge:
Cambridge University Press.
Costa, Paul T., Terracciano, Antonio, & McCrae, Robert R. (2001). Gender differences in
personality traits across cultures: Robust and surprising findings. Journal of Personality
and Social Psychology, 81, 322–331.
Dantzer, R., Bluthe, R. M., Laye, S., Bret-Dibat, J. L., Partnet, P., & Kelley, K. W. (1998).
Cytokines and sickness behaviour. Annals of the New York Academy of Sciences, 840,
586–590.
Dayan, Pater, & Abbott, L. F. (2001). Theoretical neuroscience: Computational and mat-
hematical modelling of neural systems. Boston: MIT Press.
Davidson, Richard J., Pizzagalli, Diego, Nitschke, Jack B., & Putnam, Katherine (2002).
Depression: Perspectives from affective neuroscience. Annual Review of Psychology, 53,
545–574.
Davis, Kenneth L., Panksepp, Jaak, & Normansell, Larry (2003). The Affective Neuroscience
Personality Scales: Normative Data and Implications. Neuro-Psychoanalysis, 5, 57–69.
Damasio, Antonio (1995). Descartes’ error: Emotion, reason, and the human brain. New York:
Avon Books.
Damasio, Antonio (2000). The feeling of what happens: Body, emotion and the making of
consciousness. London: Vintage.
Damasio, Antonio (2003). Looking for Spinoza: Joy, sorrow, and the feeling brain. London:
Harcourt.
Deacon, Terrence (1997). The symbolic species: The co-evolution of language and the human
brain. London: Penguin.
Deary, Ian J. (2001). Intelligence: A very short introduction. Oxford: Oxford University Press.
Devlin, K. (2000). The math gene: How mathematical thinking evolved and why numbers are
like gossip. New York: Basic Books.
Donald, Merlin (1991). Origins of the modern mind. Boston: Harvard University Press.
Donald, Merlin (2001). A mind so rare: The evolution of human consciousness. New York: W.
W. Norton.
Edelman, Gerald M. (1989). Neural Darwinism: The theory of group neuronal selection.
Oxford: Oxford University Press.
Edelman, Gerald M. (1992). Brilliant air, brilliant fire: On the matter of mind. New York:
Basic Books.
Edelman, Gerald M., & Tononi, Giulio (2001). Consciousness: How matter becomes ima-
gination. London: Penguin Books.
Ekman, Paul, & Davidson, Richard J. (Eds.). (1994). The nature of emotion: Fundamental
questions. New York: Oxford University Press.
Ellis, George F. R. (2003). On the nature of emergent reality. http://www.mth.uct.ac.za/∼
ellis/emerge.doc. To appear in P. Clayton (Ed.), The re-emergence of emergence.
Cambridge: Cambridge University Press.
Evans, Dylan (2001). Emotion: A very short introduction. Oxford: Oxford University Press.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.37 (1981-2075)

Neural development 

Frost, Joe L., Wortham, Sue C., & Reifel, Stuart (2000). Play and child development. London:
Prentice Hall.
Fullerton, Jan, Cubin, Matthew, Tiwari, Hemant, Wang, Chenxi, Bomhra, Amrjit, Davidson,
Stuart et al. (2003). Linkage analysis of extremely discordant and concordant sibling
pairs identifies Quantitative-trait loci that influence variation in the human personality
trait neuroticism. American Journal of Human Genetics, 72, 879–890.
Goleman D. (1996). Emotional intelligence: Why it can matter more than IQ. London:
Bloomsbury.
Hauser, Mark D., Chomsky, N., & Fitch, W. Tecumseh (2002). The faculty of language: How
did it evolve? Science, 298, 1569–1579.
Hughes, Fergus P. (1998). Children, play, and development. Cape Town: Allyn & Bacon.
Huh, Gene S., Boulanger, Lisa M., Hongping, Du, Riquelme, Patricia A., Brotz, Tilmann M.,
& Shatz, Carla J. (2002). Functional requirements for class I MHC in CNS development
and plasticity. Science, 290, 2155–2160.
Kiecolt-Glaser, Janice K., Mc Guire, Lynanne, Robles, Theodore F., & Glaser, Ronald (2002).
Emotions, morbidity, and mortality: New perspectives from psychoneuroimmunology.
Annual Review of Psychology, 53, 83–107.
Kingsley, Robert E. (2000). Concise text of neuroscience. Philadelphia: Lippincott, Williams
& Wilkins.
Krueger J. M., & Majde, J. A. (1994). Microbial products and cytokines in sleep and fever
regulation. Critical Reviews in Immunology, 14, 355–379.
Lai, Cecilia S. L., Fisher, Simon E., Hurst, Jane A., Vargha-Khadem, Fareneh, & Monaco,
Anthony P. (2001). A forkhead-domain gene is mutated in a severe speech and language
disorder. Nature, 413, 519–523.
Larson, Susan J. (2002). Behavioral and motivational effects of immune-system activation.
Journal of General Psychology, 129, 401–414.
Lane, Richard D., & Lynn Nadel (2000). Cognitive neuroscience of emotion. Oxford: Oxford
University Press.
Le Doux, Joseph (1998). The emotional brain: The mysterious underpinnings of emotional life.
New York: Simon & Schuster.
LeDoux, Joseph (2002). Synaptic self: How our brains become who we are. New York: Viking.
Lynam, Donald R., & Widiger, Thomas A. (2001). Using the five-factor model to represent
the DSM-IV personality disorders: An expert consensus approach. Journal of Abnormal
Psychology, 110, 401–412.
Maier, Steven F. (2003). Bi-directional immune-brain communication: Implications for
understanding stress, pain, and cognition. Brain, Behavior, and Immunity, 17, 69–85.
Maier, Steven F., & Watkins, Linda R. (1998). Cytokines for psychologists: Implications of bi-
directional immune-to-brain communication for understanding behavior, mood, and
cognition. Psychological Review, 105, 83–107.
Maier, Steven F., & Watkins, Linda R. (2000). The immune system as a sensory system:
Implications for psychology. Current Directions in Psychological Science, 9, 98–102.
Maier, Steven F., & Watkins, Linda R. (2003). Immune-to-central nervous system
communication and its role in modulating pain and cognition: Implications for cancer
and cancer treatment. Brain, Behavior and Immunity, 17 (Supplement 1), S125–131.
 JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.38 (2075-2172)

 George F. R. Ellis and Judith A. Toronchuk

Manning, Aubrey, & Dawkins, Marian S. (1998). An introduction to animal behaviour.

Cambridge: Cambridge University Press.
McCrae, Robert R., & Costa, Paul T. (1997). Personality trait structure as a human universal.
American Psychologist, 52, 509–516.
Miller, A. H. (1998). Neuroendocrine and immune system interactions in stress and
depression. Psychiatric Clinic North America, 21, 443–463.
Milsum, John H. (1966). Biological control systems analysis. New York: McGraw Hill.
Mulla, Abeda, & Buckingham, Julia C. (1999). Regulation of the hypothalamo-pituitary-
adrenal axis by cytokines. Best Practice & Resarch Clinical Endocrinology & Metabolism,
13, 503–521.
Norman, Donald (2003). Emotional design: Why we love (or hate) everyday things. New York:
Basic Books.
Panksepp, Jaak (1998). Affective neuroscience: The foundations of human and animal
emotions. Oxford: Oxford University Press.
Panksepp, Jaak (2001a). The neuro-evolutionary cusp between emotions and cognitions:
Implications for understanding consciousness and the emergence of a unified mind
science. Evolution and Cognition, 7, 141–149.
Panksepp, Jaak (2001b). The long-term psychobiological consequences of infant emotions:
Prescriptions for the 21st century. Infant Mental Health Journal, 22 (1–2), 132–173.
Panksepp, Jaak (2002). On the animalian values of the human spirit: The foundational
role of affect in psychotherapy and the evolution of consciousness. European Journal
of Psychotherapy, Counseling, & Health, 5, 225–245.
Panksepp, Jaak (2003a). Can anthropomorphic analyses of separation cries in other animals
inform us about the emotional nature of loss in human beings? Comment on Blumberg
and Sokoloff (2001). Psychological Review, 110, 376–388.
Panksepp, Jaak (2003b). At the interface of the affective, behavioural and cognitive
neurosciences: Decoding the emotional feelings of the brain. Brain and Cognition, 52,
4–14.
Panksepp, Jaak (2003c). Feeling the pain of social loss. Science, 302, 237–239.
Pinker, Steven (1994). The language instinct. London: Penguin Books.
Pinker, Steven (2002). The Blank Slate. London: Penguin Books.
Reuter, M., & Henning, J. (2003). Die Affective Neuroscience Personality Scales (ANPS)
– ein neues Instrument zur Persönlichkeitsdiagnostik. Zeitschrift für Differentielle und
Diagnostische Psychologie, 23, 223.
Rinaldi, Wendy, Law, James, & Larcher, Janet (2000). Language difficulties in an educational
context. London: Whurr Publishers.
Rothwell, N., Allan, S., & Toumond, S. (1997). The role of interleukin 1 in acute neu-
rodegeneration and stroke: pathophysiological and therapeutic implications. Journal
of Clinical Invest., 100, 2648–2652.
Rothwell, Nancy J., & Luheshi, Giamal N. (2000). Interleukin I in the brain: Biology,
pathology and therapeutic target. Trends in Neurosciences, 23, 618–625.
Salovey, Peter, & Sluyter, David (Eds.). (1997). Emotional development and emotional
intelligence: Educational implications. New York: Basic Books.
JB[v.20020404] Prn:23/11/2004; 13:46 F: Z13104.tex / p.39 (2172-2243)

Neural development 

Schneider, H., Pitossi F., Balschun D., Wagner A., del Rey A., & Besedovsky H. O. (1998).
A neuromodulatory role of interleukin-1beta in the hippocampus. Proceedings of the
National Academy of Sciences, USA, 95, 7778–7783.
Schore, Allan N. (1994). Affect regulation and the origin of the self: The neurobiology of
emotional development. Hillsdale, NJ: Lawrence Erlbaum.
Schwartz, M., & Cohen, I. R. (2000). Autoimmunity can benefit self-maintenance.
Immunology Today, 21, 265–268.
Siegel, Daniel J. (2001). The developing mind: How relationships and the brain interact to
shape who we are. New York: Guilford Press.
Slater, P. J. B. (1999). Essentials of animal behaviour. Cambridge: Cambridge University Press.
Solms, M., & Turnbull, O. (2002). The brain and the inner world: An introduction to the
neuroscience of subjective experience. New York: Other Press.
Sternberg, Esther (2000). The balance within: The science connecting health and emotions.
New York: W. H. Freeman.
Tomasello, Michael (2003). Constructing a language: A usage-based theory of language
acquisition. Boston: Harvard University Press.
Toronchuk, Judith, & Ellis, George F. R. (2005). In preparation.
Trapnell, P. D., & Wiggins, J. S. (1990). Extension of the interpersonal adjective scales
to include the Big Five dimensions of personality. Journal of Personality and Social
Psychology, 59, 781–790.
Watkins, Linda R., & Maier, Steven F. (2000). The pain of being sick: Implications
of immune-to-brain communications for understanding pain. Annual Review of
Psychology, 51, 29–57.
Watt, Douglas F. (1999). Consciousness and emotion: Review of Jaak Panksepp’s Affective
Neuroscience. Journal of Consciousness Studies, 6, 191–200.
Webster, Jeanette I., Tonelli, Leonardo, & Sternberg, Esther M. (2002). Neuroendocrine
regulation of immunity. Annual Review of Immunology, 20, 125–163.
Wicker, Bruno, Keysers, Christian, Plailly, Jane, Royet, Jean-Pierre, Gallese, Vittorio, &
Rizzolatti, Giacomo (2003). Both of us disgusted in my insula: The common neural
basis of seeing and feeling disgust. Neuron, 40, 655–664.
Wolfsberg, Tyra G., McEntyre, Johanna, & Schuler, Gregory D. (2001). Guide to the draft
human genome. Nature, 409, 824–826.
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.1 (41-108)

Consciousness, emotion and face

An event-related potentials (ERPs) study

Michela Balconi* and Claudio Lucchiari

Catholic University of Milan / Neurological National Institute “C. Besta”

Are conscious and unconscious elaboration of emotional facial stimuli

qualitatively different processes? To answer to these questions event-related
potentials elicited by supraliminal and subliminal stimuli were recorded when
subjects viewed emotional facial expressions of three emotional or neutral
stimuli. ERP effect (specifically N2 negative deflection, temporal interval
180–300) were analyzed in terms of their peak amplitude and latency
measures. First, an emotional face specificity was observed for the negative
deflection N2, with a consistent difference between ERP elicited by emotional
and neutral stimuli. Secondly, the unaware information processing was
revealed quite similar to aware one in terms of peak amplitude but not of
latency. In fact, unconscious stimulation produced a more delayed peak
variation than conscious. Third, cortical lateralization (right/left) was not
correlated to conscious/unconscious distinction. On the contrary a more
posterior distribution of the ERP was found for N2 as a function of emotional
content of the stimulus. Functional significance of our results is underlined in
terms of subliminal stimulation and emotional face decoding.

Keywords: Consciousness, emotion and face; event-related potentials (ERPs)

. Introduction

Facial expressions of emotions are social and communicative tools, since hu-
mans use facial expressions to interpret the intentions of others. Previous
paradigms of analysis have focused on conscious elaboration of emotions, leav-
ing out other interesting cognitive processes not at the level of consciousness.
In contrast with most of the previous literature, in this research we focus on
the “other side” – on unconscious psychological mechanisms and their role in
emotion recognition. There is now considerable evidence supporting the no-
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.2 (108-152)

 Michela Balconi and Claudio Lucchiari

tion that significant affective processing happens outside conscious awareness

(Bunce et al. 1999; LeDoux 1996; Shevrin et al. 1996). In fact, in addition to
consciously-perceived stimuli, there are a great number of signals from with-
out and within the subject which are perceived and processed without any
reportable awareness. The term implicit perception was suggested for this pro-
cess, as an analogue of implicit memory (Kihlstrom et al. 1992). An obvious
example, well known from experimental psychology, is the phenomenon of
subliminal perception.
Unconsciously mediated psychological phenomena could also be demon-
strated in neuropsychological contexts. For example, prosopagnosia, an inabil-
ity to recognize people from their visual appearance, may result from damage
to visual association cortices. In some cases it has been demonstrated that
prosopagnosics appear to recognize familiar faces even though they completely
fail to verbally identify the persons (Tranel & Damasio 1985). Thus the patients
show evidence of unconscious recognition that cannot be accessed consciously.
Conceptually similar findings are reported by Öhman (1999) from condition-
ing experiments with human subjects, conditioned to facial stimuli paired with
electric shock. When previously conditioned angry faces are presented masked
by neutral faces, which block their conscious recognition, they nevertheless
elicit conditioned responses. Moreover, research on attention has introduced a
distinction between automatic and conscious (Posner 1978), or automatic and
controlled (Shiffrin & Schneider 1977), information processing. This concept
was developed to account for the fact that the selectivity of attention is better
described in terms of a flexible and strategic distribution of limited processing
resources across stimuli and tasks. In addition, LeDoux (1990) has argued that
the core of the emotional system is a brain mechanism that computes the affec-
tive significance of a stimulus. He argues that this brain mechanism is part of
what gives rise to the conscious experience of emotion, and that it necessarily
operates outside of conscious awareness.
Even though the existence of unconscious effects now has been widely ac-
cepted, the question is still open concerning their importance. For example
we can ask how complicated an unconscious system is, or whether it oper-
ates in an analogous manner to the conscious system. The main purpose of
this study is to elucidate the relationship between conscious and unconscious
decoding of emotional facial expressions. First of all, we consider previous re-
search on emotional facial expressions that have used a neuropsychological
paradigm of analysis, and, specifically, ERP (event related potential) measures.
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.3 (152-204)

Consciousness, emotion and face 

An increasing number of studies has analyzed the cognitive and neuropsycho-

logical features of face recognition (Balconi & Pozzoli 2003a; Posamentier &
Abdi 2003). Specifically, PET studies (Bernstein et al. 2002; Haxby et al. 2000),
fMRI (Adolphs et al. 1998; Grelotti et al. 2002; Kanwisher et al. 1997) and ERPs
(Eimer and McCarthy 1999; Herrmann et al. 2002) have underlined the brain
specificity of emotion decoding.
ERP studies in humans have provided evidence consistent with this view.
In particular, they have revealed the early emergence of recognition processes
and their distinctiveness from other cognitive processes. Streit et al. (2000)
evaluated differences in ERPs in emotional and structural face processing. They
found an early negative deflection at about 240 ms post-stimulus (N2 effect),
related to emotional faces more than to neutral faces, for an explicit (recogni-
tion of emotional expression) task. Successively, a similar ERP effect was found
with an implicit task (Sato et al. 2000). Two theoretical positions were proposed
to explain this ERP effect: the first interpretation supposes that N2 could be a
cognitive marker of the complexity and the relevance of the stimulus (Carretié
& Iglesias 1995). A second position underlines the emotional-face specificity of
N2 (Balconi & Pozzoli 2003b; Herrmann et al. 2002).
Secondly, from the viewpoint of cognitive neurophysiology, subliminal per-
ception has been studied only in a limited number of cases (Shevrin & Fritzler
1968; Wong et al. 1994). Some investigations were applied to the classical odd-
ball paradigm (Batty & Taylor 2003; Bernat et al. 2001; Brázdil et al. 2001);
a P300 ERP effect (a positive deflection) was found for unconscious stimuli,
similar to a supraliminal condition. This result might demonstrate that ERPs
can index unconscious mental process (Shevrin 2001). Nevertheless, no pre-
vious study has explored the relationship between unconscious elaboration of
emotional faces and specific ERP effects.
In the research reported here, we adopted a model of analyses that postu-
late quite homogeneous cognitive processes for both subliminal and supralim-
inal stimulation. Specifically, we hypothesized that:
– Subliminal ERPs have a component structure similar to conventional
supraliminal ERPs.
– Subliminal ERP components have similar psychological properties to supral-
iminal ERPs.

In line with this model, we suppose that the wave profiles elicited by conscious
and unconscious decoding might be similar with respect of their peak profiles.
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.4 (204-260)

 Michela Balconi and Claudio Lucchiari

Objectives and hypotheses

Some main questions remain to be answered:

1. First of all, the cognitive nature of the negative ERP variation N2 must be
clarified, in order to analyze its specificity for emotional facial expression
decoding. The comparison of facial expressions with a neutral condition
(neutral facial expression) might be useful, in order to characterize the
emotional value of this early peak. About this point we suppose that the
ERP profile for emotional expression should be differentiated from the
wave profile of a neutral stimulus. Specifically, a negative ERP variation
(N2) should signal the emotional content of faces, with a higher peak
variation if compared to a neutral stimulus;
2. Subliminal and supraliminal stimulation have been hypothesized to have
a quite similar effect on ERP correlates, in terms of wave variations and,
more specifically, for peak emergence. Moreover, the ERPs were predicted
to discriminate between the type of the stimulus (emotional or neutral),
whether the stimuli were processed in conscious awareness (supraliminal
ERPs) or below the objective detection threshold (subliminal ERPs). At the
same time, this study aims to contribute to the understanding of emotional
processes by demonstrating that ERPs are sensitive to the affective valence
of the stimulus, whether processed in awareness or outside of conscious
awareness;
3. A third interesting point of discussion is about variations in the time of
emergence of ERP as a function of the level of consciousness. In fact, as
was previously observed, delayed unaware information processing repre-
sents a distinctive feature of implicit visual perception (Bernat et al. 2001;
Junghöfer et al. 2001), and it was represented as a consequence of a more
complex cognitive process underlying unconscious stimulation. Neverthe-
less, some other studies have underlined an anticipated peak variation for
subthreshold stimuli compared to sovrathreshold ones (Brázdil et al. 1998,
2001). Thus, this theoretical point has to be explored more exhaustively.
4. Moreover, a cortical localization effect was analyzed. We expect a more pos-
terior distribution of negative peak variation (N2), as revealed by previous
studies (Sato et al. 2001). In fact, as pointed out by some experimental
results, emotional stimuli should elicit higher activation of visual areas,
covering a broad range of the occipito-temporal cortices relative to the
neutral stimuli. For this reason, we expected a quite differentiated cortical
distribution of N2.
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.5 (260-309)

Consciousness, emotion and face 

5. In addition to the localization effect for emotions, we aim to analyze the

lateralization effect due to the levels of consciousness. Specifically, Gaz-
zaniga (1993) suggested that the left hemisphere is crucial for conscious-
ness, because its dominance in response to conscious stimuli is reflected
in terms of the left-brain decoder. Nevertheless, some discrepancies from
this theoretical perspective were observed, and an opposite right lateral-
ization was revealed for conscious awareness (Brázdil et al. 2001; Henke et
al. 1993).

Methodology

Participants
Twenty three healthy volunteers took part in the study (eighteen women, age
21–25, mean = 22.17) after giving informed consent. They were students of
Psychology at the Catholic University of Milan, all right-handed and with
normal or corrected-to-normal visual acuity. They were invited to participate
to an experiment on cognitive processes related to stimulus perception and
comprehension.

Materials
Stimulus materials were taken from the set of pictures of Ekman and Friesen
(1976) (see Fig. 1). They were black and white pictures of a male actor, present-
ing respectively a happy, sad, angry or neutral face. Each face was presented ten
times, resulting in a total of 40 stimuli. Pictures were presented in a randomised
order in the center of a computer monitor.

Figure 1. The three prototypical facial expressions of emotions (sadness, happiness,

and anger) and neutral face
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.6 (309-375)

 Michela Balconi and Claudio Lucchiari

Table 1. Percentage values for the evaluation of the subliminal/supraliminal stimuli

Stimuli
% %
Supraliminal Correctly recognized Uncorrectly recognized
Anger 91 9
Happiness 93 7
Sadness 88 12
Neutral 92 8
Subliminal

Anger 44 56
Happiness 37 63
Sadness 41 59
Neutral 35 65

At the end of the experimental section, each subject was requested to

evaluate previous stimuli viewed, for both subliminal and supraliminal exper-
imental conditions (see next paragraph). As shown in the next table (Table
1), only the supraliminal condition subjects were able to recognize the emo-
tional stimuli, whereas, on the contrary, the subliminal condition subjects did
not correctly recognize the emotional stimuli previously viewed (target) with
a probability more than chance. In fact, the χ2 analysis applied to each type of
emotion in the supraliminal condition showed a correct identification for sad-
ness (χ2 1, 22 = 5.67, P = .02), anger (χ2 1, 22 = 8.12, P = .001) and happiness
(χ2 1, 22 = 6.05, P = .01), as well as neutral stimulus (χ2 1, 22 = 8.52, P =
.001). On the contrary, the subliminal subjects showed casual responses for
each type of stimulus (respectively, for sadness χ2 1, 22 = 1.57, P = .19;
anger χ2 1, 22 = 1.04, P = .25; happiness χ2 1, 22 = 1.84, P = .10; neutral
χ2 1, 22 = 1.77, P = .13).

Procedure
The inter-stimulus fixation point was projected at the center of the screen (a
white point on a black background). Subjects were seated comfortably in a
moderately lighted room with the monitor screen positioned approximately
100 cm in front of their eyes. During the examination, they were requested
to continuously focus their eyes on the small fixation point in the center of
the screen and to minimize blinking. The subject was told to observe the faces
carefully for a successive recognition task. This task was finalized to keep the
attentional level high during decoding of faces, since an implicit or explicit
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.7 (375-414)

Consciousness, emotion and face 

response to some features of the stimulus was not required. This was done
for three main reasons: to avoid confounding motor potentials in addition to
brain potentials; to avoid causing subjects to be more attentive to the emo-
tional stimuli than the neutral ones; finally, because asking for a response to
stimuli that are not being consciously seen can be nonsensical to participants.
Prior to recording ERPs, the subject was familiarized with the overall procedure
(training session), where every subject saw in a random order all the emotional
stimuli presented in the successive experimental session (a block of 10 trials,
each facial expression repeated twice).

Subliminal/supraliminal stimulation
Two experimental conditions were carried out (between-subjects independent
variable). In the supraliminal condition, twelve subjects consciously saw the
stimulus, which was presented for 500 ms on the monitor with an interstimulus
interval of 1500 ms. On the contrary, the other eleven subjects saw subthresh-
old stimuli. For the subliminal condition the stimulus duration was of 1 ms,
with an interstimulus interval of 1500 ms. Subliminal stimuli in the current
study meet objective detection threshold criteria. According to the signal de-
tection theory, when detection sensitivity is at chance, it is unlikely that there
is conscious awareness of the stimulus (Macmillan 1986; Snodgrass 2000).
The methods of measuring unconscious processes by presenting visual stim-
uli subliminally has a long history marked by methodological difficulties and
advances. Snodgrass (2000) argues that detection is sufficiently exhaustive to
result in a conscious perception, on the basis of signal detection theory (SDT)

EEG Registration parameters

The EEG was recorded with a 32-channel DC amplifer (SYNAMPS system) and
acquisition software (NEUROSCAN 4.0) at twelve electrodes (four central, Fz,
Cz, Pz, Oz; eight lateral, F2, F3, T2, T3, P2, P3, O1, O2) (international 10–
20 system) with reference electrodes at the mastoids. Electroculograms (EOG)
were recorded from electrodes lateral and superior to the left eye. The signal
(sampled at 256 Hz) was amplified and processed with a pass-band from .01 to
50 Hz and was recorded in continuous mode. Impedance was controlled and
maintained below 5 KΩ. An averaged waveform (off-line) was obtained from
about ten artifact-free (trials exceeding 50 µV in amplitude were excluded from
the averaging process) individual target stimuli for each type of emotion. Peak
amplitude measurement was quantified relative to 100 ms prestimulus. The
noise in the signal was low (7% epochs were rejected). To evaluate differences
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.8 (414-493)

 Michela Balconi and Claudio Lucchiari

in ERP response for the five facial expressions, we focused data analysis within
the time windows of 180–300 ms.

Data analysis

A three-way mixed design analysis of variance (ANOVA) condition (conscious

vs unconscious – between subjects) × type (emotional vs neutral – within-
subjects) × electrode sites (within-subjects) was applied to the peak amplitude
measure. Type I errors associated with inhomogeneity of variance were con-
trolled by decreasing the degrees of freedom using the Greenhouse-Geiser
epsilon. The analysis showed significant differences only for the main effect
of type F(1, 22) = 11.07, P = 0.001. A higher negative peak was revealed
for emotional faces (M = 1.56; Ds = 0.21) when compared to neutral
(M = 1.15; Ds = 0.28) faces. On the contrary, profiles of ERP for conscious
and unconscious conditions were undifferentiated F(1, 22) = 1.04, P = 0.20
(respectively for conscious M = 1.28; Ds = 0.19, and unconscious M = 1.38;
Ds = 0.15).
With respect of the site effect, an interaction type × site effect F(11, 22) =
10.20, P = 0.001 was observed, with a more temporal distribution of the peak
for emotional stimuli. In fact, the post hoc comparisons (Dunnet test) revealed
a significant difference between emotional (M = 1.88; Ds = 0.25) and neutral
(M = 1.37; Ds = 0.25) stimuli in T1 F(1, 22) = 6.61, P = 0.001 and T2
F(1, 22) = 4.58, P = 0.001. The wave profiles for temporal sites T1 and T2 are
represented in Figure 2a and 2b, as a function of the type of stimulus.
No other comparison was significant. Left and right brain differences as
a function of consciousness were explored in a second ANOVA (condition ×
side). In this case, no main effect was significant to the analysis (respectively for
condition F(1, 22) = 1.11, P = 0.25, and site F(11, 22) = 1.33, P = 0.30, ), nor
their interaction F(11, 22) = 1.48, P = 0.22. The next figure (Fig. 3) represents
left/right lateralization as a function of condition effect.
The ANOVA applied to the second dependent measure (latency) showed
a significant main effect of condition F(1, 22) = 7.98, P = 0.001). On the
contrary, type F(1, 22) = 1.01, P = 0.40 and electrodes F(11, 22) = 1.17, P =
0.20 were not significant, as well as their two- and three-way interactions. The
unconscious stimuli were temporally delayed (M = 220 ms; Ds = 17) when
compared to conscious ones (M = 200; Ds = 26). The absence of significant
result for type effect makes more evident that the temporal retardation of the
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.9 (493-493)

Consciousness, emotion and face 

N2 P3
+

2.0 ìV
ms

0 100 200 300 400 500

neutral facial expressions

emotional facial expressions

Figure 2a. Grand-average ERPs elicited by neutral and emotional facial expressions
in T1

N2 P3
+

2.0 ìV
ms

0 100 200 300 400 500

neutral facial expressions

emotional facial expressions

Figure 2b. Grand-average ERPs elicited by neutral and emotional facial expressions
in T2

180

160
left right
140
conscious unconscious

Figure 3. Lateralization effect as a function of condition (grand-average ERP) for N2

 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.10 (493-540)

 Michela Balconi and Claudio Lucchiari

peak appearance is related to the unconscious elaboration of the stimuli and

not to its emotional value.

Discussion

First of all, no other study has exhaustively analyzed the N2 ERP effect, with the
aim to analyze its functional significance for emotional face recognition pro-
cesses. The data support the view that emotion discrimination occurs at a first
stage of conceptual stimulus processing, and the current results indicate that
emotional facial expressions induce greater activation of the posterior tempo-
ral areas, with a latency of about 210 ms from stimulus onset. Therefore, the
first main point of the current research is that the N2 deflection appears strictly
related to the emotional value of faces (Streit et al. 2000). The negative deflec-
tion is not only due to global processing of the facial stimuli, but N2 is more
specifically related to the emotional expression decoding (Streit et al. 2000),
since N2 could represent an explicit marker of emotional facial expressions,
and not a generic detector of facial stimulus elaboration
Nevertheless, a successive positive ERP deflection (P300) has been moni-
tored by some authors after an emotional stimulation, even when it does not
seem to be exclusive for faces, since it has been observed even in response to
adjectives or objects with an emotional content (Bernat et al. 2001). P3 effect
could be more generally related to emotional content of the stimulus (pictures
with emotional content) or to the stimulus complexity, and this effect is viewed
as reflecting decision or cognitive closure of the recognition processing (Brázdil
et al. 2001; Iragui et al. 1999). In line with this hypothesis, previous results
on P3 applied to facial stimuli are contradictory: while in some studies the
neutral faces evoke lower amplitude than emotional ones (Carretié & Iglesias
1995), some other studies have found that neutral stimuli evoke the highest
peak (Vanderploeg et al. 1987).
As previously emphasized, differences in lateralization were found for N2.
The negative variation was heterogeneously distributed on the scalp, and the
temporal (both right and left) lateralization was emerging. In line with pre-
vious results, in fact, the posterior sites (specifically temporal left and right
sites) were observed as much more involved in emotional facial expressions
than neutral stimuli. In previous studies, neural networks have been demon-
strated for processing specific facial emotions, with the implicated regions
including cortical (mainly prefrontal and occipito-temporal junction) and sub-
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.11 (540-598)

Consciousness, emotion and face 

cortical structures (amygdala, basal ganglia and insula) (Damasio et al. 2000;
Gorno-Tempini et al. 2001).
Secondly, as observed in our results, ERP profiles were unaffected by the
supraliminal/subliminal stimulation. Similar ERP effects were observed, with
analogous morphological peak profiles, except for the temporal feature of the
ERP. Based on this analysis, similarities in processing between supraliminal and
subliminal stimulation can also be assessed: substantial analogies in the sub-
liminal and supraliminal ERP component structure were well-founded, sug-
gesting that similar neural activity is involved (Shevrin 2001; Snodgrass 2000);
and all these similarities between the classical N2 and its subliminal analogue
suggest that they are of a similar origin in both experimental conditions. By
obtaining subliminal and supraliminal ERPs to the same stimuli, the relation-
ship between conscious and unconscious affective processes can be examined.
Responses to stimuli in the subliminal and supraliminal conditions were simi-
lar in waveform, as well as sharing differences between emotional and neutral
stimuli. These similarities suggest that similar neutral pathways were engaged
during the two conditions, implying that some mental processess may oper-
ate on the same basis whether or not conscious perception is involved. More
generally, it seems evident that the information presented to a subject under
subliminal conditions may be perceived and processed on a higher level even if
the subject is not aware of this information.
Nevertheless, temporal retardation of the peak appears to distinguish be-
tween subliminal and supraliminal information processing. Peak latencies of
the corresponding deflections elicited by subliminally presented stimuli were
clearly distinct when compared to the latencies of the supraliminal N2. Sub-
liminally, the N2 peak is produced later.
In order to explain the temporal effect observed, we hypothesized that the
time required to elaborate the emotional information might be conditioned
by the type of stimuli viewed. The differences found between the two experi-
mental consciousness conditions would depend on the type of processing and,
more specifically, on the level of complexity and/or of the attentional effort
of the two kinds of cognitive processing implicated. As suggested by Brázdil
and coll. (1998), who found different patterns in mutual time relations in a
part of their sample (in some cases delayed and in the others anticipated), the
attention level differences of the subjects could be a valid explanation of the
heterogeneous results. But, since ERP provides information regarding the tem-
poral sequence of human information processing, further analysis of the time
relations between the supraliminal and the subliminal N2 peak latencies might
be fruitful in the future.
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.12 (598-643)

 Michela Balconi and Claudio Lucchiari

What are the implications of the reported research? First from a method-
ological standpoint, the research provides strong evidence for the assertion that
unconscious processess are instantiated in electrophysiological events. Although
often smaller in amplitude (Shevrin 2001), subliminal ERPs appear to have the
same component structure as supraliminal ERPs and may serve as markers for
unconscious process. Secondly, from the present results we can infer that sub-
liminal components are correlated with psychological process that are quite
similar to those occurring supraliminally. Therefore, our findings suggest that
at least some of our understanding of the supraliminal N2 effect may also apply
to unconscious N2 effect.
Finally, even if in the present research we have no results in favour of lat-
eralization of conscious processing, the question of possible lateralization of
awareness to the right side emerges. In fact, some authors suggest the crucial
role of the right hemisphere for consciousness, whereas others favor the left
hemisphere. The lateralization findings for the supraliminal ERPs call attention
to possible qualitative differences between conscious and unconscious affec-
tive processing, and this topic has to be explored systematically. Nevertheless,
it would be speculation to form a theory on the lateralization of consciousness
from sparse experimental data only. Clinical neurology can make an important
contribution to the discussion. For example, the dependency of neglect syn-
drome development accompanying right hemisphere lesions in most patients is
very well known. This syndrome represents a typical disorder of conscious per-
ception of stimuli in the contralateral hemispace. Furthermore, several papers
have confirmed preserved implicit perception in some patients with neglect
syndrome. Similar clinical experiences with dissociations between perception
and consciousness after brain damage support the crucial role of right-sided
structures in conscious processing of perceived visual signals.
On the whole, concerning our initial twofold question (structural and psy-
chological similarity of subliminal and supraliminal ERP profiles), on one hand
we can answer in favor of the structural similarity of the two ERPs. On the other
hand, the psychological homogeneity of subliminal and supraliminal stimula-
tion is a main point that requires more attention, and which we must explore
in the future.

Note

* Addres enquiries to Michela Balconi, Ph.D., Department of Psychology, Catholic Univer-

sity, L. Go Gemelli, 1, 20123 Milan, Italy; e-mail: michela.balconi@unicatt.it
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.13 (643-737)

Consciousness, emotion and face 

References

Adolphs, R., Tranel, D., & Damasio, A. R. (1998). The human amygdala in social judgment.
Nature, 393, 470–474.
Balconi, M., & Pozzoli, U. (2003a). ERPs (event-related potentials), semantic attribution,
and facial expressions of emotions. Consciousness & Emotion, 4, 63–80.
Balconi, M., & Pozzoli, U. (2003b). Face-selective processing and the effect of pleasant
and unpleasant emotional expressions on ERP correlates. International Journal of
Psychophysiology, 49, 67–74.
Batty, M., & Taylor, M. J. (2003). Early processing of the six basic facial emotional ex-
pressions, Cognitive Brain Research, 17, 613–620.
Bernat, E., Bunce, S., & Shevrin, H. (2001). Event-related brain potentials differentiate
positive and negative mood adjectives during both supraliminal and subliminal visual
processing. International Journal of Psychophysiology, 42, 11–34.
Bernstein, L. J., Beig, S., Siegenthaler, A. L., & Grady, C. L. (2002). The effect of encoding
strategy of the neural correlates of memory for faces. Neuropsychololgia, 40, 86–98.
Brázdil, M., Rektor, I., Daniel, P., Dufek, M., & Jurák, P. (2001). Intracerebral event-related
potentials to subthreshold target stimuli. Clinical Neurophysiology, 112, 650–661.
Brázdil, M., Rektor, I., Dufek, M., Jurák, P., & Daniel, P. (1998). Effect of subthreshold
stimuli on event-related potentials. Electrophysiology and Clinical Neurophysiology, 107,
64–68.
Bunce, S., Bernat, E., Wong, P. S., & Shevrin, H. (1999). Further evidence for unconscious
learning: Preliminary support for the conditioning of facial EMG to subliminal stimuli.
Journal of Psychiatry Research, 33, 341–347.
Carretié, L., & Iglesias, J. (1995). An ERP study on the specificity of facial expression
processing. International Journal of Psychophysiology, 19, 183–192.
Damasio, A. R., Grabowski, T. J., Bechara, A., Damasio, H., Ponto, L. L., Parvizi, J. (2000).
Subcortical and cortical brain activity during the feeling of self-generated emotions.
Nature Neuroscience, 3, 1049–1056.
Eimer, M., & McCarhty, R. A. (1999). Prosopagnosia and structural encoding of face:
Eveidence from event-related potentials. Neuroreport, 10, 255–259.
Ekman, P., & Friesen, W. V. (1976). Pictures of facial affect. Palo Alto: Consulting Psychologist
Press.
Gazzaniga, M. S. (1993). Brain mechanisms and conscious experience. Experimental and
theoretical studies of consciousness. In CIBA Fundation Symposium 174 (pp. 247–262).
Chichester: Wiley.
Gorno-Tempini, M. L., Pradelli, S., Serafini, M., Baraldi, P., & Porro, C. (2001). Explicit and
incidental facial expression processing: An fMRI study. Neuroimage, 14, 465–473.
Grelotti, D. J., Gauthier, I., Schultz, R. T. (2002). Social interest and the development
of cortical face specialization: What autism teaches us about face processing.
Developmental Psychobiology, 40, 213–225.
Haxby, J. V., Hoffman, E. A., & Gobbini, I. M. (2000). The distributed human neural system
for faces perception. Trends in Cognitive Science, 4, 223–233.
Henke, K., Landis, T., & Markowitsch, H. J. (1993). Subliminal perception of pictures in the
right hemisphere. Consciousness and Cognition, 2, 225–236.
 JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.14 (737-825)

 Michela Balconi and Claudio Lucchiari

Herrmann, M. J., Aranda, D., Ellgring, H., Mueller, T. J., Strik, W. K., Heidrich, A., &
Fallgatter, A. J. (2002). Face-specific event-related potential in humans is independent
from facial expression. International Journal of Psychophysiology, 45, 241–244.
Iragui, V. J., Kutas, M., Mitchiner, M. R., & Hillyard, S. A. (1993). Effect of aging
on event-related brain potentials and reaction times in an auditory oddball task.
Psychophysiology, 30, 10–22.
Junghöfer, M., Bradley, M. M., Elbert, T. R., & Lang, P. J. (2001). Fleeting images: A new
look at early emotion discrimination. Psychophysiology, 38, 175–178.
Kanwisher, N., McDermontt, J., & Chun, M. M. (1997). The fusiform face area: A module
in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17,
4302–4311.
Kihlstrom, J. F., Barnhardt, T. M., & Tataryn, D. J. (1992). Implicit perception. In R. F.
Bornstein & T. F. Pittam (Eds.), Perception without awareness: Cognitive, clinical, and
social perspectives (pp. 17–54). New York: Guilford.
LeDoux, J. E. (1990). Fear pathways in the brain : Implications for a theory of the emotional
brain. In Brain, P. F., Parmigiani, S., Balnchard, R. J., & Mainardi, D. (Eds.), Fear and
defence (pp. 112–139). London: Harwood.
LeDoux, J. E. (1996). The emotional brain: The mysterious underpinning of emotional life.
New York: Simon and Schuster.
Macmillan, N. (1986). The psychophysics of subliminal perception. Behavioral Brain
Sciences, 9, 38–39.
Öhman, A. (1999). Distinguishing unconscious from conscious emotional processes:
Methodological considerations and theoretical implications. In Dalgleish, T., & Power,
M. J. (Eds.), Handbook of cognition and emotion (pp. 112–128). Chichester: Wiley.
Posamentier, M. T., & Abdi, H. (2003). Processing faces and facial expressions. Neuro-
psychology Review, 13, 113–143.
Posner, M. I. (1978). Chronometric explorations of mind. Erlbaum: Hillsdale.
Sato, W., Takanori, K., Sakiko, Y., & Michikazu, M. (2000). Emotional expression
boosts early visual processing of the face: ERP reecording and its decomposition by
independent component analysis. Neuroreport, 12, 709–714.
Shevrin, H. (2001). Event-related markers of unconscious processes. International Journal of
Psychophysiology, 42, 209–218.
Shevrin, H., Bond, J. A., Brakel, L. A. W., Hertel, R. K., & Williams, W. J. (1996). Con-
scious and unconscious processess: Psychodynamic, cognitive, and neurophysiological non–
vergences. New York: Guilford Press.
Shevrin, H., & Fritzler, D. (1968). Visual evoked response correlates of unconscious mental
processes. Science, 161, 295–298.
Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information
processing II. Perceptual learning, automatic attending, and a general theory.
Psychological Review, 84, 127–190.
Snodgrass, J. M. (2000). Unconscious perception: Theory, method, and evidence. Amsterdam:
John Benjamins.
Streit, M., Wölwer, W., Brinkmeyer, J., Ihl, R., & Gaebel, W. (2000). Electrophysiological
correlates of emotional and structural face processing in humans. Neuroscience Letters,
278, 13–16.
JB[v.20020404] Prn:23/11/2004; 10:06 F: Z13105.tex / p.15 (825-844)

Consciousness, emotion and face 

Tranel, D., & Damasio, A. R. (1985). Knowledge without awareness: An autonomic index of
facial recognition. Science, 228, 1453–1454.
Vanderploeg, R. D., Brown, W. S., & Marsh, J. T. (1987). Judgments of emotion in words
and faces: ERP correlates. International Journal of Pshycophysiology, 5, 193–205.
Wong, P. S., Shevrin, H., & Williams, W. J. (1994). Conscious and nonconscious processes:
An ERP index of an anticipatory response in a conditioning paradigm using visual
masked stimuli. Psychophysiology, 31, 87–101.
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.1 (41-120)

Phenomenal consciousness, sense

impressions, and the logic of ‘what it’s like’

David Beisecker
University of Nevada, Las Vegas

Many philosophical puzzles about phenomenal consciousness are raised in

terms of “what it’s like” to have certain experiences: to see red or taste a
pomegranate, to be in love or to be afraid, to wake up with a hangover,
maybe even to be a bat. To the extent that such discourse is dismissed as
unintelligible, one fails to have a satisfactory theory of phenomenal
consciousness. Yet such discourse notoriously resists analysis in purely
physiological or functional terms. The “hard problem” of consciousness
arises because statements involving the qualitative character of experience –
or “qualia” – bear no obvious logical connections to descriptions couched in
physical or physiological terms. Nevertheless, we shouldn’t abandon all hope
for unraveling the mystery of phenomenal consciousness. In this paper, I
argue that the appearance of an explanatory gap arises because talk about
“what it’s like” performs an important, yet perfectly unmysterious, epistemic
function that cannot be played by ordinary physical vocabulary. Once this
function is understood, we can readily account for many of the puzzling
things we are tempted to say about phenomenal consciousness, including
those statements taken as evidence for the hard problem – namely,
attributions of knowledge of what it’s like to have various experiences as well
as inter- and intra-personal comparisons of the qualitative character of
experience.

Keywords: hard problem, qualia, knowledge argument, sense impressions

What it’s like: The vulgar and the philosophical

Many philosophical puzzles about phenomenal consciousness are raised in

terms of “what it’s like” to have certain experiences: to see red or taste a
pomegranate, to be in love or to be afraid, to wake up with a hangover, maybe
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.2 (120-159)

 David Beisecker

even to be a bat. To the extent that such discourse is dismissed as unintelligible,

one fails to have a satisfying theory of phenomenal consciousness. Yet such dis-
course notoriously resists analysis in purely physiological or functional terms.
The so-called “hard problem” of consciousness arises because statements in-
volving the qualitative character of experience – or “qualia” – bear no evident
logical connections to descriptions couched in physical, physiological, func-
tional, or even intentional terms. Nevertheless, we shouldn’t abandon all hope
for unraveling the mystery of phenomenal consciousness. Just as so-called se-
mantic accounts of truth aim to make sense of proper applications of the truth
predicate, a semantic account of phenomenal consciousness would attempt to
regiment talk about “what it’s like” to have certain experiences. The task of
providing a logic for such talk hasn’t been seriously pursued, and too quickly
dismissed. This paper attempts to address that lacuna. In it, I argue that the
appearance of an explanatory gap arises because talk about “what it’s like”
performs an important, yet perfectly unmysterious, epistemic function that
cannot be played by ordinary physical vocabulary. Once this function is under-
stood, we can readily account for many of the puzzling things we are tempted to
say about phenomenal consciousness, including those statements taken as evi-
dence for the hard problem – namely, attributions of knowledge of what it’s like
to have various experiences as well as inter- and intra-personal comparisons of
the qualitative character of experience.
Before we begin, however, we need to distinguish the philosophical con-
ception of what it’s like from more ordinary notions. For while the former, so-
phisticated idea is presumably an extension of less sophisticated counterparts,
everyday talk about what it’s like differs from rarefied philosophical parlance.
When the vulgar (or those not trafficking in the consciousness industry) ask
one another to describe “what it’s like” to have a certain experience, they’re
generally interested in the experience’s effect upon the subject’s psychologi-
cal and emotional makeups. Inter- and intra- personal comparisons of what
it’s like are thus relatively unproblematic. Your reactions to skydiving might
be similar to my reactions to snowboarding. And your first taste of Arrogant
Bastard Ale (the bottle says you’ll probably not like it) might be very differ-
ent from a subsequent taste, if and when you grow to appreciate an exquisitely
hopped beer.
However, the experiences that form the stock and trade of philosophical
discussions of phenomenal consciousness (e.g., seeing red, smelling creosote,
or hearing a buzzing sound) are often too thin to elicit distinctive reactive atti-
tudes. Furthermore, we are tempted to say of those very reactive or emotional
attitudes (anxiety, excitement, and the like), that there is something it is like
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.3 (159-199)

The logic of ‘what it’s like’ 

to have them. So when those in the consciousness biz talk about what it’s like
to have a certain experience, they’re apparently after something else: the qual-
itative character of an experience, or its so-called “qualia.” Such a conception
is meant to answer to widely-held intuitions that the “subjective feeling” of
experiencing a particular property might have been other than what it actu-
ally is, and that how it feels for one to have a certain experience could differ
from how that same experience feels to another. Rather than ridiculing this so-
phisticated conception as sophistical philosophical nonsense, in the hopes that
it will eventually just go away (an attitude copped, for instance, by Dennett
1988), the aim of this paper is to clarify many of the funny things we in the biz
are tempted to say about this qualitative dimension of phenomenal experience.
That is, I think an account of qualia (or “what it’s like”) can be constructed,
which respects these intuitions, or at least renders them intelligible.

Sellarsian sense impressions

Let’s begin with an almost embarrassingly mundane observation. Competent

speakers of natural languages must learn how to apply observation vocabulary
in experience.
We all need to be trained to make even the most basic observation reports.
Our ability to classify things as red, or even as looking red is not innate. While we
have biologically innate predilections for certain classification schemes, other
speakers of our language must teach us how to make observation reports that
accord with the specific classificatory dispositions of our fellows. Subjects thus
face the task of coordinating or calibrating states of themselves with the ap-
plication of observation concepts in experience. Simply put, speakers must
learn to report the presence of a certain property (e.g., red) whenever they
are struck in certain fashions – that is, whenever they are in certain internal
discriminatory states. And they must further learn when to restrain their ac-
quired dispositions to report such a presence when circumstances are such that
the subject’s being in a particular state is not a reliable indicator of something’s
actually exhibiting that property. In those circumstances, subjects learn to say
that it only “looks” or “seems” as if that property is present. They learn to re-
port merely that they are stimulated in a way that, under normal circumstances,
would reliably indicate that property’s presence.
Now even though these internal discriminatory states are presumably
physiological states (and states of the nervous system in particular), subjects
are typically not able to identify them in physiological terms. Even our most ac-
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.4 (199-274)

 David Beisecker

complished neuroscientists have trouble identifying the neural correlates of the

simplest sensations. The point here is that ordinary observation vocabulary –
including ‘looks’-talk – is conceptually prior to a developed neuroscience. In
speaking about the task facing speakers as they learn to apply observation con-
cepts in experience, we should remain theoretically non-committal regarding
the underlying physiological substrate.
In the final episode of his “Myth of Jones,” Wilfrid Sellars (2001, Part XVI)
showed us how the philosophical notion of a sense impression could be in-
troduced as a theoretically silent way to refer to such discriminatory states of
subjects. Roughly, Sellars regarded a sense impression of some particular per-
ceptible property as the imprint that is characteristically left upon us by the
presence of that property under normal circumstances, which can then be used
by that subject to elicit non-inferential judgments of its presence.1 And so the
task described above is that of subjects learning how to coordinate sense im-
pressions with the application of appropriate observation concepts. To borrow
a term from Dretske (1988), one might say that subjects must face the task of
“recruiting” appropriate internal states of theirs to play the role of indicators of
particular observable features of the world.
One of Sellars’ chief accomplishments was showing how once sense im-
pressions have been introduced into our vocabulary, it turns out to be a trivial
trick for subjects to come to “observe” when they’re having certain sense im-
pressions. Following Lycan (1996), one might say that subjects acquire a sort
of “higher-order perception.” To report the presence of a red sense impression,
subjects merely need to detect when they are disposed to report when some-
thing is or looks red to them. Thus subjects have a readily explicable authority
over reports about their own sense impressions that they lack in the case of
reports about external qualities and the sense impressions of others. More-
over, since subjects can be trained to make observation reports of their own
sense impressions, one might be tempted to say that those impressions leave
their own “higher-order” impressions. Curiously enough, since the conditions
in which one is disposed to report the presence of a red sense impression are
precisely those in which something is or looks red to that subject, the sense im-
pression of red turns out to be equivalent to the impression of that impression.
It pleasantly follows that reports about sense impressions are, what you might
say, “phenomenally transparent.”
The overall Sellarsian strategy is to work from the outside in, rather than
the inside out. Our ideas of inner sense impressions derive from our concepts of
the properties they are impressions of. To believe otherwise – to think that our
taxonomies of sense impressions are conceptually prior to external properties –
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.5 (274-307)

The logic of ‘what it’s like’ 

is, like classical empiricism, to subscribe to the “myth of the given.” Though
Sellars doesn’t do so in Empiricism and the Philosophy of Mind, this account
of sense impressions extends nicely to the self attribution of other psychologi-
cal states said to have phenomenal content. I’m thinking here of inner sensory
states like pain, brute cravings or urges, moods, and of course, emotional states
like anger, fear or exhilaration. When acquiring the concepts of such states, we
learn that they have behavioral and expressive cores; their originary application
is to subjects reacting or expressing themselves in certain ways to specific types
of situations. As Darwin famously advanced, there appear to be characteris-
tic responses and facial expressions associated with, and indicative of, many
of the basic emotions. Put much too crudely, fearful creatures typically react
to situations they find threatening by cowering or fleeing, while angry crea-
tures typically bare their teeth and show aggression. By the same token, there
are characteristic reactions to bodily harm that we associate with creatures in
pain. We then discover that just as in the case of sensing external qualities, we
can in our own case associate such inner sensory, emotional, and psychological
states with particular Sellarsian sense impressions. They each have their dis-
tinctive “feels” for us, such as the characteristic flush of anger or the trembling
and quickening of the heart and respiratory rates associated with fear. Using
these sense impressions, we can effortlessly and reliably report the applicabil-
ity of such concepts to our own person without having to observe our own
outward behavior, and can even begin to apply them in our own case to those
non-standard situations in which such feelings fail to produce the standard
responses (i.e., fear that doesn’t result in flight). And we can begin to under-
stand similar self-reports in others. Consequently, we can readily account for
the privileged access we appear to have with respect to many of our own psycho-
logical and emotional states that we don’t have when we consider those states
in others, and in our own case, we begin to apply these concepts in experience,
and so make non-inferential “observations” of our own states of mind.2
For our purposes, several features of this broadly Sellarsian account of
sense impressions warrant further comment:
(1) To begin with, sense impressions are functionally specified, understood as
internal discriminatory states that dispose subjects to make certain observa-
tion reports. As such, they are not identified in physical or physiological terms
and could be realized in myriad ways. Different sense impressions of the same
perceptual quality might have vastly different intrinsic constitutions. A sense
impression of red for a typical human might be realized in a wholly different
manner in a bug-eyed alien . . . or a bat (if indeed, bats have such impressions at
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.6 (307-355)

 David Beisecker

all). Sense impressions do not have to be similar even between members of the
same species. To take a striking example, persons with synaesthesia appear to
have recruited sense impressions governing their application of various obser-
vation concepts, which are less discriminatory than normal folk, and which are
subject to an unusual range of non-standard conditions. Conditions in which
they are inclined to say something “looks red” to them (such as the presence
of particular numbers or letters) can be quite different, and slightly more ex-
tensive, than the conditions in which I’m apt to say something looks red to me.
Since they are functionally specified, sense impressions are not to be identified
with states of the central nervous system, although they might be realized by
such states. The precise manner in which different sense impressions are re-
alized in creatures like us is a matter of empirical investigation and discovery.
Though it might seem a bit weird, there is nothing in the bare notion of a sense
impression that would prohibit non-material realizations of sense impressions.
The bare notion of a sense impression is thus neutral between materialism
and dualism.
(2) Sense impressions do not have to possess the qualities that they are impres-
sions of. A sense impression of red is most likely not itself red.3 Nevertheless,
sense impressions stand (or are supposed to stand) in containment and exclu-
sion relationships to one another in much the same way that the features that
they are impressions of stand in relation to one another. This thought is all we
need to capture the truth behind the idea that sense impressions are (or at least
should be) images of the external world. Just as an instance of some determi-
nate shade of red, such as scarlet or crimson, is at the same time an instance
of the determinable red, the particular internal state that realizes (or plays the
role of) a sense impression of scarlet at the same time realizes a sense impres-
sion of red. That is, the class of red sense impressions includes that of scarlet
sense impressions. Similarly, red and green sense impressions are supposed to
exclude one another just like properties of redness and greenness presumably
exclude one another in the external world.4
(3) The richness of our sensory manifold far outstrips the observational dis-
criminations we are able to draw in natural language. Our sense impressions
are variegated in ways that we cannot express through our concepts. And it is
evident we could recruit our sense impressions to indicate features of ourselves
and the world other than those that we actually do recruit them to indicate.
Subjects can thus have sense impressions of particular perceptible and emo-
tive qualities without yet having acquired the concepts of those qualities. Even
non-linguistic creatures could be understood (by us) to have sense impressions.
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.7 (355-410)

The logic of ‘what it’s like’ 

They just don’t recognize them as such. Insofar as there is some characteristic
way in which red things in normal circumstances strike a dog, allowing them to
discriminate red things, we can regard dogs as having sense impressions of red
(they might of course turn out to be color-blind). But dogs are not conscious
of their sense impressions; they aren’t aware of them as such and certainly don’t
report their presence. So while we might credit them with the sentience pos-
sessed by any creature in sensory contact with its environment, we need not
credit them with the full-blown sapience enjoyed by concept-mongering crea-
tures. And insofar as dogs do not endeavor to coordinate their own sensory
states with the (verbal) application of observation concepts, we need not con-
clude that they have any idea of what it’s like for them to see red (though there
might be something that it’s like for us to view red things through doggy eyes).
If you happen to think that dogs do apply observation concepts, then I kindly
ask you to slide further on down the great chain of being.

(4) Finally, although we classify sense impressions in terms of particular sense

contents, we do not have to regard these contents as part of their intrinsic
natures. There is no need to assign sense impressions any “non-conceptual in-
tentional content” (Tye 2000), or meaning that isn’t parasitic upon conceptual
meaning. As mentioned above, though some non-linguistic creatures might
have red sense impressions, there is no need to think that there is any sense
in which these impressions are red to them. More significantly, the physiolog-
ical states that turn out to realize my sense impressions of red might not have
done so. Had the world been other than the way it is (or had I been outfit-
ted with those color-inverting lenses of philosophical legend), then that state
might instead have been recruited as an indicator of green. In fact, the phys-
iological state that serves now as my sense impression of red might not serve
that function in the future. We must continue to recalibrate our dispositions
to report when things are or seem red to us (a monitoring capacity that would
seem to be unavailable to so-called “blindsighted” individuals). As my percep-
tual equipment ages, it might undergo a “red shift.” The physiological state that
once was recruited to be my sense impression of red might no longer be a re-
liable indicator of redness in my environs, in which case I would be obliged to
recruit a different state (should one still be available, and I not have become
color-blind).
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.8 (410-445)

 David Beisecker

The function of ‘what it’s like’

With those remarks about Sellarsian sense impressions behind us, we can now
proceed to the punch line. I submit that many of the curious things philoso-
phers have said about phenomenal consciousness or the qualitative character
of conscious experience can be understood as making perfectly straightforward
assertions about sense impressions understood in this Sellarsian fashion. In
particular, the expression ‘what it’s like’ (in its philosophical sense) functions
as an elliptical and pre-theoretical means of picking out the particular manner
in which a subject realizes a certain kind of sense impression. Such a proposal
makes evident and intuitive sense of inter- and intra-personal comparisons of
the qualitative dimension of experience. Due to differences in our perceptual
constitutions and the structures of our perceptual modalities, what it’s like for
us to see red might be completely different from what it’s like for a bat or bug-
eyed alien to see red, so much so that we could not understand what it would be
like to be either. And while what it’s like for me to see red is presumably pretty
much what it’s like for most everyone else, this generalization can break down
in the case of folk like synaesthetes, who have recruited non-standard sense im-
pressions to indicate externally observable properties. Furthermore, since my
perceptual apparatus could change over time, even though this change is so
gradual that I fail to realize it, what it’s like for me to see red now might well
not be what it’s like for me to see red in the distant future or past. It would even
seem to be possible (albeit highly improbable) for what it’s like for me to see
red eventually to shift all the way across the spectrum and become what it’s like
for me to see green. In effect, I would have been outfitted over time with those
color-inverting lenses of philosophical legend.
One appealing aspect of this proposal is that it allows certain internalist
intuitions or prejudices to take hold, while at the same time enabling us to
be externalists about intentional or representational content. On this account,
internally indistinguishable subjects (those “molecule-by-molecule” duplicates
of philosophical fantasy) will have experiences with similar phenomenal char-
acters, even though external considerations dictate that the representational
contents of their experiences are radically different.5
On this account, it is the functional unspecificity of sense impressions that
allows one to suspect that what it’s like to have a particular kind of experience
could have been other that what it actually is. This is precisely the kind of in-
tuition that allows so-called “modal arguments” for the explanatory gap to get
off the ground (Chalmers 1996; Levine 1983). And the present proposal nicely
affords two distinct ways to cash it out. On the one hand, ‘what it’s like’ can
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.9 (445-480)

The logic of ‘what it’s like’ 

be taken to refer rigidly to the indicated sense content, allowing the internal
discriminatory state to vary. Read in this fashion, the thought that what it’s
like to have a certain sort of experience could have been otherwise means that
it is conceivable for one to have recruited some other internal discriminatory
state to indicate that sense content. Alternately, ‘what it’s like’ can be taken to
rigidly pick out the internal discriminatory state, allowing the indicated sense
content to vary. On this interpretation, the thought above asserts that that in-
ternal discriminatory state could have been recruited to indicate an entirely
different sense content. Taken together, these separate thoughts generate the
idea that ‘what it’s like’ is determined neither by a subject’s physiology nor by
the contents of its intentional states. But now we can see that these intuitions
in favor of an explanatory gap are motivated by different ways in which one
can pick out the referent of ‘what it’s like.’ In other words, the appearance of an
explanatory gap trades on a readily explicable referential ambiguity.6
Perhaps more significantly, this story about sense impressions can help
untangle the issues surrounding the so-called “knowledge argument.”7 Sellars
himself famously thought that sense impressions play little role justifying em-
pirical knowledge. He also had little truck with the skeptical worries generated
by absent and inverted qualia scenarios. Nevertheless, it’s clear Mary (the color
benighted cognitive scientist in Frank Jackson’s oft-discussed thought exper-
iment) has not come to face the task that the language of sense-impressions
has been introduced to describe – namely that of coordinating her own in-
ternal states with the application of particular observation concepts in ex-
perience. The knowledge argument gains its force, because it’s unclear how
Mary’s assumed vast knowledge of physiological facts ever could help to over-
come this task. Doing brain science by itself will not tell Mary when to report
non-inferentially the internal occurrence of a red sense-impression.
When Mary escapes her black and white environment (a place I like to
think of as the Kansas depicted in The Wizard of Oz) and spies a ripe tomato,
which she somehow already knows to be red, she exclaims, “Ah, so this is what
it’s like to see red.” What has she learned, and what kind of thought does
her exclamation express? It seems reasonable to suppose that her exclamation
amounts to the claim that she is now having a red sense impression. If that is
the case, then it would be true enough, and it would be a piece of knowledge
unavailable to her till now, simply because she’s never had the opportunity to
have such an impression. But there’s nothing in this that can’t be cashed out in
materialist terms. So if that was the only kind of knowledge that Mary gains,
then the so-called knowledge argument would lack the anti-materialist force
its proponents take it to have. Instead, Mary is claimed to gain some further
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.10 (480-537)

 David Beisecker

knowledge: namely “knowledge of what it’s like” to see red. What, then, does
that amount to?

Knowing what it’s like

I propose that attributions of knowledge of what it’s like to have a certain

kind of experience should be understood as claims that a subject has a justi-
fied ability to apply corresponding observation concepts in experience. While
most speakers would qualify as capable enforcers of the norms governing color
reports, it would be irresponsible to extend this authority to just anyone. A
color-blind person would obviously be an incompetent teacher of color terms,
even if he knows a great deal about the human visual system, as well as the infer-
ential connections between colors and other empirical concepts. It would also
be reasonable to withhold this authority from those, such as young children,
who haven’t been sufficiently indoctrinated into our reporting practices. Un-
like the color blind, Mary (we suppose) has the potential to make accurate color
discriminations. She also knows all the inferential connections color terms bear
to the other terms in our language. However, until she justifiably demonstrates
that she can apply color concepts in experience as reliably as competent speak-
ers, we can reasonably deny that she truly knows what it’s like to have perceptual
experiences of color.8 She clearly lacks a justificatory status, which manifests it-
self in our reasonable reluctance to grant her authority enforcing the norms
governing our observational vocabulary. And this would be so, even if she hap-
pens to possess an uncanny ability to make accurate color discriminations. As
a result, we can say that knowing what it’s like requires more than just having
a perceptual or recognitional ability; one must be in a position to justify this
discriminative capacity as well.9
Notice crucially that I haven’t claimed that Mary is unable to entertain
any specific beliefs. Even during her stint in Kansas, she might suspect that
something looks red to her yet fail to know this, for she fails to have the ap-
propriate experience to justify this suspicion. Thus we can hang onto Jackson’s
conclusion that pre-release Mary lacks propositional knowledge that we would
normally express with observation reports. Rather than missing the ability to
form certain beliefs, she lacks the history or experience required for her to en-
tertain those beliefs responsibly. And it is her assumed epistemic responsibility,
not simply her lack of experience, which really prevents her from ever having
entertained beliefs that things look red to her.
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.11 (537-581)

The logic of ‘what it’s like’ 

Consider how this proposal applies to those ever-popular subjects of philo-

sophical fantasy: our physical and functional duplicates spontaneously gener-
ated out of swampmuck. Such abominations might make all sorts of claims
about how things look to them, and they might try to convince us that they
have the requisite experience and know-how to enforce the norms governing
our observation vocabulary. But the justifications they give for entitlement to
this authority would fail, simply on the grounds that they would be false. So
even though, by some remarkable coincidence, some such beings happen to
have the discriminatory capacity and classificatory dispositions to be compe-
tent reporters, it still would be inappropriate for us to so treat them. For if it
truly were a cosmic coincidence that they have this gift, then we would be in no
position to responsibly believe this to be the case.10
Insofar as their perceptual apparatus differs from our own, we’d also be
justifiably reluctant to grant perceptually exotic creatures – bats, bots, or bug-
eyed aliens, for instance – the authority to govern the use of our color terms.
Lacking the perceptual capacities to employ our observation concepts in expe-
rience, they might not ever attain the status of full-fledged (norm-enforcing)
members of our linguistic community. In particular, their different physiol-
ogy might well prevent such beasts from being able to tell when things are
likely only to “look red” to a human observer. That is, their different perceptual
equipment might prevent them from anticipating our justifiable perceptual er-
rors. Likewise, we would be unable to master fully a bat’s observation concepts.
Moreover, their different responsive repertoires might preclude them from hav-
ing anything more than a rudimentary grasp of what it’s like to have our
emotional attitudes (and vice-versa). Hence we can respect the intuition that
we are unable to know what it’s like to be a bat, without having to claim we
can’t so much as entertain the same beliefs or that there is some sort of special
phenomenal content wholly unavailable to us. Some perceptually exotic crea-
tures might even make the same color discriminations that we do (in their own
terms, of course). The conditions of proper application for some of their ob-
servation concepts would mirror those of our own color concepts. Still, if we
lack sufficient contact with these creatures to justifiably believe this extensional
equivalency, we can reasonably deny them the authority to enforce the rules
governing the use of our color concepts. So while they would know what it’s
like for them to see red, they might not know what it’s like for us to see red.
This last point shows how we can deal with an objection that could be
raised to approaches like the one advanced here, which tie knowing what it’s
like to have certain experiences with linguistic mastery or grasp of particu-
lar perceptual concepts. The objection has us suppose that Mary, while still in
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.12 (581-622)

 David Beisecker

Kansas, has gained access to several unlabeled paint chips, at least one of which
happens to be red. With that chip in view, Mary comes to have her first red
visual sense impression, and so one might be tempted to say that at that point
she comes to know what it’s like to see red. Yet she fails to realize that it is a red
sense impression that she has come to experience. So while she comes to know
what it’s like to see red, she nevertheless does not yet have a justified ability
to apply the concept red in experience. It might seem, then, that our proposed
analysis of knowing what it’s like fails.
But this objection ignores the perspectival nature of knowledge attribu-
tions generally. While I grant that Mary has failed to demonstrate mastery over
our concept of red, presumably she can still classify future visual experiences
as being of roughly the same type as she has when she views the red paint chip.
Thus we might still claim that she has acquired a justified ability to apply an
observation concept which she could demonstratively identify as “the shade of
that chip,” and which turns out to be more or less extensionally equivalent to
our concept of red. And so, in a de re sense (or from our perspective), we might
say of the property red, that Mary has learned what it’s like to see it. But in a de
dicto sense (or from the perspective of her concepts), we can reasonably deny
that she knows what it’s like to see red. Not until Mary comes to realize that her
experience is one that we would classify as a seeing of red, would she character-
ize herself as knowing what it’s like to see red. Once we register that attributions
of knowledge of what it’s like admit to the same de dicto/de re distinctions as
attributions of knowledge more generally, we can see that the objection fails
to provide a true counterexample to the proposed analysis. Indeed, I take this
consistency with other types of knowledge attributions to be a great virtue of
the present account.
In sum, the knowledge argument rests on the fact that one cannot discern
whether subjects know what it’s like to have an experience just by examining
the causal transactions inside their heads as they have those experiences. But
that merely shows that such a narrow focus abstracts away from the epistemi-
cally significant, historical facts required for them to have such knowledge. The
proper moral isn’t that facts about phenomenal consciousness must remain
objectively ineffable, for these further social and historical features are by no
means inaccessible from a third-person perspective. Moreover, we can finally
see why subjects would find it important to possess knowledge of what it’s like.
For justifiably applying observation concepts in experience entitles subjects to
do things forbidden to the inexperienced. Though “swampzombies” might be
inclined to behave as I do, others would be disposed to treat them differently.
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.13 (622-670)

The logic of ‘what it’s like’ 

Hence having knowledge of what it’s like really can matter or make a difference
to conscious subjects.11

Conclusion

By making sense of the puzzling things philosophers are inclined to say about
“what it’s like” to have certain experiences, I think that I’ve begun to sketch
a perfectly unmysterious account of phenomenal consciousness or “qualia.”12
Now some might object that such an approach is too deflationary to explain
phenomenal consciousness (Chalmers 1996: 186–189). While it might explicate
our judgments about the qualitative character of experience, it fails to address
the puzzling features of phenomenal consciousness itself. It’s no more com-
mitted to the genuine existence of qualia than explications of religious belief
must be committed to the existence of deities. But this objection ignores a re-
spect in which this account is not eliminativist. I haven’t simply told a story
about how we come to believe in phenomenal consciousness; I’ve also shown
how free-standing claims about what it’s like to have certain experiences can be
true. Thus this account licenses beliefs in phenomenal consciousness itself, not
just beliefs in those beliefs.13
Indeed, a semantic account of phenomenal consciousness such as the one
advanced here, would seem tailor-made to bridge the explanatory gap. Those
who subscribe to the gap occasionally bid us to imagine beings who are phys-
iologically and psychologically similar to us, and who live in physically similar
environments. They then claim that despite these similarities, it is conceivable
for such beings to differ from us phenomenally. Since there could be phe-
nomenal differences without any corresponding differences in the physical,
physiological, and psychological circumstances, it is argued that physical or
physiological explanations of consciousness are not in the offing.
So are there any unmysterious differences between us and the denizens of
those imagined worlds, which might explain the appearance of an explanatory
gap? I think so. Despite our similarities, we belong to distinct linguistic com-
munities, defer to different experts, and thus speak distinct languages. Despite a
hypothesized remarkable degree of intertranslatability between our languages,
their concepts are not really our concepts, and like the bug-eyed aliens envi-
sioned above, it would be irresponsible for us to grant them authority over our
observation terms. Since they don’t apply our particular observation concepts
in experience, they wouldn’t really know (in the de dicto sense) what it’s like to
see red. Indeed, we may say that seeing red is nothing to them, even though they
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.14 (670-725)

 David Beisecker

might apply some extensionally equivalent concept. Hence it seems natural to

try to trace the mystery of phenomenal consciousness to logical peculiarities of
the discourse we use to describe it, much as I’ve attempted to do here.14

Notes

. As Brian Coates (2004) usefully explains, Sellars initially introduced the notion of a
sense impression to distinguish sensations from thoughts. The sense impression is the non-
conceptual “descriptive residue” that distinguishes a perceiving that something is the case
from a mere thinking that it is the case.
. Curiously, I think the case of doxastic attitudes is somewhat different. Instead of associ-
ating particular beliefs with characteristic feels or sense impressions, we learn to ascribe to
ourselves a particular belief by looking to the strength of the evidence that we have that that
belief is true. For that reason, ascribing to oneself the belief that p is pragmatically equiva-
lent to asserting p. For that reason as well, beliefs are not said to have much of a phenomenal
component. For more discussion of doxastic first-person authority, see Beisecker (2003).
. Note that this idea is not at all compelling when applied to the other phenomenal states
we’ve been considering. An impression of anger or fear is not itself angry or fearful, just as
an impression of pain would seem to be the wrong sort of thing to be in pain. Instead, it
makes more sense to ask whether we react to these impressions like we react to the objects of
our emotions. Must an impression of pain itself be painful, or must we fear the impression
of fear? Such questions strike me to be very much like asking whether a sense impression of
red would have to look red to its subject. According to the account on offer here, it clearly
would not. I suspect, then, that it would be a mistake to identify fear, pain, and the like with
their impressions.
. It will be obvious here (and throughout) that I appeal to a naïve realism about colors (and
other perceptual properties), according to which such properties are primarily attributed to
(or possessed by) everyday objects in the external world, and not sensations. I’m afraid that
this paper is not the place to defend such a quotidian position.
. These internalist intuitions pose one of the greatest challenges to intentionalist (or rep-
resentationalist) theories of phenomenal content (another being that phenomenal qualities
attend other psychological states, such as emotions, which aren’t clearly representational).
By most accounts, intentional (representational) content is externally determined. Conse-
quently, it seems that one can dream up cases in which undetectable changes in a subject’s
environment can generate changes in intentional content without producing any corre-
sponding phenomenal changes (see Block 1990). Moreover, since most theories of inten-
tional content are also functional, subjects with seemingly distinct phenomenal profiles
could have equivalent intentional profiles. Thus intentional and phenomenal distinctions
wouldn’t seem to track one another, making it unlikely that phenomenal content could be
reduced or understood in terms of intentional, representational content. In response, some
representationalists (Tye 2000) have sought to invent a distinct notion of “non-conceptual”
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.15 (725-785)

The logic of ‘what it’s like’ 

or “perceptual” intentional content that cleaves more closely to these internalist intuitions.
The account on offer here shows why such a maneuver is unnecessary.
. This type of referential ambiguity is characteristic of definite descriptions in modal con-
texts. For instance, the statement, “My wife could have been in pictures.” similarly admits
of two distinct readings. It could either be about Monica (my actual, current wife) in
particular, or it might be understood as making a claim about the range of my erstwhile
marital prospects. Outside the philosophy of language, such referential ambiguity appears
to generate little trouble, and even less excitement.
. See Jackson (1982). The knowledge argument continues to be one of the most vivid il-
lustrations of the hard problem of consciousness. For the uninitiated, here’s a little bit of
background: Jackson’s initial aim was to draw out an intuition that there is some sort of
epistemic gap between phenomenal and non-phenomenal facts. To do so, he invited us to
consider the celebrated case of Mary, a neuroscientist who is supposed to know everything
there is to know about the mechanics of the human visual system, but for some fantastic rea-
son (typically imprisonment in a wholly black-and-white environment), she has never had
a red sense impression. Most are inclined to agree that despite her vast knowledge of neu-
romechanics, Mary nevertheless lacks “knowledge of what it’s like” to see red. So the thought
experiment suggests that phenomenal knowledge cannot be reduced to, or derived from,
theoretical knowledge of physical, physiological, or even functional and representational
facts. Originally, Jackson went on to elevate this epistemic gap into a metaphysical one. That
is, he took the thought experiment to support the thesis that phenomenal facts are ontolog-
ically distinct from the mundanely physical, physiological or functional. Although fanciful
and woefully underdescribed, it seems hard to resist the intuition that Mary learns some-
thing when she escapes her black and white environment. The challenge for materialistically
inclined philosophers of mind is to explain (or explain away) her post-release enhanced
epistemic standing without invoking mysterious, non-material “phenomenal” facts.
. Observe that simply having an experience need not be sufficient for knowing what it
would be like to have that kind of experience. One can see this most clearly in the case of
the perceptually subtle qualities attributed, for instance to wine, beer, or chocolate. Utterly
lacking a connoisseur’s palate, a single passing acquaintance with an expensive wine will
most likely not be enough for me to claim that I truly know what it’s like to experience its
finer characteristics.
. For this reason, we need to distinguish this analysis of knowing what it’s like from the
various versions of the “ability hypothesis” advanced by Lewis (1990). Perhaps the most
popular type of response to the knowledge argument, Lewis suggested that the knowledge
Mary gains is not factual knowledge at all (or knowledge that), but rather some sort of
ability (or knowledge how). Originally, Lewis proposed that upon having her first red vi-
sual impressions, Mary gains new imaginative capacities – e.g., the ability to conjure up
a red impression in memory. More sophisticated versions of this strategy hold that Mary
gains recognitional capacities or something like the ability to access physical facts in a new
“quasi-indexical” fashion (see Loar 1990; Carruthers 2000; Perry 2001; and Papineau 2002).
Without going into great detail, the trouble with these proposals is that it is hard to pinpoint
exactly what the ability or abilities in question are, for it seems that one can pry them apart
from the knowledge Mary gains upon her first red sense impression. For instance, one could
 JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.16 (785-855)

 David Beisecker

reasonably suppose that Mary learns what it’s like to see red, even if she couldn’t later come
to envisage it in imagination.
. Moreover, if you’re of a frame of mind to believe that such beings don’t genuinely apply
any observation concepts at all, on the grounds that they lack the requisite history to be true
participants in a linguistic community, then you might conclude that there is nothing it is
like for them to see red. For in that case the expression ‘what it’s like for them to see red’ fails
to determine any referent. So in a certain sense, they’d be “zombies,” at least for a time, even
though they’d try to convince us otherwise.
. I am aware that some critics will contend that such extrinsic, normative differences are,
from a narrowly scientific perspective, explanatorily otiose. I think that such criticisms rest
upon implausibly scientistic (perhaps physicalistic) assumptions that systematically exclude
the rational types of explanations, in which attributions of phenomenal consciousness (and
other intentionally-freighted concepts) most naturally find their home. See my “Function-
alism and Folk Psychology: How the Mental Earns its Keep” (forthcoming 2004)
. To be sure, I cannot pretend to have offered a complete account of consciousness, for I
have not addressed the issue of what would make a mental state conscious, as opposed to
unconscious.
. Notice as well that such attributions are objective or attribution-transcendent in that it
can be proper to attribute to one knowledge of what it’s like to see red without anyone actu-
ally doing so, and that everyone’s attributing such knowledge to another (a “swampzombie,”
perhaps) doesn’t make it the case that it would be proper to do so. So while knowledge of
what it’s like makes sense only in the context of attributing such knowledge to others, it
doesn’t follow that facts involving phenomenal consciousness are merely attributed.
. Earlier versions of this paper were presented at the 2001 Tucson Conference on Conscious-
ness, the XXIst World Congress of Philosophy in Istanbul, and at a group session meeting of
the Society for Philosophy in the Contemporary World at the 2003 Eastern Division Meeting
of the American Philosophical Association. Many thanks to the organizers, participants, and
audiences of each of these sessions for their endurance and the help they provided at each
stage along the way.

References

Beisecker, David (2003). Interpretation and first-person authority: Davidson on self-

knowledge. Southwest Philosophy Review, 16 (1), 89–96.
Beisecker, David (forthcoming, 2004). Functionalism and folk psychology: How the mental
earns its keep. Southwest Philosophical Studies, 26.
Block, Ned (1990). Inverted earth. Philosophical Perspectives, 4, 52–79.
Carruthers, Peter (2000). Phenomenal consciousness: A naturalistic theory. Cambridge:
Cambridge University Press.
Chalmers, David (1996). The conscious mind: In search of a fundamental theory. Oxford:
Oxford University Press.
JB[v.20020404] Prn:23/11/2004; 10:08 F: Z13106.tex / p.17 (855-905)

The logic of ‘what it’s like’ 

Coates, Paul (2004). Wilfrid Sellars, perceptual consciousness, and theories of attention.
Essays in Philosophy, 5 (1), http://www.humboldt.edu/∼essays/coates.html
Dennett, Daniel (1988). Quining qualia. In A. Marcel & E. Bisiach (Eds.), Consciousness in
contemporary science (pp. 43–77). Oxford: Oxford University Press.
Dretske, Fred (1988). Explaining behavior: Reasons in a world of causes. Cambridge, MA:
MIT Press.
Jackson, Frank (1982). Epiphenomenal qualia. Philosophical Quarterly, 32, 127–136.
Levine, Joseph (1983). Materialism and qualia: The explanatory gap. Pacific Philosophical
Quarterly, 64, 354–361.
Lewis, David (1990). What experience teaches. In William Lycan (Ed.), Mind and cognition:
A reader (pp. 447–61). Oxford: Blackwell.
Loar, Brian (1990). Phenomenal states. Philosophical Perspectives, 4, 81–108.
Lycan, William (1996). Consciousness and experience. Cambridge, MA: MIT Press.
Papineau, David (2002). Thinking about consciousness. Oxford: Clarendon.
Perry, John (2001). Knowledge, possibility, and consciousness. Cambridge, MA: MIT Press.
Sellars, Wilfrid (1963/2001). Empiricism and the philosophy of mind. Reprinted in W.
DeVries & T. Triplett (Eds.), Knowledge, mind, and the given. Indianapolis, IN: Hackett.
Tye, Michael (2000). Consciousness, color, and content. Cambridge, MA: MIT Press.
JB[v.20020404] Prn:13/09/2004; 14:49 F: Z131P2.tex / p.1 (40-66)

P II

Agency and choice

JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.1 (41-116)

Exposing the covert agent

Anton Lethin
San Francisco, California

The sense of self as a covert agent is a key component to the sense of self. This
paper focuses on covert action as a preparation to interact. This activates the
entire motor system, including the gamma motoneurons innervating the
muscle spindles. The propioceptive stimulation is fed back to the network of
origin, contributing to a sense of self as generating the covert activity.
A study of motivated behavior in the rat is presented to clarify how the
motivation potentiates actions in the body. This is set into Panksepp’s
subcortical action system of emotional circuits, where the motivation arises.
This sets the motor tone for the planned action.
This picture is interpreted with Ellis’s and Newton’s model portraying
how emotional motivation can lead to phenomenal consciousness. It is
proposed that no motor imagery occurs without involving the body. The
higher levels of awareness depend on the subcortical bodily intentionality.

Keywords: periaqueductal, covert, gamma, facilitation, somatosensory.

The sense of agency is a key component to the sense of self. There is evidence
that this sense depends on events preceding action that prepare for movement
(Haggard & Libet 2001: 47–63). It is now realized that actions involve a covert
stage. “The covert stage is a representation of the future, which includes the
goal of the action, the means to reach it, and the consequences on the organism
and the external world” (Jeannerod 2001: S103). Does the sense of self depend
on feedback from this covert activity? A sense of self as covert agent? If so, what
is the nature of this feedback? This paper proposes that this covert preparation
for an interaction with the affordance leads to somatosensory feedback from
the body, and that this is necessary for a sense of self grounded in the body and
the world. I believe this preparation is emotionally motivated.
Gallagher discusses a minimal self- “a pre-reflective point of origin for ac-
tion, experience and thought . . . a consciousness of oneself as an immediate
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.2 (116-166)

 Anton Lethin

subject of experience, unextended in time. The minimal self almost certainly

depends on brain processes and an ecologically embedded body . . .” (Gallagher
2000: 15). There are “two closely related aspects of minimal self-awareness: self-
ownership- the sense that it is my body that is moving; and self-agency- the
sense that I am the initiator or source of the action.” “. . . experimental research
on normal subjects suggests that the sense of agency is based on that which pre-
cedes action and translates intention into action” (ibid.: 16; my emphasis added).
This would include planning and preparing to act. The feedback from this to
the network of origin leads to self-awareness.
Libet and Haggard, in separate experiments, showed that electroen-
cephalographic events precede the conscious awareness of deciding to move
(Haggard & Libet 2001: 47–63). Haggard introduced a new element by includ-
ing situations wherein a subject must choose between a left- and a right-handed
action. “Since the motor system must have selected which specific movement
to perform by the time that the readiness potential lateralises, we concluded
that conscious intentions were related to specific rather than general prepara-
tion for action” (ibid.: 51). Their finding suggests that awareness of intention is
associated with this lateralized readiness potential, a stage of action preparation
known to be devoted to the selection of the specific movement to be made.
What is the nature of this preparation? I want to focus on the preparations
to move the body.

Bodily intentionality

I would describe this as a bodily intentionality. “It is my body as a sensorimotor

organism perceiving and acting in the world that first expresses intentional-
ity. There is a bodily intentionality, on Merleau-Ponty’s view, on which all
other forms of intentionality rest” (Wider 1997: 135). I would suggest that the
body’s preparing to interact expresses this intentionality. “. . . intention does not
directly generate behaviour, rather it modulates response activation within a sys-
tem that is sensitive to environmental (bottom-up) factors as well as intention
(top-down control)” (Humphrey & Riddoch 2003: 203; emphasis added). Jean-
nerod discussed the stage of “covert actions”: “The hypothesis that the motor
system is part of a simulation network that is activated under a variety of condi-
tions in relation to action, either self-intended or observed from other individ-
uals, will be developed. The function of this process of simulation would be not
only to shape the motor system in anticipation to execution, but also to pro-
vide the self with information on the feasibility and the meaning of potential
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.3 (166-208)

Exposing the covert agent 

actions” (2001: S103). Gallagher interviewed him and responded: “You suggest
that goal-directedness is a primary constituent of action . . . This means, I think,
that the motor system is not simply a mechanism that organizes itself in terms
of what muscles need to be moved, but it organizes itself around intentions”
(Gallagher & Jeannerod 2002: 13; emphasis added). The pre-movement activ-
ity identified by Jeannerod in the motor system represents bodily intentionality.
This involves the whole organism in an integrated action – visual, autonomic
including blood flow, neuropeptide, neurotransmitter, and hormonal changes.
(Damasio 1999: 145–149; Decety 1993: 549–563).
Newton’s definition of representation is similar to Jeannerod’s definition
of the covert stage: “Representation is the process of performing goal-directed
activity in a manner that allows the activity to be rehearsed and optimized in
advance of the realization of the goal. This realization (whether planned or
simply hypothesized) is what is represented by the activity,” (Newton 2003).
She has moved beyond just including the goal in the planning of the action to
providing for rehearsal and optimizing.
How can the organism rehearse and optimize an activity before the perfor-
mance? In order to optimize this covert pre-movement activity, the organism
will need afferent input from both the salient environmental condition (which
we will refer to as the “affordance”) and from the body. The organism needs
to know what posture the body is starting from, and where the affordance is
in relation to the body. The physiological system for propioception involves
the muscles and tendons, the visual apparatus, and the labyrinths of the inner
ear. Based on these studies and others, Marcel radically proposed that what he
calls “. . . a minimal sense of ownership is provided by the spatial content of
movement specifications . . . The . . . body parts that are to implement the ac-
tion must be specified in a common reference frame with the targets . . . The
only spatial description common for all body parts and for external locations is
an egocentric one . . . The self enters the representational scene as the origin of
the egocentric frame of reference utilized in movement specifications . . . This
would give, in the normal phenomenology, a perspectivalness of the source
of the action, that is, spatial points of origin and intention.” He views the ex-
perience of the self during action as being immersed, meaning that the self
is perceptually recessive – there is no reflective consciousness of self (Marcel
2003: 43). Newton has a similar view: “While bodily awareness seems to be a
background element in all conscious experience, it can and frequently does fall
outside the focus of attention. In those cases we may be said to be conscious
‘only’ of external sensory input. But while we may speak in that way, the very
notion of externality presupposes some awareness of one’s own body and its
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.4 (208-255)

 Anton Lethin

boundaries, as does any awareness of possible interactions with the external

world – its ‘affordances’” (Newton 2001: 57; original emphasis).

Neuromuscular basis of bodily intentionality

I want to focus now on the neuromuscular aspect of bodily awareness. The

activity in the brain can facilitate a movement by stimulating the gamma mo-
toneurons in the spinal cord (Eldred 1953). These stimulate the muscle spindles
in the skeletal muscles, which are tiny stretch receptors that also contain a
tiny muscle. When a muscle is stretched, the spindles sense this and increase
the propioceptive afference to the alpha motoneurons and to the spinal cord
and brain. (See Fig. 1). When the gamma motoneurons stimulate the spindles,
the spindle muscles contract and increase the propioceptive stimulation, just
as though the muscles had been stretched. This strongly facilitates the alpha
motoneurons to the muscles (see appendix for more details). I call the loop
from the brain to the gamma motoneurons through the spindles back to the
brain the “body gamma loop.” The “peripheral loop” describes only the loop
from the gamma motoneurons to the alpha motoneurons. How does this re-
late to preparing to move? Jeannerod investigated the relation of simulation of
action to preparing to act. Functional brain imaging by magnetic resonance
(fMRI) has shown activation of sensorimotor cortex during imagined action
in multiple studies (Jeannerod 1999: 8). Experiments on motor imagery sup-
ported “the hypothesis that mental simulation of action is assigned to the same
motor representation system as preparation to execution” (ibid.: 4). If so, “mental
simulation should activate motor pathways . . . The main result of this exper-
iment was that motoneurons excitability, as tested by the amplitude of spinal
monosynaptic reflexes, was increased during mental simulation . . . Insofar as
the sensitivity of the neuromuscular spindles is under the control of gamma
motoneurons, the increase in excitability of the (muscle stretch reflex) suggests a
selective increase in gamma motoneurons activity during mental simulation of a
movement” (ibid.: 6–7; emphasis added). Muscle stretch reflex results from the
muscle spindles’ stimulation of the alpha motoneurons. The muscles can be
prepared to move by increasing the tension of the muscle spindles inside the
muscles. This increases the propioceptive afferent feedback both to the spinal
cord and to the network originating the preparation of the body’s readiness for
the action (see below). Referring to Merleau-Ponty’s emphasis on body inten-
tionality and reflexivity, Wider quotes Dillon: “It is his ‘discovery of corporeal
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.5 (255-270)

Exposing the covert agent 

B
R
A
I
N

gamma
S
P
I
Spindle N
A
L

propio C
O
R
D
alpha
muscle

Figure 1.

reflexivity’ . . . (that is) the means to overcome ontological dualism” (Wider

199: 138; Dillon 1993: 79).
Preparing to act is very complex. In my view it is a preparation of the entire
organism, involving the autonomic nervous system, including hormonal, neu-
rotransmitter, and neuropeptide changes (Damasio 1999: 59–62). Even mental
simulation of action activates cardiac and respiratory control mechanisms (De-
cety et al. 1993: 549–563). They suggest that autonomic activation during imag-
ined action is part of the more general phenomenon of preparation for action
of the entire nervous system. Even so, the motor preparation all ends up facil-
itating the alpha motoneurons to the muscles to be contracted and inhibiting
the antagonists. Part of the facilitation is caused by the gamma motoneurons’
stimulation of the muscle spindles.
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.6 (270-321)

 Anton Lethin

Panksepp’s emotional circuits

The present account of preparation will be clarified if we transpose it to the

network of emotional circuits described by Panksepp in his book (1998a).
His seminal work summarizes decades of his and others’ experimental stud-
ies about the seven basic motivational-behavioral systems that converge on
the periaqueductal gray. I propose that the covert preparation to interact orig-
inates here. He believes these seven basic circuits are primitive networks that
have persisted during evolution in mammalian brains, and are the roots of
emotion in primitive “value generators.” Each system emphasizes a different
combination of neurotransmitter pathways, and is highly reactive to modula-
tion by neuropeptides. These all project to a diencephalic-midbrain area with
an epicenter in the periaqueductal gray, and they are expressed by activat-
ing innate motor programs. Their “initial adaptive functions were to initiate,
synchronize, and energize sets of coherent physiological, behavioral, and psy-
chological changes” (ibid.: 123). He emphasizes the role of motor functions: “In
affective experience, a direct motor preparatory linkage appears to be especially
evident . . . The intrinsic neurodynamics of such affective, motor-tone setting
circuits, along with various converging somatic and visceral inputs, may cre-
ate pervasive and fractally propagated feeling of self-ness within the organism”
(Panksepp 1998b: 574–575; emphasis added). This suggests that the sense of
self as covert agent may originally have developed with the adjusting of mo-
tor tone as a movement is prepared. Motor tone is the tension in the muscles,
which can be perceived by the resistance a muscle offers to passive stretching.
It is maintained by a low-frequency, asynchronous discharge of impulses in a
small fraction of the motor nerve fibers supplying a muscle, and can be modu-
lated by supraspinal facilitation or inhibition (Bard 1961: 1117–1118) (see Fig.
1 in this paper). “In the present context, primary process intentionality is envi-
sioned to be the natural action readiness that is intrinsically coded within the
interaction of emotional processes with the neural representation of the SELF”
(Panksepp 1998b: 573). This is an emotionally motivated preparation to move
to interact with the affordance.
He uses the capitalized SELF to refer to this fundamental neural substrate as
a distinct brain system among all the associated psychological states that coa-
lesce during development. He refers to this as a ‘Simple Ego-type Life Force.” It
is important to bring out the primary role he envisages for this system. This
“is the lowest region of the brain to orchestrate various coordinated emo-
tional responses via a variety of motor outputs” (ibid.: 570). In discussing the
evolutionary development of this network, he concludes: “I assume the SELF
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.7 (321-386)

Exposing the covert agent 

provides the first executive mechanism for behavioral coherence and bodily
awareness” (Panksepp 1998a: 311).

Motivated behavior in the rat

His emphasis on motor-tone setting as part of readiness can be looked at more

closely by using an example: a study of motivated behavior in the rat. Stellar’s
classic work proposed that the hypothalamus was the center of motivated be-
havior (Stellar, E. 1954: 5–11). Pfaff did a thorough investigation of the neural
pathway from the hypothalamus involved in the execution of the lordosis reflex
in the female rat (Pfaff 1980: 1–281; Stellar, J. & Stellar, E. 1985: 73). The rat was
motivated to get ready for copulation by preparing to hyperextend her lower
back. With lordosis the female becomes sexually receptive to the male. They
administered the estrogenic substance estradiol to the female rat to facilitate
this reflex.
The pathway represents a convergence of hypothalamic influences on
spinal cord reflexes via brainstem mechanisms. The output of ventromedial
hypothalamic cells, excited by estradiol, converges on midbrain central gray
neurons with afferent sensory input and on midbrain reticular formation neu-
rons. From there, the descending influences are relayed through the reticular
formation of the medulla to ventral horn cells of the spinal cord via the reti-
culospinal tract. Other descending influences from the lateral vestibular nuclei
are necessary to elevate the tone of the axial muscles. Here, between T-12 and
S-1, the reflexes of vertebral dorsiflexion are executed in response to flank, rump,
and perineal stimulation by the male rat. Thus lordosis and a sexually recep-
tive posture are executed by the female rat (See Fig. 2). (Pfaff 1982: 290–292;
Pfaff 1980: 190–191, 239–241). The estrogenic stimulation of the hypothala-
mus does not elicit a lordosis reflex. It only prepares the rat to interact. The
cutaneous stimulation either by the male rat or administered experimentally is
necessary for lordosis to occur (Sakuma & Pfaff 1979a, b).
I interpret the process as follows. The estrogenic stimulation has activated
the latent innate goal to receive the male rat’s advances. This motivates the
preparation of the lordosis behavior by potentiating the muscles’ contraction.
This can occur by tensing the spindle muscles within the extensor spinal mus-
cles. As this occurs, there is increased propioceptive stimulation both to the
alpha motoneurons in the spinal cord and back up to the brain stem and hy-
pothalamus in the areas originating the preparation of the muscles (see Fig. 1).
“Descending tracts might have either of two modes of action in facilitating lor-
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.8 (386-386)

 Anton Lethin

medial preoptic
med. ant. hypothal
ventromedial
Estradiol nucl. hypothal.

midbrain
central midbrain reticular
gray form

lat. vestib. nucl.

medullary
reticular
formation

lateral vestibulosp.
and reticulosp.
Spinal tracts
cord

Dorsal Roots
Stimuli L1, L2

L5, L6, 51

Flanks
ep e
rec ssur
s
tor

Skin of
pre

rump Lat. longissimus & Lordosis

tailbase transverso – Response
perineum spinalis

Figure 2. Reprinted by permission from Pfaff (1980).

dosis. One is a tonic effect, in which spinal circuits relevant for lordosis would
be prepared for reflex execution before the onset of the adequate peripheral
stimuli. For instance, tonic facilitation might result in a subliminal amount
of background activity in motoneuron pools which supply muscles impor-
tant for lordosis and a corresponding reduction in activity in motoneuron
pools for muscles antagonistic to lordosis. Against this prepared background,
cutaneous stimuli adequate for lordosis would be able to trigger the behav-
ioral response” (Pfaff 1980: 190). He goes on to describe a possible second
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.9 (386-436)

Exposing the covert agent 

mode of action which would involve spinobulbospinal reflex loops, initiated

by cutaneous afferent activity. This would not be a major mode of action.
Preparatory action occurs at several different levels. How does the motor image of
the planned action arise? There is a hierarchical organization of the planned in-
teraction, with source-schemas, component schemas, sequencing schemas and
motor schemas representing different degrees of integration of the same action
(Jeannerod 1997: 173).
“The hypothalamus is only part of an extensive limbic system involving the
brainstem, diencephalon and forebrain. In many cases it may be the major in-
tegrator in a hierarchy of integrators along the neuraxis” (Stellar, J. & Stellar, E.
1985: 27). The “most immediate behaviorally relevant action would be at lum-
bar spinal cord level. Preparatory action with a longer time course appears to
take place in the lower brainstem. Some components of tonic facilitation may
also occur in the central gray of the mesencephalon. Finally, the longest time
course of cellular preparation for lordosis behavior must occur in and around
the ventromedial nucleus of the hypothalamus, where long-acting estrogenic
effects are registered” (Pfaff 1980: 191). This picture of preparing for an action
raises the question of whether planning for an action at a high level can lead to
the experience of self as agent. Pfaff ’s work suggests that the preparation at the
high level integrates the preparing at all the levels of the nervous system. This
implies that the experience will be based on feedback to all levels of the self-
from the periaqueductal gray to the prefrontal cortex. Gallistel pointed out
that “. . . higher-level circuitry coordinates the activities of the lower circuits
primarily by regulating their potential for activation, a phenomenon I term
selective potentiation and depotentiation” (1980: 398). He finds that these pro-
cesses of potentiation and depotentiation are the means by which motivational
processes influence behavior. On the spinal level the alpha motoneurons to the
lateral longissimus and transverse spinalis muscles are facilitated. This can also
be described as potentiated. The muscle contraction will be much stronger if
and when it occurs. Any afferent stimulation from the appropiate cutaneous
nerves will also directly stimulate both the gamma and alpha motoneurons as
well as the central nervous system.
With the propioceptive afferent feedback the rat experiences her motor in-
tentionality to assume a lordotic posture in interaction with a male rat. Can
we assume that she experiences the preparing with her self as the agent? The
goal-oriented activity is sensed in the networks of origin, so it may well be that
the rat feels this as an action of her SELF, as defined by Panksepp. The latent
disposition to interact sexually with a male rat is innate, and obviously includes
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.10 (436-493)

 Anton Lethin

the goal. When this disposition is activated, the preparing to interact adjusts to
this actual male rat moving now in this corner of this cage or room.
As mentioned above, the estrogenic stimulation of the hypothalamus does
not elicit a lordosis reflex. It only prepares the rat to interact. The cutaneous
stimulation either by the male rat’s mounting or administered experimentally
is necessary for lordosis to occur. The female rat was motivated to lordotically
interact with a male rat. The cutaneous stimulation would amplify the moti-
vation and give it priority over other dispositions. In the report by Pfaff he did
not describe the gamma motor stimulation to the muscle spindles. He inter-
preted the descending influences from the hypothalamus as being below the
threshold for activating muscle contraction. “The parameters of motoneuron-
muscle dynamics most clearly controlled by lateral vestibular and lateral reticu-
lar influences (for instance, “muscle stiffness” and the threshold for the stretch
reflex . . .) remain to be determined” (Pfaff 1980: 206). The gamma motor ac-
tivity and propioceptive response could be studied directly in the female rat to
confirm my interpretation. This is an example of how motor intentionality is
embodied, in this instance activated by estradiol administration. The goal and
motor program are innate.

Seeking system activates preparation

I want to place this lordosis “reflex” into the picture visualized by Panksepp.
He proposes that the primitive self is based on a subcortical network that is
the primitive initiator and motivator for the organism. As mentioned above,
this is in the diencephalon and mesencephalon centering in the periaqueductal
gray, and with adjacent networks it integrates the emotional circuits, sensory
input, motor activity, and autonomic nervous system. This network is respon-
sive to the body’s needs and functions in the same way as Damasio’s proto-self
(1999: 153–160).
The emotional-motivational circuits organize diverse behaviors by activat-
ing or inhibiting motor plans. One of these is the SEEKING system. Panksepp
uses capitals when referring to one of the genetically ingrained emotional op-
erating systems. Previously he called this system the “appetitive motivational
seeking system which helps elaborate energetic search and goal-directed behav-
iors in behalf of any of a variety of distinct goal objects” (Panksepp 1998a: 52).
“. . . the emotive tendencies aroused by this type of brain stimulation most
clearly resemble the normal appetitive phase of behavior that precedes consum-
matory acts” (ibid.: 147, emphasis added). Richard Depue called the seeking
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.11 (493-536)

Exposing the covert agent 

system the behavioral facilitation system (Depue 1989: 458). “. . . when fully
aroused, it helps fill the mind with interest and motivates the organism to
move their bodies effortlessly” (Panksepp 1998a: 52). As for subjective expe-
rience, he suggests “that “intense interest,” “engaged curiosity,” and “eager
anticipation” are the types of feelings that reflect arousal of this system in
humans” (ibid.: 149). Watt points out that the seeking system “probably un-
derpins a most basic emotional capacity . . . – the capacity to experience hope”
(1999: 195). There are two components to this arousal. For example, when
the organism espies and prepares to move toward the food nearby, his expe-
rience of eager anticipation occurs without any movement. Then when he starts
moving toward the food, he can sense the movement with continuing eager
anticipation of the consummation.
The seeking system is “quite motivationally and goal non-specific, facilitat-
ing only the relative activation of other potentially affectively rich interactions
with others and the environment, mediated by activations of the other primes
or prototypes, and of course, the hypothalamic mediation of homeostasis in
the seeking of food, drink, and other biological requirements” (ibid.: 196). One
example would be the specialized neurons that are sensitive to the various hor-
mones that control sexual tendencies, one of which is estrogen. When Pfaff
administered estradiol to the female rat, this sexual urge was expressed through
the SEEKING system. “Critical circuits that sensitize the lordotic spinal reflex
via tonic descending influences arise from the central gray of the midbrain and
the ventromedial hypothalamus . . .”(Panksepp 1998a: 240). The most effective
way to sensitize the spinal reflex is to increase the tension in the muscle spindles
by stimulating the gamma motoneurons. The cutaneous stimulation affects the
gamma and alpha motoneurons in the lumbosacral spinal cord, and this elic-
its a lordosis spinal reflex without needing further supraspinal involvement. As
Depue said, the behavior is facilitated by the SEEKING system.
Panksepp points out that “the core of the SEEKING system is remarkably
well highlighted by the trajectory of brain DA systems, especially the mesolim-
bic and mesocortical components which ascend from the A10 DA neurons of
the VTA to the shell of the nucleus accumbens, and areas of the frontal cor-
tex and amygdala” (ibid.: 156). (DA refers to dopamine and VTA to ventral
tegmental area in the mesencephalon). The energizing aspects of dopamine
are well documented. He adds, however, that the ascending DA systems are
only one link within the complex chains of electrophysiological and neuro-
chemical events, and it is certain that the system also has important descending
components. The reticulospinal tract would be one of these.
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.12 (536-598)

 Anton Lethin

Emotionally motivated preparing to interact

We are phenomenally conscious of our sense of self as agent. Ellis and Newton
proposed a characterization of consciousness, taken from their phenomenal
experience of it, which they broke down into three elements (Ellis & New-
ton 1998: 439). I want to apply each of these to the lordosis reflex viewed in
Panksepp’s SEEKING system.
1. “[The first element is] an emotional motivation which grounds an interest
in anticipating the future.”

The hypothalamic stimulation of the SEEKING system in our example acti-

vates the motivation for the preparation. As mentioned above, the experience
of the SEEKING system being activated can be described as one of eager an-
ticipation. The anticipation facilitates the lordosis response when the affor-
dance appears. The female rat is emotionally motivated to copulate with a
male rat, and seeks this interaction. The motivation grounds the interest as
it embodies it.
2. “[The second element is] sensory, sensorimotor or propioceptive imagery
(which, by itself, can occur preconsciously) activated by this emotional
motivation”.

They emphasize that it is the emotionally motivated anticipation of input that

leads to such “imagery”. Ellis and Newton describe an image as “the felt sense
that one is looking for (or listening for, tasting for, propioceptively feeling for,
etc.) some object or state of affairs that would take the form of an intentional
object. The role of imagery in action-planning involves forming (not neces-
sarily a visual or auditory image but) a sensory and/or propioceptive and/or
sensorimotor image of oneself performing the action in the way planned” (New-
ton 1996; Ellis & Newton 1998: 435; emphasis added). The image is represent-
ing “not just sensory data, but also the execution of bodily actions” (Ellis &
Newton 1998: 436).
I want to focus on the process of “forming” in their description of imaging.
It would appear that this is part of intentionality, since it involves seeking an
intentional object. One aspect of this would be the motor intentionality that
includes preparing a movement interaction with the affordance. If we look at
the rat’s motivation as potentiating the action at many levels, then this would
take place in the experiment described in the hypothalamus and descending
through the midbrain and brain stem to the spinal cord. To prepare to hyper-
extend the lower back would include stimulating the gamma motoneurons in
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.13 (598-628)

Exposing the covert agent 

the spinal cord, according to my hypothesis. This results in propioceptive feed-

back from the muscle spindles, creating a propioceptive pattern of the planned
bodily action. Can this be described as the felt sense that one is “propioceptively
feeling for” a state of affairs (receiving the male rat)? This assumes that the pat-
tern of gamma motor stimulation persists in memory. Then when the afferent
propioceptive pattern returns to the network of origin, the self can sense that
one is seeking this state of affairs. There is reflexive self-reference. This can oc-
cur pre-consciously, as Ellis and Newton say: “To image something is to have an
internally generated experience similar to the actual perceptual experience of
that thing” (ibid.: 439). The generating is the expression of intentionality. The
female rat may also form a visual image of the male rat as part of the seeking, as
well as an olfactory image and a cutaneous image. She is “looking for” a specific
state of affairs. Searle has proposed that visual experiences have intentionality
(Searle 1992: 195), and I agree. There is an expected pattern that is prepared
for, which may be pre-conscious. The intentional object mentioned above as-
sumes the goal is part of the preparation to interact. To adjust the preparation
to fit the goal requires that the sensory input from the affordance is integrated
with the propioceptive input and the efferent preparatory activity.
3. “[The third element is] a resonating between the activity of emotionally-
motivated imagery and the activity stimulated by incoming sensory data
and data reactivated through memory”.

Here we move to a higher level. “This core system of the SELF interacts closely
with other nearby components for exteroceptive consciousness such as the
Extended Reticular and Thalamic Activating System (ERTAS) . . . Thus, the
PAG-centred emotional SELF system may be seen as the very core of the
visceral-hypothalamic-limbic axis (which is essential for affective, interocep-
tive consciousness), while the ERTAS is the core of the adjacent somatic-
thalamic-neocortical axis (which is essential for exteroceptive consciousness)”
(Panksepp 1998b: 571). “PAG may even be essential for the arousal or main-
tenance of a conscious state, or at least for virtually all behavioural intention-
ality . . .” (Watt 1999: 196). As the estradiol-stimulated female rat is seeking to
copulate with a male rat, she forms a propioceptive image of the sought inter-
action. Then she sees and smells a male rat in the room. The sensory data from
the male rat elicits a new instant of preparing to copulate, anticipating the fu-
ture. At the same time, remembering previous copulation reactivates a latent
disposition to copulate. These both interact with the emotionally-motivated
imagery of copulating existing in its seeking activity. As Newton describes it,
the past, present and future are blended. There is temporal thickness, and phe-
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.14 (628-681)

 Anton Lethin

nomenal consciousness emerges (Newton 2001: 54–55). This is awareness of

the experience, and does not imply any reflective consciousness.

Higher levels depend on lower

Where does the experience of motor intentionality take place? It is controver-

sial as to whether this occurs solely in the mind/brain, or in a brain-body-brain
loop. For example, does the preparation to move generate the “as-if sensa-
tion” of propioceptive activity in the brain – propioceptively “imaging” what
it would feel like to move in this way? Alternatively, does the organism pre-
pare the entire body to move in a way that generates propioceptive activity
from the muscles themselves, as well as the rest of the organism? Looking first
at high-level activity, I turn to Newton’s description: ‘So what I’d say is that
the intentional experience of planning (or just thinking about) an action with
oneself as the agent, IS the experience of sensori-motor/propioceptive goal-
oriented activity. In that activity, the goal is part of the activity; forming action
images or high-level representations of that activity for planning purposes is
being an agent” (Newton 2003). What is the experience of sensori-motor/ pro-
pioceptive goal-oriented activity? One imagines the activity (e.g., reaching to
grasp an object), forming a dynamic pattern of the movement, including the
expected propioceptive and sensory feedback. Is this reflexive self-referring on
a high-level? Is it somehow dependent on lower levels? “Although a high level of
awareness is certainly not a local property of the PAG itself, such functions may
emerge from the higher brain areas that are recursively linked to the PAG, es-
pecially in cingulate, frontal, and temporal lobes” (Panksepp 1998b: 578; PAG
refers to the periaqueductal gray). There are supporting studies by Damasio
1994, Mantyh 1982, and Sesack et al. 1989. “Obviously, affective feelings, as all
other forms of consciousness, are hierarchically organized in the brain, with the
higher functions being decisively dependent on the integrity of the lower functions,
but not vice versa” (Panksepp 1998b: 578; emphasis added). He summarizes ev-
idence from studies with several approaches to validate his conclusion. The rat’s
preparing to hyperextend her lower back is part of an affective feeling, part of
the primitive emotional circuitry. The emotional motivation toward a specific
goal activates the preparation for a movement interaction. The higher levels
influence the selection, modulation and the inhibition of the lower levels, but
this activity is all organized toward the goal generated in the subcortical area.
Ellis and Newton agree: “From an empirical standpoint, afferent process-
ing, e.g., in the occipital lobe, never results in conscious awareness of the object
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.15 (681-724)

Exposing the covert agent 

unless accompanied by corticothalamic loops instigated by midbrain motiva-

tional activity, especially frontal-limbic activity (Ellis & Newton 1998: 433; em-
phasis added; Posner & Rothbart 1992). As Merleau-Ponty has said, all higher
levels of intentionality are dependent on a basic motor intentionality. This is
where the motivation toward the goal arises. The prefrontal cortex may in-
tegrate and control the behavior, but the impetus toward the behavior comes
from the PAG. “Eager anticipation and intense interest” describe the experience
when the seeking system is aroused.

Propioceptive imaging

Can there be propioceptive images created in the mind/brain without involv-

ing the body itself? “This imagining of action affordances is facilitated by the
motorically-initiated efferent brain commands that Damasio calls the ‘as if
body loop’ ” (Ellis 2000: 45). This “as if body loop” bypasses the body. Dama-
sio has “suggested, and LeDoux agrees, that some emotional responses can
change somatosensory representations in the brain directly, “as if ” the latter
were receiving signals from the body, although in effect the body is bypassed.
People probably have both body-loop and as-if-body-loop mechanisms to suit
diverse processing conditions. The critical point, however, is that both mecha-
nisms are body-related” (Damasio 1997: 141; Damasio 1999: 280–281; emphasis
added). It appears necessary that if an emotional response is to create a pro-
pioceptive image, it will result from the emotion evoking an efferent motor
image. The body loop as described by Damasio did not specify a brain-gamma
motoneuron-muscle spindle-propioceptive afferent loop. I propose that this is
part of the body loop.
Naito et al. investigated whether motor imagery contains kinesthetic sen-
sations which are a substitute for the sensory feedback that would normally
arise from the overt action (Naito 2002: 3683–3691). They induced an illu-
sion of palmar flexion by vibrating a wrist extensor at a frequency known to
stimulate the muscle spindles. They found that motor imagery of wrist move-
ment influenced this sensory experience of kinesthetic illusion. They found
“no peripheral inputs or minimal inputs, if any” during the motor imagery.
They based this on the PET scans measuring regional cerebral blood flow. Also
they found no EMG activity during imagery. This contrasts with other studies
mentioned below. They did train their subjects to image without producing
any EMG activity, so that could account for this difference. It is conceivable
that training to suppress EMG activity during imagery might inhibit the alpha
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.16 (724-776)

 Anton Lethin

motoneuron activity, as Requin found in his studies (1984: 277). They con-
cluded by agreeing with Frith and Dolan (1997) that “mental imagery reflects
the effects of previous knowledge about the predicted sensory effects of the
subject’s own actions on sensory processing areas in the absence of sensory
input” (Naito 2002: 3689).
Jeannerod’s experiments produced different evidence on this question. He
found that “. . . even if you are simply imagining the action in terms of its
goal, in simulating it you also rehearse all the neuronal circuitry. . . . if you
examine the brain activity during motor imagination, you will find activation
of the motor cortex, the cerebellum, etc.” (Gallagher & Jeannerod 2002: 14).
As mentioned above, the increase in excitability of the muscle stretch reflex
in his studies “suggests a selective increase in gamma motoneurons activity
during mental simulation” (Jeannerod 1999: 6–7). There are, however, many
influences on the excitability of the alpha motoneurons included in the muscle
stretch reflex path. Fadiga found that corticospinal excitability was specifically
modulated by motor imagery in a magnetic stimulation study, recording mo-
tor evoked potentials (Fadiga 1998: 147–158). Corticospinal excitability refers
to the excitability of the alpha motoneurons stimulated by the cortical neurons.
Was this increase due to gamma motor stimulation to the muscle spindles,
leading to more propioceptive stimulation of the alpha motoneurons, or due to
direct alpha stimulation? There are several other reports of an increase of EMG
activity in muscles involved in the imagined motor act (Porro et al. 1996: 7696;
Jacobson 1930; Wehner et al. 1984; Harris & Robinson 1986). Some studies
have not shown an EMG increase. The results of the studies can reflect dif-
ferences in consciously preparing a prescribed movement, motor imaging, and
mental simulation, and in how the activity is studied. “During preparation there
is concurrent massive inhibition of the alpha motoneurons, which ‘prevents any
premature triggering of action’ ” (Jeannerod 1997: 118; Requin 1977: 139–174).
This can explain the absence of EMGs, H-reflex changes, and motor-evoked
potentials in some of the studies (Naito 2002: 3683–3691; Kasai 1997: 147–
150). Jeannerod concluded that the inhibition occurring with simulation is less
intense, and there is only a “partial block of the motoneurons, as shown by
residual EMG and increased reflex excitability” (Jeannerod 1997: 118).

Selective fusimotor activity

Fusimotor is the same as gamma motor (L. fusus, a spindle). Selective refers to
activation of fusimotor neurons without simultaneous alpha motor activation.
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.17 (776-816)

Exposing the covert agent 

Prochazka proposed that “the fusimotor system plays a role independent of the
alpha motoneurons system, associated with arousal and expectancy” (Proc-
hazka 1989: 281–307). Taylor recorded a spindle afferent from a jaw closing
muscle in a chronically prepared cat while it was lapping milk. It became sat-
isfied and stopped. After “a short pause the animal took interest in the milk
again and there ensued a period of rhythmically modulated spindle discharge
without EMG or jaw movement before lapping recommenced. . . Evidently the
central pattern generator started working and sending an output to the (static)
fusimotor neurones before the excitability of the alpha motoneurones was suffi-
cient to make an overt expression of the rhythm” (Taylor, Durbaba, & Rodgers
1995: 374).
Gandevia and others in 1997 recorded muscle spindle afferents directly
during mental rehearsal of movements. They found that no spindle activity
occurred unless there was increased EMG activity. “Mental rehearsal did acti-
vate alpha motoneurons, and if this was sufficiently strong, the skeletomotor
discharge was accompanied by recruitment of spindle afferents” (1997: 264).
They theorized that liminal contractions occurred without overt movement,
and that this involved unintentional performance of the planned motor task.
They concluded that there was no selective fusimotor activation during imag-
ined movement. “ ‘Anticipation’ is associated with changes in gain of spinal
reflexes and muscles are often tensed unintentionally in preparation for the
command signal to move.” (ibid.: 265; also see Burke 1980).

Ascending propioception during peripheral inhibition

So we have evidence of efference to the body for both gamma and alpha.
What about afferent feedback from the body of this efference? It is possible
that the central stimulation of the gamma motoneurons occurs with concur-
rent pre-synaptic inhibition of the propioceptive afferents’ synapses with the
alpha motoneurons (to prevent premature action). There are also spindle af-
ferents synapsing with interneurons in the spinal cord which synapse with
somatosensory tracts to the brain. I propose that these would not be inhibited.
The differential control of collaterals of sensory fibers by GABAergic interneu-
rons results in varying levels of synaptic effectiveness (Rudomin 2002: 167).
GABA is a neurotransmitter released by most presynaptic terminals. This pro-
pioceptive feedback to the network originating the gamma motor stimulation
can be the basis for the organism being aware of how it is preparing for the
intended movement via the body-loop through the muscle spindles. Gallagher
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.18 (816-873)

 Anton Lethin

has proposed that somatic propioception, in its most typical form, provides
a sense of ownership for the body and its movements (2003).This is a pre-
reflective, non-perceptual bodily awareness. It refers to the subjective experience
of ownership of embodiment. He is referring to posture and movement, but I
find that his analysis fits the process of preparing to move.

No motor imaging without involving the body

It is possible that the artificial circumstances of experimental studies asking for

motor imaging result in very minimal emotional motivation, particularly when
required to relax.
I propose that normally there is no motor imaging without involving the
body. When the subcortical areas originate the emotionally-motivated activ-
ity, this activates the preparation in the entire organism for an interaction.
The motor system, including the autonomic nervous system, is involved in the
activation. Propioceptive feedback from the body would occur. An exception
would occur with pathology- for example, patients who have lost their propio-
ceptive afference have a markedly distorted sense of self (Cole 1998: 261–262).
As for persons with quadriplegia, Damasio points out that “all the surveys of
patients with spinal cord damage . . . have revealed some degree of impaired
feeling, as one should have expected given that the spinal cord is a partial con-
duit for relevant body input” (1999: 289). Normally, the goal of the imagined
movement is expressed in patterns of activity in the higher levels of the motor
system and in the body. Even with patients who have lost their propioceptive
afference, the subcortical area will still motivate the preparation in the higher
levels. In this case, there would be no propioceptive afference from the body
loop, so that the body would be by-passed for this mode, except for visual and
vestibular. The as-if body loop would still occur.
In discussing how covert actions simulate actions, Jeannerod says: “. . . acti-
vation of the motor cortex and of the descending motor pathways seems to
fulfill several critical functions. First, this activation contributes to generating
corollary signals that propagate upstream to parietal and premotor cortex. This
mechanism would allow evaluation of the potential consequences of the future
action. It could also provide the subject with information for consciously mon-
itoring his (simulation states) and realizing that he is the agent of this covert
activity, in spite of absence of overt behavior” (Jeannerod 2001: S108; empha-
sis added). What is the source of this information for monitoring? The phrase
“descending motor pathways” could refer only to the subcortical motor nu-
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.19 (873-907)

Exposing the covert agent 

clei of the thalamus and basal ganglia or also to the pathways descending to
the gamma motor neurons, which would generate afferent signals from the
muscle spindles. The innervation of the gamma motoneurons occurs over four
pathways: corticobulbospinal, rubrospinal, reticulospinal, and vestibulospinal.
The corticobulbar fibers synapse in the red nucleus and motor areas of the
reticular formation in the brain stem, where the rubrospinal and reticulospinal
tracts arise (Kingsley 2000: 241–247). This means that the activation of the cor-
tex in covert actions may reflect the activity of the corticobulbar neurons to
the rubrospinal and reticulospinal tracts, as well as the activity of the corti-
cospinal tract itself. The vestibulospinal tract originates in the vestibular nuclei
of the brain stem. Their various effects are facilitatory or inhibitory of the alpha
and gamma motoneurons. When he realizes that “he is the agent of this covert
activity”, we suggest that he experiences the activity as a motor intention.
So there is evidence that the organism prepares to interact by activity in the
entire motor system, possibly including the gamma motoneurons to the mus-
cle spindles. The organism is also directly preparing the alpha motoneurons
for the action as the final common pathway. This exemplifies the hierarchy of
potentiation described above by Gallistel. All of the supraspinal networks that
influence the gamma motoneurons may also affect the alpha motoneurons’ ex-
citability. Most of these act through interneurons, allowing for integration of
influences.

Conclusion

The primary importance of the motivation originating from the subcortical

emotional circuits rests on the fact that this is the area responding to the biolog-
ical needs of the organism. These are the value-generators. “Without emotion
allowing and informing a central representation of value, executive and atten-
tional functions are collapsed at their base, as are personal meaning and any
viable image of agentic active self . . .”(Watt 1999: 193). The activities of the
higher levels of the brain are organized around the emotional motivation orig-
inating here. Speaking of the needs and values of the organism, it might be well
to add that one of the emotional neural circuits that Panksepp delineates deals
with the social needs for bonding, sex, and nurturance.
Ellis and Newton proposed that when the three elements described above
interact in a certain way, they are inevitably accompanied by consciousness. “If
the organism’s knowledge of its environment is to involve a ‘felt’ dimension, in
the sense that there is ‘something it feels like’ to have a state of consciousness,
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.20 (907-959)

 Anton Lethin

the conscious processing must first flow from an emotional process within the
organism, which pre-exists any particular input, and puts its qualitative stamp
on each selected input” (Ellis & Newton 1998: 431). The phrase “interact in a
certain way”, of course, raises many intriguing questions.
The sense of self as covert agent originates in the subcortical area cen-
tered on the periaqueductal gray. The primitive emotional circuits motivate the
behavior by potentiating a movement interaction with the affordance. It is pro-
posed that this potentiation stimulates the gamma motoneurons to the muscle
spindles in the muscles (and possibly the alpha motoneurons). The resulting
propioceptive activity is fed back to the network of origin, which leads to an
awareness of self as generating the potentiation. This aspect of the sense of self
thus depends on somatosensory feedback. “Propioceptive awareness thus pro-
vides an immediate experiential access to my pre-reflective, embodied self . . .”
(Gallagher & Marcel 1999: 21).

Appendix

The muscle spindles are tiny sensory receptors embedded in the skeletal mus-
cles. They sense the steady-state length of the muscle as well as dynamic
changes in length and tension. When the muscle is stretched, the spindle is also
stretched, and the sensory nerve activity increases. When the muscle contracts,
the spindle length decreases, and the sensory nerve activity decreases (unless
overridden by the central nervous system – see below).
There is, however, a unique feature of this sensory receptor. It contains tiny
muscle fibres (called intrafusal). When the gamma motor nerve to the spindle
causes the intrafusal muscle fibres to contract, the spindle’s sensitivity to stretch
increases. This has important ramifications. Even though the muscle itself re-
mains at the same length and is not stretched, the gamma motor stimulation
has increased the stimulation of the sensory nerves just as though the muscle
had been stretched. There is no movement of the muscles, but there is increased
propioceptive activity (Kingsley 2000: 217; Matthews 1972). When this occurs,
impulses travel up the muscle afferent nerves to the spinal cord. There they
synapse with the alpha motoneurons that can contract the muscle itself with
added stimulation from the brain. The increase of stimulation from the muscle
spindles exerts a powerful facilitation of these neurons (they are more sensitive
to stimulation). The muscle tone is increased (Taylor & Prochazka 1981; Taylor
et al. 1995). The gamma and alpha motoneurons both originate in the spinal
cord and send axons to the muscles. Their activity is influenced by supraspinal
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.21 (959-1037)

Exposing the covert agent 

and propiospinal activity, by cutaneous afferent and autonomic input (Glad-

den 2000: 213), and by interneurons in the spinal cord. The subject’s motor
set alters the response to these influences. The muscle afferent nerves stimulate
neural tracts going to the brain. (Fig.1) The central stimulation of the gamma
and alpha motoneurons can be independent or integrated. The central nervous
system stimulation of the gamma motoneurons may precede the alpha stimu-
lation, thus facilitating their activity, or the gamma stimulation may increase
simultaneously with the alpha when a muscle contraction is initiated (Kingsley
2000: 221–223).

References

Bard, Philip (Ed.). (1961). Medical Physiology. St. Louis: C.V. Mosby.
Burke, David, McKeon, Brian, Skuse, Nevell F., Westerman, Roderick A. (1980). Anti-
cipation and Fusimotor Activity in Preparation for a Voluntary Contraction. Journal
of Physiology, 306 (1), 337–348.
Cole, Jonathan, & Paillard, Jacques (1998). Living without Touch and Peripheral
information about Body Position and Movement: Studies with Deafferented subjects.
In Jose Luis Bermudez, Anthony Marcel, Naomi Eilan (Eds.), The Body and the Self (pp.
245–266). Cambridge, MA and London: MIT Press.
Damasio, Antonio (1994). “Descartes” Error. New York: Putnam.
Damasio, Antonio (1997). Thinking and Feeling. Scientific American. June 1997, 140–141.
Damasio, Antonio (1999). The Feeling of What Happens. New York: Harcourt.
Decety, Jean, Jeannerod, Marc, Durozard, D., & Bavarel, G. (1993). Central activation of
autonomic effectors during mental simulation of motor actions in man. Journal of
Physiology, 461 (1), 549–563.
Depue, Richard. A., & Iacono, William. G. (1989). Neurobehavioral aspects of affective
disorders. Annual Review of Psychology, 40, 457–492.
Dillon, M. C. (1993). The Unconscious Language and World. In Patrick Burke and Van der
Veken, Jan (Eds.), Merleau-Ponty in Contemporary Perspective. Dordrecht, Netherlands:
Kluwer Academic Publishers.
Eldred, E., Granit, R., & Merton, P. A. (1953). Supraspinal Control of the Muscle Spindles
and its Significance. Journal of Physiology, 122, 498.
Ellis, Ralph, & Newton, Natika (1998). Three Paradoxes of Phenomenal consciousness:
Bridging the Explanatory Gap. Journal of Consciousness Studies, 5 (4), 419–442.
Ellis, Ralph (2000). Efferent Brain Processes and the Enactive Approach to Consciousness.
Journal of Consciousness Studies, 7 (4), 40–50.
Fadiga, Luciano, Buccino, Giovanni, Craighero, Laila, Fogassi, Leonardo, Gallese, Vittorio,
& Pavesi, Leonardo. (1998). Corticospinal excitability is specifically modulated by
imagery: A magnetic stimulation study. Neuropsychologia, 37, 147–158.
Frith, Chris, Dolan, R. J. (1997). Brain mechanisms associated with top-down processes in
perception. Philos Trans R Soc Lond B Biol Sci., 352, 1221–1230.
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.22 (1037-1134)

 Anton Lethin

Gallagher, Shaun (1998). Body Schema and Intentionality. In Jose Luis Bermudez, Anthony
Marcel, & Naomi Eilan (Eds.) The Body and the Self (pp. 225–244). Cambridge, MA:
MIT Press.
Gallagher, Shaun, & Marcel, Anthony (1999). The Self in Contextualized Action. Journal of
Consciousness Studies, 6 (4), 4–30.
Gallagher, Shaun (2000). Philosophical conceptions of the self: Implications for cognitive
science. Trends in Cognitive Science, 4 (1), 14–21.
Gallagher, Shaun, & Jeannerod, Marc. (2002). From action to interaction. Journal of
Consciousness Studies, 9 (1), 4–30.
Gallagher, Shaun (2003). Bodily self-awareness and object perception. Theoria et Historia
Scientarium: International Journal for Intedisciplinary Studies, 7 (1), 53–68.
Gallistel, C. R. (1980). From muscles to motivation. American Scientist, 68, 398–409.
Gandevia, Simon C., Wilson, Linda R., Inglis, J. Timothy, & Burke, David (1997). Mental
rehearsal of motor tasks recruits alpha-motoneurones but fails to recruit human
fusimotor neurons selectively. Journal of Physiology, 505 (1), 259–266.
Gladden, M. H., Maxwell, D. J., Sahal, A., & Jankowska, E. (2000). Coupling between
serotoninergic and noradrenergic neurons and gamma-motoneurones in the cat.
Journal of Physiology, 527, 212–223.
Haggard, Patrick, & Libet, Benjamin (2001). Conscious Intention and Brain Activity. J.
Consciousness Studies, 8 (11), 467–463.
Harris, K. S., & Robinson, W. J. (1986). The effect of skill level on EMG activity during
internal and external imagery. Journal of Sport Psychology, 8, 105–111.
Humphreys, Glyn W., & Riddoch, M. Jane (2003). Fractionating the intentional control of
behavior. In Johannes Roessler & Naomi Eilan (Eds.), Agency and self-awareness (pp.
201–217). Oxford: Clarendon Press.
Jacobson, E. (1930). Electrical measurements of neuromuscular states during mental
activities. I. Imagination of movement involving skeletal muscle. American Journal of
Physiology, 91, 567–608.
Jeannerod, Marc. (1997). The Cognitive Neuroscience of Action. Cambridge, MA: Blackwell.
Jeannerod, Marc. (1999). To Act or Not to Act: Perspectives on The Representation of
Actions. Quarterly Journal of Experimental Psychology, 52A (1), 1–29.
Jeannerod, Marc. (2001). Neural Simulation of Action: A Unifying Mechanism for Motor
Cognition. Neuroimage, 14, S103–S109.
Kasai, T., Kawai, S., Kawanishi, M., & Yahagi, S. (1997). Evidence for facilitation of motor
evoked potentials (MEPs) induced by motor imagery. Brain Res., 744 (1), 147–150.
Kingsley, Robert (2000). “Concise Text of Neuroscience”. (2nd ed.). Lippincott, Williams
and Wilkins, Philadelphia.
Lethin, Anton (2002). How do we embody intentionality? Journal of Consciousness Studies,
9 (8), 36–44.
Mantyh, Patrick W. (1982). Forebrain projections to the periaqueductal gray in the monkey,
with observations in the cat and rat. The Journal of Comparative Neurology, 206, 146–
158.
Marcel, Anthony (2003). The Sense of Agency: Awareness and ownership of Action. In
Roessler, Johannes, & Eilan, Naomi (Eds.), Agency and Self-Awareness (pp. 48–93).
Oxford: Clarendon Press.
JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.23 (1134-1253)

Exposing the covert agent 

Matthews, P. B. C. (1972). Mammalian Muscle Receptors and Their Central Actions. London:
Arnold.
Naito, Eiichi, Takanori Kochiyama, Ryo Kitada, Satoshi Nakamura, Michikazu Matsumura,
Yoshiharu Yonekura, & Norihiro Sadato (2002). Internally simulated Movement
Sensations during Motor Imagery Activate cortical Motor Areas and the Cerebellum.
Journal of Neuroscience, 22 (9), 3683–3691.
Newton, Natika. (1996). Foundations of Understanding. Amsterdam: John Benjamins.
Newton, Natika (2001). Emergence and the uniqueness of consciousness. Journal of
Consciousness Studies, 8 (9–10), 47–59.
Newton, Natika (2003). Personal communication.
Newton, Natika (2004). The art of representation: Support for an enactive approach.
Commentary on Rick Grush. Behavioral and Brain Sciences, 27 (3).
Panksepp, Jaak (1998a). Affective Neuroscience: The foundations of human and animal
emotions. New York: Oxford University Press.
Panksepp, Jaak (1998b). The periconscious substrates of consciousness: Affective states and
the evolutionary origins of the Self. Journal of Consciousness Studies, 5 (5–6), 566–582.
Pfaff, Donald W. (1980). Estrogens and Brain Functions: Neural Analysis of a Hormone-
Controlled Mammalian Reproductive Behavior. Springer-Verlag: New York.
Pfaff, Donald W. (Ed.). (1982). The Physiological Mechanism of Motivation. Springer-Verlag:
New York.
Porro, Carlo A., Francescato, Maria Pia, Cettolo, Valentina, Diamond, Mathew E., Baraldi,
Patrizia, Zuiani, Chiava, Bazzocchi, Massimo, & di Prampero, Pietro E. (1996). Primary
motor and sensory cortex activation during motor performance and motor imagery: A
functional magnetic resonance imaging study. Journal of Neuroscience, 16 (23), 7688–
7698.
Posner, Michael I., & Rothbart, Mary K. (1992). Attentional mechanisms and conscious
experience. In A. D. Milner & M. D. Rugg (Eds.), The Neuropsychology of Consciousness.
London: Academic Press.
Prochazka, Arthur (1989). Sensorimotor gain control: A basic strategy of motor systems?
Progress in Neurobiology, 33, 281–307.
Requin, Jean, Bonnet, Michel, & Semjen, A. (1977). Is there a specificity in the supraspinal
control of motor structures during preparation? In S. Dornic (Ed.), Attention and
Performance, VI (pp. 139–174). Erlbaum: Hillsdale.
Requin, Jean, Lecas, Jean-Claude, & Bonnet, Michel (1984). Some experimental evidence for
a three-step model of motor preparation. In Sylvan Kornblum & Jean Requin (Eds.),
Preparatory states and processes. Hillsdale, N.J. and London: Erlbaum.
Rudomin, Pablo (2002). Selectivity of the central command of sensory information in the
mammalian spinal cord. In Simon Gandevia, Uwe Proske, & Douglas G. Stuart (Eds.),
Sensorimotor control of movement and posture. New York: Kluwer Academic/Plenum.
Sakuma, Y., & Pfaff, D. W. (1979a). Facilitation of female reproductive behavior from
mesencephalic central grey in the rat. American Journal of Physiology, 237 (5), R278–
284.
Sakuma, Y., & Pfaff, D. W. (1979b). Mesencephalic mechanisms for integration of female
reproductive behavior in the rat. American Journal of Physiology, 237 (5), R285–290.
Searle, John R. (1992). The Rediscovery of the Mind. Cambridge: MIT Press.
 JB[v.20020404] Prn:23/11/2004; 10:18 F: Z13107.tex / p.24 (1253-1295)

 Anton Lethin

Sesack, Susan R., Deutch, Ariel Y., Roth, Robert H., & Bunney, Benjamin (1989). To-
pographic organization of the efferent projections of the medial prefrontal cortex in the
rat: An anterograde tract-tracing study with Phaeolus vulgaris leucoagglutinin. Journal
of Comparative Neurology, 290, 213–242.
Stellar, Eliot (1954). The physiology of motivation. Psychological Review, 61, 5–22.
Stellar, James R., & Stellar, Eliot (1985). The Neurobiology of Motivation and Reward. New
York: Springer-Verlag.
Taylor, A., Durbaba, R., & Rodgers, J. F. (1995). The evolving gamma loop. In Anthony
Taylor, Margaret Gladden, & Rade Durbaba (Eds.), Alpha and Gamma Motor Systems.
New York and London: Plenum Press.
Taylor, A., & Prochazka, A. (1981). Muscle receptors and movement: Proceedings of a sym-
posium held at the Sherrington School of Physiology. New York: Oxford University Press.
Watt, Douglas F. (1999). Consciousness and emotion. Review of Jaak Panksepp’s “Affective
Neuroscience”. Journal of Consciousness Studies, 6, 191–200.
Wehner, T., Vogt, S., & Stadler, M., (1984). Task-specific EMG characteristics during mental
practice. Psychological Reearch, 46, 369–401.
Wider, Kathleen V. (1997). The Bodily Nature of Consciousness. Ithaca and London: Cornell
University Press.
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.1 (41-102)

Doing it and meaning it

And the relationship between the two

Marek McGann and Steve Torrance

Centre for Research in Cognitive Science (COGS), University of
Sussex/DBS School of Arts, Dublin / School of Health & Social Sciences,
Middlesex University

A number of related approaches to cognition and consciousness have been

gaining momentum in the literature in recent years, emphasizing the active,
embodied nature of the mind and grounding mental states and processes in
the interaction between mind, body and world. These approaches are here
gathered under the title “enactive.” Given a working hypothesis, that there is a
meaningfulness to our conscious states that seems fundamental to those
conscious states, implications of this enactive approach to mind are outlined.
It is argued that taking such an enactive approach implies a fundamental role
for motivational states and goals, which currently lack an explicit explanation
in consciousness studies or cognitive science more generally. An enactive
framework for considering goals is sketched, and two of the more dramatic
implications for our understandings and investigations of consciousness
are outlined.

Keywords: Enaction, enactive theories of mind and consciousness,

embodiment, intentionality, conceptions and theories of goals

Introduction: A working hypothesis

Consciousness is many things to many people. Definitions are usually futile,

and sweeping statements are invitations to dismissal as much as to argument.
Progress must be made, however, and to this end let us begin with a claim, a
sweeping statement, a launching point for an argument about the nature and
structure of consciousness. The hypothesis is this: There is a meaningfulness
to our conscious states that seems to be fundamental to those conscious states.
Whatever else it may be, consciousness, being aware of ourselves or the world,
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.2 (102-152)

 Marek McGann and Steve Torrance

seems meaningful. Being conscious of something seems, at least in part, for

that thing to have some meaning for us. Our relationship with the world seems
mediated by this relationship. It is not one of simple occurrence, but one of
interaction. We are not rocks in a stream of experience, but agents actively
involved with our environment, our world.
This claim, if sensible, has some important implications for theories of
consciousness. It is not necessary to be insolubly wedded to it, but it would
seem fair to say that it is sufficiently prevalent in the literature on conscious-
ness to warrant investigating its implications in some depth. In particular, this
claim, coupled with a view on mind and consciousness fast gaining momen-
tum in the literature, has some important and fundamental ramifications for
theory and research. The view on mind and consciousness is the enactive ap-
proach, a perspective on cognition that makes central our nature as agents in
the world. There are a number of facets to this approach, many of which arose
independently within subdisciplines of cognitive science. Though there is no
united front to this perspective put forward in the literature, the present pa-
per will try to draw out the common themes that bear implications for our
investigation and understanding of consciousness.
In what follows, we will first very briefly deal with two possible counterex-
amples to the claim that consciousness is bound up with meaningfulness, then
outline the form of enactive approach that will be dealt with here, finding com-
monalities in the work of Varela, Thompson and Rosch (1991), Hurley (1998),
O’Regan and Noë (2001) and others. This will include drawing out implica-
tions concerning the kinds of concepts that must be clearly described if an
enactive approach is to hold together. In particular, the concept of perspective,
which is defining of an agent and subject in the world, will be examined and
given a basic evolutionary account. In doing so, the work of John Collier will be
used to explicate a difficulty faced by some such evolutionary accounts, and a
solution to that problem. An appreciation of the power of enaction in our un-
derstandings of the mind will throw into relief the question of motivations and
goals. A brief foray into the cognitive scientific literature provides no extant
theory of goals which might be used successfully to describe the agentic nature
of our existence, so a sketch of an enactive framework is drawn before some
of the more potent implications of the enactive approach for consciousness
studies are outlined. But first things first.
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.3 (152-192)

Doing it and meaning it 

Rejecting two potential counterexamples to meaningful consciousness

There are two possible counterexamples to the importance of meaningfulness

in consciousness, which need to be addressed before we can get to the issue at
hand. Firstly, is to point out that there are times or situations when we might
be fully aware of something, but fail to grasp any meaning in it. One might
read an article, for instance, but fail utterly to comprehend it. You might suffer
some experience, but see no point to it, become thoroughly confused by it and
be left unable to understand it, even at a very basic level.
A simple response to this argument is that such situations do not really in-
volve a complete absence of meaning. Rather, while there is meaning in it, it
is not the kind of meaning that we want. One may read a sentence and fail to
understand it, but could say many things about it - what it sounds like when
read aloud, whether it is a long or short sentence, if it is in a language we would
normally understand. What is not available to us is its meaning in the con-
text of the text in which the sentence is embedded. There might be a whole
set of similar aspects of a confusing event that could be described and under-
stood, but could not be fixed in any importantly relevant context. What is so
frustrating about such situations is that we are aware of them, aware of some
meaning in them, but are somehow prevented from grasping the meaning we
feel is actually important, that is relevant to us and our goals.
Another way of arguing with the claim that meaning is fundamental to
consciousness is to say that basic perceptual experiences might have no real
meaning in and of themselves. What, after all, is particularly meaningful about
a simple perception of redness? Or of pain? Is it not true, however, that a per-
ception of red entails “there is something red,” or, at the very least, “ there is
some redness”? This would be a very minimal kind of meaning, to be sure, but
even this minimal form requires explanation if consciousness is to be grasped
in its entirety.

Meaning what you do and doing what you mean: The enactive approach

The enactive approach has yet to put forward a united front in the literature
on consciousness and cognitive science. The term is used here in a somewhat
sweeping fashion to refer to a related set of views and beliefs on the mind
that have been gathering momentum in cognitive science over the past ten
to fifteen years. Though there are differences in the details, and some strong
disagreements amongst some of the proponents, certain themes and beliefs
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.4 (192-243)

 Marek McGann and Steve Torrance

are common. Chief amongst these, and fundamental to the perspective, is the
relationship between mind, body and world.
Varela, Thompson, and Rosch (1991), argue that mind and world are en-
acted through the interaction between subject and object, observer and world.
They reject the possibility of a completely pregiven observer or a completely
pregiven world. There can be no conception of one without the other. The
structures of the world allow the structures of the observer to exist, while the
structures of the observer allow the structures of the world to be conceived and
perceived. It is this complex interplay between the world and the subject which
gives rise to meaning, the understanding of the world.
The enactive approach to cognitive science makes use of a quite different
conception of the mind than more traditional information processing mod-
els. The mind, rather than receiving information, building representations and
producing new knowledge, or outputting commands to the body, is intricately
interwoven with its embodiment and its world. The enactive mind is not a
passive recipient of information from the world, but actively engages with
its environment, unbuffered by separable functions of perception and action
(Hurley 1998). This approach would draw upon and endorse a number of dif-
ferent research agendas becoming prevalent across the discipline. Perception,
for instance, is not conceived as the transmission of information but more as
an exploration of the world by various means (such as the active vision de-
scribed by O’Regan & Noë 2001). Cognition is not tied into the workings of
an “inner mind,” some cognitive core, but occurs in directed interaction be-
tween the body and the world it inhabits. The enactive mind is thus also an
embodied mind, the enactive approach ties in with the growing literature on
the role and value of embodiment. That the cognitive system owes more to
its physical instantiation than previously thought is a conception given clear
exposition by Clark (1997). He argues that the operation of the cognitive sys-
tem is not general, and potentially arbitrary, information processing but is
grounded in the details of the agent’s embodiment. The enactive view means
that thinking and action are radically goal-directed and constrained. This is not
a medium-neutral mind. The specifics of embodiment matter for the specifics
of cognition. The enactive approach would thus agree with the thinking of
Glenberg (1997), when he argues that the human memory system is not for
rote learning, but for guiding our actions in a complex three-dimensional and
subtly social world. Similarly a change in thinking about working memory is
called for, and might be found in the work of Ballard, Hayhoe, Pook, and Ra-
jesh (1997). Under their account, working memory is not necessarily about
holding representations active in mind but about selecting and targeting em-
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.5 (243-287)

Doing it and meaning it 

bodied operations in the service of present goals. Finally, learning a skill is not
about learning the relevant facts about the environment, but is about learning
to control the dynamics of our relationship with that environment, here again
other work in cognitive science finds just that (see Clark’s (1997) analysis of
Thelen and Smith’s (1994) work on children learning to get around different
kinds of obstacles while crawling, and then later when learning to walk).
Many of the cognitive functions, so often seen as requiring detailed “inner”
representations need not rely on such representations, but can be successfully
guided by the constraints imposed by the world itself (Hutchins 1995; O’Regan
1992; Clark 1997; O’Regan & Noë 2001). The moral of the tale is that the
mind is not removed from the world, tucked away inside the body, thinking its
thoughts in some private office, receiving reports and sending out instructions.
Rather, the mind exists in the interaction between embodiment and world. Its
operations may cross the boundaries of the organism, become extended into its
environment, and loop back in tighter or looser feedback relationships (Hurley
1998; Clark 1997; Clark & Chalmers, 1998).

Enaction and the importance of perspective

The enactive approach binds meaning into the mutual constraint of subject
and world, a complex feedback dynamic. Such feedback dynamics, however,
cannot be sufficient for descriptions of mental content and thinking. Our intu-
itions regarding the mind require that the set of systems with complex feedback
dynamics be much larger than the set of cognitive systems. The weather can-
not think, surely, nor do the dynamics of population densities in predators and
their prey deliberate. Hurley (1998) suggests that fundamental to the concept
of cognition and consciousness is the concept of a perspective. While meaning
may be a function of the complex causal relations in a dynamical system, what
fixes that function (what effectively disambiguates the set of relations in some
way) is a perspective. Hurley’s purpose in her discussions of perspective (see
Hurley 1998, particularly Essays 2, 4, 5 and 9) is an analysis of the interdepen-
dence of perception and action in the mind, and she does not commit herself
to any particular explanation of the phenomenon of perspective, though she
suggests (Hurley 1998: 7) that an evolutionary approach may be employed to
get a grip on things. Her purposes simply require that something be capable
of fixing the inter-relationship between causal processes at a subpersonal level,
and mental content at the personal level at particular times.
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.6 (287-337)

 Marek McGann and Steve Torrance

A complete dynamic theory of meaning, built on the likes of Susan Hurley’s

approach, accepts an active nature for the cognitive system. It argues for an
interdependence between perception, action and cognition and requires some
explanation for the existence of perspective to value the complex of feedback
that is the embodied organism.

An appeal to evolution, a problem and a solution

A simple way of addressing the question of perspective might be to appeal to

evolution, claiming that natural selection made it that way. This will in fact be
the approach taken in the present paper. However, following a line of argument
put forward by John Collier (2000), it can be suggested that to focus on natural
selection alone will leave us with an incomplete or at least a less intuitively
correct explanation of perspective.
Collier (2000) argues that the typical stance taken in dealing with evo-
lution (to trot out the unassailable logic of differential reproduction) misses
something important. Under this typical view, function is defined in terms of
its ultimate value to differential reproduction. Collier (2000), however, argues
that this view focuses broadly on the general case of the lineage and plays down
the importance of the individual case, the function in terms of a specific ani-
mal. It is at this specific individual level that functions must operate in order
to be selected. Collier thus suggests that some evolutionary accounts confuse
consequences with causes. The cause is the usefulness to the individual animal,
the consequence, selection.
Collier claims that etiological accounts of function (those which emphasize
pure selection) mostly ignore the organizational requirements of biological en-
tities, even when this organization may play an important role in the function
in question. Such purely selectionist accounts can lead us into an erroneous
modular account that fails to appreciate the importance of the configuration
and flexibility of capacities. Selection acts as a limiting but not determining
factor on functionality. In order to be selected, a trait must already be func-
tional - that is why it is selected in the first place. The functionality of a trait can
often be determined without any knowledge of the origins of that trait. Thus,
the perspective of the individual animal is important in the identification and
analysis of characteristics (including, presumably, cognitive characteristics).
However, this is not to deny the validity of the evolutionary approach, but
simply to shift the focus of its analytical spotlight. Collier’s (2000) point is
that functions should be understood within the framework of a set of orga-
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.7 (337-396)

Doing it and meaning it 

nizational requirements which themselves may be best understood in terms of

natural selection. To elide this individual level of value of a trait, and jump
straight to the value for the lineage, will be to miss something non-trivial. Col-
lier is thus arguing that an understanding of the emergence of an individual
perspective is essential to an understanding of the kinds of functions, cognitive
functions, that we are interested in.
Collier’s (2000) aim is to provide some means of allowing traits to be val-
ued (and thus considered functional or not functional) in the individual case.
To this end, he employs the concept of autonomy. In a 2002 paper, he defines
autonomy as follows:
A system is autonomous if and only if the organization of internal aspects
of system processes is the dominant factor in the system’s self-preservation
making itself and the processes that contribute to autonomy functional.
(Collier 2002)

This view is informed by Maturana and Varela’s (1980, 1987) concept of au-
topoiesis, but is less absolute, admitting of gradations where autopoiesis is all
or nothing. Nevertheless, Collier’s approach can be put to use, allowing us to
drive an individual system into a gauntlet where it must sacrifice its autonomy
and become indistinguishable as an entity independent of its environment or
risk disintegration in being insensitive and unadaptable to its environment.
Life is dynamic, but those dynamics are not open-ended. A system which does
not in some way maintain some kind of boundary between itself and the dy-
namics of the world surrounding it is essentially indistinguishable from that
world. The weather is not alive, because its open dynamics do not distinguish
it in any real way from the broader physical system in which it is embedded.
On the other extreme, however, a rock may have a much more constrained dy-
namics, but these are constrained almost to the point of non-existence. A rock
has no dynamics at a global level of description which might allow us to distin-
guish it as a entity as anything more then the sum of its physical parts. In such
a gauntlet of distinction and unity do life and cognition develop.

The centrality of goals

The enactive approach encourages a more holistic view of the mind and cogni-
tion than more traditional perspectives, placing cognitive functions in a con-
text of embodiment, evolutionary history, or personal and physical constraint.
The cognitive system is conceived as active rather than reactive.1 Meaning
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.8 (396-441)

 Marek McGann and Steve Torrance

is bound up in the dynamics of the system and those dynamics, under the
enactive approach, describe an agent. These are by implication dynamics of
goal-directedness. This is a concept which the different aspects of the enactive
literature have orbited, but which has not yet been given the central place in
our thinking about thinking that it deserves. The mind is engaged in satisfy-
ing the needs, desires and goals of the organism. This simple fact appears to
run behind all of the aspects of the enactive approach. In satisfying these needs
universal computation and representation are not generally necessary. In order
to understand an agent’s behavior you must deal with the specific context of
the limits and demands of its embodiment. The fact that these demands and
constraints are not arbitrary but directed at some end is skirted continually in
enactive writings, but no one, as yet, seems to have driven this point home: To
explain what it is the mind does we must offer an explicit account of motiva-
tion and goals. More so, as these operate at all levels of cognitive function, this
account cannot assume conceptual capacities or representational functions of
the kind in which cognitive theories typically traffic. The enactive approach
plays down the omnipresence of such representational capacities in cognition.

Conceptions of goals extant in psychology

As we have seen, other aspects of the enactive approach were presaged or inde-
pendently proposed within different domains of cognitive science. Might this
be the same for a theory of goals? Psychology, as representative of cognitive sci-
entific work, includes many theories that make reference to motivational states
or goals. However, those theories extant within psychology tend to hold def-
initions of goals as either explicitly representational or implicit and assumed.
Social and personality theories such as Bandura’s (1992, 1997), or Mischel and
Shoda’s (1995) have increasingly emphasized the manner in which actions are
driven and directed by an agent’s goals, but the level of description of these
goals tends to be at the level of personhood, self-esteem and self-efficacy. A
range of such evaluative concepts are described by these and similar theories,
and while they may count as a challenging particular case to a general theory of
motivation, they resist easy integration into a perspective on the mind which
denies the prevalence of conceptually driven thought. Social and personality
theories are description at too high a level, then, for what we need. We might
turn to more analytic, cognitive approaches.
The torrent of research over the past decade on executive function is much
more usefully pitched at a subpersonal level of explanation. However, it appears
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.9 (441-471)

Doing it and meaning it 

to take the concept of goals as a given, a starting point for theory and experi-
ment. This research has seemed to focus on the processes involved in adopting,
following and switching between goals, but the concept of goals themselves is
implicit. Selective attention, choosing to focus consciousness on certain stim-
uli rather than others, is typically evidence of goal-directed cognition (Duncan
2001). However, research on selective attention tends to stipulate or assume the
goal to be followed. We have significant control over what we pay attention to,
and attention focuses cognitive effort, but precisely what the goal in question is
or how it is instantiated in the system is far from clear. Research into selective
attention’s counterpart, inhibition (Wegner 1997; Wegner & Wenzlaff 1996),
suffers from a similar problem, as do monitoring (Norman & Shallice 1986)
and task-switching (Monsell 2003, 1996). These different streams of research
highlight the control that we human beings are capable of exerting over our
cognitive functions, but either seek a mechanistic account where goals are un-
necessary constructs or assume the existence of goals and work at higher levels
of description, attempting to explain how they are followed.
More venerable research traditions in cognitive science have more to say on
goals, but for the main part fail as an adequate explanation in one of these two
ways. Newell and Simon’s (1972) problem-space theory of problem solving ef-
fectively uses goals as representational reference values. These do not drive the
system, but act as data for some comparator function to determine whether
or not the goal state is present or absent. The rather rigid contexts required
to assess this theory (which appears to remain without significant challenge as
a framework for problem solving theory: Hunt 1994, Eysenck & Keane 2000)
predetermine the goals of participants. A problem is often presented as a puzzle
to solve, where the problem dimensions and valid operations are few. More
naturalistic problem solving, in more fluid environments, slides into the arena
of naturalistic decision making. A new process-tracing research paradigm (see
Crozier & Ranyard 1997) has moved substantially away from normative ac-
counts (such as Tversky & Kahneman’s prospect theory 1979). This new ap-
proach has given rise to a change in description of the entire process. No longer
is described an early information gathering stage at the end of which some eval-
uative function divides the best (or least worst) response from the also-rans.
Decisions are dynamic negotiations between the agent and world in which
some early choice is made and then used as the launch point for a series of
interdependent evaluations, option-differentiation and decision-consolidation
processes, wrapped in an action-feedback cycle that the new approach con-
siders central to the way decisions are made. Here too, however, goals remain
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.10 (471-518)

 Marek McGann and Steve Torrance

described at a fairly personal and intuitive level, with no real means available
of analyzing the concept in terms amenable to an enactive perspective.

An enactive framework for understanding goals

The enactive approach, under the view presented here, demands a means to un-
derstand ends. Contemporary cognitive science can offer some suggestions but
few real answers. What follows will roughly sketch the outline of a framework
that might be useful in structuring our thinking on this issue, a framework that
might successfully bridge the space between a minimal case of enactive poten-
tial and the complex self-governed dynamics that we humans seem to enjoy.
The minimal case must occur at the origin of perspective. A result of evo-
lution, we might expect this minimal case to be impossible to define precisely,
but to be the low end of a gradient of entities for whom the world holds some
implication. At this simple level we might expect there to be some system which
operates to maintain itself, has autonomy, and therefore impacts on the system
from its environment will affect it in positive and negative ways (no simply
neutral occurrences here, the world implies something for this system). The
system will have basic interests that must be maintained, and if it is not struc-
tured in such a way as to serve those interests, it will not survive for long. The
minimal case of enaction, then, is governed by interests that are basic concerns
of an autonomous living system. Life regulatory processes such as homeostasis
might be considered the basic instantiation of interests. Being basic, these val-
ues are also immediate and inflexible. A creature that operates only according
to interests will have no possibility of either prediction or memory, slaved to
the immediate state of its basic motivations. Some simple single cell organisms
and some plants can be considered in this category, where the current impact
of the environment on the organism controls its movements, but does so in the
interests of that organism. Think, for instance, of the manner in which a glu-
cose gradient drives the flagella of a bacterium so that it will move into richer
fields of nutrients.
Given some flexibility, interests and the actions they produce might be-
come sensitive to context, so that the relationship between stimulus and re-
sponse is less rigid. Flexibility will be constrained by the fundamental organi-
zation of the creature around its interests, but even a limited flexibility may
allow contexts and actions to be loosely configurable. Following Susan Hur-
ley (2003), such flexibility may provide for a coarse combinatorial structure
between means and ends and the possibility of action in error (hence norma-
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.11 (518-577)

Doing it and meaning it 

tivity). Hurley argues that these two criteria are sufficient for describing the
agent in question as acting according to reasons (and that these reasons are not
simply attributed, but exist at the appropriate level of description of the agent).
The development of symbol use and voluntary memory allows human be-
ings to control their cognitive context in an ad-hoc and arbitrary manner. The
world becomes suffused with meaning layered ever-thicker by every new com-
mitment we make. You may look at your watch and say that it is eight o’clock.
This has a number of possible meanings, but if you are intending to catch a
plane at half past nine it becomes even more loaded with meaning for you. We
weave rich tapestries of habitual goals over our life-span, but it is in the light
of having such self-generated goals, purposes, that our world is so rich with
implication.
This gradient of interests, with limited body-specific implications, to rea-
sons, with context-dependent but flexible action-specific implications, to pur-
poses with context-configurable but commitment-specific implications is of-
fered as a first-step to scaffolding a general understanding of the goal-directed
structure of enaction. Goal-directedness, under the enactive approach, is fun-
damental to and constitutive of the mind. This has dramatic ramifications for
consciousness.

What it means for consciousness

Theories of consciousness often take awareness to be prior to motivation. Mo-

tivation and goals may be accepted as having some organizing and structuring
role in consciousness, but play no constitutive part. The radically goal-directed
nature of mind, an implication, it is argued here, of the enactive view, turns
this view on its head. The meaningfulness that is so central to consciousness
depends on a bedrock of motivation.
Antonio Damasio (1999) has made emotions the core of his theory of
consciousness. His account is integrable within an enactive framework, ac-
knowledging as he does the importance of embodiment (Damasio 1994) and
the grounding of consciousness in life regulatory processes (Damasio 1999,
1994). His Descartes’ Error puts forward a theory emphasizing the organiza-
tional role of emotions over basic and higher cognitive processes. He developed
from this his view of emotions and feelings as underpinning the very basis of
consciousness. From an enactive perspective this would seem to be a promising
road to follow. Though Damasio has laid the foundation for thinking in these
terms his own principal aim has been to identify the neural circuitry respon-
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.12 (577-624)

 Marek McGann and Steve Torrance

sible for maintaining these functions. A complete enactive theory of content

and consciousness will be broader than this, encompassing all possible forms
of instantiating awareness. Damasio also relies on concepts of “maps” and rep-
resentations which, while not a major difficulty for integration, require some
explication in more dynamic terms. Also, the relationship between emotions
and goals (which may be one of identity) will need further analysis to follow
this line of investigation. There are numerous accounts to choose as a start-
ing point: Damasio’s own, those of Panksepp (1998), Lazarus (1991), Frijda
(Frijda & Moffat 1993; Frijda & Swagerman 1987), or Power and Dagleish
(1999), to name but a few. Each provides challenges for integration into the
enactive framework, but each maintains a prominent role for emotion and
motivation in cognitive function and consciousness. This foundation of evalu-
ative and motivational issues underlying consciousness potentially has a more
wide-sweeping implication for our investigations, however.
Many theories of consciousness in contemporary consciousness studies
seem to stress the “how” question of consciousness. It is this determined focus:
“how could any physical system be conscious?”, “how can it feel like some-
thing to ....?” that stands behind the so-called hard problem of consciousness
(Chalmers 1995). The goal-focused enactive approach presented here sidesteps
this issue somewhat, suggesting that the more sensible question is not “how?”
but “why?”. The points put forward in this paper suggest that the kinds of
meaning that are constitutive of our conscious experiences are formed on the
basis of sensitivity to relevant relationships between subjects and world. Per-
spective is formed of dynamic relationships that define a teleological gradient
which in turn defines a gradient of implications. Given a system defined by an
agent and the implications of the world for that agent, it may seem more appro-
priate to ask how could it not feel like something. To claim that consciousness
can somehow be conceived of independently of this relationship of meaning,
steeped in the facts of embodiment and dynamics, challenges the claim that
consciousness is meaningful. Such views, attempting to explain how there can
be such things as “qualia” or “subjective feels” are forced to make the further
step of explaining the relationship between subjective feels and the meaning-
fulness that consciousness seems to enjoy. If the contents of consciousness are
indeed meaningful, an assumption of the arguments of the present paper, then
this might be to take a divide-and-conquer strategy to explaining conscious-
ness a step too far. Perhaps there has been a little too much dividing, and not
enough conquering.
The enactive perspective, then, will find it more productive to focus on
why something is experienced in this way rather than that. This will be to move
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.13 (624-660)

Doing it and meaning it 

the emphasis, as Hurley and Noë (2003) do, from an absolute explanatory gap
(“how can physical systems be conscious at all?”) to a comparative one (“how
do different forms of consciousness differ from one another?”). Addressing
such questions will involve not just a first-person approach to conscious states,
but also a richly context sensitive approach in which the interests, reasons and
purposes of the subject are given account. The functions of consciousness and
its structure are the target of such a context-sensitive approach. Merlin Don-
ald (2001) has argued vehemently for a move to such a “why”-focused view.
Donald arrived at this conclusion not from a theoretical standpoint, not via
an enactive perspective, but through survey of a wealth of data concerning
cognition and neural function. The enactive approach may provide a fresh per-
spective on old data, and more productive approaches to future research; fresh
questions from a new set of assumptions.
How we view consciousness, as passive froth on cognitive function, or ac-
tive and goal-directed “all the way down”; as abstract information processor or
grounded in real world embodied constraints: these things are once more in
review. The dynamic and deliberate nature of consciousness has been thrown
into stark new relief by the enactive approach.

Acknowledgments

Both authors are grateful to Dr. Ron Chrisley at COGS, who supervises the first
author. A debt of gratitude is also due to the E-Intentionality seminar group,
particularly Mike Beaton, Rob Clowes, Tony Morse and Hanne De Jaegher at
Sussex for their interest and comments on different issues raised here. The
authors are also grateful for discussions with Alva Noë, Evan Thompson and
Erik Myin.

Note

. Reactive conceptions of mind are prevalent in both behaviorist and many cognitivist ac-
counts. The cognitive system seems often given the job of keeping track of the world in a
dispassionate fashion, developing representational perceptions and drawing inferences in
algorithmic determinism, driven by data.
 JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.14 (660-777)

 Marek McGann and Steve Torrance

References

Bandura, A. (1997). Self-efficacy: The exercise of control. Cambridge: Cambridge University

Press.
Bandura, A. (1992). Self-efficacy mechanisms in psychobiological functioning. In R.
Schwarzer (Ed.), Self-efficacy: Thought control of action. Washington, DC: Hemisphere.
Chalmers, D. (1995). The conscious mind. Oxford: Oxford University Press.
Clark, A. (1997). Being there: Reuniting brain, body and world. Cambridge, MA: MIT Press.
Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58, 7–19.
Collier, J. (2002). What is autonomy? International Journal of Anticipatory Systems, 12, 212–
221.
Collier, J. (2000). Autonomy and process closure as the basis for functionality. Annals of the
New York Academy of Science, 901, 280–291.
Crozier, R., & Ranyard, R. (1997). Cognitive process models and explanations of decision
making. In R. Ranyard, W. R. Crozier, & O. Svenson (Eds.), Decision Making: Cognitive
Models and Explanations. London: Routledge.
Damasio, A. R. (1999). The feeling of what happens. London: Willian Heineman.
Damasio, A. R. (1994). Descartes’ error. New York: Papermac.
Duncan, J. (2001). Frontal lobe function and the control of visual attention. In J. Braun,
C. Koch, & J. Davis (Eds.), Visual attention and cortical circuits. Cambridge, MA: MIT
Press.
Eysenck, M. J., & Keane, M. T. (2000). Cognitive psychology: A student’s handbook. Hove:
Psychology Press.
Frijda, N., & Moffat, D. (1994). Modeling emotion. Cognitive studies. Bulletin of the Japanese
Cognitive Science Society, 1 (2), 5–15.
Frijda, N., & Swagerman, J. (1987). Can computers feel? Cognition and Emotion, 1, 235–258.
Glenberg, A. (1997). What memory is for. Behavioral and Brain Sciences, 8, 1–55.
Hunt, E. (1994). Problem solving. In R. Sternberg (Ed.), Thinking and problem solving. Hove:
Academic Press.
Hurley, S. L. (2003). Animal action in the space of reasons. Mind and Language, 18, 231–256.
Hurley, S. L. (1997). Consciousness in action. Cambridge, MA: Harvard University Press.
Hurley, S. L., & Noë, A. (2003). Neural plasticity and consciousness. Biology and Philosophy,
18, 131–168.
Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision making under
risk. Ecomometrica, 47, 263–291.
Lazarus, R. S. (1991). Emotion and adaptation. Oxford: Oxford University Press.
Maturana, H. R., & Varela, F. J. (1987). The tree of knowledge: The biological roots of human
understanding. Boston: New Science Library.
Maturana, H. R., & Varela, F. J. (1980). Autopoiesis and cognition. Boston: D. Reidel.
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7, 134–140.
Monsell, S. (1996). Control of mental processes. In V. Bruce (Ed.), Unsolved mysteries of the
mind: Tutorial essays in cognition. Oxford: Erlbaum.
Newll, A., & Simon, H. (1972). Human problem solving. Englewood Cliffs, NJ: Prentice Hall.
JB[v.20020404] Prn:23/11/2004; 10:19 F: Z13108.tex / p.15 (777-806)

Doing it and meaning it 

Norman, D., & Shallice, T. (1986). Attention to action: Willed and automatic control of
behaviour. In R. J. Davidson, G. Schwartz, & D. Shapiro (Eds.), Consciousness and self-
regulation. New York: Plenum.
Panksepp, J. (1998). Affective neuroscience. Oxford: Oxford University Press.
Power, M., & Dalgleish, T. (1999). Two routes to emotion: Some implications of multi-level
theories of emotion. Behavioral and Cognitive Psychotherapy, 27, 129–141.
Thelen, E., & Smith, J. (1994). A dynamic systems approach to the development of cognition
and action. Cambridge, MA: MIT Press.
Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind. Cambridge, MA: MIT
Press.
Wegner, D. M. (1997). When the antidote is the poison. Psychological Science, 8, 148–150.
Wegner, D. M., & Wenzlaff, R. M. (1996). Mental control. In E. T. Higgins & A. Kruglanski
(Ed.), Social psychology: Handbook of basic principles. New York: Guildford Press.
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.1 (41-104)

Anticipatory consciousness, Libet’s veto and

a close-enough theory of free will

Azim F. Shariff and Jordan B. Peterson

University of Toronto

Benjamin Libet concluded that consciousness operates under severe

constraints, given its relationship to the timing of neural events. He therefore
proposed that the power of consciousness is limited to veto: an action
sequence can be stopped, consciously, once unconsciously initiated, but
real-time actions cannot be initiated, consciously. However, it is possible that
consciousness has the power to initiate the ballistic operation of previously
automatized responses. Under such conditions, it could still possess
substantial causal power, except at the smallest time-frames. We therefore
suggest that conscious attention is typically focused on the future, rather than
the present, and that this focus on the future allows it to remain causally
effective in real-time, despite the time-lag necessary for the initiation of
voluntary actions. This reformulation of the relationship between time and
consciousness allows for the existence of free will, within the confines of a
model of action that remains in accordance with the tenets of modern
neuroscience.

Keywords: Free will, conscious volition, Benjamin Libet, time-lag, backwards

referral, supervisory attentional system

Introduction

The outcome of Benjamin Libet’s experiments on the timing of conscious

awareness put a new twist on the long-standing debate about the existence and
nature of free will. His most heavily cited experiment (Libet 1985) aimed to
determine where in the sequence of a voluntary action conscious awareness
enters and acts. Libet demonstrated that awareness of an action (in this specific
case, a wrist flick) occurred 250–350 ms after EEG-measured reaction poten-
tial technology indicated its onset or preparation. Libet logically concluded that
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.2 (104-155)

 Azim F. Shariff and Jordan B. Peterson

consciousness could not be causally involved in such action, since it arises af-
ter its initiation: “The initiation of the freely voluntary act appears to begin in
the brain unconsciously, well before the person consciously knows he wants
to act!” (Libet 1999: 51). If consciousness is acting at all, therefore, it must be
doing so at least 350 ms before voluntary acts occur.
Further experiments by Libet and others (Libet, Alberts, Wright, & Fein-
stein 1967; Keller & Heckhausen 1990) have provided support for the con-
cept of a conscious time-lag that would preclude the real-time response of
consciously-derived actions. Norretranders (1991) has argued that this time
lag is a necessary consequence of the immense amount of stimulus informa-
tion that must first be processed and then discarded to provide a coherent and
concise story that can be consciously grasped. Such conclusions appear log-
ically sound. However, their implications still seem difficult to accept. If we
cannot respond in real time with consciously determined actions, then our ac-
tions are unconsciously determined, by necessity. Our perception of conscious
self-efficacy – conscious free will – is illusory.
However, the hypothesis of free will as an illusion appears to inexorably
relegate consciousness to the realm of epiphenomenalism. Without any force
of its own, consciousness must merely be a parallel phenomenon that occurs
alongside our actions – a spectator, not a participant. John Searle (2000) has
suggested, nonetheless, that the epiphenomenalist explanation of conscious-
ness is unsatisfying – in no small part because of its incompatibility with what
we understand about evolutionary processes. First, consciousness is a complex
and metabolically expensive biological process, as revealed by demonstration of
the differences in the glucose consumption of unconscious and conscious brain
states (Nofzinger, Mintun, Wiseman, Kupfer & Moore 1997). Second, the rapid
increase in human brain size likely contributed to emergence of consciousness
in humans. This encephalization process has resulted in significant evolution-
ary costs in the form of danger to both baby and mother during childbirth
(Travathan 1987) and the prolongation of infantile dependence. The adaptive
value of consciousness would therefore logically need to outweigh the costs that
accompany it. But how much adaptive value could a system that is unable to
make any changes to the actions of the organism have? As far as survival or re-
production is concerned, an epiphenomenal consciousness must be completely
ineffectual. So why incur the costs? Thus the problem remains unsolved and a
matter of heated debate.
Libet tried to fit the possibility of an active consciousness within his the-
ory by arguing that the purpose of consciousness was its ability to veto any
unconsciously derived action. Various members of Libet’s group (for example,
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.3 (155-201)

A close-enough theory of free will 

Libet, Gleason, Wright & Pearl 1983) have shown that it takes 50 ms to deliver a
synaptic message to move a peripheral body part, such as the wrist. Conscious
awareness of such an action appears to emerge about 200 ms prior to the ac-
tion. This means that only 150 ms remain for the subject to consciously choose
whether or not to proceed with the action. According to Libet’s hypothesis,
consciousness serves as a mechanism of last resort, capable of selecting which
unconsciously prepared decisions result in action.
This veto argument, although influential, is not without its problems. Pri-
marily: shouldn’t this conscious choice, concerning veto, be unconsciously ini-
tiated as well? Libet (1999) tries valiantly to indicate that this is not necessary.
However, he runs into the following problem: the veto choice is conscious and
immediate, and must therefore occur in a window of 100 ms. Other decisions,
as described previously, cannot be conscious, because 300 ms are required for
awareness of choice. In order to reconcile these seemingly incompatible ideas,
Libet constructs a complex logical case, describing the veto as “a control func-
tion, different from simply becoming aware of the wish to act” (p. 53). Finding
no direct empirical support for this position, he is forced in the end to rely on
a lack of counter-evidence: “And, there is no experimental evidence against the
possibility that the control process may appear without development by prior
unconscious processes” (p. 53).
Instead of a conscious intervention that affects the action outcome after
it has been initiated, however, it still appears reasonable to posit that the ac-
tive role of consciousness comes beforehand – the role that temporality plays
in the process just has to be clarified further. We therefore propose that con-
sciousness acts in an indirect and more temporally distal role, anticipating and
perceiving upcoming stimuli (more than 400 ms into the future), and mod-
ulating the subject’s largely unconscious preparatory responses, in accordance
with Norman and Shallice’s (1986) model of Supervisory Attention. Norman
and Shallice’s model, our contributions to it, and a reinterpretation of Libet’s
veto will be subsequently examined in more detail.

The routinization of behaviour

Shallice has provided an information-processing model of attention, based

on a distinction between automatic and deliberate action that goes back at
least as far as James (1890). In Shallice’s view, automatic and deliberate ac-
tions are handled by two separate systems: the Contention Scheduling System
and the Supervisory Attentional System. Routine, automatic, well-learned be-
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.4 (201-259)

 Azim F. Shariff and Jordan B. Peterson

haviours that do not require significant attentional resources are handled by

the former system, while novel and non-routine behaviours are facilitated by
the latter. The notion of Contention Scheduling is predicated on the assump-
tion that complex motor programs (schemas, in Shallice’s terminology) can
be unconsciously initiated. This assumption is also shared by Libet, who holds
that all action is initiated without conscious involvement. Direct evidence for
unconscious action initiation has come from many sources.
Australian physiologists Taylor and McCloskey (1990, 1996) have used
backwards masking, for example, to elicit a complex and fully voluntary motor
program, without triggering conscious attention of the eliciting stimulus. As
the subjects of these experiments were not consciously aware of the stimulus
that triggered their behaviour, so the argument proceeds, they could not have
consciously initiated their actions. The validity of this argument is predicated
on the assumption that the motor program elicited by the trigger must have
existed prior to its initiation, as the processes underlying its construction are
so complex that they could not occur (1) without conscious mediation and (2)
in the interval between stimulation and action.
In a follow-up experiment, Taylor & McCloskey (1996) showed that sub-
jects can manifest two pre-programmed (previously learned) voluntary motor
responses to two different masked stimuli. Taylor & McCloskey have neither
yet established an upper limit to the number of pre-programmed response
patterns that could be held in readiness, nor determined how complex these
programs might be. However, the authors do suggest a primarily unconscious
and pre-programmed interaction with the world, in a manner that would allow
for Libet’s time-lag of consciousness:
Almost all motor reactions and many other motor performances must occur
before conscious perception of their triggering stimulus. Such a model could
allow the stimulus to act as a trigger when only a small amount of sensory
data has been processed, although conscious perception would require further
sensory input. (Taylor & McCloskey 1990: 445)

Recently, Ann Graybiel (1998) has suggested that many motor actions may be
grouped – or chunked – together to form structures similar to the schemas
described by Shallice.1 This chunking (a term borrowed from Miller’s (1957)
description of the compression of discrete bits of information in memory) re-
codes mental representations of action sequences by reducing the number of
distinctively represented units – essentially collapsing several motor programs
into one. Thus, these action repertoires, once elicited, run both automatically
and ballistically. From Shallice’s perspective, action schemas compete actively
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.5 (259-295)

A close-enough theory of free will 

for activation. When a given schema exceeds a particularly activation threshold,

it runs, and laterally inhibits the operation of all related schemas, to avoid out-
put conflict during implementation (see Cooper, Shallice, & Farringdon 1995
for a discussion and simulation of the contention scheduling system). Recent
work reviewed by Swanson (2000) lays a plausible hierarchically-structured
neurophysiological platform underneath these speculations. Over-learned or
reflexive action schemas are neurally instantiated at specialized, lower levels
of the nervous system hierarchy. This specialized low level instantiation allows
the rapid implementation of well-practiced or reflexive action (at the cost of
flexibility, which requires more processing and is therefore slower). Sensory
input splits into two main streams, early in the processing chain. Some input
is relayed directly to hypothalamic and lower spinal motor control systems,
allowing for very rapid but “unconscious” responses. Other input is relayed
upwards, into higher order cortical centers concerned primarily with complex
but time-consuming high-resolution perception and subsequent action.
The more complex but slower processing/action systems are particularly
useful in non-routine or novel situations, where deliberate attention is re-
quired, or when existing schema produce unexpected errors in outcome. The
Supervisory Attentional System, responsible for processing under such condi-
tions, does not directly select actions, according to Norman & Shallice (1986),
but applies extra activation or inhibition to the Contention Scheduling system
in order to bias the selection of particular schemas. Thus, attention directed
towards a certain plan of action greatly increases the likelihood of that plan
being selected by the Contention Scheduling system. While at bat in baseball,
for example, an individual’s attentional resources are typically directed towards
hitting an oncoming pitch, rather than catching it. Simple conscious concen-
tration on this particular goal biases the selection of the automatized “hit the
ball” schema, which is itself a hierarchy of efferent muscle commands, control-
ling aim, step and swing. When one of the automatic sequences proves faulty
(in case of “schema malfunction”) it cannot typically be fixed in real time,
partly for reasons of time constraint, and partly because the individual com-
ponents of the sequence have been chunked into a whole and are no longer
available to voluntary regulation at the level of detail required. The program
must therefore be brought effortfully into voluntary attention and unpacked
into its constituent elements for the purposes of repair. The processes underly-
ing playing a musical instrument, such as a guitar, provide a relevant example.
Individual muscle movements are chunked into the fingering and plucking or
strumming of strings. These notes and chords are themselves chunked into
progressions. The underlying motor “macros” operate ballistically, increas-
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.6 (295-341)

 Azim F. Shariff and Jordan B. Peterson

ingly, as expertise with the musical piece is developed. Once a bad habit is
established, the attempt to modify it requires the unpacking, or unchunking,
of the part of the motor hierarchy that is now underneath the conscious level.
What this means, in essence, is that “consciousness” moves up the motor hi-
erarchy, as each level of that hierarchy becomes automated. Once automated,
however, the level is “unconscious,” and has to be unpacked and re-presented
to consciousness before it can be repaired or changed.

Indirect conscious control through the biasing of attentional resources

There is strong clinical evidence that conscious willing of actions and the ac-
tual execution of those actions are handled separately within the brain. Such
evidence has been derived primarily through analysis of “double dissociation”
disorders – disorders where one process but not the other appears damaged
(see Appendix). Some of this evidence suggests that automatic behavioural
“macros” extend beyond motor sequences, into the domain of object per-
ception – which in turn appears as something very much predicated on the
pragmatics of use, rather than material or objective feature (see Gibson 1979).
Indeed, Shallice and his colleagues (Cooper, Shallice & Farringdon 1995), de-
scribe the process of “environmental activation.” Though the presence of a
single object or “stimulus” can initiate a specific schema, they explain, it is
more likely that complex and sophisticated schemas will be activated by com-
plex and evocative environmental situations (such as combinations or arrays
of objects such as a lit match and a dangling cigarette) (see Jeannerod 1997, for
a similar discussion).
However, the idea that such automated perception-motor spanning pro-
cesses exist can also be integrated productively with the idea of the Supervisory
Attentional System. Selective attention can clearly alter the perception of visual
stimuli – from the simple and meticulously studied reversibility of the Necker
cube (Orbach, Ehrlich & Haith 1963), to the complex ability to discriminate
“objects” from the surrounding sensory clutter. Consider the well-known cock-
tail party effect (Cherry 1953) or, more recently, the results of fMRI studies
conducted on the nature of visual attention by Kastner, De Weerd, Desimone,
& Ungerleider (1998). Kastner et al. demonstrated that mental representations
of various stimuli in a cluttered array interact in competitive and mutually in-
hibitory ways. Each stimulus expresses a suppressive effect on other neighbour-
ing stimuli in the visual field. However, the focus of attention on one particular
stimulus offsets the competing suppressive effects induced by these neighbour-
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.7 (341-407)

A close-enough theory of free will 

ing stimuli. This mechanism of selective attention appears analogous to the

lateral inhibition theoretically characteristic of Shallice’s Contention Schedul-
ing processes (Cooper, Shallice & Farringdon 1995). Thus, the Supervisory
Attentional System may also be directed at the sensory field, biasing certain per-
ceptions, in addition to or supplementing activation of their associated motor
responses. It does not seem too much to say, in fact, that the act of perception
and the process of biasing motor outcome are in some manner inseparable
phenomena.
Interesting similarities exist between the properties that serve as action cues
in the previous accounts and the concept of affordances characteristic of J. J.
Gibson’s theory of direct perception (Gibson 1979). According to Gibson, an
affordance is what an object in the environment offers an organism – what it
invites from the perceiver (a chair, for instance, affords sitting; a glass of wa-
ter, drinking; air, walking through, and so on). From Gibson’s perspective, the
perception of the affordance is integrally linked to the selection of appropriate
action. The three fundamental properties of Gibsonian affordances are
1. An affordance exists relative to the action capabilities of a particular actor.
2. The existence of an affordance is independent of the actor’s ability to per-
ceive it.
3. An affordance does not change as the needs and goals of the actor change.

Though the idea of affordance offers a useful way of thinking about the action
cues offered by the environment, Gibson’s definition appears too limited – par-
ticularly with regards to his final two points. In keeping with his over-arching
theory of the directness of visual perception, he maintains that affordances are
invariant properties of an “object.” However, the direct perception hypothesis,
which essentially presumed the absence of any thoroughly mediated percep-
tual processing, has never gathered mainstream support (see Fodor & Pylyshyn
1981 for an early criticism). If affordances are conceptualized more precisely as
emergent properties of the interaction between subject and object, however,
the situation may be easily rectified: the affordances of an object can not only
change relative to the action capacities of a particular subject (as Gibson sug-
gests), but also change between subjects with the same capabilities and – more
importantly – within subjects, as the goals of the perceiving individual change.
There is no reason why a knife cannot reveal the affordance of cutting
bread, for example, or spreading butter, depending on the current goals and at-
tentional focus of a subject. From such a perspective, voluntary attention serves
in part to alter the perceptual interpretation of information. For all intents and
purposes, therefore, the “same” sensory information may precipitate into one
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.8 (407-438)

 Azim F. Shariff and Jordan B. Peterson

of several potentially “different” objects, with different affordances – linked, as

a result, to different action responses. Indeed, what is seen as an “object” may
be better understood as a pattern of environmental information that, over time,
has distinguished itself from a broader and unspecified background, because
of its particular, goal-directed utility (or utilities). A given pattern comes to
evoke increasingly automatized attention, from such a perspective, because of
its repeated linking through action with a given desired outcome. Furthermore
– even in the absence of currently incoming sensory information – “objects”
and their affordances experienced in the past can be reinterpreted by direct-
ing attention to the standing mental representations accessible in memory.
Identified objects or situational patterns, previously understood in a particular
context with an associated action, can therefore be re-examined, in abstrac-
tion, for further and previously unconsidered implications for behavior. Such
re-examination of previously familiar things allows for the reconstrual of their
significance and alteration of the motor actions associated with them for fu-
ture encounters. This process of reconstrual can be done rapidly, at least in
principle, but not instantaneously (not under 350 ms). Such complex processes
require conscious attention, and conscious attention takes time.
It therefore appears plausible, at least, that conscious volitional control
might operate indirectly, by modulating the mental environment – by bias-
ing one perceptual interpretation of the current environment over another,
or several others. From such a perspective, the motor habits associated with
any given perceptual interpretation actually constitute part of that interpre-
tation. Once something is seen in a particular way, therefore (in a manner
in accordance with current consciously held goals), the probability that cer-
tain patterns of action will be implemented will be increased, as a matter of
course. The phenomenon of “utilization behaviour” (Lhermitte et al. 1986),
described in the appendix, provides the clearest evidence for the existence of
such automated perceptual-motor response. Utilization behaviour is associ-
ated with prefrontal damage, sustained in adulthood. Individuals who manifest
utilization behaviour react “automatically” with action to the presence of fa-
miliar objects, and cannot inhibit that automatic response. Such individuals
will, for example, grasp a drinking glass and drink from it, in the absence of
expressed or experienced thirst, or will turn through an open door, without
conscious desire to take that path. Under normal conditions, the prefrontal
cortex, whose inhibitory capacity is well-documented (Fuster 1993) apparently
determines which perceptual-motor programs are relevant and appropriate to
current voluntary behaviour and specifically disinhibits them, allowing them
to pass through a narrow inhibitory gate. Once that decision has been made,
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.9 (438-493)

A close-enough theory of free will 

all other perceptual possibilities and competing affordances are laterally inhib-
ited in to avoid conflicts or perception or action. When the prefrontal cortex
sustains a particular sort of damage, the automated actions objects “afford”
grip behaviour involuntarily, either because inhibition disappears, or because
voluntary goal-direction is no longer extant.
Under normal circumstances of subjective thirst, a glass of water elicits
a liquid-container perceptual scheme, and a drinking motor-program. Un-
der abnormal conditions, however – such as those obtaining during a sud-
den and unexpected kitchen fire – the affordance-interpretation of a glass of
water might shift to “fire extinguisher,” and activate (disinhibit, more accu-
rately) a very different motor program. Even under more prosaic conditions,
the perceptual-motor scheme activated by the pattern that constitutes the glass
could manifest itself in a number of different manners. If an uncooperative
staple needed to be flattened, or a stack of windblown papers subdued, the
perceptual scheme could easily be “manipulable solid heavy surface, near at
hand” and the affordance-interpretation “hammer” or “paperweight”, respec-
tively. In this manner, the attentional environment, under the modulation of a
goal-directed consciousness, is the determinant of how objects are perceived,
and which motor programs are likely to be initiated. The conscious subject
needs only to hold his or her goal in consciousness to align automatic percep-
tions and actions in accordance with that goal.2 Is this free will? If “will” means
the ability to elicit certain perceptions and their attendant action patterns as a
consequence of directed attention, then the answer is “yes.” If, however, “will”
means complete, deliberate, direct control of those perceptions and actions, at
the highest-resolution, most detailed level of analysis, then the answer is “no.”
In our opinion, the low-resolution capability seems close enough to free will to
qualify. As conscious beings, we do not force our own hands. Instead, we whis-
per suggestions, by modulating our intentions and perceptions. This capacity
to suggest and direct appears as one of the primary functions of consciousness,
allowing us to interact with the “same” world and “same” situation in multi-
ple ways, given the plethora of perceptual and action programs afforded by the
flexibility of attention and the complexity of objects.
One problem remains: can consciousness affect real-time activity in any
meaningful manner, given the time-lag necessary for complex conscious pro-
cessing? Shouldn’t the attentional environment necessarily be delayed, in com-
parison to the real world? Shouldn’t this delay doom the organism towards
anachronistic reactions, when consciousness operates? We present a tentative
solution to this problem in the next section.
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.10 (493-532)

 Azim F. Shariff and Jordan B. Peterson

Conscious attention, detached from and unable to control real-time

responses, focuses on events that will occur within a time frame to which
it can react

In order for us to consciously modulate our responses, we need to be able to

process the sensory field, and implement the appropriate actions. However, it
takes a relatively long time to elicit conscious sensation – and even more time
to consciously plan any relevant response. Doesn’t this necessarily mean that
the causal role consciousness plays must be characterized by severe limitations?
Not if attention is typically directed to the future, rather than the present.
When an individual is involved in a fast-paced sport, such as skiing, is it
not probable that he directs attention not to the space immediately in front of
his boots (since it is too late to consciously affect any outcome relevant to that
space) but to the terrain a few meters ahead? Attention focused on the future,
in this manner, could clearly allow for conscious modulation of perception and
reaction. A skier looks ahead, and modifies his perception and actions accord-
ingly – for the anticipated future. We learn to take into account how long is
required for this time-lagged conscious response. At time-frames smaller than
the temporal gap occupied by our consciousness, our responses are already
fixed, and we are impotent to change them. Such unconscious responses run
automatically. We thus have the capacity for free will, but not over very short
durations. In consequence, we fix our conscious attention where we will be,
not where we are.3
Eye-tracking studies that examine the gaze of piano players on musical
notation have clearly revealed this future-predicated fixation. Skilled pianists
consciously observe the musical notation before them to prime, and then elicit,
previously automatized, complex repertoires of finger movements (“macros,”
as described previously). After a given action repertoire has been initiated,
however, any changes to that particular “macro” cannot be implemented –
at least in accordance with the theory we are presenting. Thus, attending to
information within the temporal span of that macro would be useless. We
would therefore predict that pianists would attend to the bars of music suit-
able for priming the next action macro, instead of the current one. Indeed,
skilled pianists look between two and five fixations (the length of a “look”)
ahead (Gilman & Underwood 2003; Truit, Clifton, Pollastek & Rayner 1997;
Rayner & Pollastek 1997; Goolsby 1994). Furthermore, experienced musi-
cians look farther ahead than inexperienced ones. This is precisely what would
be expected if more experienced players likely develop larger-scale and more
deeply automatized and therefore unconsciously implementable macros. Simi-
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.11 (532-603)

A close-enough theory of free will 

lar behaviours, modified similarly by skill difference, characterize touch-typists

(Inhoff & Wang 1992; Salthouse 1984)
Beyond the recognition that consciousness simply attends to the future,
we propose – in keeping with Libet’s original suggestion, described later –
that our subjective experience construes this future as the present, at least for
short-duration decisions. In other words, the anticipated near future is referred
backwards, and interpreted as the present. Nijhawan (1994) demonstrated, for
example, that experimental subjects can be induced to misperceive a set of
moving dots as further ahead of a set of flashing but stationary dots when the
two sets were actually aligned. Subjects typically perceive the moving dots on a
trajectory line 100 ms ahead of where they actually are. Thus their perception
of the present actually reflects their prediction of the near future. This phe-
nomenon, known as the flash-lag effect, has been demonstrated by Nijhawan
and his colleagues in a variety of permutations of the original experiment (eg,
Khurana, Watanabe & Nijhawan 2003). The phenomenal experience of mov-
ing objects truly appears to be predicated on extrapolation of where the object
will be (given its current trajectory).
This adjustment of temporality makes sense, if perception is considered
as anticipatory preparation for action. To account for the time it takes to be-
come aware of useful information and to modulate an appropriate reaction,
it is necessary to focus attention on the future. By perceiving the immediate
future as the present, we can actually and simply believe that consciousness is
acting in real time on the (misconstrued) present, when it is actually prepar-
ing for the (likely) future (see Figure 1). This is not a processing error, or a
form of self-deception. It is merely the simplest solution to a complex prob-
lem (how to respond optimally to the ever-changing flow of environmental
patterns or events). Such notions are entirely consistent with Libet’s notion of
“backward referral.” Libet posits that the present is interpreted as the past, and
has demonstrated that a skin prick, consciously experienced 500 ms later than
its occurrence, is then referred back in time in conscious experience to when it
actually occurred (so that it appears to the subject that the sensation was con-
sciously experienced just 20 ms after the skin was actually pricked). Thus, in
order for consciousness to seem efficacious, despite its time-lag, it appears that
the entire subjective interpretation of time is shifted ahead of the actual time.
Despite its apparent peculiarity, such time-shifting is neither a process that
is rare nor strange. Klein (1999) and Hameroff (1999) explain, for example,
that mental juggling of temporal states is a frequent occurrence, necessary for
the brain to coordinate sensory information arriving at different times from
different parts of the body. The brain is therefore perfectly capable of reor-
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.12 (603-619)

 Azim F. Shariff and Jordan B. Peterson

Time of Awareness Attended to

stimulus of stimulus time frame
Conscious
time
Backwards
Libet’s
referral of
backwards
anticipated
referral
future

Real time

Past Present Future

Figure 1. The conceptual relation between Libet’s backwards referral and anticipatory
consciousness.

ganizing information to create a smooth and consistent experience. Because

of our capacity to time-shift, it is indeed possible for us to perceive the pre-
dicted future as the present. Because we can predict reasonably accurately, this
shift generally causes us no trouble, and functionally simplifies our interactions
with the world. However, the cost of the gap between prediction and actual-
ity is a certain vulnerability to error. When the actual present does not match
the immediately predicted “future,” we place ourselves at risk. Our accuracy is
clearly compromised during situations that are very rapidly and unpredictably
changing. Since our conscious attention is directed towards a future that hasn’t
happened yet, it can only make predictions to the best of its ability. Ballistic
motor programs are selected with slightly dated information, spuriously pre-
sented as up-to-date. This is why not all reactions are perfectly accurate, and
why that inaccuracy surprises us.

Further implications of this theory for Libet’s notion of the veto

So far in this discussion we have borrowed from, expanded upon and modified
parts of Libet’s original theory of conscious action and free will. We have also
described a number of mechanisms that would allow, hypothetically, for the
existence of constrained free will, and dealt with the experimental results that
would call the existence of such will into question. We have taken particular is-
sue with Libet’s notion that the causal role of consciousness must necessarily be
reduced to that of veto, positing that the direction of attention into the future
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.13 (619-667)

A close-enough theory of free will 

allows consciousness to play a determining role, despite the time necessary for
its operation.
It is also worth noting that the veto power of consciousness, re-interpreted,
may also be broader and more powerful than previously presumed. Ongoing
operations may be halted not so much by simple commands of inhibition, but
by competition: a new perceptual motor schema, initiated by attention directed
consciously to new goals and aims, may over-ride the current schema, rather
than merely bringing it to a halt (Cooper, Shallice & Farringdon 1995). Simple
inhibition is more likely to be characteristic of “unconscious” veto. More reflex-
ive perceptual schemas and motor reflexes, such as those characterizing startle
and freezing, for example, are those that most likely have enough power to
rapidly overcome currently initiated macros. The neurocircuitry that mediates
such responses is certainly characterized by the underlying structure that would
enable such displacement (Swanson 2000). However, even more conscious
operations might still produce similar consequences, even if they are imple-
mented unconsciously. To better understand how this could work, you might
raise your fist and slam it down on a table top. Do it a second time – but this
time stop your fist an inch above the table, just before it strikes the surface. You
can do this, but your conscious inhibition of contact with the table does not oc-
cur during the downward action. It can’t. There simply isn’t time. Instead, the
“inhibition” comes prior to the action: the almost-strike-the-table-schema is
biased, through direction of attention, over the strike-the-table-schema. This
is not a “veto” in the manner that Libet proposed, since consciousness must
offer its input before, rather than after, the initiation of such a rapid action.

Conclusion

The concept of anticipatory consciousness – the idea that our conscious per-
ception of the world is shifted forward to take into account the processing time-
lag – constitutes a reasonably plausible explanation for how a time-intensive
consciousness could still remain effective in a world that often progresses faster
than the speed of thought. The theory has another non-trivial advantage, too:
it allows for the unification in conception of two aspects of consciousness
already united in phenomenology – awareness and volition. From our perspec-
tive, goal-directed, voluntary direction of attention (awareness) sets the stage
for consciously biased selection of ballistic and automatized action patterns.
Awareness is thus necessary for free will – and, although that will is limited
(because previously automatized actions are not under direct, high-resolution,
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.14 (667-721)

 Azim F. Shariff and Jordan B. Peterson

conscious control) it is not trivial. We can still determine what we see, albeit in-
directly, and we can still choose the path we will take in the future, although not
in the quarter-second just prior to the choice. It is thus clear that the presump-
tion that consciousness has necessary temporal limitations does not necessarily
invalidate any claims to its causal efficacy.

Appendix: The clinical neuroanatomy of volitional and automatic action

The separation of the willing, initiation and execution of motor action has been
well established, mostly as a consequence of the study of clinical disorders, but
also from directed animal and human neuroimaging studies. Researchers have
attempted to compile this evidence into an anatomical model of the processes
involved in such actions. We might well begin by examining some disorders
of the basal ganglia, whose analysis sheds particular light on the differentiated
phases of motor action.
Both Huntington’s and Parkinson’s disease involve damage to different
parts of the basal ganglia, implicated as physiological key to the execution, but
not the conscious selection, of action schemas (Spence & Frith 1999). Hunt-
ington’s disease, caused by atrophy of the striatum, impairs the acquisition of
new motor skills, and results in unwanted movements (Hiendel, Butters, &
Salmon 1988). Hikosaka and colleagues (Hikosaka, Miyashita, Miyachi, Sakai,
& Lu 1998) used a GABA-agonist to create reversible lesions in parts of monkey
brains in order to observe the differential effects on the acquisition and perfor-
mance of visuomotor sequences. They found that the pre-supplemental motor
area and the caudate in the basal ganglia are likely involved in learning new
sequences, while the putamen and the cerebellar dentate nucleus are involved
either in the storage or retrieval of these sequences. The differential roles of
the caudate and putamen help explain why Huntington’s disease, a disorder of
the entire striatum, affects both motor acquisition and performance. Parkin-
son’s disease, characterized primarily by loss of dopaminergic neurons in the
substantia nigra pars reticulata, appears to have a more specific consequence.
Parkinson’s sufferers often find themselves unable to perform the actions pat-
terns they intend to, although the intention itself remains intact. As Spence
and Frith (1999) indicate, a Parkinson’s sufferer often freezes up altogether,
knowing “precisely what action he wants to perform, but unable to initiate it”
(p. 20).
Human functional neuroimaging in the Hikosaka study confirmed the role
of the pre-supplemental motor area in learning, and also added evidence for
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.15 (721-772)

A close-enough theory of free will 

the involvement of the dorsolateral prefrontal cortex in the acquisition of novel

motor sequences. As sequences became more learned and routinized, however,
activation moves posterior, to parietal regions. Hikosaka et al. concluded that
the prefrontal cortex, the pre-supplemental motor area, and the anterior re-
gions of the basal ganglia initiate the learning of sequences which, once learned
and automatized, involve the posterior basal ganglia, parts of the cerebellum
and parietal lobe. Matsumoto and colleagues (Matsumoto, Hanakawa, Maki,
Graybiel, & Kimurai 1999) followed up on this research, expanding the role of
the striatum to include both the acquisition and retrieval of learned motor pro-
grams. They suggested that the nigrostriatal dopamine system played a central
role in the entire process of learning, storing and retrieving procedural mem-
ory. The dopamine system is certainly capable of strengthening connections
reinforced by positive outcome.
Frank, Loughery and O’Reilly (2001) built on this cursory model of action-
selection with reference to the “gating” nature of the basal ganglia. The projec-
tions within the basal ganglia, from the striatum to the globus pallidus internal
segment (GPi) or substantia nigra pars reticulata (SNr) are inhibitory, as are
those from the GPi/SNr to the thalamus. In the latter case, it is the constant fir-
ing of the GPi/SNr neurons that serves to inhibit the thalamic neurons. Thus,
the firing of the striatal neurons inhibits the constantly firing GPi/SNr neurons.
This, in turn, disinhibits the thalamus (Chevalier & Deniau 1990). The firing
of the striatum has therefore been referred to as something that “releases the
brakes” for motor actions. Plans are instantiated in cortical regions projecting
to the striatum which, in conjunction with the rest of the basal ganglia, “re-
leases” the motor plan associated with those plans. The thalamus projects to
the motor cortex, and the disinhibited plan is run.
Spence and Frith (1999) break down the action-selection system into three
parts. The dorsolateral prefrontal cortex, in conjunction with the anterior cin-
gulate cortex, is involved in the selection and creation of action patterns. Both
regions were shown to be differentially activated when subjects were asked
to pay attention to their actions (Passingham 1997). Ingvar and Philipson
(1977) have noted similar activation when subjects were told to simply imag-
ine making movements (incidentally, Shallice (1988) localizes his Supervisory
Attentional System squarely in the frontal lobes). Spence and Frith contrast the
prefrontal system with the subcortical system comprising the basal ganglia and
the cerebellum (which are more involved in the execution than the selection
of action). The parietal cortex, in turn, appears responsible for the storage of
previously practiced and chunked motor programs.
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.16 (772-815)

 Azim F. Shariff and Jordan B. Peterson

These suggestions are consistent with those the idea that there is a distinc-
tion between action intention and action execution. Those with Parkinson’s
disease suffer an inability to execute, but can still intend. The opposite pattern
should logically characterize those with prefrontal damage, but who have intact
basal ganglia. Though the frontal regions are much more complex, this expec-
tation is confirmed by clinical evidence derived from at least two prefrontal
lobe disorders. Graybiel (1998) describes obsessive-compulsive disorder pa-
tients compelled to perform action sequences without their explicit intention.
Similarly, patients suffering from Utilization Behaviour (Lhermitte, Pillon,
& Serdaru 1986) manifest “stimulus-bound” over-reliance on environmental
stimuli for action. Upon exposure to the relevant cues, these patients can-
not help performing the associated motor plan, even when that performance
is contextually inappropriate. Lhermitte described this process as something
stemming from a loss of intellectual control, due to impaired frontal lobe inhi-
bition, resulting in unrestricted release of parietally-mediated motor activities.
Norman and Shallice (1986) suggest, similarly, that this disorder involves dys-
function in the Supervisory Attentional System, and review relevant evidence
in a more exhaustive fasion (see also Shallice 1988).

Notes

. Many terms have been used to describe the collection of motor responses, including be-
havioral macros (Graybiel 1998), scripts (Schank & Abelson 1977), memory organization
packets (MOPs) (Shank 1982) and schemas (Norman & Shallice 1986). These terms can be
considered largely synonymous.
. A series of experiments by John Bargh and his colleagues (Bargh, Gollwitzer, Lee-Chai,
Barndollar & Trotschel 2001; Chartrand & Bargh 1996) have revealed that even the goal
behind a pattern of activity may remain outside of conscious awareness. Subjects were un-
consciously primed with certain words that affected their goals and consequent strategies
in an ambiguous situation. These results raise an important note: Consciousness is a suffi-
cient but not necessary condition for the occurrence of the attentional biasing. One could
just as easily allow attention to be drawn to things automatically, as a result of the current
motivation. However, volitional power comes from the ability to purposefully direct this
attention, from, for example typing on a computer screen to the paper on my left to the
paper’s affordance to be picked up and read.
. This conscious biasing can be done more distally and broadly in order to prime certain
fast reactions later. This might help explain why mental visualization improves performance
in skilled activities. By running a mental “simulation” of a certain activity, attention is
primed to be directed at the perceptual-motor macros, hastening their unconscious elici-
tation and improving accuracy.
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.17 (815-912)

A close-enough theory of free will 

References

Baldo, M. V. & Klein, S. A. (1995). Extrapolation or attention shift? Nature, 378, 565–6.
Bargh, J. A., Gollwitzer, P. M., Lee-Chai, A., Barndollar, K. & Trotschel, R. (2001). The
automated will: Nonconscious activation and pursuit of behavioral goals. Journal of
Personality and Social Psychology, 81, 1014–1027.
Chartrand, T. L., & Bargh, J. A. (1996). Automatic activation of impression infomation
and memorization goals: Nonconscious goal priming reproduces effects of explicit task
instructions. Journal of Personality and Social Psychology, 71, 464–478.
Cherry, E. C. (1953). Some experiments on the recognition of speech, with one and with
two ears. Journal of the Acoustical Society of America, 25, 975–979.
Chevalier, G. & Deniau, J. M. (1990). Disinhibition as a basic process in the expression of
striatal functions. Trends in Neurosciences, 13, 277–280.
Cooper, R. P., Shallice, T., & Farringdon, J. (1995). Symbolic and continuous processes in the
automatic selection of actions. In Hallam, J. (Ed.), Hybrid problems, hybrid solutions:
Frontiers in artificial intelligence and applications (pp. 27–37). Amsterdam: IOS Press.
Fodor, J. & Pylyshyn, Z. (1981). How direct is visual perception?: Some reflections on
Gibson’s “Ecological Approach”. Cognition, 9, 139–196.
Frank, M. J., Loughry, B., & O’Reilly, R. (2001). Interactions between frontal cortex and basal
ganglia in working memory: A computational model. Cognitive, Affective & Behavioral
Neuroscience, 1, 137–160.
Gibson, J. J. (1979). Ecological approach to visual perception. Boston: Houghton Mifflin.
Gilmann, E., & Underwood, G. (2003). Restricting the field of view to investigate perceptual
spans of pianists. Visual Cognition, 10, 201–232.
Goolsby, T. (1994). Profiles of processing – eye-movements during sight-reading. Music
Perception, 12, 97–123.
Graybiel, A. M. (1998). The basal ganglia and chunking of action repertoires. Neurobiology
of learning and memory, 70, 119–136.
Fuster, J. M. (1980). The prefrontal cortex: Anatomy, physiology, and neuropsychology of the
frontal lobe. New York: Raven Press.
Hameroff, S. R. (1999). The timing of conscious experience – introduction. In Hameroff, S.
R., Kaszniak, A. W., & Chalmers, D. J. (Eds.), Toward a Science of Consciousness III: The
Third Tucson Discussion and Debates. Boston: MIT press.
Heindel, W. C., Butters, N., & Salmon, D. P. (1988). Impaired learning of a motor skill in
patients with Huntington’s disease. Behavioral Neuroscience, 102 (1), 141–7.
Hikosaka, O., Miyashtia, K., Miyachi, S., Sakai, K., & Lu, X. (1998). Differential roles of
the frontal cortex, basal ganglia, and cerebellum in visuomotor sequence learning.
Neurobiology of Learning and Memory, 70, 137–149.
Ingvar, D. H. & Philipson, L. (1977). Distribution of cerebral blood flow in the dominant
hemisphere during motor ideation and motor performance. Annals of Neurology, 2,
230–237.
Inhoff, A. W., & Wang, J. (1992). Encoding of text, manual movement planning and eye-
hand coordination during copy typing. Journal of Experimental Psychology: Human
Perception and Performance, 18, 437–448.
James, W. (1890). The principles of psychology. New York: Henry Holt.
 JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.18 (912-1006)

 Azim F. Shariff and Jordan B. Peterson

Jeannerod, M. (1997). The cognitive neuroscience of action. Oxford: Blackwell.

Kastner, S., De Weerd, P., Desimone, R., & Ungerleider, L. G. (1998). Mechanisms of directed
attention in the human extrastriate cortex as revealed by functional MRI. Science, 282,
108–111.
Keller, I., & Heckhausen, H. (1990). Readiness potentials preceding spontaneous motor acts:
Voluntary vs. involuntary control. Electroencephalography and Clinical Neurophysiology,
76, 351–361.
Klein, S. A. (1999). Do apparent temporal anomalies require nonclassical explanation?
In Hameroff, S. R., Kaszniak, A. W., & Chalmers, D. J. (Eds.), Toward a Science of
Consciousness III: The Third Tucson Discussion and Debates. Boston: MIT press.
Khurana, B., Watanabe, K., & Nijhawan, R. (2003). Flash lag effect: Speeding up to get
ahead? Journal of Vision, 3 (9), 394a.
Lhermitte, F., Pillon, B., & Serdaru, M. (1986). Human autonomy and the frontal lobes.
part 1: Imitation and Utilization behavior: A neuropsychological study of 75 patients.
Annals of Neurology, 19, 326–334.
Libet, B. (1985). Unconscious cerebral initiative and the role of conscious will in voluntary
action. Behavior and Brain Sciences, 8, 529–566.
Libet, B. (1999). Do we have free will? Journal of Consciousness Studies, 6 (8–9), 47–57.
Libet, B., Alberts, W. W., Wright, Jr. E. W. & Feinstein, B. (1967). Responses of human
somatosensory cortex to stimuli below threshold for conscious sensation. Science, 158,
1597–1600.
Libet, B., Gleason, C. A., Wright, E. W. & Pearl, D. K. (1983). Time of unconscious intention
to act in relation to onset of cerebral activity (readiness potential): The unconscious
initiation of a freely voluntary act. Brain, 106, 623–642.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our
capacity for processing information. Psychological Review, 63, 81–97.
Nijhawan, R. (1994). Motion extrapolation in catching. Nature, 370, 256–7.
Nofzinger, E. A., Mintun, M. A., Wiseman, M. B., Kupfer, D. J. & Moore, R. Y. (1997).
Forebrain activation in REM sleep: An FDG PET study. Brain Research, 770, 192–201.
Norman, D. A., & Shallice, T. (1986). Attention to action: Willing and automatic control
of behavior. In R. J. Davidson, G. E. Schwartz, & D. Shapiro (Eds.), Consciousness and
self-regulation. Vol. 4. New York: Plenum Press.
Norrertranders, T. (1990/1998). The user illusion: Cutting consciousness down to size. New
York: Penguin Books.
Orbach, J., Ehrlich, D., & Haith, H. A. (1963). Reversibility of the Necker cube: I. An
examination of the concept of “satiation of orientation”. Perceptual and Motor Skills,
17, 439–458.
Passingham, R. E. (1997). Functional organization of the motor system. In R. S. J.
Frackowiak, C. Mazziotta, K. J. Friston & D. Frith (Eds.), Human brain function. San
Diego: Academic Press.
Rayner, K., & Pollastek, A. (1997). Eye movements, the hand eye span and the perceptual
span during sight reading of music. Current Directions in Psychological Science, 6, 49–53.
Salthouse, T. A. (1984). Effects of age and skill in typing. Journal of Experimental Psychology:
General, 113, 345–371.
JB[v.20020404] Prn:23/11/2004; 10:21 F: Z13109.tex / p.19 (1006-1049)

A close-enough theory of free will 

Searle, J. R. (2000). Consciousness, free action, and the brain. Journal of Consciousness
Studies, 7 (10), 3–22.
Schank, R. C. (1982). Dynamic memory. Cambridge: Cambridge University Press.
Schank, R. C., & Abelson, R. (1977). Scripts, plans, goals, and understanding. Hillsdale, NJ:
Erlbaum.
Spence, S. A. & Frith, C. D. (1999). Towards a functional anatomy of volition. Journal of
Consciousness Studies, 6, 11–29.
Swanson, L. W. (2000). Cererbral hemisphere regulation of motivated behavior. Brain
Research, 886, 113–164.
Taylor, J. L. & McCloskey, D. I. (1990). Triggering of preprogrammed movements as
reactions to masked stimuli. Journal of Neurophysiology, 63, 439–446.
Taylor, J. L. & McCloskey, D. I. (1996). Selection of motor responses on the basis of
unperceived stimuli. Experimental Brain Research, 110, 62–66.
Trevathan, W. R. (1987). Human birth: An evolutionary perspective. New York: Aldine De
Gruyter.
Truitt, F. E., Clifton, C., Pollastek, A., & Rayner, K. (1997). The perceptual span and eye-hand
span in sight-reading music. Visual Cognition, 4, 143–161.
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.1 (41-102)

Freud’s phenomenology of the emotions

Thomas Natsoulas
Department of Psychology, University of California, Davis

This article embarks on a close explication of Sigmund Freud’s

phenomenology of emotion and an effort to give it a unified expression.
Most of the article consists of detailed study of a succession of specific texts of
Freud’s and shows how these support and only seem, in places, to contradict
what I understand to be Freud’s account of emotional experience from the
first-person perspective. Because of my approach, I have space for just three
points of access to Freud’s phenomenology of the emotions. I focus especially
upon the section of his metapsychological essay “The Unconscious” titled
“Unconscious Emotion” and upon parts of The Ego and the Id. Also, before
embarking on my main task, I comment on the notion of a phenomenology
of unconscious mental occurrences and related matters. Freud (but not
William James) would strongly oppose the idea of such a phenomenology: in
his view, inner awareness of any unconscious mental occurrence is
impossible, whether repressed or merely descriptively unconscious (i.e.,
preconscious).

Keywords: emotion, phenomenology, Sigmund Freud, conscious mental

occurrences, unconscious mental occurrences, inner awareness, repression,
“misconstrued” emotions, affect, cognitive content, William James

Introduction

Sigmund Freud’s phenomenology of emotion is the main topic of the present

article. My sources are very largely Freud’s writings in an English translation,
namely, the twenty-three volumes, of his articles, books, and letters, edited
by James Strachey for the Hogarth Press. Phenomenology, the word, is not
included in the additional, index volume of this Standard Edition, nor did I
come upon the word elsewhere in Freud’s works. Nevertheless, in that sense
which I have in mind (see below), his writings do contain a sophisticated phe-
nomenology of emotion. The main purpose of this article is to start to show
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.2 (102-157)

 Thomas Natsoulas

the latter is the case. Another purpose is to increase the currency of Freud’s
phenomenology of the emotions by rendering it more accessible.

A definition of phenomenology

The sense of phenomenology I have in mind is the one that the prominent
American philosopher and phenomenologist David Woodruff Smith explicates
in The Circle of Acquaintance (1989). Here is his definition, not including a
qualification that he appends and to which I shall come to very soon.
What is phenomenology? Phenomenology may be defined as the study of the
structures of consciousness or experience, including the “ways” things “ap-
pear” or are presented in consciousness (“phenomena,” in one technical sense
of the term). The study of intentional characters, or contents, of experiences
is thus a part of phenomenology. Phenomenology is sometimes taken to be a
movement, featuring the works of Husserl, Heidegger, Merleau-Ponty, Sartre
and others – an historical definition that allows for wide differences in the-
ory and method within the movement. Or sometimes phenomenology is itself
identified with a philosophic method, especially Husserl’s method of epoché
or Heidegger’s hermeneutic method. However, I prefer to take phenomenol-
ogy to be a field of study defined by its subject matter, which is human expe-
rience, or consciousness. The differences between, say, Husserl and Heidegger
I see as deep differences about how even to refer to the subject matter.
(Woodruff Smith 1989: 13–14)

The above definition appears in a book of phenomenological investigations

into perception, consciousness, and empathy. Effectively distinguished therein
are experience, phenomenology, and metaphenomenology. None other than
experience per se is the monograph’s subject matter. To consider the volume,
rather, as a study of the logic of phenomenological description would be error.
Woodruff Smith emphasizes that the book reports an extended piece of phe-
nomenological research, an empirical study of experience from the first-person
perspective. His interest lies here in the accurate description of experience; ex-
plication of the phenomenological structures of basic kinds of experience is his
aim. Woodruff Smith’s method includes, perforce, verbal expression of con-
tents and other features of typical experiences of certain sorts. But he does
not attempt in this place to identify what phenomenology involves logically,
linguistically, and so on.
Below are four brief comments on Woodruff Smith’s definition of phe-
nomenology. The first comment brings William James into the picture. The
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.3 (157-203)

Freud’s phenomenology 

second distinguishes phenomenology from a kind of research on experience

that would not be phenomenological. The third comment states a qualification
which Woodruff Smith himself appends to his definition of phenomenology.
The fourth is an objection of mine to his qualification. Following these four
comments, the rest of this Introduction introduces aspects of Freud’s and
James’s thought that pertain to Woodruff Smith’s qualification of his definition
of phenomenology.
1. In the field of consciousness, giving some attention to James’s (1890/1950)
views often can help one to bring clarification to the matter under consider-
ation. He is an excellent example of a leading psychologist who engaged in
phenomenology in a self-conscious and sophisticated way. His famous account
of the stream of consciousness from the subject’s perspective preceded the phe-
nomenological movement to which Woodruff Smith adverts. Indeed, James’s
subjectively based characterizations of the states of consciousness were a major,
acknowledged influence upon that movement.
2. Woodruff Smith’s definition may be taken to imply the following proposi-
tion: “Anyone who is engaged in empirical research on experience, its nature
or character, is ipso facto engaged in phenomenology.” However, this propo-
sition is neither implied nor factually the case. Thus, a scientist engaged in
brain research who fails to examine the experiences of his or her subjects is not
involved therein in any kind of phenomenological research, even if (a) every
experience that transpires is a brain process, and (b) the specific brain pro-
cesses under study are in fact identical to experiences. A completely objective
phenomenology is impossible because it would perforce omit how the expe-
riences are subjectively. As Woodruff Smith rightly asserts, phenomenology’s
subject matter (i.e., the structures of experience) encompasses the way that
things are presented or appear to consciousness. And this “way” is not just one
part of what gets studied, implying it might be ignored in studying another
part of the total subject matter. Rather, an essential concern of every piece of
phenomenological research is how things are presented to the consciousness of
one or more subjects.
3. A certain aspect of Woodruff Smith’s understanding of phenomenology as
field of study will surely be judged as incompatible with the first-person re-
quirement I am emphasizing. It will soon be seen that Freud, for one, would
so judge; for Freud insisted that the unconscious mental occurrences are never
objects of inner awareness. The questionable feature of Woodruff Smith’s con-
ception of phenomenology comes along with his justified conviction that un-
conscious mental occurrences do take place and are no less mental than the
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.4 (203-254)

 Thomas Natsoulas

conscious mental occurrences to which he refers in his definition. In further

claiming that the unconscious mental occurrences, too, are a part of the subject
matter of phenomenology, Woodruff Smith carries well beyond his definition:
What about unconscious mental activities, which are so central in psycho-
analytic theory, cognitive psychology and neuropsychology? Basically, phe-
nomenology is intentional psychology, dealing with meaningful structures
of mental activity, so the term “phenomenology” would be well extended
to cover unconscious as well as conscious mental activities – “surface” phe-
nomenology dealing with the conscious and “depth” phenomenology with the
unconscious. (Woodruff Smith 1989: 14)

4. My objection to the above is this: Even though phenomenology is cer-

tainly as Woodruff Smith asserts – “intentional psychology, dealing with mean-
ingful structures of mental activity” – it is not the case, as well, that all inten-
tional psychology is phenomenological. Many cognitive psychologists assume
that a mental occurrence has content whether or not it is an object of inner
awareness; and an unconscious occurrence’s having content is not owed to its
appearing before a second consciousness, as James is readily understood to
hold. Moreover, some research that is currently proceeding in cognitive psy-
chology is not occupied with “the ‘ways’ things ‘appear’ or are presented in
consciousness”; some cognitive psychologists seek to explain behavior through
reference to cognitive processes that are going on externally to the stream of
consciousness.

Freud’s unconscious mental occurrences

Freud holds that all the unconscious mental occurrences are intrinsically so
that none of their instances can ever be conscious (Natsoulas 2001c, and ref-
erences to Freud given there). This is a general statement that has reference
to both the repressed, or dynamically unconscious, and the preconscious, or
merely descriptively unconscious. As will be seen, these two subcategories of
Freud’s unconscious mental occurrences do not differ with respect to instanti-
ating in themselves the property of consciousness. Neither of them ever does.
In order for an instance of any one of the unconscious mental occurrences to be
conscious, the mental occurrence would have to be, per impossibile, a different
kind of mental occurrence, an intrinsically conscious mental occurrence.
Contrary to how it may at first appear, this denial of consciousness to both
the preconscious and the repressed mental occurrences is wholly consistent
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.5 (254-304)

Freud’s phenomenology 

with the familiar Freudian thesis that an unconscious occurrence may “become
conscious.” The latter phrase has a special, technical sense. Because, according
to Freud, no instance of an unconscious mental occurrence can itself literally
come to consciousness, I insert a hyphen whenever I use becomes conscious in
Freud’s special sense. I next explain that special sense.
Consider the preconscious mental occurrences. Let me state, first, a con-
ception of them that is not consistent with Freud’s:
Thus, it is frequently thought that, when instances of a preconscious men-
tal occurrence take place, they may be objects of inner awareness. That is,
frequently, there exist conditions in which the instances of a preconscious
mental occurrence are conscious in the same sense the instances of a conscious
mental occurrence are conscious. According to this non-Freudian conception,
a preconscious mental occurrence transpires unconsciously at some times
and at some times consciously. A conscious mental occurrence, in contrast,
takes place consciously on every occasion of its occurrence. And a repressed
mental occurrence, whenever it takes place, does so unconsciously unless it
becomes transformed, via psychoanalytic therapy or in another way, into a
preconscious mental occurrence.

This is not how Freud conceives of the property of preconsciousness. What

then is it in Freud for a preconscious mental occurrence to become-conscious,
if it is not that one or more instances of it transpire consciously? Freud defines
“the preconscious” to consist of those many unconscious mental occurrences
that readily can become-conscious in a different sense from that of the view
described in the preceding paragraph. Whereas it is true that any preconscious
mental occurrence can become-conscious, its doing so is always in the form of
a certain kind of proximate effect of an instance of that preconscious occur-
rence, an effect that is distinct from the instance itself. On all such occasions,
both transpire, the preconscious instance and its special, distinct effect. This
effect is an instance of a conscious occurrence and matches a specific descrip-
tion that reflects its cause. In Freud’s theory, a preconscious mental occurrence
does not, as frequently is supposed, somehow itself become transformed into
a conscious mental occurrence. Nor does it get altered so that it resembles a
conscious occurrence by occurring consciously, albeit only sometimes.
The distinct, occurrent effect that is a preconscious mental occurrence’s
becoming-conscious for the moment takes place in a different subsystem of
the mental apparatus from there where the preconscious-occurrence instance
transpires that is its cause. The special effect, being conscious, must take place
in the perception–consciousness subsystem. According to Freud’s theory, this
is the only mental-apparatus subsystem in which the instances of any conscious
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.6 (304-359)

 Thomas Natsoulas

mental occurrence transpire. A preconscious occurrence’s becoming-conscious

is the taking place of an instance of it and this instance’s evoking a conscious
counterpart of itself with corresponding content. Some of the implications of
this view are the following.

1. It means, inter alia, that no conditions ever come about that would per-
mit an unconscious mental occurrence of either kind (dynamically or just
descriptively unconscious) to be an object of inner awareness. The latter point
is expressible as follows using some terms from Woodruff Smith’s definition
of phenomenology: According to Freud, in the case of all of the unconscious
mental occurrences, an appearance or presence to consciousness is impossible;
there is no way in which such a mental occurrence, whether it is dynamically
unconscious or preconscious, may so appear. Thus, the sense in which an un-
conscious wish or an unconscious thought instantiates the property of having
a content or meaningful structure does not include that the wish or thought is
itself directly apprehended. Freud’s unconscious processes are not to be under-
stood as states of a secondary consciousness, which James (1890/1950) would
want to propose that they are.

2. Of course, thinking and believing correctly that a certain mental occur-

rence is now taking place in oneself is not equivalent to its appearing to one’s
consciousness. It was not Freud’s view that, if a patient accepts that a certain
unconscious wish is significantly responsible for certain of his or her actions,
then this wish has gotten itself transformed into a conscious occurrence. In
this regard, the most that can happen is this: Owing to the success of the psy-
chotherapeutic process, the patient may come consciously to wish in a way
well resembling in content that unconscious wish which is responsible for the
occurrence of the conscious wish.

3. And surely, it cannot be successfully maintained that a conscious effect of a

preconscious mental occurrence is the latter’s conscious presence, in the same
sense as a conscious occurrence is said to appear to consciousness. Someone
may want to argue along such lines, but this would mean putting a metaphor-
ical sense of “conscious presence” to use, a sense allowing that, under certain
conditions, something (X) is present to consciousness if there is something else
(Y) that has such presence and Y adequately resembles X.
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.7 (359-411)

Freud’s phenomenology 

Relevancies from James

The unconscious mental occurrences are not experiences; none of them are
components of a stream of consciousness – unless, instead, they are conceived
of, along with James (1890/1950) and contrary to Freud’s explicit refusal to
do so, as the basic durational components of a second stream of conscious-
ness proceeding in the same person. As James characterized such “secondary
personal selves,” they are able, inter alia, to “form conscious unities” and to
possess “continuous memories.” The form of a secondary self “tends to per-
sonality, and the later thoughts pertaining to it [i.e., the states of consciousness
that constitute it] remember the earlier ones and adopt them as their own”
(James 1890/1950: 227).
Clearly, James has in mind, as Freud does not, states of consciousness that
are much like the basic durational components of the primary stream. Those
secondary states, being components of a stream of states of consciousness, can
be the objects of inner awareness in the form of states of consciousness fol-
lowing them in their stream. As does not exist between streams, there is some
“insight,” so to speak, among the states of the same stream.
Given James’s theory, an inner awareness can suitably be called an “ap-
pendage” to the state of consciousness that is its object. Appendage was the
word Freud (1895/1966) used to contrast his own understanding of conscious-
ness to the alternative sort of view of consciousness as being something ap-
pended to, associated with, not an intrinsic feature of any conscious occurrence
as it always is according to Freud.
Assume that James’s appendage account of inner awareness is true. A re-
cent paper of mine (Natsoulas 2003) claims it would be consistent with his
account to expect inner awareness of an experience sometimes to occur before
the experience itself does. In the interest of bringing out the main features of
that account, I explain, in two steps, how this surprising implication follows.

1. According to James’s view of the normal case, the total brain process on
its own is what directly produces any inner awareness that occurs, just as it is
the total brain process that brings any other experience into being.1 This in-
cludes, of course, any experience of which one has inner awareness. Although
it can affect the course of the brain process, an experience is not necessarily
related causally to the inner awareness that apprehends it. Insofar as an expe-
rience does serve as such a cause, it is through that experience’s affecting the
brain process.
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.8 (411-455)

 Thomas Natsoulas

2. Thus, since the ongoing total brain process contains all of the “information”
that is needed to produce both experience and inner awareness of the latter,
these two could come into being in either order relative to each other. For that
matter, it is unclear why James insists inner awareness must be in the form of a
distinct state of consciousness; that is, he holds that inner awareness cannot be
intrinsically a feature of the state of consciousness that is its object, although it
can be such a feature of a subsequent state and have the earlier state as its object.
After all, James conceives of every state of consciousness as a unitary awareness
and each such state typically possesses, in his view, several intentional objects.
All of the above also apply to the Jamesian secondary streams that I have
mentioned except that, of course, the presence of two streams in the one in-
dividual requires that the total brain process gets divided up into two large
brain processes that respectively produce two distinct successions of states of
consciousness.

An intrinsic theory of inner awareness

Freud’s posited unconscious mental occurrences resemble conscious mental

occurrences in possessing contents and being meaningful. But, as is the case
for any feature of an unconscious mental process, these features of theirs are
themselves not “knowable” except indirectly, through inference from some-
thing or other else that itself does have a presence to consciousness. Thus, the
effects that an unconscious mental occurrence may have on occurrences in
the perception–consciousness subsystem are not to be construed as having the
unconscious mental occurrence as their object. Indeed, any conscious men-
tal occurrence that is evoked by an unconscious mental occurrence, however
similar the conscious mental occurrence may be to its unconscious cause, the
conscious mental occurrence is not an inner awareness of this cause.
Nor is this conscious mental occurrence an inner awareness of any other
mental occurrence, the sole exception being itself, since it is a conscious mental
occurrence. For, in clear contrast to James, Freud’s view of inner awareness is
of the intrinsic sort (Natsoulas 2001c). This means that Freud would tend to
hold as Woodruff Smith (1989) does:
A conscious mental event must include an occurrent awareness of its own tran-
spiring; the subject must experience the event as it transpires. Furthermore,
this awareness must be an occurrent part of the given mental event itself, for
reasons that will emerge [cf. Natsoulas 2001a, b, and other installments in that
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.9 (455-513)

Freud’s phenomenology 

series]. Thus, it is an inner awareness, a constituent awareness of a part of

the subject’s own stream of consciousness. Evidently, the “immediate knowl-
edge” required for consciousness will be some kind of acquaintance: a direct
cognitive awareness of the given mental event , though not a separate presen-
tation thereof. The awareness will be cognitive in that it itself justifies a claim
to know the mental event is occurring. It will be epistemically direct, involving
no inference. (p. 81)

Already, from this passage’s first sentence, the contrast to James is noticeable.
The sentence says that a conscious mental event of one’s own cannot occur
without one’s experiencing it and experiencing it involves having inner aware-
ness of it. In James’s contrasting view, experiencing a conscious mental event
does not require having inner awareness of it. Whether awareness of it ac-
companies a state of consciousness or not, the state is an experience. Such
an accompaniment does not entail, according to James’s appendage theory,
any difference in its object from how the latter would be if no awareness of
it accompanied it.
That the above is the case with respect to James’s view can readily be seen
if one considers a stream of consciousness that has come to a sudden, complete
stop. This stream ceased its flow; its last component state (socL ) went out of
existence without being succeeded by another state of consciousness, as did oc-
cur to the states that comprised the stream in an uninterrupted period of flow.
Thus, socL took place without its becoming an object of inner awareness. In or-
der for socL to have been such an object, a state of consciousness with socL for
one of its intentional objects would have had to follow socL in its stream. Based
upon James’s (1890/1950) discussion of the specious present, it would seem in-
ner awareness transpires either immediately or very soon after its object comes
and goes. But, as we have seen, an inner awareness of a state of consciousness
cannot be a feature of the state itself that is the object of inner awareness; more
accurately stated, a state cannot be an inner awareness of itself.
For James, then, “inner” awareness is a kind of “outer” awareness, whereas
for Freud, a state of consciousness has itself among its objects in every instance
of its occurrence. And it is not an error to hold that, for James, those mental oc-
currences transpiring within one without being parts of one’s primary stream
would be included in phenomenology’s subject matter, since those, too, are
states belonging to a (secondary) stream of consciousness. They are experiences
and objects of inner awareness.
Freud held otherwise: There is only one stream of consciousness per hu-
man being. It flows, as it were, through the perception–consciousness sub-
system of the mental apparatus in the brain. No preconscious or dynamically
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.10 (513-545)

 Thomas Natsoulas

unconscious mental occurrence is a component of any stream of consciousness.

Unconscious mental occurrences are not experiences in any of their instances
of occurrence. They are not objects of inner awareness. They have, as it were,
no phenomenology. The unconscious mental occurrences are purely objec-
tive. There is no “way” that they are for the subject, for consciousness. They
do not have presence to consciousness, although they resemble the conscious
mental occurrences in having intentionality and meaning. The unconscious
mental occurrences resemble the conscious occurrences in content, but they
differ from the conscious in being non-qualitative. In all of their instances,
the conscious mental occurrences are qualitative; their contents are not merely
meaningful; there is also somehow these mental occurrences “feel.” Our inner
awarenesses apprehend them as having a particular qualitative character, which
James, too, attributed to all of our experiences, to every state of consciousness,
to every basic durational component of any stream of consciousness.

Close examination of relevant texts

In this article, my mode of achieving access to Freud’s phenomenology of the

emotions consists in the close examination of texts. My exposition here is, I
believe, more explicit than frequently encountered. From this point on, each
of my subsections is focused on a relevantly enlightening section from Freud’s
work. The reader will thereby become well informed concerning the bases of
my interpretations, and he or she will be more prepared to evaluate them than
is usual. Thus, I do not limit myself to stating conclusions that I have reached
and appending selected, supporting quotations and paraphrases. For example,
I give explicit attention to sentences that some readers may understand to con-
tradict my account of Freud’s theory. However, some of the usual must go on
as well. I must apply conclusions that I have reached from other portions of
Freud’s works to the passage that I am currently explicating.

Access Point No. 1

One point of access to Freud’s phenomenology of emotion begins on page 21
of The Ego and the Id (1923/1961).2 In the last full paragraph on the page, he
embarks on a discussion of “sensations and feelings.” For now, let me just assert
that much of what Freud tells us concerning the mental occurrences which he
includes under sensation and feeling is pertinent to grasping his conception
of the emotions qua experiences. Although argument will be forthcoming in
subsequent articles, it appears safe to state at this point that, for Freud, the
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.11 (545-599)

Freud’s phenomenology 

emotions are kinds of feeling, whereas, of course, in his view, not every feeling
qualifies as an emotion.
On page 21, Freud asks whether he has been right to hold that, when-
ever a conscious mental occurrence transpires, it does so only in perception–
consciousness. Doubt may develop regarding consciousness’s unilocular na-
ture, as Freud proceeds to consider how the sensations and feelings come to
take place owing to “internal perceptions.” Freud has already stated, in the same
chapter, that the sensations and feelings – which are “received from within”
the mental apparatus – are among the conscious mental occurrences. In their
instantiation of the property of consciousness, they are like the perceptual
experiences – which are “received from without” – and unlike the thoughts
that occur externally to the perception–consciousness subsystem. All the lat-
ter thoughts are unconscious mental occurrences, whereas thoughts that are
among the conscious do exist too. In contrast, the sensations, the feelings,
and the perceptual experiences are not ever unconscious mental occurrences.
If Freud were to accept the thesis that sensations and feelings sometimes do
take place outside perception–consciousness, then his theory of the mental
apparatus would have to be fundamentally revised.
Just before Freud turns to sensations and feelings, he is saying how un-
conscious occurrences such as thoughts become-conscious. Because of this
account, question might arise as to whether every instance of a conscious
mental occurrence occurs in perception–consciousness. According to Freud,
an unconscious thought, if it becomes preconscious, may become-conscious –
that is, it may readily evoke adequate conscious counterparts of itself. While
sensations and feelings are consequences, too, of certain unconscious men-
tal occurrences, they travel, as it were, a different causal path to their evo-
cation within perception–consciousness. Indeed, many mental paths exist by
which the sensations or feelings come to occur; “processes . . . in the most di-
verse and certainly also in the deepest strata of the mental apparatus” (Freud
1923/1961: 21–22) produce them. As is true of the perceptual experiences, sen-
sations and feelings “may come from different places simultaneously and may
thus have different or even opposite qualities” (p. 22). Freud says they are “mul-
tilocular,” but he means how they come to occur, not their transpiring in a
subsystem of the mental apparatus other than perception–consciousness.
Freud again asks: Might the unconscious mental determinants of sensa-
tions and feelings be sensations and feelings themselves? Might they cause
sensations and feelings to arise right there where the determinants themselves
take place? Thus, sensations and feelings could transpire sans their unconscious
determinants “reaching” perception–consciousness. Freud poses his rhetori-
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.12 (599-648)

 Thomas Natsoulas

cal question in terms of that unconscious “something” which becomes con-

scious; he must mean here becomes-conscious, in that special sense identified
in the Introduction of the present article, just as he does when he speaks of
the becoming conscious of unconscious thoughts. In both cases, his posited
“somethings” cannot literally become conscious. If they cannot evoke in the
perception–consciousness subsystem a conscious mental occurrence that cor-
responds to them, they may yet, and at most, get to be preconscious through
psychoanalysis or the like.
Freud’s clinical experience leads him to avow only conscious occurrences
of sensations and feelings. He claims, on the basis of this work, that the un-
conscious determinant of a sensation or of a feeling “behaves like a repressed
impulse [or] compulsion” (p. 22), or like tensions owing to physical need that
remain unconscious. That is, the unconscious “something,” whereof one has
no inner awareness, may “exert driving force.” And can therefore contribute
to the explanation of behavior (action and bodily reaction) as well as certain
sensations or feelings in the perception–consciousness subsystem.
Accordingly, it is by “reaching” the latter subsystem that the unconscious
determinants of a feeling become-conscious. Reaching there and the like (e.g.,
their being blocked from reaching there) are, of course, figurative expressions
for their causing (or their failing to cause) conscious counterparts of them-
selves in the mental-apparatus locus where all conscious mental occurrences
transpire. Much the same is proposed for the relation of conscious thoughts
(and wishes) to their proximate unconscious determinants.
In the course of this discussion, Freud mentions “pain” as an example of
“something” which can remain unconscious and behave like a repressed com-
pulsion. However, from the context, it is evident that he means not the pain
experience but the unconscious “something” that normally produces that ex-
perience; it does not produce the experience if it remains repressed. Those who
attribute to Freud the thesis that emotions can occur unconsciously, as well as
consciously, fail to take fully into account in their interpretations important
statements of Freud’s such as this one:
If the way forward [to consciousness] is barred, they do not come into being as
sensations, although the “something” that corresponds to them in the course
of excitation is the same as if they did. We then come to speak, in a condensed
and not entirely correct manner, of “unconscious feelings,” keeping up the
analogy with unconscious ideas which is not altogether justifiable. (p. 22)

But, analogously to Freud’s account of unconscious thoughts, a feeling’s prox-

imate unconscious mental cause, if not repressed, may justifiably be thought
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.13 (648-699)

Freud’s phenomenology 

of as “preconscious” if one means that this mental cause readily becomes-

conscious, readily evokes the corresponding feeling. When its way forward
is barred, it is prevented from performing its normal function as otherwise
it would. If no longer repressed, the cause has a certain specific feeling kind
of effect.
A repressed mental cause’s way may not be completely barred, according to
Freud, and thus it excites a different feeling effect in perception–consciousness.
An unconscious compulsion that is obstructed in its normal effects will, in-
stead, “become conscious as unpleasure” (p. 22). This is a kind of feeling and,
therefore, a conscious mental occurrence. But note: the repressed mental oc-
currence cannot, based on the unpleasure it produces in consciousness, be
rightly described as becoming-conscious, in the sense that something precon-
scious becomes-conscious.
Our experiences of unpleasure frequently are, according to the theory, con-
scious effects of repressed mental occurrences. Indeed, unconscious mental
occurrences have many effects on the perception–consciousness subsystem.
Not all of these, however, are of the kind that qualify their respective un-
conscious causes as becoming-conscious. Not all their conscious effects are
cases of what Freud spoke of as an unconscious occurrence’s “reaching” the
perception–consciousness subsystem. Indeed, as it produces the unpleasure of
which Freud speaks, the unconscious mental occurrence remains repressed. So
long as it remains repressed, it does not, as Freud expresses it, “come into being
as feeling.” For an unconscious mental occurrence to “reach” consciousness is
not for it to arrive their literally, nor is it for it simply to make a difference in
what is going on in the perception–consciousness subsystem.
From Freud’s perspective, what is it that is unjustifiable and not cor-
rect regarding the notion of unconscious feelings? Answer: the referents of
the notion are not actually themselves feelings. Feelings transpire only in
perception–consciousness and, as are all of the mental occurrences that tran-
spire there, they are conscious in every instance of their occurrence. As we
have seen, while “unconscious feelings” are not feelings, they do evoke real
feelings. First, under repression, they evoke unpleasure. And, second, if they
are not dynamically prevented from doing so, they evoke feelings specifically
corresponding to them.
Even when their way to consciousness is not barred, it is not possible for
such mental causes as the “unconscious feelings” to become converted into
conscious mental occurrences. Then they merely become-conscious; that is,
they unconsciously cause to transpire a corresponding feeling in perception–
consciousness. This feeling “corresponds” to its unconscious mental cause in
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.14 (699-748)

 Thomas Natsoulas

the sense of its being what is naturally to be expected when its causal relation-
ship to the perception–consciousness subsystem has not been distorted by the
intervention of other unconscious mental causes. Unconscious thoughts and
wishes do take place, but feelings and the like are intrinsically such as to re-
quire what only perception–consciousness can provide among the subsystems
of the mental apparatus.
By way of stating how the unconscious thoughts and the “unconscious feel-
ings” differ from each other, the analogy between them being a significantly
limited one, the above indented quote from Freud (1923/1961: 22) continues
as follows:
Whereas with Ucs. ideas [i.e., unconscious thoughts] connecting links must
be created before they can be brought into the Cs. [i.e., become-conscious],
with feelings, which are themselves transmitted directly, this does not occur.
In other words: the distinction between Cs. and Pcs. has no meaning where
feelings are concerned; the Pcs. here drops out – and feelings are either con-
scious or unconscious. Even when they are attached to word-presentations,
their becoming conscious is not due to that circumstance, but they become so
directly. (pp. 22–23)

As seen from the just preceding portion of this passage, it would be error to un-
derstand Freud to be speaking here literally of unconscious feelings. When he
states “feelings are either conscious or unconscious,” he must mean that either
they occur or, because of repression, they do not occur notwithstanding that
their normal unconscious cause is now operative. There is no contradiction of
his thesis that unconscious feelings do not exist. Rather, he is talking loosely
here, in the very “condensed and not entirely correct manner” that he has just
stated is “not altogether justifiable” (p. 22).
The sense in which the “unconscious feelings” become conscious “directly,”
as Freud states they do, is by unmediatedly causing the corresponding feeling
within perception–consciousness. Comparing all repressed feelings with all re-
pressed thoughts, one can say the kind of mediation the repressed thoughts
require to be capable of becoming-conscious, namely, transformation into pre-
conscious mental occurrences, is not necessary for the unconscious mental
cause of a feeling. Simply owing to the lifting of repression, an “unconscious
feeling” becomes preconscious in the sense of its now readily producing the
corresponding feeling.
In the above indented quotation, the sense of the epithet “preconscious”
includes more than just the described mental occurrence’s causal relation to
perception–consciousness, namely, an unconscious thought’s connection with
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.15 (748-809)

Freud’s phenomenology 

word presentations. Such connection is essential in order for an unconscious

thought to “reach” consciousness, whereas word presentations are not needed
in order for an “unconscious feeling” to “come into being as feeling.”
There are two modes of expression that Freud uses in the passage under
scrutiny that could lead to an understanding of his conception of consciousness
that would be erroneous. He makes reference at the start to certain “internal
perceptions” of certain unconscious mental occurrences as being that which
gives rise in consciousness to the sensations and feelings. And at the end of
the passage, he speaks of the “unconscious feelings” as “becoming conscious
directly.” These statements may tempt a reader to take an alternative view to
mine. According to the alternative view, unconscious mental occurrences can
be for Freud objects of a kind of direct apprehension.
However, the first page of Freud’s chapter contains a rejection of this sort
of conception of the unconscious mental occurrences. He states that an ac-
curate understanding of the unconscious mental occurrences’ relation to con-
sciousness calls for a “third alternative” to their advancing to consciousness or
consciousnesss’s making its way to them. Both are “equally unimaginable” for
“those internal processes which we may – roughly and inexactly – sum up un-
der the name of thought-processes” (p. 19). Notwithstanding Freud’s applica-
tion of the latter name, he means his reference to be to all unconscious mental
occurrences. Both the “unconscious feelings” and the unconscious thoughts
“represent displacements of mental energy which are effected somewhere in
the interior of the [mental] apparatus as this energy proceeds on its way toward
action” (p. 19). An “unconscious feeling” is no more capable than an uncon-
scious thought of “advanc[ing] to the surface, which causes consciousness to
be generated”; nor can it ever become an object of direct apprehension.
Whether repressed or not, the “unconscious feelings” are what they are, not
something else such as an actual feeling. Those unconscious mental “some-
things” that correspond to feelings and bring them about in the absence of
repression are “the same” whether they are repressed or not (p. 22). Part of
their being the same is being unconscious mental occurrences whether or not
they succeed in causing a corresponding feeling to occur. Freud concludes his
comments about the becoming-conscious of unconscious thoughts as follows:
“Consciousness remains where it is, therefore; but, on the other hand, the Ucs.
does not rise into the Cs.” (p. 20). Then, immediately, he moves on to the sen-
sations and the feelings with respect to, as seen in this subsection, whether
their unconscious mental causes might themselves be sensations and feelings
although they occur in a different subsystem of the mental apparatus.
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.16 (809-857)

 Thomas Natsoulas

Access Point No. 2

In the final paragraph of his chapter, Freud (pp. 26–27) mentions a “strange”
phenomenon which his work with psychoanalytic patients has forced upon his
attention, and expresses bewilderment, for the reason that the phenomenon
“compels us, in spite of our better critical judgment, to speak of an ‘un-
conscious sense of guilt”’ (p. 27). Even such higher mental functions as self-
criticism and conscience can go on entirely outside perception–consciousness.
But some discussants (e.g., Schur 1969) interpret Freud’s reaction to be a con-
sequence, rather, of the evident necessity that he include in his account of the
mental apparatus unconscious guilt feelings, which is contrary to his thesis that
all feelings are conscious mental occurrences.
In a later chapter, Freud comes back to the concept of an unconscious
sense of guilt. Given the present purpose, a passage there requires close atten-
tion. When Freud states the following, he is attributing the negative therapeutic
reaction of some of his patients largely to their unconscious sense of guilt:
But as far as the patient is concerned this sense of guilt is dumb; it does not tell
him he is guilty; he doe not feel guilty; he feels ill. This sense of guilt expresses
itself only as a resistance to recovery which it is extremely difficult to over-
come. It is also particularly difficult to convince the patient that this motive
lies behind his continuing to be ill. (pp. 49–50)

Finding Freud’s explanation of these patients’ behavior convincing is, of course,

not necessary to a comprehension of his concept. Our interest lies in de-
termining the sort of mental occurrence an unconscious sense of guilt is
proposed to be.
Does it consist of guilt feelings? Evidently not, judging from Freud’s ob-
servation that such a patient does not feel guilty. However, a reader could
understand Freud’s statement to mean: only conscious guilt feelings, not all
guilt feelings, are missing. For the context of Freud’s statement is the patient’s
lack of an awareness. It could be suggested that a patient who sincerely reports
not feeling guilt may still be unconsciously undergoing feelings of guilt. If it
is repressed, a sense of guilt cannot elicit, as it otherwise could, corresponding
feelings in the perception–consciousness subsystem, wherein the patient would
have inner awareness of feeling guilty. Those are the lines along which a reader
may reason.
In what, according to Freud, does an unconscious sense of guilt consist?
Freud states that a patient with such a sense feels ill, not guilty. This statement
refers, surely, to that unpleasure he mentions earlier in the book. It will be re-
called about repressed occurrences whose way forward is not wholly barred
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.17 (857-907)

Freud’s phenomenology 

that, among their other effects upon the perception–consciousness subsytem,

they produce unpleasure. Thus, they are causes of some conscious effects but
not the same such effects as they cause to occur when not repressed, for ex-
ample, those conscious counterparts that qualify them as preconscious. And
where it is an unconscious sense of guilt that is producing the feelings of un-
pleasure, these are not to be confused with their cause, which, as argued, does
not consist of feelings.
At this point, Freud appends a footnote concerning his efforts to deal with
the obstruction of therapeutic progress which arises from an unconscious sense
of guilt. The only method available to him for this purpose, he states, is “the
slow procedure of unmasking its unconscious repressed roots, and thus grad-
ually changing it into a conscious sense of guilt” (p. 50). I must suggest that
Freud is expressing himself in an abbreviated manner that he would consider
incorrect if it were meant literally. As we have seen, no unconscious mental oc-
currence can ever, in his view, rise up to perception–consciousness and occur
there. The gradual change Freud refers to in the passage (pp. 26–27) is the sense
of guilt’s becoming such as to evoke readily a conscious counterpart of itself, as
transpires when an unconscious sense of guilt occurs that is not repressed.
Yet, such statements as the above one from Freud are used (e.g., Green
1977) to argue that Freud’s unconscious sense of guilt does involve unconscious
guilt feelings, feelings of which the person has no inner awareness. This inter-
pretation would mean Freud was contradicting himself in the same short book.
And one has to wonder, if indeed unconscious guilt feelings did seem to Freud,
as Green states, “incontestable,” then why Freud does not say, “These patients
are feeling guilty but have no inner awareness of so feeling,” just as he does fre-
quently assert our inability to apprehend firsthand our unconscious thoughts
and desires. Not only does Freud (1923/1961) say nothing like this, but in a
paper which he published the next year (Freud 1924/1961), he does not mod-
ify his treatment of the unconscious sense of guilt, his treatment of it as being
something that fails to qualify as a feeling or emotion.
Thus, Freud (1924/1961) mentions his prior discussion and describes the
patients who show that negative therapeutic reaction as unable to accept the
idea of their having a sense of guilt sans any awareness of it. About them, he re-
ports, “They know only too well by what torments – the pangs of conscience –
a conscious sense of guilt, a consciousness of guilt, expresses itself ” (p. 166).
Their current condition seems to them not to include guilt feelings. Interest-
ingly, Freud is prepared to meet their objection; they are correct to insist they
do not feel guilty; he is willing to replace the term unconscious sense of guilt with
need for punishment. Freud again states that the former term is not psycholog-
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.18 (907-956)

 Thomas Natsoulas

ically correct; and adds that the latter term will serve just as well for referring
to the mental occurrences of interest. At the same time, he insists this uncon-
scious need must be understood to be much like a conscious sense of guilt. As
we have seen, the unconscious need for punishment differs from a conscious
sense of guilt in, at least, two respects: (a) It is not an instance of feeling. (b)
When it is not repressed, it is a mental cause of guilt feelings.
Schur (1969) argues that unconscious guilt feelings are part of Freud’s the-
ory. To show that this is true, Schur calls the reader’s attention to a paragraph
from Civilization and its discontents (Freud 1930/1961: 134–136). But only if
the paragraph can rightly be construed as drawing relevant differences between
unconscious anxiety and unconscious guilt does it support Schur’s claim. In
fact, the paragraph brings out, instead, certain similarities between these two
kinds of unconscious happenings, in the process of seeking to demonstrate that
a better understanding of the unconscious sense of guilt can be reached by
treating of it as being strongly analogous to that which is loosely describable as
anxiety that is unconscious. Here is part of what Freud states in the paragraph:
Anxiety is always present somewhere or other behind every symptom; but at
one time it takes noisy possession of the whole of consciousness, while at
another time it conceals itself so completely that we are obliged to speak of
unconscious anxiety or, if we want to have a clearer psychological conscience,
since anxiety is in the first instance simply a feeling, of possibilities of anxiety.
Consequently it is very conceivable that the sense of guilt produced by civiliza-
tion is not perceived as such either, and remains to a large extent unconscious,
or appears as a sort of malaise, a dissatisfaction for which people seek other
motivations. (p. 135)

Note two things:

1. The concept of unconscious anxiety is not theoretically acceptable. The so-
called unconscious anxiety does not consist of feelings. Rather, it is a mental
cause of the sort of conscious mental occurrence that we call anxiety with a
clear psychological conscience.
2. The “malaise” Freud mentions would seem to be equivalent to the unplea-
sure he attributes as an effect to unconscious mental causes that are repressed
but not completely. A repressed sense of guilt may cause one to feel dissatis-
fied, even to feel ill, without its thereby revealing anything of what the matter
is. In the same paragraph, Freud also speaks of a “tormenting uneasiness” that
some of his patients with a repressed sense of guilt suffer if they are prevented
from carrying out certain actions. As evidenced by inner awareness, neither of
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.19 (956-1010)

Freud’s phenomenology 

these sets of feelings has a cognitive content which conforms to its respective
unconscious mental cause.

Access Point No. 3

The next text to be examined, for what it tells us about Freud’s phenomenology
of the emotions, is “Unconscious Emotions,” the third section of his essay “The
Unconscious”(1915/1957b). He starts this section by inquiring as to whether,
in addition to unconscious thoughts, there also exist unconscious instinc-
tual impulses, unconscious emotions, and unconscious feelings. Freud deals
with the instincts first, proposing about them that “the antithesis of conscious
and unconscious is not applicable”(p. 177). That is, while instinctual impulses
surely exist, it is meaningless to use the phrase unconscious instinct or uncon-
scious instinctual impulse because this would imply that there are conscious in-
stinctual impulses as well. If a conscious instinctual impulse did occur, it would
be, by virtue of its nature, an object of consciousness (i.e., inner awareness).
But, an instinctual impulse is intrinsically a different kind of occurrence
which always takes place in the mental apparatus externally to the perception–
consciousness subsystem. Therefore, it is not a conscious mental occurrence
and cannot ever be transformed into a conscious mental occurrence. All in-
stinctual impulses and mental occurrences that transpire in the unconscious
are by their intrinsic nature not conscious. Their transformation into conscious
mental occurrences would require fundamental change in what they are.
Without an instinctive impulse’s relations to other sorts of mental oc-
currences, we could not know anything at all about it. (Freud would revise
this view could he witness current advances in neurophysiological observation
and measurement.) Moreover, the instinctual impulses are unable to become-
conscious either. They cannot, as unconscious thoughts can if they are not re-
pressed, evoke in the perception–consciousness subsystem a (conscious) coun-
terpart of themselves. Only the unconscious thought can become-conscious to
which an instinctual impulse is attached.
Of course, it is entirely possible to have conscious thoughts concerning
one or more of one’s instinctual impulses, even conscious thoughts concern-
ing an instinctual impulse that is right now active within one, perforce outside
perception–consciousness. However, having such thoughts is not to be con-
fused with inner awareness of an instinctual impulse. Compare thinking about
one’s instinctual impulses with conscious thoughts concerning processes tran-
spiring now in one’s liver. Surely, the occurrence of such thoughts would not
support the thesis that the liver’s processes are conscious occurrences. X’s being
accompanied by a conscious thought about X is surely not the equivalent of X’s
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.20 (1010-1063)

 Thomas Natsoulas

being a conscious occurrence. So too, no matter how much or in what way one
thinks about any instinctual impulse, it is not transformed into a conscious
mental occurrence. It remains what it is: something that occurs where nothing
conscious ever happens, namely, within the unconscious system of the mental
apparatus.
From the Freudian perspective, it would be mistaken to conceive of the
emotions or feelings, in any instance, as instinctual impulses. As the instinc-
tual impulses are not, the emotions and the feelings are qualitative (affective)
occurrences; and, therefore, they are all necessarily conscious occurrences. In
contrast to thoughts, there does not exist any unconscious version of an emo-
tion or feeling. One way that Freud puts this point is to say that “affects
and emotions correspond to processes of discharge, the final manifestation of
which are feelings”(p. 178). The emotions and affects are outcomes of certain
processes arising elsewhere in the mental apparatus and yielding specific excita-
tions of, mental occurrences within, the perception–consciousness subsystem.
And these excitations are mental occurrences that are consciously experienced
as feelings. Thus, the “correspondence” to which Freud makes reference above
is a relation existing between two quite distinct items: as effects, certain con-
scious occurrences and, as their causes, certain occurrences in the same mental
apparatus that are not conscious, namely, instinctual impulses.
Here is another way in which Freud puts the point that, unlike the case of
thoughts, there are no unconscious versions of the emotions or feelings: “Un-
conscious ideas continue to exist after repression as actual structures in the
system Ucs., whereas all that corresponds to unconscious affects is a potential
beginning which is prevented from developing” (p. 178). When conscious af-
fects are prevented by repression from occurring, no unconscious affects exist.
What does exist at such times in the unconscious is the corresponding instinc-
tual impulse which otherwise would have brought those conscious affects into
existence in the perception–consciousness subsystem.
Instinctual impulses contrast with unconscious thoughts. The latter can
become-conscious; they evoke conscious thoughts with cognitive content that
they themselves possess. In contrast, an instinctual impulse cannot evoke, anal-
ogously, a conscious instinctual impulse; no instinctual impulse exists that is
conscious, according to Freud. With respect to consciousness, instinctual im-
pulses are capable of producing (a) emotions or feelings, which are always
conscious, and (b) unconscious thoughts, which can become-conscious in that
technical sense that I have identified in the present article.
Thus, Freud (1915/1957b) states, “When we. . .speak of a repressed in-
stinctual impulse. . .we can only mean an instinctual impulse the ideational
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.21 (1063-1114)

Freud’s phenomenology 

representative of which is unconscious, for nothing else comes into consid-

eration” (p. 177). The sense of unconscious in the latter statement is the same
as repressed; the unconscious thought associated with a repressed instinctual
impulse is repressed. However, this is not the only matter that is altered by
the repression of an instinctual impulse. On the next page, Freud expresses the
following about
the vicissitudes undergone, in consequence of repression, by the quantitative
factor in the instinctual impulse. We know that three such vicissitudes are
possible: either the affect remains, wholly or in part, as it is; or it is trans-
formed into a qualitatively different quota of affect, above all into anxiety; or
it is suppressed, i.e., it is prevented from developing at all. (p. 178)

The quantitative factor in the instinctual impulse is its energy, which produces
feeling or affect in the perception–consciousness subsystem. Repression pre-
vents entirely, partially, or not at all the conscious expression of this energy
in the form of conscious affect. And it may cause the energy of the instinc-
tual impulse to operate upon the perception–consciousness subsystem in such
a way that there is produced affect qualitatively different from what it would
produce in the absence of this repression. With reference to this instinctual
vicissitude, Freud (1915/1957a: 153) speaks of the “transformation” of instinc-
tual energy into affects, and he suggests, in the section of “The Unconscious”
which we are scrutinizing here, that this transformation frequently depends
upon the “substitutive idea,” that is, the unconscious thought which the re-
pressed instinctual impulse gets itself attached to instead of the unconscious
thought that would be the instinctual impulse’s “representative” absent the re-
pression. Freud (1915/1957b) states, “The nature of that substitute determines
the qualitative character of the affect” (p. 179).
For Freud it is of the essence of the emotions, feelings, and affects, evoked
by the instinctive impulses, that the one in whom they occur has awareness
of them; these mental occurrences must become known to consciousness. In
every instance of emotion, feeling, or affect, “the possibility of the attribute
of unconsciousness” is “completely excluded.” Intrinsically, all of these mental
occurrences are such that, per impossibile, to expunge their property of con-
sciousness would be to expunge the occurrences themselves. Freud continues:
But in psycho-analytic practice we are accustomed to speak of unconscious
love, hate, anger, etc., and find it impossible to avoid even the paradoxical con-
junction, “unconscious consciousness of guilt,” or a paradoxical “unconscious
anxiety.” Is there more meaning in these terms than speaking of “unconscious
instincts”? (p. 177)
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.22 (1114-1168)

 Thomas Natsoulas

Freud’s acknowledgment at this point of “misconstrued” emotions, feelings,

and affects suggests that there might be forthcoming from him an affirmative
answer to his latter question.
However, in Freud’s theory, the so-called “misconstrued,” too, are no less
conscious mental occurrences. The repression involved in their case, which is
responsible for their “misconstrual,” has not prevented their occurring. The
consequence of repression is, rather, a difference between the mental occur-
rence that they are and the mental occurrence they would be without repres-
sion. The present kind of “misconstrued” occurrence can be just as it would be
except for its cognitive content. Its quota of affect is the same qualitatively as
it would be without the repression, and it may be the same in degree or quan-
tity. Freud holds that no more than an emotion’s “idea” need be repressed in
the instances he calls “misconstrued.” But he does not mean a “misconstrued”
emotion, feeling, or affect does not possess a cognitive content; rather, it does
not have its proper such content.
At this point, Freud seems to imply that an emotion or affect must have
a cognitive content. He states, “It may happen that an affective or emotional
impulse is perceived, but misconstrued. Owing to the repression of its proper
representative it has been forced to become connected with another idea, and
is now regarded by consciousness as the manifestation of that idea” (p. 178).
Of course, Freud is speaking of the emotion or affect; this, not the instinctual
impulse, is the object of consciousness; the instinctual impulse is never con-
scious. Freud has in mind, it would seem, that evoking a misconstrued emotion
or affect involves as key causes both an instinctual impulse and an unconscious
(preconscious) thought substituting for a repressed thought.
A very short digression is needed at this point, pertaining to the question
implicitly raised in the paragraph right before this one, although this question
concerning Freud’s phenomenology of the emotions shall not be pursued in
the present article. The question is whether for Freud all of the occurrences of
emotion possess cognitive content.
If emotions are, as Freud holds, necessarily conscious, then every occurrent
instance of any emotion must have a content that has reference to the instance
itself. But, setting this category of cognitive content aside, we need to inquire
into whether an occurrent emotion always possesses an intentional object not
counting itself, or the emotion is, at least, as though it does possess such an
intentional object, the latter in those cases in which its apparent object does
not possess any actual existence. Relevantly to the question, Freud states, “It
is possible for the development of affect to proceed directly from the system
Ucs.; in that case the affect always has the character of anxiety, for which all
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.23 (1168-1223)

Freud’s phenomenology 

‘repressed’ affects are exchanged” (p. 178). He adds that the more frequent case
is (“as a rule”) the one in which a repressed instinctual impulse gets itself con-
nected with a preconscious thought that substitutes for the normal but now
repressed thought. But what if it does not? What is the content like of the anx-
iety which is produced by a repressed instinctual impulse on its own? Such
questions must be pursued in other textual loci than those chosen for scrutiny
in the present article.
Repression can be such that an emotion or affect does not occur, but a mis-
construed emotion is an occurrent emotion no less so than the emotion that
would have transpired consciously were it not for the operation of repression.
It is not the same emotion; it should be called a “substitute” emotion. It pos-
sesses both cognitive (ideational) and qualitative (feeling, affective) content,
just as those emotions do whose occurrences involve no repression. In my opin-
ion, the emotions that Freud calls “misconstrued” should not be so described,
for the adjective suggests erroneous inner awareness. The “error” in this case
precedes consciousness; it does not take place in perception–consciousness; it
takes place unconsciously, specifically, where the instinctual impulse is forced
to connect with an unconscious thought that is not its “proper representa-
tive.” At the same time, in the perception–consciousness subsystem, both the
“misconstrued” emotion’s qualitative and cognitive contents are taken by in-
ner awareness to be as they actually are. Of course, this is not to say that they
are apprehended with reference to their unconscious causes. It is to say, sim-
ply, that they are apprehended as they themselves are, their actual qualitative
and cognitive content. Non-cognizance of unconscious mental causes is not
the equivalent of misconstruing conscious effects of theirs.
I need also to bring out that the present Freudian notion of “misconstrual”
should not lead to the understanding that the very undergoing of any emotion
involves a process of interpretation, a construing of its affective component
perhaps. This is further good reason to view the emotions Freud calls miscon-
strued as being, in my term, substitute emotions. Their becoming “connected
with an idea” is a characterization of their unconscious causes, not some-
thing transpiring within consciousness. Like other emotions, a misconstrued
or substitute emotion emerges in the stream of consciousness with a certain
qualitative and cognitive content.
Rejecting the existence of unconscious affects, contrasting them with the
actual occurrence unconsciously of unconscious thoughts, Freud adds, “But
there may very well be in the system Ucs. affective structures, which, like
others, become conscious” (p. 178). This statement should not be taken to
include the notion of the presence outside of perception–consciousness but
 JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.24 (1223-1272)

 Thomas Natsoulas

in the mental apparatus of something affective and analogous to unconscious

thoughts. Those affective structures to which Freud is referring are not emo-
tions, feelings, or affects. And their becoming conscious is not a matter of
their producing in consciousness a counterpart of themselves having simi-
lar content. Freud goes on to say that the affective structures mentioned are
“processes of discharge,” whereas unconscious thoughts are “cathexes–basically
of memory-traces.” As such, unconscious thoughts are capable of becoming-
conscious in Freud’s technical sense. In contrast, the becoming conscious
of an unconscious affective structure is this: its producing in perception–
consciousness certain feelings combined with cognitive content that is owed
to the becoming-conscious of an unconscious thought which “represents” that
affective structure.
As I mentioned at the start, my present article only begins to lay out Freud’s
phenomenology of the emotions. Some loose ends left for later will certainly
have been noticed. In addition, much more than I have been able to investigate
here is contained in Freud’s work and bears upon my topic. I plan to return to
these matters in subsequent articles.

Notes

. I speak here of experiences rather than of states of consciousness merely to make exposi-
tion of James’s conception a little easier. But, in his view, every state of consciousness is an
experience, albeit often a very complex one with many objects. Any state of consciousness
that takes place is the total experience of the moment, assuming no secondary consciousness
stream is also flowing. Whenever two streams of consciousness flow simultaneously within
a single individual, there are produced for that duration two distinct total experiences of the
moment, as it were.
. From here on in the text, bare page references will be to this same source, until I make
it clear that I have now turned to a new text of Freud’s. This practice will continue for the
remainder of the article as further texts are considered.

References

Freud, S. (1957a). Repression. In S. Freud, Standard Edition, Vol. 14 (pp. 141–158). London:
Hogarth. (Original work published 1915)
Freud, S. (1957b). The unconscious. In S. Freud, Standard Edition, Vol. 14 (pp. 166–204).
London: Hogarth. (Original work published 1915)
Freud, S. (1961). The ego and the id. In S. Freud, Standard Edition, Vol. 19 (pp. 12–66).
London: Hogarth. (Original work published 1923)
JB[v.20020404] Prn:23/11/2004; 10:23 F: Z13110.tex / p.25 (1272-1339)

Freud’s phenomenology 

Freud, S. (1961). The economic problem of masochism. Standard Edition, Vol. 19 (pp. 159–
170). London: Hogarth. (Original work published 1924)
Freud, S. (1961). Civilization and its discontents. In S. Freud, Standard Edition, Vol. 21 (pp.
64–145). London: Hogarth. (Original work published 1930)
Freud, S. (1966). Project for a scientific psychology. In S. Freud, Standard Edition, Vol. 1 (pp.
295–387). London: Hogarth. (Original work published 1895)
Green, A. (1977). Conceptions of affect. International Journal of Psycho-Analysis, 58, 120–
156.
James, W. (1950). The principles of psychology (Volume 1). New York: Dover. (Original work
published in 1890)
Natsoulas, T. (2001a). The case for intrinsic theory: V. Some arguments from James’s
Varieties. The Journal of Mind and Behavior, 22, 41–68.
Natsoulas, T. (2001b). The case for intrinsic theory: VI. Incompatibilities within the stream
of consciousness. The Journal of Mind and Behavior, 22, 119–146.
Natsoulas, T. (2001c). The Freudian conscious. Consciousness & Emotion, 2, 1–28.
Natsoulas, T. (2003). The case for intrinsic theory: VIII. The experiential in acquiring
knowledge firsthand of one’s experiences. The Journal of Mind and Behavior, 24, 284–
316.
Schur, M. (1969). Affects and cognitions. International Journal of Psycho-Analysis, 50, 647–
653.
Woodruff Smith, D. (1989). The circle of acquaintance: Perception, consciousness, and
empathy. Dordrecht, Netherlands: Kluwer.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.1 (41-111)

Verbal expressions of self and emotions

A taxonomy with implications for Alexithymia and
related disorders

Louise Sundararajan and Lenhart K. Schubert

Regional Forensic Unit, Rochester, NY / Computer Science,
University of Rochester

A taxonomy of verbal expressions of self and emotions is proposed and tested

empirically with a computerized language analysis program, SSWC
(Sundararajan-Schubert Word Count). As an extension of the semiotics of
Charles Peirce, this taxonomy is based on the following hypothesis: there is a
“reflexive undertow” inherent in any sign or thought that fulfills the triadic
relations of the sign as delineated by Peirce. Lack of reflexivity is therefore the
liability of signs that fail to traverse the triadic path of its development. This
liability can be assessed on the basis of the distinctions made by Peirce
between sufficiently complete and incomplete signs and their corresponding
cognitive styles. The sufficiently complete sign corresponds to the
consciousness of Thirdness, which is characterized by thought processes that
are representational and reflexive. The incomplete signs correspond to the
consciousness of Secondness, which is reactive, and non-reflexive. This
hypothesis is applied to the alexithymia construct, on the one hand; and on
the other, tested by two empirical studies that analyze over 700 writing
samples with SSWC. Preliminary results show that the proposed taxonomy
has not only construct validity but also potential for making a contribution
to research in individual differences.

Keywords: taxonomy of affective lexicon, semiotics, Charles Perice,

reflexivity, Secondness, Thirdness, alexithymia, lack of introspection,
language analysis programs, individual differences in information processing

This paper proposes and offers empirical support for a taxonomy of verbal
expressions of self and emotions. Theoretical foundation of this taxonomy
rests upon the semiotics of Charles Peirce. The advantages of the Peircean
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.2 (111-183)

 Louise Sundararajan and Lenhart K. Schubert

framework lie in its systematic approach to self and thought as signs, and in
its clear delineation of the triadic relations of the sign. By assessing whether
and where a verbal expression is arrested in the triadic path of its develop-
ment as sign, a taxonomy of affective lexicon and personal pronouns can be
developed. This hypothesis is presented in four parts. Part one gives a brief
exposition of the ideal toward which and the norms by which sign/thought de-
velops, according to the Peircean semiotics – the triadic relations of the sign
with corresponding modes of consciousnesses; the difference between suffi-
ciently complete and incomplete signs, and the application of this analysis to
alexithymia. Part two presents a pattern matching language analysis program,
the SSWC (Sundararajan-Schubert Word Count), that translates this taxonomy
into 15 scales of verbal expressions of self and emotions. Part three presents two
empirical studies of SSWC to provide supporting evidence for the construct va-
lidity of the proposed taxonomy. Directions for future research will be explored
in the final section.

. Peircean semiotics and its application to Alexithymia

Triadic Relations of the Sign

What is a sign? According to Peirce, anything that represents another thing to
a mind is a sign: “A sign is an object which stands for another to some mind”
(Peirce in Fisch 1982, Vol. 3: 66). This perspective differs from the usual dyadic
signifier-signified formulation of signs by the inclusion of a third element, the
interpretation of signs by a mind. This third element – the “mind” – is referred
to by Peirce as the “Interpretant.” Thus we have the triadic relations of signs:
Object (signified)→ Sign (signifier)→ Interpretant (interpretation)

The third element, the mind, is pivotal in the Peircean semiotics: it is through
the medium of the mind that connections between the signifier and the sig-
nified are established. Furthermore, in the Peircean analysis, the development
of the sign and that of thought have coalesced: “Thought is in signs that at-
tain meaning through the triadic relation: object sign interpretant” (Hoopes
1991: 8). This Peircean equation of sign and thought (which includes feelings)
has important methodological implications for our purposes. It makes it pos-
sible for us to understand different types of signs in terms of the corresponding
cognitive styles, or “modes of consciousness” in the Peircean parlance.
Corresponding to the three elements of the sign, there are three modes
of consciousness: firstness, secondness, and thirdness. “Firstness is the sheer
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.3 (183-222)

Verbal expressions of self and emotions 

thisness, or existence, of things. Secondness is dyadic, or reactive, relations be-

tween things. And thirdness is triadic, or representational, relations among
things” (Hoopes 1991: 10). Since firstness – which corresponds to the signi-
fied (Object) – is the ineffable experience, the burden of further developments
of thought rests with Secondness and Thirdness. The Peircean Secondness and
Thirdness correspond roughly to Pennebaker’s low level and high level thinking
(1989). The Peircean distinction between Secondness and Thirdness also par-
allels with what is known in consciousness literature as the difference between
first-order experience and second-order awareness. The first-order experience
lacks the mental distance of reflection – it is a state of immersion in the here and
now factuality such that one’s affectivity is absorbed in action and may not be
recognized as emotion, as Frijda (in press) points out. This state of immersion
corresponds to the Peircean Secondness, which is a dualistic state of conscious-
ness, in which the third element – the Interpretant – is lacking. In contrast is the
second-order awareness, also known as “reflexive consciousness,” which refers
to a state of reflection and elaboration in feeling and thought. It is awareness of
awareness, a detached reflection that converts experiences into thoughts (Lam-
bie & Marcel 2002). This corresponds to the Peircean Thirdness. But Peirce
has much more to say about the reflexive consciousness. Indeed, the Peircian
framework can shed some light on the seemingly paradoxical observation of
Lambie and Marcel (2002) and others (Bucci 1995; Rosen 2004) that whereas
the first-order consciousness is that of immersion, the second-order conscious-
ness is capable of both detached and immersed intentional stance at once.
Peirce shows that the difference between levels of consciousness can best be
understood in terms of levels of complexity in structure – structures of higher
complexity can encompass that of lower complexity, but not vice versa. Thus
Peirce differentiates between Secondness (first-order experience) and Third-
ness (the second-order awareness) in terms of whether the triadic path of the
sign has been traversed or not, which is another way of asking whether a sign
has the third dimension (Interpretant) or not. The three dimensional con-
sciousness (Thirdness) and two dimensional consciousness (Secondness) can
be differentiated along the following registers – continuity, representation, and
reflexivity – which we discuss in turn.

Continuity
A sign is “Anything which determines something else (its interpretant) to refer
to an object to which itself refers (its object) in the same way, the interpre-
tant becoming in turn a sign, and so on ad infinitum” (Peirce in Hoopes
1991: 239). This condensed, but precise, formulation is rich in implications:
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.4 (222-263)

 Louise Sundararajan and Lenhart K. Schubert

a. It explains why the ideal thought process (thirdness) is referred to as “syn-

thetic consciousness,” because thought that traverses the triadic path of the
sign necessarily brings two facts (Object and Interpretant) into one single fact
(Sign) (Hoope: 182). b. Thirdness is necessarily constructive and generative,
because the Interpretant in turn becomes a sign to be interpreted by another
Interpretant, and the triadic process continues ad infinitum. In other words,
the movement of thought is a potentially unbroken cycle of signs and their
interpretations. Thus Peirce claims that the mind is a sign generator. This ob-
servation is consistent with the reportability of experience in second-order
awareness, as Frijda (in press) points out that attending to one’s experiences
tends to boost the impetus to categorize and use words. c. In the meaning
making process, it is essential for a feeling or thought to be followed by another
thought, as Hoopes points out, “A feeling is a mere sign, awaiting interpretation
in its relation with a subsequent thought or feeling before it can have meaning”
(Hoopes 1991: 10). To put it metaphorically, in the forest of signification, a tree
(thought) that falls without being heard (presented to and interpreted) by an-
other mind (thought) falls in vain. In sharp contrast is Secondness, which is
a stagnant thought process, because it lacks the Interpretant, the interpreting
thought that generates new interpretations/signs. In this respect, the difference
between the Peircean Thirdness and Secondness corresponds to some extent
to that between “substantive processing” which is constructive and generative,
and closed information search strategies which apply existing rules but do not
generate new ones (Forgas 2001). This “substantive processing” in Thirdness is
referred to by Peirce as a “sense of learning” or simply “thought.” The synthetic
or integrative nature of thought also has been noted by other theorists as “high
level thinking” (Pennebaker 1989) or as a high level of “referential activity”
(Bucci 1995).

Representation
Besides the generation of interpretations, the third element of sign – the mind –
is necessary for representation, for Peirce insists that representation is always
representation to a mind. Thus in the universe of Thirdness, all relations are
mediated (with the signs serving as medium). In contrast, Secondness is the
“dual consciousness” that governs relations without mediation, such as the
“process of action and reaction when one object strikes a second” (Hoopes
1991: 10). Or as Lambie and Marcel (2002) point out with regard to the first
order experience that the emotional states experienced in an immersed (un-
mediated) way is a world of sensations and objects, of “affected self,” and
“affecting world” (p. 244).
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.5 (263-299)

Verbal expressions of self and emotions 

An example of reactive or unmediated relation (Secondness) is a simple

action such as “I moved the table.” In contrast is the command of the offi-
cer “Ground arms!” which entails the mediated (representational) relation of
Thirdness. “The command mediates between the officer and the soldiers in
such a way as to put the conduct of the soldiers in line with the intention of the
officer. Just as we might say that the piece of music represents the thoughts of the
composer, we might say that the command represents the will of the officer [to
the soldiers]” (Colapietro 1989: 13–14). Persons who move the table need not
appeal (via a representation of their intention, i.e., sign) to the “mind” of the
table for their action to be effected. Thus relations in Secondness are dichoto-
mous: subject versus object; activity versus passivity; agent versus patient, etc.
Only one of the terms of this binary opposition has agency; and a lack thereof
inevitably marks the object of action (such as the table). This dichotomy does
not hold in the mediated relation of Thirdness, where agency is not confined to
the subject alone. The command “Ground arms!” is dependent on the soldiers
to exercise their agency to interpret it. Imagine the unlikely scenario of the offi-
cer resorting to direct action such as drugging the soldiers to disarm them, and
we can see how the “language of forces” referred to by Peirce is characteristic of
Secondness, a frame of mind that approaches the object of action as something
mindless, something that “forms material for the exercise of my will” (Peirce
cited in Hoopes 1991: 190).
Particularly pertinent to our discussion is the fact that the notion of emo-
tion as passion is susceptible to the language of forces. Peirce notes that in the
language of motives and desires, the self is rendered an object being compelled
from within: “we speak of allurements and motives in the language of forces,
as though a man suffered compulsion from within” (Peirce cited in Hoopes
1991: 195). Furthermore, when emotions are represented as actions, they also
fall under the category of Secondness. More precisely, such cases are consid-
ered by Peirce as “degenerate secondness” or “internal” seconds in contrast
to “external seconds, which are constituted by external fact, and are true ac-
tions of one thing upon another” (Peirce cited in Hoopes 1991: 195). A case in
point is the verb form of words such as love or hate, known as “noncausative
verbs” (Clore, Ortony, & Foss 1987: 756). These “noncausative verbs” have one
peculiar feature, namely, that they function as emotion terms only in the ac-
tive voice such as “I hate winter,” and not in the passive voice such as “I was
hated.” This subject-object, or activity-passivity dichotomy is characteristic of
the “dual consciousness” of Secondness, where one is either the subject acting
on an object, or the object being acted upon, as we may recall.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.6 (299-342)

 Louise Sundararajan and Lenhart K. Schubert

Reflexivity
Finally, thinking at the level of thirdness is intrinsically dialogical – thinking
is thought talking to itself, as Peirce and others have noted – hence reflexive.
Consider the scenario of a child touching the hot stove, a scenario given by
Peirce himself (1931–58, Vol. 5, paragraph 233): When the child feels the pain,
“ he becomes aware of ignorance, and it is necessary to suppose a self in which
this ignorance can inhere” (Fisch 1982, Vol. 2: 202). Hoopes (1991) explains:
“From the resulting feeling (sign), the child arrives at the conclusion (Interpre-
tant) that there is such a thing as error that it inheres in its self (Object)” (p. 8).
The development of thought in this scenario can be captured by the triadic sign
relations:
Object (signified) Sign (signifier) Interpreptant (interpretation)

Thus:
Object (tissue damage) Sign (burn sensation) Intrepretant (error)

Object (updated self)

Wiley (1994) has noted the “reflexive undertow” of communication (p. 27), in
which the speaker is “counted twice, once as communicator and once as re-
flexive communicatee” (p. 27). Applied to the intrasubjective communication,
the “speaker”(Object of the sign) is the self that first communicates through
the sign (the burning sensation) to the mind (Interpretant) for an interpre-
tation, which in turn feeds back to the self, resulting in a revised version of
the self as one capable of ignorance. Note that the self (Object) is constructed
along the way, with elements of novelty (revisions) not found before the sig-
nifying process. Lambie and Marcel (2002) have also pointed out that aware-
ness/representation of emotion and that of the self go hand in glove: “what
is fully meant by awareness of emotion is awareness of it as part of oneself ”
(p. 253). Furthermore, self-reflexivity has a moral dimension: it is the self to
self or thought to thought dialogue that makes self control possible. More on
this later.
Self-reflexivity is precisely what is missing in the unmediated relation of
Secondness. Peirce has noted that:
When a child wants to move a table, he is likely to be so absorbed in what he
wills as to be oblivious to himself: “Does he think of himself as desiring, or
only of the table as fit to be moved?” (1931–1958, Vol. 5, paragraph 230)
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.7 (342-384)

Verbal expressions of self and emotions 

The same applies, when emotion is processed in an immersed (unmediated)

mode. For instance, in “I hate him,” the emoter is not self-reflexive, as his or
her attention is absorbed by the qualities of the person as “fit” to be hated,
no less than a table as “fit to be moved.” Frijda (in press) has made a similar
observation concerning infatuation as an instance of the first-order experience,
where one is enthralled by a person’s attractiveness, and considers “I love her”
as an objective fact. Frijda goes on to say that in the immersed consciousness of
the first-order experience, no subjectivity, no reference to the self is involved.

Application to Alexithymia
Alexithymia is a “multifaceted personality construct that has been associated
with various medical and psychiatric disorders (Taylor 2000: 134). The salient
features of alexithymia are: 1. “difficulty identifying and describing subjective
feelings,” 2. “difficulty distinguishing between feelings and bodily sensations
of emotional arousal,” 3. “constricted imaginal capacities, as evidenced by a
paucity of fantasies,” and 4. “an externally oriented cognitive style” (Taylor
2000: 135). These deficits may be understood as a lack, in varying degrees,
of “inwardness”: 1 pertains to difficulty with processing one type of internal
stimuli–subjective feelings; 2 pertains to difficulty in differentiating between
two types of internal stimuli – physical sensations versus emotional arousal;
3 pertains to the inability to inhabit an inward–mental, symbolic–space; and
4 refers to a “lack of introspection” (Taylor et al. 1997: 56). This lack of in-
wardness has been described in various terms ranging from a lack of “psy-
chological mindedness” (Conte et al. 1990; Fonagy et al. 1991; Taylor & Bagby
2000), to a “bypass of the psychic” (Greco 1998: 133). Such descriptions are too
vague for further development in theory and measurement. The semiotics of
Peirce can help us with more precise formulations of both “inwardness” and its
lack thereof.
As Sartre points out, humans differ from objects in that whereas the latter
are being in itself, humans are being for itself. This self-reflexivity (for itself)
of the human consciousness is reflected in the dialogical nature of thought
(thought talking to itself). One consequence of self-reflexivity is a “subjectiv-
ity” rich in “inner life” in which the self is sustained by an ongoing internal di-
alogue, rather than dictated by external circumstances. This ongoing dialogue
between self and self, or thought and thought, opens up a region of freedom,
a “psychological waystation” which allows possibilities and constraints to be
negotiated such that the individual is able to respond rather than simply re-
act to any stimuli. This inwardness plays an important role in the thinking of
Peirce, who explains its importance in terms of consciousness (cited in Colapi-
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.8 (384-458)

 Louise Sundararajan and Lenhart K. Schubert

etro 1989: 114): “the true definition of consciousness is connection with the
internal world”; and “ the function of consciousness is to render self-control
possible and efficient.”
These advantages of the “inner space” or “subjectivity” are denied the
consciousness of Secondness. Not mediated by the dialogue of thought, trans-
actions of self to self as well as self to other in the universe of Secondness
become rigid and mechanical, as if facts are knocking upon facts with lit-
tle room for negotiation. This “object mode” is characteristic of the so-called
“Pensée opératoire” (Marty & de M’Uzan 1963), which refers to the tendency
for alexithymics to approach self and other as objects (Taylor et al. 1997; Krystal
1988). Greco notes that:
. . . the alexithymic apprehends society . . . as a system of determinisms rather
than a system of constraints . . . because one pole of the dialectic which allows
for negotiation with society itself, the pole of subjectivity, is absent. There is
no margin for negotiation because the two poles of the relationship are indis-
tinguished. This indistinction, and the lack of a margin for negotiation, means
that no response-ability is possible: what responds is the autonomic nervous
system left to its own devices, unable to discriminate between an actual need
or value and a symbolic one, between the organic self and the psychological
self. (1998: 142)
The self in this “object mode” is the victim of circumstances, a condition
consistent with Marty and de M’Uzan’s observation of alexithymia (1963): “Ev-
erything happens as if it were imposed on these individuals” (p. 348). As much
as they perceive themselves as being acted upon like objects, alexithymics tend
to perceive others as objects as well. Krystal observes that these patients tend
to use people as objects for “manipulative or exploitative purposes . . . . there is
no personal investment in these objects as unique individuals . . . ” (1988: 247).
This tendency of the alexithymics to perceive self and other in the same
light as objects finds an elegant, i.e., parsimonious, explanation in the frame-
work of Peircean semiotics. In his stipulation of the “Interpretant,” Peirce in-
sists on the sign being read/interpreted by “a certain mind,” without making
any distinction between the mind of other’s or that of one’s own. This equa-
tion would predict the self to self and self to other transactions to be based on
similar schemas. Those who ignore their own thoughts and feelings can be ex-
pected to fail in acknowledging the subjective “inner space” in others as well.
This is borne out by the “Pensée opératoire” phenomena in alexithymia. Taylor
cites clinical observations of “brittle” diabetics who “fail to create mental rep-
resentations of [both] the self and primary caretakers as thinking, feeling, and
experiencing persons” (Taylor et al. 1997: 237).
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.9 (458-495)

Verbal expressions of self and emotions 

. A taxonomy of self and emotions and its implementation by SSWC

Based on the foregoing analysis, we have developed a taxonomy of personal

pronouns and affective lexicon to categorize verbal expressions of self and emo-
tions. This taxonomy capitalizes on two practical implications derived from the
Peircean semiotics: a. self-reflexivity, the feedback loop that helps to update the
experiencing self, is a hallmark of the fully developed signs; and conversely, b.
lack of reflexivity is characteristic of signs or thought that fail to complete the
triadic/reflexive path of its development. Using self-reflexivity as the gold stan-
dard for the representation of personal experiences (see Sundaratrajan 2001),
verbal expressions of self and emotions can be categorized on the basis of their
varying degrees of deficiency in self-reflexivity. We have accordingly developed
a pattern matching language analysis program SSWC (Sundararajan-Schubert
Word Count) with 15 scales, to be presented below. Three additional scales –
one redundant (the Arousal scale combining the High Activation and the Low
Activation scales) and two ad hoc scales – will not be considered here. In com-
piling the dictionaries for each scale, we have incorporated the affective lexicon
in Ortony et al. (1987), Averill (1975), and Johnson-Laird and Oatley (1989).
We have also consulted Lane’s LEAS Scoring Manual and Glossary (1991).

A taxonomy of personal pronouns

The self is from the start implicated in the sign relation – it is the Object (the
signified), as Hoopes points out, “the self is the sign relation, since feeling is
meaningless unless it is interpreted as the sign of an object” (1991: 8). Thus
various signs of the self – personal pronouns – can be categorized according
to their implied distance from the experiencing self (in descending order), as
follows: Detached Self, Affected Self, Focal Self, and Reflexive Self.
Detached self : this scale – marked by terms such as “everyone” or “they” – is
an index of the third person perspective, which reflects a detached intentional
stance toward personal experiences.
Affected self : Marked by words and expressions such as “me,” and “make
me feel,” this scale indexes the perception of the self not as doer, but as being
done to. It reflects an immersed intentional stance, in which “what one experi-
ences is an affected self or an affecting world” (Lambie & Marcel 2002: 244).
Focal self : this scale consists of three terms – “I,” “myself,” and “my own.”
It is indicative of focal attention to the self. The focal self is not to be confused
with the experiencing self–the ultimate point of reference of all emotions (see
Dorsey 1971). Used in moderation, self-focus is associated with self-reflection,
a definitely self-reflexive turn. However, heightened self-focus tends to objec-
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.10 (495-600)

 Louise Sundararajan and Lenhart K. Schubert

tify the self, rendering it “an object in the world distinct from others, an object
with boundaries, fixed properties, and the capacity to be controlled” (Silvia &
Gendolla 2001: 243). This tendency to turn the self into an object of experi-
ence is antithetical to the reflexive awareness of the self as origin and subject of
experience.
Reflexive self : The reflexive self refers to an awareness of the experiencing
self (Sundararajan 2001), which unfortunately cannot be represented directly.
The experiencing self is a perceptually recessive self embedded in action and
experience (Gallagher and Marcel 1999). Silvia and Gendolla (2001) refer to
this perceptually recessive self as “subjective self awareness . . . when attention is
focused away from the self and the person experiences himself as the source of
perception and action. The person in this state feels active and agentic, experi-
ential, and unified with the environment through the medium of activity” (p.
243, emphasis added). Thus the experiencing self is the de-centered self that is
theoretically antithetical to the focal self (the “I”). Furthermore, since reflex-
ivity is dialogical (thought talking to itself), the reflexive self is not an “I” so
much as a “we” (see Wiley 1994). This is evident in the interpersonal context:
the officer may derive a sense of “we” when the soldiers show their understand-
ing (Interpretant) of his intention (Object) conveyed through his command
(Sign) “ground arms!” The same is true with communication in the intraper-
sonal context: The self reaches a sense of coherence, analogous to the sense of
“we” in social exchange (Wiley 1994), when tissue damage (Object), its Sign
(pain), and its Interpretant (error) are in good accord with one another, as is
the case with the scenario of the child touching the stove. That the sense of “we”
applies to both the intra- as well as interpersonal community of thought finds
an eloquent expression in the equation that Peirce draws between self as society
and society as self: on the one hand, an individual’s “ thoughts are what he is
‘saying to himself,’ that is, is saying to that other self that is just coming into
life in the flow of time. . . .” On the other hand, “the man’s circle of society . . .
is a sort of loosely compacted person, in some respects of higher rank than the
person of an individual organism” (Peirce 1931–1958, Vol. 5: paragraph 421).
This sense of “we,” however, cannot be directly represented by the linguistic
“we.” The “we” in English has been co-opted often times to mean just the oppo-
site – an impersonal stance. The scale of Reflexive Self, therefore, excludes “we”
but includes expressions such as “our own.” Also included in this scale are ex-
pressions of self-referentiality such as “oneself,” or “ourselves.” Excluded again
is “myself ” which, with its heightened self focus as an atomic self, is antithetical
to the extended self behind the sense of “we.”
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.11 (600-616)

Verbal expressions of self and emotions 

Table 1. Taxonomy of affective lexicon and corresponding sign deficiencies

Modes of Deficiencies in sign function SSWC Categories

Consciousness in descending order
Secondness Infidelity in reporting
Object mode–deficiency
in symbolic elaboration
Thirdness Potentially complete signs with
varying degrees of deficiency
in inward attention
Externally oriented interpretation
Emotion management
Attention to affect
Note. In parenthesis are token expressions.
Iconic use of affective lexicon:
Words in a dictionary
Denial (“Doesn’t bother me.”)
Violent Words (kill, profanities)
Emotion as Event (“I hate . . .”)
Somatic (breathless)
External Attribution (sexy)
Suffering (insulted)
Affect Non-Focal (confused)
Low Activation (relaxed)
High Activation (nervous)
Valence Focus (miserable)
Affect-Focal (happy, sad)

A taxonomy of affective lexicon

To reiterate the proposed gold standard of signs in the domain of affective
expressions, a sufficiently complete sign (signifier) is one in which the inter-
pretation (Interpretant) refers back reflexively to the emotional experience (the
signified or Object). Thus the more a verbal expression of emotion approxi-
mates this ideal, the higher degree of fidelity it will have as a representation/sign
of the emotional experience. The notion of fidelity can be further broken down
into different requirements at each phase of the triadic trajectory of the sign.
The path from Object to Sign requires fidelity in reporting; that from the Sign
to the Interpretant requires symbolic elaborations to turn events or physical
sensations into mental representations; and that from the Interpretant to the
Object requires inward attention to complete the reflexive feedback loop. On
the basis of this understanding, we propose a taxonomy of affective words and
expressions in terms of their deficiency in fulfilling the essential functions of
the sign (see Table 1).
In the following discussions, each type of deficiency, in descending order, is
illustrated with a diagram of the triadic relation, with an “x” indicating where
the development of the sign is arrested.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.12 (616-687)

 Louise Sundararajan and Lenhart K. Schubert

1. Signs that are de-coupled from the signified:

Object x Sign Interpretant

Denial: The scale of Denial is composed of expressions such as “doesn’t

bother me,” or “feel numb.” It refers to signs that fail to meet the require-
ment of fidelity in representation by minimizing or repressing the impact of
the experience (the Object or signified).

2. Signs characteristic of Secondness

Object Sign x Interpretant

Signs in this category are characteristic of Secondness, the universe of facts

and actions, and are generally deficient in their impetus toward symbolic
elaborations.
Violent words: This scale is composed of words of violent action, such as
“kill,” “rape,” obscenities (which are verbally acting out behaviors), and words
denoting the notion of emotion as energy and discharge, such as “letting off
steam.” As signs, the referential focus of these words is on action and reaction –
the universe of Secondness – thereby detracting the emoter from the “inner
space” of feelings and interpretations.
Emotion as event: Dictionary for this scale consists of 21 “noncausative
verbs” (such as love, hate, used as verb, in active, not passive, voice) from Clore
et al. (1987: 763–765). Statements such as “I love her” are non-reflective, be-
cause the experience is reported as a fact or event. Otherwise put, the emoter is
immersed in the experience such that his or her affect is absorbed by the event,
leaving little impetus for symbolic elaboration. What does it mean or what is it
like to love someone? “My love is like a red red rose.” Note that this symbolic
elaboration of experience is possible only when one leaves the world of facts
and actions and enters the world of concepts (as denoted by the noun form of
love here). Thus the noun and adjective forms of the terms in this scale belong
elsewhere – they fall under the scale of Affect-Focal words instead (see below).
Somatic: Dictionary for this scale is based primarily on the word list of
“physical and bodily states” in Clore et al. (1987). We also added psycho-
somatic expressions such as “heart racing.” Words under this category are
instances where the referential focus is primarily on the physical rather than
mental or psychological conditions. Such signs are what Peirce refers to as
indices. “An index is a sign which would, at once, lose the character which
makes it a sign if its Object were removed, but would not lose that charac-
ter if there were no Interpretant” (Peirce cited in Hoopes 1991: 239). Peirce
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.13 (687-723)

Verbal expressions of self and emotions 

gives the examples of the bullet-hole as a sign of a gun shot to show that
indices have direct physical connections to the signified, a connection inde-
pendent of an Interpretant: “for without the shot there would have been no
hole; but there is a hole there, whether anybody has the sense to attribute it
to a shot or not” (Peirce cited in Hoopes 1991: 240). By the same token, the
assumption behind somatic complaints such as “fatigue” is that they are indi-
cations of some physiological change, a condition that “truly” exists, regardless
of whether it is recognized/interpreted as such or not. From the Peircean per-
spective, this alleged independence from the Interpretant explains why words
denoting somatic concerns are usually deficient in their impetus for symbolic
elaborations.

3. Signs with external referential focus

Object Sign Interpretant

Signs in this category are interpretations with an external orientation that

detracts them from the inward attention required for the completion of the
feedback loop from the Interpretant to the Object.
External attribution: This scale is based primarily on terms referred to by
Clore et al. (1987) as “external conditions.” These are words that refer to ei-
ther (a) “opinions, evaluations, or reactions that the person elicits in others”
(p. 348), such as “sexy,” or (b) “the described person’s state of being in the
world” (p. 350), such as “abandoned.” The referential focus of these words is
on the external attributions of the emotional states, rather than the emotional
states per se.
Suffering: This is a subscale of External Attribution. It consists of verbs in
passive voice, such as “violated” or “insulted,” words that designate the extreme
pole of the affected self – the victim stance. Words in the suffering scale lack
self-reflexivity, as their referential focus is on the cause of the emotion (being
insulted), rather than the emotion itself (anger, etc.).

4. Signs with potential for thirdness

Object Sign Interpretant

Object
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.14 (723-776)

 Louise Sundararajan and Lenhart K. Schubert

Signs in this last category have the potential to traverse the triadic path of the
sign. Verbal expressions in this category can be further divided into two types
of signs showing either inattention or attention to affect.

4a. Signs focusing on management of emotion

Affect non-focal: Dictionary for this scale is based primarily on the affect
non-focal terms in Clore et al. (1987). Words such as “understanding,” or “cry”
are signs in which the referential focus has shifted from affectivity to cogni-
tive and behavioral components of the experience. Also included in this scale
are clichés, such as “depressed.” As Clore et al. (1987) have pointed out, affect
non-focal and affect-focal expressions are not always sharply differentiated, the
former can be expected to be found in the penumbra of affect focal lexicon as
its spillover. However, we hypothesize that a preponderance of affect non-focal
terms are the result of heuristic information search strategies and motivated
processing as an attempt to control and limit the scope and impact of one’s
affective experience (Forgas 2001).
Low activation: This scale is based primarily on the word list of deactivated
state in Barrett and Russell (1998). It consists of words such as “relaxed,” words
that denote the absence of affective arousal. Elevation of this scale is hypothe-
sized to be indicative of motivated processing to keep emotional stimulation at
the minimum.

4b. Signs showing attention to affect

High activation: This scale is composed of words that denote an arousal
focus with high activation, such as “nervous.” Dictionary for this scale is based
primarily on the word list of activated affect in Barrett and Russell (1998).
Valence focus: Words such as “miserable” are signs that show a referential
focus on the pleasant-unpleasant aspect of the experience. Dictionary for this
scale consists of the word list of pleasant and unpleasant affect in Barrett and
Russell (1998). Also included are word lists with highest scores on the Depth
and Evaluation dimensions in Averill (1975: 17).
Affect focal: This scale is based primarily on the affect-focal scale of Clore et
al. (1987). Words such as “happy” and “sad” meet the conditions for bona fide
emotion words, according to clore et al. (1987): “the best examples of emotion
words would be ones that refer to internal (as opposed to external) conditions,
those that refer to mental (as opposed to physical) conditions, and those that
have a significant focus on affect [as opposed to cognition or behavior]” (p.
752). Items for inclusion in this scale also meet Lane’s criteria (1991) for level
three emotional awareness. Because of their referential focus on the mental
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.15 (776-818)

Verbal expressions of self and emotions 

and internal, Affect Focal words have the best potential to traverse the triadic
path of the sign. However, possibilities for short-circuiting the triadic trajectory
always exist.
Affective terms lose their reflexivity when used as icons. An icon is a rep-
resentation that is independent of its Object. According to Peirce, “An icon
is a sign which would possess the character which renders it significant, even
though its Object had no existence” (cited in Hoopes 1991: 239), for instance
a diagram of something that does not exist. The iconic representation cor-
responds to What Gendlin refers to as “conceptualization”: “When symbols
conceptualize or represent, they themselves ‘mean’ what they represent. We
might say that they mean independently” (Gendlin 1997: 96, emphasis in the
original). Because of its tendency to mean “independently,” the icon can be-
come de-coupled from its Object, and thereby losing its reflexive thrust. A case
in point is the way “semantic construers” process emotion concepts (Robin-
son et al. 2002; Robinson et al. 2003). In contrast to “episodic construers”,
for whom emotion concepts depend for their meaning on accessibility to per-
sonal experiences, semantic construers seem to approach the same as icons that
have meaning independent of experience. Thus semantic construers may not
be able to differentiate between affect-laden words–which require the reflexive
awareness of the match between the signifier (word) and the signified (personal
experience), and neutral words–which can say what they mean independent
of personal experiences. This has been found to be the case with psychopaths
(Herpertz, & Sass 2000; Williamson, Harpur, & Hare, 1991). The fact that it
is possible for some individuals to crank out affective lexicon as if reading off
the dictionary has methodological implications for emotion research, an issue
which will be addressed later.

. Two empirical studies of SSWC

Study 1

In Study 1, a total of 586 texts were collected for analysis by SSWC (Sundararajan-
Schubert Word Count) from 121 participants recruited from 3 different popu-
lations: college students, psychiatric inpatients, and professionals. The samples
were chosen to provide a wide spectrum of age, race and gender distribution,
and educational background. The purpose of the study was to test construct
validity of the SSWC categories.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.16 (818-867)

 Louise Sundararajan and Lenhart K. Schubert

Method
Participants
A total of 121 participants were recruited from 3 different populations:
Sample1 consisted of 69 undergraduate and graduate students from local
colleges, (55% female), with a mean age of 22.5 years (SD = 4.0) and 15.4 years
(SD = 1.6) of education completed. There were 1 (1.45%) African American, 1
(1.45%) Native American, and 67 (97.1%) Caucasians. Five dropped out from
the initial pool of 74.
Sample 2 consisted of 42 inpatients from a psychiatric hospital (33% fe-
male), with a mean age of 38.36 years (SD = 13.47), and mean years of edu-
cation 11.21 (SD = 2.72). There were 1 (2.38%) Asian,18 ( 42.86%) African
Americans, and 23 (54.76%) Caucasians. Majority of the patients were diag-
nosed with schizophrenia; referral was based on alexithymia-related symp-
toms, such as eating disorder (7.14%), substance abuse (71.43%), problems
with impulse control (52.38%), sexual offense (7.14%), PTSD (4.76%), and
borderline personality disorder (2.38%). Out of 70 referrals, 28 either dropped
out or were disqualified due to their mental status.
Sample 3 consisted of 10 humanistically oriented professionals–9 psy-
chotherapists at the doctoral level (except for one at the master’s level), and
1 philosophy professor. This group had equal distribution of gender (5 men
and 5 women), with a mean age of 49 years (SD = 11.3). There were 1 (10%)
Asian, and 9 (90%) Caucasians. One dropped out from the initial pool of 11.
The humanistically oriented professionals were recruited because of their al-
leged potential for “self-introspection.” None of the participants in this group
were familiar with the measures used in the study. Due to the small sample size,
data from this group will not be used for separate analysis.

Materials and procedure

Participants completed a battery of questionnaires and writing tasks in 4 to
5 thirty-minute sessions, which were spread out in 2 to 3 weeks. Order of
presentation of the measures was randomized within each session. Samples 1
and 2 completed the study in groups; sample 3 did it individually via the in-
ternet. Voluntary participation and confidentiality of the data were reiterated
in the informed consent, which was signed by all participants. Sample 2 re-
ceived payment ($3.00 dollars per session) for their participation; samples 1
and 3 did not.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.17 (867-932)

Verbal expressions of self and emotions 

Battery of questionnaires: TAS-20, LEAS-A, APRQ, SCS.

TAS-20: Toronto Alexithymia Scale-20 (Bagby, Parker, & Taylor 1994; Bagby,
Taylor, & Parker 1994) has three subscales: EOT (Externally-Oriented Think-
ing), DIF (Difficulty Identifying Feelings), DDF (Difficulty Describing Feel-
ings). The TAS-20 is the standard measurement of alexithymia.
LEAS-A (Levels of Emotional Awareness Scale-A) (Lane et al. 1990). This
is a performance based measure of emotional awareness, which is relevant to
alexithymia but not a direct measure of it. LEAS-A is the shorter form of LEAS
with 10 items instead of the original 20. The shorter form was used in order
to accommodate the limited attention span of the psychiatric inpatients (sam-
ple 2). LEAS-A was scored by 2 raters, with inter-rater correlations at r = .97,
p < .000.
APRQ: The Alexithymia Provoked Response Questionnaire (Krystal et al.,
1986). This is a performance based measurement of alexithymia. The use of
APRQ was modified from its standard use as structured interview in a one to
one setting to a paper and pencil questionnaire administered in groups (Sam-
ples 1 and 2), and individually via the internet (Sample 3). APRQ was scored
by 2 raters, with inter-rater correlations at r = .85, p < .000.
SCS: Self-Consciousness Scale (Fenigstein et al. 1975), and one of its sub-
scales, P-SCS (Private Self-Consciousness Scale). The SCS total and P-SCS
scores were included to see if “self consciousness” had any overlap with the
varieties of self concepts covered by our model.

Four essay writings, 10 minutes each: Each testing session was randomly as-
signed one of the following writing tasks: a. the most stressful event, b. the most
enjoyable event, c. gender differences, d. feelings about the war with Iraq.

Feeling words: Each testing session ended with the task of writing 6 feeling
words to indicate the current mood, with a total of 24 terms produced by each
participant who attended all sessions. Participants were given a word list from
which to choose 6 feeling words, except for the last session in which no word
list was provided. The word list consisted of 33 words based on PANAS mood
scale (Watson et al. 1988) and valence and arousal terms adapted from Barrett
& Russell (1998). The feeling words generated by the participants were meant
to measure not their actual mood states, so much as their cognitive style –
whether the individual preferred to use “upset” (an Affect Non-Focal term) or
“disappointed” (an Affect Focal term) in describing his or her mood state.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.18 (932-1057)

 Louise Sundararajan and Lenhart K. Schubert

Two language analysis programs: Written texts of the participants were col-
lected from the sessions, transcribed into computer text files, and processed by
two language analysis programs, LIWC and SSWC.
LIWC 2001: Linguistic Inquiry and Word Count (Pennebaker et al. 2001).
This program contains multiple word category dimensions, and calculates the
percentage of total words represented in each category.
SSWC: Sundararajan-Schubert Word Count (Sundararajan & Schubert
2002, 2003). This pattern matching language analysis program counts occur-
rences of words and phrases in a dictionary of close to 2000 entries. Technically,
the allowable occurrence patterns follow a regular-expression syntax whose
primitives are words, word stems, and parts of speech, with allowance for nega-
tion. The word count gives percentage of production of any SSWC category
out of the total words of a particular text. For instance, if “happy” appears in a
text of 100 words, the Affect Focal score of that text would be .01; whereas in
the case of the same term in a text of 10 words, the Affect Focal score would
be .10. The rationale behind this calculation is based on the assumption (see
“componential analysis” in Ortony et al. 1987) that different components of
the semantic universe compete for expression – the percentage gives an indica-
tion of the extent to which one particular component has out-competed others
for the cognitive resources allotted to the production of a text of certain length,
be it 100 or 10 words.

Results and discussion

Pearson correlational analyses were performed to assess the relationship among
all the measures. SCS showed no significant correlation with measures of emo-
tion, as can be expected, but was correlated with some of the SSWC categories
of self expression. APRQ showed no significant correlation with SSWC vari-
ables, probably due to our modification of its format from interview to paper
and pencil test. LEAS correlated highly and positively with APRQ (r = .61,
p < .0001), and moderately and negatively with TAS-20 total (r = –.34, p <
.0001), but had few correlations with SSWC variables. TAS-20 had significant
correlations with SSWC variables in some but not all samples. The correlations
between LIWC and majority of SSWC variables, in contrast, were robust across
samples. Table 2 is a summary of the correlations with relevant LIWC variables
across all samples, including Study 2 to be examined later.
The following SSWC correlations with external measures will focus pri-
marily on the LIWC variables; other external measures will be included when
appropriate.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.19 (1057-1057)

Verbal expressions of self and emotions 

Table 2. SSWC and LIWC Correlations: Study 1 (A, B) and Study 2 (C) combined
SSWC Categories Relevant LIWC Categories

Affective I We OTHREF POSEMO NEGEMO AFFECT BODY SENSES SWEAR

Lexicon

Denial A, B, C
Violent A
Words B
C .28* –.33** .47**** .35** .74****
Emotion A
as Event B .42*** .34** .29**
C .46**** .27* .25* .38**
Somatic A .38**
B .30** .47****
C .49**** .25*
External A .51*** .64****
Attribution B .28*
C .41*** .44***
Suffering A .30* .37**
B
C .31** .36**
Affect A .55**** .38**
Non-focal B .39*** .39*** .46****
C .37** .27* .46****
Low A .48***
Activation B .47****
C
High A .33* .55***
Activation B –.30** .49****
C .27*
Valence A .35* .40**
Focus B .43*** .47**** .47****
C .27* .28*
Affect A .37* .46** .57**** .51***
Focal B –.31** .45**** .58**** .50****
C .39** –.24* .56**** .54****

Personal Pronoun
Detached A .56****
Self B .50**** .24* .25*
C .26* .30** .42***
Affected A .62**** .31*
Self B .36** .40***
C .31** .73**** .40*** .26* .30**
Focal A .79****
Self B .85****
C .88**** –.39*** –.30* –.26*
Reflexive A .75**** .46**
Self B .52**** .35** –.29* –.25*
C –.34* .56**** .40*** .26*

Note. A = Study 1, patients (n = 42); B = Study 1, students (n = 69); C = Study 2, prison inmates (n = 65).
*p < .05; **p < .01; ***p < .001; ****p < .0001, two-tailed; empty cells = no significant correlations.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.20 (1057-1114)

 Louise Sundararajan and Lenhart K. Schubert

Correlations with LIWC and other External measures

Denial: This scale had no correlation with any LIWC scales under consid-
eration.
Violent words: This scale had no correlation with any LIWC scales under
consideration. Significant and negative correlation between Violent Words and
LEAS was found in the patient sample (r = –.31, p < .05). This is consistent
with the finding of Barrett (unpublished data, cited in Lundh 2002) that LEAS
correlated positively with self restraint and impulse control.
Emotion as event: In the student sample this scale had multiple positive
correlations with the LIWC scales of positive emotion (POSEMO), affect, and
feelings and sensations ( SENSES).
Somatic: This category in the patient sample correlated positively with
body states and symptoms (BODY), and similarly in the student sample with
both BODY and feelings and sensations (SENSES).
External attribution: In the patient sample this scale had strong correlations
with positive emotion (POSEMO), and affect (r = 64, p < .0001). In contrast,
student sample showed a weaker correlation with affect (r = .28, p < .05). In
addition to the LIWC, positive correlations were found in the student sample
with Private Self Consciousness (r = .27, p < .05), and Self Consciousness total
scores (r = .29, p < .05), suggesting that, for the student sample at least, an
external referential focus can result in looking at the self through the eyes of
the other, resulting in heightened self consciousness.
Suffering: This category in the patient sample showed positive correlation
with sensations and feelings (SENSES), and affect. No correlation was found in
the student sample.
Affect non-focal: This category in both patient and student samples showed
positive correlations with positive emotions (POSEMO), and affect. In addi-
tion, student sample showed a positive correlation with feelings and sensations
(SENSES). In addition to the LIWC, the Affect Non-focal category in the stu-
dent sample correlated negatively with TAS-20 scores: TAS total(r = –.27, p <
.05), DIF (r = –.25, p < .05), and DDF(r = –.25, p < .05). To the extent
that, as Lundh et al. (2002) suggested, the DIF (Difficulty Identifying Feel-
ings) and DDF (Difficulty Describing Feelings) “measure something like the
lack of meta-emotional self-efficacy concerning the identification and commu-
nication of emotion” (p. 376), negative correlations with these two factors of
TAS-20 are consistent with our hypothesis that Affect Non-Focal is an aspect of
emotion management, which necessarily contributes to one’s “meta-emotional
self-efficacy.”
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.21 (1114-1169)

Verbal expressions of self and emotions 

Low activation: This category in the patient sample showed positive corre-
lation with AFFECT , but in the student sample it was correlated with physical
sensations and feelings (SENSES).
High activation: Patient sample showed positive correlations with negative
emotion (NEGEMO), and sensations and feelings (SENSES); student sample
showed positive correlation with negative emotion, and negative correlation
with positive emotion, thus contributing to discriminant construct validity.
Valence focus: Patient and student samples showed positive correlations
with affect, and sensations and feelings (SENSES). In addition, student sample
showed positive correlation with positive emotion (POSEMO).
Affect focal: Patient and student samples showed comparable positive cor-
relations with AFFECT, sensations and feelings (SENSES), and negative emo-
tion (NEGEMO). In addition, patient sample showed positive correlation with
“I”; and student sample showed negative correlation with “WE.” In addition
to the LIWC, P-SCS (Private Self Consciousness) correlated negatively with
Affect-Focal in the patient sample (r = –. 31, p < .05). This is consistent
with our hypothesis that attention to emotion entails a diffused sense of the
self, which is antithetical to the preoccupation of the self as measured by the
P-SCS scale.
Detached self : Patient sample showed positive correlation with refer-
ence to others (OTHREF). Student sample showed equally strong correlation
with OTHREF, and in addition positive correlations with positive emotion
(POSEMO), and AFFECT.
Affected self : This category has a dual referential foci, the self as being
done to and the other as affecting me. As expected, both patient and stu-
dent samples showed correlations with the LIWC scales of “I” and reference
to others (OTHREF).
Focal self : This category showed strong correlations with “I” in both pa-
tient and student samples. In addition to LIWC, this category correlated pos-
itively in the patient sample with a subscale of TAS-20, Difficulty Describing
Feelings (DDF r = .37, p < .05), suggesting that impairment in expressing
emotions may be accompanied by heightened self focus. Yet, this heightened
self focus may not be the same as self consciousness as measured by SCS. In
fact the patient sample showed a negative correlation between Focal Self and
Private Self Consciousness (r = –.35, p < .05).
Reflexive self : This scale showed strong correlation, as predicted, with
“WE” in both patient and student samples. It also correlated with reference
to others (OTHREF) in both patient and student samples. In addition, student
sample showed negative correlations with sensations and feelings (SENSES r =
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.22 (1169-1323)

 Louise Sundararajan and Lenhart K. Schubert

–.29, p < .05), and with SWEAR (r = –.25, p < .05). This is consistent with our
hypothesis that the Reflexive Self resides in meaning and coherence of the ex-
tended self, rather than in feelings and sensations of the isolated self, let alone
acting out behaviors (swear).

Factor structure of SSWC

We conducted an exploratory factor analysis on the SSWC category means
from the written texts pooled across the three samples. The principal factor
method was used to extract the factors. Next, a varimax (orthogonal) rotation
revealed 5 factors that accounted for 55% of the total variance: Affect Focal,
Somatization, Object Mode, Emotion Management, and Denial of Emotion.
The first factor names the attention to affect dimension, while the other four
factors pertain to different modalities of inattention to emotion. In addition to
the orthogonal rotation, an oblique rotation was performed, yielding similar
results. The factors are not significantly correlated–no inter-factor correlations
greater than .18 (which pertained to the correlation between Affect Focal and
Somatization). Orthogonal rotated factor loadings are presented in Table 3.
Factor correlations with relevant LIWC categories are presented in Table 4.
In the following paragraphs, we examine each factor and its correlation
with external measures.
Affect focal: This factor has to do with attention to affect. As shown in Table
3, it had high loadings on the following scales of SSWC, in descending order:
Affect Focal, Valence Focus, High Activation, Suffering, and Affect non-Focal.
Consistent with the construct, this factor correlated positively and significantly
with the following scales of LIWC (see Table 4): AFFECT, SENSES (sensations,
feelings), NEGEMO (negative emotions), POSEMO (positive emotions), and
attention to the here and now (PRESENT). Interestingly, Affect Focal also cor-
related positively with inhibition of thoughts and impulses (INHIB, r = .22,
p < .05). This is consistent with the Peircean framework that posits a dimen-
sion of inwardness (as evidenced by attention to affect) as necessary condition
for self control.
Somatization: This factor names the somatic consequences as a result of
inattention to emotion. It had high loadings on somatic concerns (Somatic),
and heightened self focus (Focal Self), both being states of high arousal. Sur-
prisingly however the highest loading on somatization was Low Activation (r =
.83), suggesting the possibility of motivated processing which can frequently
produce an opposite, affect-incongruent outcome (Forgas 2001). This possibil-
ity is borne out by the high loading (r = .58) of External Attribution, which with
its externally focused processing may have contributed to the incongruence
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.23 (1323-1323)

Verbal expressions of self and emotions 

Table 3. Rotated Orthogonal Factor Loadings: Study 1 (n = 121)

SSWC Affect Somatization Object Emotion Denial of

Categories Focal Mode Management Emotion
Affective Lexicon
Denial .09 .17 –.02 .06 .73
Violent words .04 .12 .03 –.53 –.16
Emotion as .26 –.06 .78 .05 –.12
event
Somatic .26 .66 .10 –.10 .32
External –.05 .58 .21 –.12 –.13
Attribution
Suffering .35 .26 .01 –.05 –.55
Affect .30 .08 –.17 .54 –.12
Non-Focal
Low Activation .03 .83 –.13 .21 .15
High Activation .63 –.14 .13 –.34 .32
Valence Focus .73 .19 .02 .13 –.08
Affect Focal .85 .04 –.03 .14 .09
Personal Pronoun
Detached Self –.18 .23 .77 –.03 –.04
Affected Self –.09 0 .40 .69 .24
Focal Self .15 .38 .03 .43 –.11
Reflexive Self –.21 –.08 .13 –.07 –.35
% Common 25.62% 22.28% 18.15% 17.60% 16.35%
Variance
Explained
Note. Boldfaced values are correlations greater than .30.

Table 4. Factor Correlations with LIWC Categories: Study 1 (n = 121)

Relevant LIWC Affect Somatization Object Emotion Denial of

Scales Focal Mode Management Emotion
PRONOUN .48**** .41****
I .20* .18* .44****
POSEMO .24** .22* .28** .27**
NEGEMO .33*** –.18*
SENSES .49**** .22*
AFFECT .50**** .32*** .24**
BODY .36****
SWEAR
INHIB .22* –.23** –.21*
PRESENT .24** .22** .41****
Note. *p < .05; **p < .01; ***p < .001; ****p < .0001, two-tailed
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.24 (1323-1368)

 Louise Sundararajan and Lenhart K. Schubert

in awareness. Kitayama and Howard have suggested that attentive processing

of target code results in accurate and efficient perception, whereas attentive
processing “misdirected” to nontarget perceptual code will result in invalid in-
formation and impairment in the “efficiency of correct perception” (1994: 47).
The mood incongruence in turn can be expected to further contribute to som-
atization. Not surprisingly, the somatization factor correlated negatively with
Levels of Emotional Awareness Scale (LEAS r = –.19, p < .05).
Correlations with LIWC: As shown in Table 4, the Somatization factor had
positive correlations with the following LIWC scales: Affect, sensations and
feelings (SENSES), positive emotions (POSEMO), the here and now perspec-
tive (PRESENT), and “I.” The correlation of this factor with positive and not
negative emotions is consistent with our speculation of motivated processing
behind somatization.
Object mode: This factor names the mode of thinking referred to as “Pensée
opératoire” in the literature of alexithymia. As shown in Table 3, it had high
loading on an immersed mode of information processing – Emotion as Event
(r = .78). Other ramifications of the object mode include objectification of
one’s experiences by looking at them from the third person perspective, thus
high loading on Detached Self (r = .77), as well as perceiving oneself as being
the object of other’s actions, consistent with moderate loading on Affected Self
(r = .40). As predicted, this factor correlated positively with TAS-20 total scores
(r = .27, p < .01), and with the subscales of Difficulty Identifying Feelings (DIF
r = .24, p < .01), and Externally-Oriented Thinking (EOT r = .25, p < .01).
Correlations with LIWC: As shown in Table 4, The factor of Object Mode
correlated positively and strongly with the use of personal pronouns (PRO-
NOUN, r = 48, p < .0001), but only modestly with the focal self (I r = .18,
p < .05). This is consistent with the finding of Masley et al. (2002) that “exter-
nally oriented thinking” were outwardly focused with a corresponding trend
toward less self focus. This factor also correlated positively and strongly with
immersion in the here and now (Present, r = .41, p < .0001). Its correlation
with affective categories were of lower magnitude by comparison (Affect r =
.24, p < .01; positive emotion r = .28, p < .01). It is interesting to note the neg-
ative correlation with inhibition (INHIB r = –.23, p < .01), which is consistent
with the claims of Peirce and others that psychological “inwardness” is needed
for self control.
Emotion management: This factor names an assortment of motivated pro-
cessing to control and limit the impact of affective experiences. The particular
version of emotion management that emerges here seems to be self control.
As shown in Table 3, this version of emotion management had high loadings
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.25 (1368-1417)

Verbal expressions of self and emotions 

on Affect Non-Focal , and on perceptions of the self both as patient, and as

agent. The negative loadings of Violent Words (r = –.53) and High Activa-
tion (r = –.34) are consistent with the profile of emotion management through
self control.
Correlations with LIWC: As shown in Table 4, this factor correlated posi-
tively and significantly with both Pronoun, and “I.” Consistent with the mo-
tivated processing of emotion management is its positive correlation with
positive emotion (POSEMO) and negative correlation with negative emotion
(NEGEMO). The negative correlation between Emotion Management and in-
hibition of thoughts and impulses (INHIB r = –.21, p < .05) is instructive: it
helps to clarify the distinction between the Emotion Management version of
self control, which is deficient in “inwardness,” and the Peircean notion of self
control which depends on psychological inwardness for its efficacy.
Denial of emotion: This factor concerns denial of emotion and its conse-
quences. It had positive loading on Denial, and negative loading on Suffering ,
which is consistent with the minimizing tendency of denial. The negative load-
ing on the Reflexive Self (r = –.35) is also consistent with our definition of
denial of emotion as lacking in reflexivity. Moderate and positive loadings on
Somatic (r = .32) and High Activation (r = .32) are consistent with the subtype
of alexithymia known as externalizing, which is characterized by arousability
on the one hand and denial of stress on the other (Masley et al. 2002). This is
borne out by the positive correlations between the Denial factor and the TAS-
20 total score (r = .21, p < .05), as well as its subscale, Externally-Oriented
Thinking (EOT r = .20, p < .05).
Correlations with LIWC: As shown in Table 4, the Denial factor correlated
positively and significantly with the LIWC scale of body states and symptoms
(BODY r = .36, p < .0001), which adds convergent validity to the alexithymia
connection mentioned above.

Study 2

Study 2 consisted of an analysis by SSWC (Sundararajan-Schubert Word

Count) and LIWC (Linguistic Inquiry and Word Count ) of 192 writing sam-
ples from a published study of psychiatric prison inmates (Richards, Beal,
Seagal & Pennebaker 2000), in which participants were randomly assigned
to experimental condition of trauma writing for three days. Sex offenders, in
comparison to non sex-offenders, were found to be more responsive to the
writing intervention, as measured by decreased infirmary visits post trauma
writing.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.26 (1417-1473)

 Louise Sundararajan and Lenhart K. Schubert

Method
Participants
Participants were a subset of 65 male psychiatric prison inmates from the orig-
inal sample of 98 inmates. With demographic information missing on two
participants, this subset had the mean age of 35.06 years, (SD = 9.38), and 12.00
years (SD = 2.13) of education completed. Of the 65 participants, 30 were serv-
ing terms for sex-related crimes (rape, etc.), whereas the other 35 were being
held on non sex-related violent crimes (murder, robbery, etc.).

Materials and Procedure Participants were randomly assigned to write for

three consecutive days about either personal traumatic experiences (n = 36)
or superficial topics (n = 29) for 20 minutes per day. These texts were analyzed
by two language analysis programs, LWC and SSWC, as in Study 1.

Results and discussion

Correlations between SSWC and LIWC
Pearson correlations between the two language analysis programs are presented
in Table 2, under “C”.
Denial of emotion: As in Study 1, no correlation with identified LIWC
variables.
Violent words: In contrast to Study 1 which showed no correlation with
relevant LIWC variables, Study 2 showed multiple correlations, possibly due
to higher productivity of this category among the prison inmate popula-
tion. A strong and positive correlation was found between Violent Words and
SWEAR (r = .74, p < .0001), as can be expected; other correlations include
Body, reference to others (OTHREF), and positive correlation with negative
emotion (NEGEMO) with corresponding negative correlation with positive
emotion (MOSEMO).
Emotion as event: Comparable correlations with LIWC categories as in
Study 1, except for the additional positive correlation with reference to others
(OTHREF), and with negative emotion.
Somatic: Similar and positive correlation with feelings and sensations
(SENSES) as in Study 1. In addition, positive correlation with SWEAR (r =
.25, p < .05). This connection between somatic concerns and verbally acting
out behaviors such as swear is well supported by the alexithymia literature. For
instance, Masley et al. (2002) reported a correlation of .30 (p < .05) between
SWEAR and TAS-20 total scores.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.27 (1473-1531)

Verbal expressions of self and emotions 

External attribution: similar positive correlations with positive emotion

(POSEMO) and AFFECT as Study 1.
Suffering: similar positive correlation with AFFECT as Study 1, with an
additional positive correlation with negative emotion (NEGEMO).
Affect non-focal: similar positive correlations with Affect and positive emo-
tion (POSEMO) as Study 1, with an additional positive correlation with nega-
tive emotion (NEGEMO).
Low activation: unlike Study 1, no correlations with identified LIWC vari-
ables.
High activation: positive correlation with negative emotion (NEGEMO) as
in Study 1.
Valence focus: positive correlation with Affect as in Study 1, with an addi-
tional positive correlation with self reference (I ).
Affect focal: similar to Study 1, with positive correlations with AFFECT,
negative emotion (NEGEMO), and self reference (I), and negative correlation
with WE. Both studies 1 and 2 suggest that Affect Focal expressions tend to
be those of negative emotions and associated with the atomic, rather than
communal, sense of the self.
Detached self : This category behaved differently than in Study 1. Instead of
showing positive correlation with reference to others (Study 1), the Detached
Self in Study 2 was positively correlated with affective categories of LIWC: Af-
fect, negative emotion (NEGEMO), and positive emotion (POSEMO). This
connection between emotion concepts (Affect) and taking the third person
perspective (the Detached Self) is consistent with the observation of Lam-
bie and Marcel (2002) that “a reflective or detached stance is an attentional
modulation that converts experiences into thoughts” (p. 223).
Affected self : similar positive correlations with self reference (I ) and other
reference (OTHREF) as in Study 1, but with a difference in emphasis: whereas
the patient sample in Study 1 showed higher correlation with self reference
than with other reference, Study 2 was the other way around. In addition, this
category in Study 2 had correlations not found in Study 1: negative emotion
(NEGEMO), Affect, and SWEAR .
Focal self : besides replicating the positive correlation in Study 1 with self
reference (I), this category in Study 2 showed three negative correlations: with
WE, reference to others (OTHREF), and body states and symptoms (BODY).
Reflexive self : Besides replicating the positive correlations in Study 1 with
We and OTHREF, this category in Study 2 showed negative correlation with
self reference (I r = –.34, p < .01), which is consistent with our construct of the
Reflexive Self as an extended rather than the atomic self. Its positive correlation
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.28 (1531-1681)

 Louise Sundararajan and Lenhart K. Schubert

with physical states (BODY r = .26, p < .05) is also consistent with the construct
of reflexivity – both the Reflexive Self and awareness of physical states have
in common an inward orientation, a correlation that helps to underline the
notion that the experiencing/reflexive self is the “embodied” self.
Overall, results in Study 1 were replicated in Study 2, except that the lat-
ter showed more correlations with negative emotion (NEGEMO), reference to
others (OTHREF), and SWEAR. Even the lack of correlation with LIWC in cer-
tain categories – Violent Words in study 1; Low Activation in Study 2; Emotion
as Event (immersion in the world of facts and action) in the (predominantly
schizophrenic) patient sample of Study 1; Suffering in the student sample of
Study 1 – seem to make sense in light of the demographics of the particular
samples. Lastly, the prison population in Study 2 was relatively more articulate
with pronouns, resulting in multiple correlations between the self concepts and
the affective categories of LIWC.

SSWC: Factor Structure and Correlations with LIWC

Exploratory factor analysis revealed 5 factors that accounted for 58% of the
variance. Orthogonal (varimax) rotation of these factors yielded the loadings
presented in Table 5. An oblique rotation yielded similar results.
These factors had correlations no greater than .15: the same modest cor-
relation (r = .15) between Somatization and Emotion Management was main-
tained as in Study 1. Factor correlations with relevant LIWC categories are pre-
sented in Table 6, and will be examined along with each factor in the following
discussion.
Affect focal: Same high loadings as in Study 1 on Affect Focal, (r = .78)
and Valence Focus (r = .77), and moderate loading on Affect Non-Focal (r =
.28). But this factor in Study 2 was the mirror image of that in Study 1. Study
2 presented a detached mode of attention to affect–evidenced by high loading
on the Detached Self (r = .73), and a significant and negative loading on High
Activation (r = –.21) – as opposed to the immersed mode of its counterpart in
Study 1 with high loadings on High Activation and Suffering. The two studies
are mirror images of each other in another sense: consistent with the significant
and positive correlation in Study 1 of Affect Focal with the inhibition category
in LIWC (INHIB, r = .22, p < .05), the same factor in Study 2 had a modest
and negative loading on Violent Words (r = –.18).
Correlations with LIWC scales: As shown in Table 6, this factor had similar
positive correlations with the affective scales of LIWC (AFFECT, NEGEMO,
and POSEMO) as in Study 1. In addition, it showed positive correlation with
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.29 (1681-1681)

Verbal expressions of self and emotions 

Table 5. Rotated Orthogonal Factor Loadings: Study 2 (n = 65)

SSWC Affect Object Somatization Emotion Denial of

Categories Focal Mode Management Emotion
Affective Lexicon
Denial .16 .01 –.10 –.09 .84
Violent words –.18 .57 .22 –.18 .36
Emotion as .08 .64 –.26 .24 –.12
Event
Somatic –.13 .15 .56 .09 .35
External –.14 –.21 .13 .78 .08
Attribution
Suffering .13 .05 0 .68 –.10
Affect .28 .25 –.25 .55 0
Non-Focal
Low Activation .12 –.01 –.08 –.03 –.47
High Activation –.21 .44 .45 .04 –.16
Valence Focus .77 –.20 .27 –.07 –.14
Affect Focal .78 .30 .17 .09 –.14
Personal Pronoun
Detached Self .73 –.05 –.15 .15 .13
Affected Self .09 .88 0 –.08 .08
Focal Self .21 .02 .67 –.30 –.06
Reflexive Self –.18 .21 –.63 –.12 –.02
% Common 23.58% 23.14% 19.51% 18.83% 14.94%
Variance
Explained
Note. Boldfaced values are correlations greater than .30.

Table 6. Factor Correlations with LIWC Categories: Study 2 (n = 65)

Relevant LIWC Affect Object Somatization Emotion Denial of

Scales Focal Mode Management Emotion
PRONOUN .26* .68****
I .27* .68****
POSEMO .26* .43***
NEGEMO .33** .52**** .32**
SENSES .27* .29*
AFFECT .44*** .36** .55****
BODY –.30*
SWEAR .42*** .26* .30*
INHIB
PRESENT .43*** .33**
Note. *p < .05; **p < .01; ***p < .001; ****p < .0001, two-tailed
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.30 (1681-1730)

 Louise Sundararajan and Lenhart K. Schubert

Pronoun use (r = .26, p < .05), and negative correlation with body states and
symptoms (Body r = –.30, p < .05).
Object mode: Same high loadings as in Study 1 on Emotion as Event , and
Affected Self . Negative connotations of the Object Mode are more accentuated
in Study 2, as evidenced by high loadings on Violent Words (r = .57), High
Activation (r = .44), and moderate loading on Affect Focal expressions (r =
.30), which consisted primarily of negative emotions as correlations with LIWC
make clear, below.
Correlations with LIWC scales: As shown in Table 6, this factor had similar
pattern of correlations with Pronoun, I, Affect, and Present as in Study 1. In
addition, it had a positive correlation with SWEAR (r = .42, p < .001). Masley
et al. (2002) found a comparable positive correlation (.42, p < .01) between
SWEAR and the DIF (Difficulty Identifying Feelings) subscale of TAS-20, sug-
gesting the possibility of an overlap between DIF and Object Mode. This is
consistent with the results of Study 1, which showed multiple positive correla-
tions between the Object Mode and TAS-20 scores, including DIF. The positive
correlation between Object Mode and SWEAR is the direct corollary of the
negative correlation between this factor and inhibition (INHIB r = –.23, p <
.01) in Study 1. The two studies differed in the valence of this factor: Study 2
had positive correlation with negative emotion (NEGEMO r = .52, p < .0001),
whereas Study 1 had positive correlation with positive emotion (POSEMO r =
.28, p < .01).
Somatization: Same high loadings as in Study 1 on Focal Self, and So-
matic. Negative loading on Reflexive Self (r = .–.63) is consistent with our
hypothesis that somatic expressions of emotions belong to the reactive con-
sciousness of Secondness in contrast to the reflexive consciousness of Third-
ness, which is characteristic of the Reflexive Self. Also consistent with the usual
co-occurrence of somatic concerns and high arousal is the high loading on
High Activation (r = .45).
Correlations with LIWC scales: As shown in Table 6, correlations with I,
SENSES, and Present were essentially the same as in Study 1. The only differ-
ence is that Study 1 had positive correlation with positive emotion, whereas
Study 2 showed positive correlation with SWEAR. This is consistent with the
Low Activation loading in Study 1 in contrast to High Activation loading of
Study 2. The lack of significant correlation between Somatization and Body
in both studies 1 (Table 4) and 2 (Table 6) is instructive: in spite of consider-
able overlap between the two concepts, the Somatic in SSWC and the BODY in
LIWC differ in emphasis – the former accentuates the psychosomatic aspects of
the physical state such as “heart racing,” whereas the latter, the less pathological
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.31 (1730-1796)

Verbal expressions of self and emotions 

and more neutral physical states, such as “cough.” This is consistent with the
fact that the Reflexive Self was correlated positively with the BODY (Table 2),
but negatively with Somatization (Table 5).
Emotion management: Very similar loading on Affect Non-Focal as in
Study 1. In contrast to the self control version of emotion management in Study
1, Study 2 presented an externalizing coping style, in which one takes the vic-
tim stance by shifting the appraisal of blameworthiness from the self to the
affecting environment. Thus high loadings on External Attribution (r = .78)
and Suffering (r = .68), and corresponding negative loading on Focal self (r =
–.30).
Correlations with LIWC scales: As shown in Table 6, Study 2 shared with
Study 1 the positive correlation with positive emotion. The two studies cor-
related differently with the negative emotion category (NEGEMO), consistent
with their difference in coping strategies: Study 1, which profiled self control,
showed negative correlation with NEGEMO (r = –.18, p < .05), whereas Study
2, which profiled a victim stance, showed positive correlation with the same
(r = .32, p < .01). The fact that Emotion Management more than Affect Fo-
cal had relatively higher correlations with Affect (r = .55, p < .0001) and its
subscales of POSEMO and NEGEMO requires explanation. Our definition of
Affect Focal is not based on the affective connotation of verbal expressions, so
much as on the dimension of reflexivity (awareness of awareness). Thus “I am
sad” is an Affect Focal statement, because of its implication of self avowal (I
own my sadness), whereas the statement “it is tragic” falls under the category
of External Attribution (which had high loading on Emotion Management),
because of the lack of self avowal. But both statements are replete with affec-
tivity, hence would score equally well on the LIWC categories of Affect and
negative emotion (NEGEMO).
Denial of emotion: Same high loadings as in Study 1 on Denial, and So-
matic. It’s negative loading on Low Activation (r = –.47) is compatible with the
positive loading in Study 1 on High Activation (r = .32). One additional fla-
vor that Study 2 contributed to this factor was the positive loading on Violent
Words (r = .36).
Correlations with LIWC scales: As shown in Table 6, correlation with feel-
ings and sensations (SENSES) is compatible with the correlation with BODY
in Study 1. Positive correlation with SWEAR (r = .30, p < .05) adds another
ramification to the denial of emotions.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.32 (1796-1828)

 Louise Sundararajan and Lenhart K. Schubert

General discussion

Over 700 writing samples by 186 participants were collected, and analyzed
with the SSWC, from two separate studies: Of the 3 samples in Study 1, the
professionals (n = 10) produced a mean of 1179.20 words (SD = 334.23); the
psychiatric inpatients (n = 42) produced less than half the amount (Mean =
577.98, SD = 208.72); and college students (n = 69) wrote slightly more than
patients, with a Mean of 662.21 words (SD = 204.03). The prison inmates in
Study 2 (n = 65) wrote more than the students, but less than the profession-
als, with a Mean of 763.15 words ( SD = 553.68). Despite the heterogeneity
of the writing samples, and despite even larger differences in demographics
and research methodology, the two studies revealed essentially the same factor
structure that organized the semantic space of the texts analyzed by SSWC. In
both studies the loadings of the SSWC categories were consistent with expec-
tations, and all items loaded above .30 with their appropriate factor. The high
number of statistically significant correlations across all samples with the af-
fective categories of the standard language analysis program, LIWC, confirms
our hypothesis that all SSWC categories fall within the purview of the affective
domain. Supporting discriminant validity of the construct is one important
exception to the rule – the scale of Denial of Emotion had no correlations with
LIWC’s affective categories in either study.
Furthermore, the factor patterns in the two studies were mirror images
of one another in a twofold sense, one study either dovetailed the other or
was its antipode and complement. On Object Mode and Denial of Emotion,
arousal and disinhibition were the two flavors that were implicit in one study
and explicit in the other. As for the rest of the factors, the two studies presented
variations that fell along the divide of internal versus external focus, a divide
fundamental to cognitive attention (Lambie & Marcel 2002), and probably has
a neurophysiological basis (Lane et al. 1997; Gusnard et al. 2001). Thus along
the internal focus axis was the immersed version of Affect Focal (Study 1), the
high arousal version of Somatic (Study 2), and the self control version of Emo-
tion Management (Study 1); and along the external focus axis, the detached
version of Affect Focal (Study 2), the low arousal version of Somatic (Study 1),
and the externalizing version of Emotion Management (Study 2). These varia-
tions of the themes were consistent with the finding that the inmate population
of Study 2 used relatively more externally focused strategies in processing emo-
tional information, expressed more negative emotions, and used more swear
words than the non-criminal populations of Study 1.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.33 (1828-1982)

Verbal expressions of self and emotions 

Can the Peircean semiotics be translated into a diagnostic tool of individ-

ual differences? Peirce is encouraging when he claims that “the word or sign
which man uses is the man himself. . . . Thus my language is the sum total
of myself ” (1931–1958, Vol.5, paragraph 314, emphasis in the original). A
similar sentiment has been expressed by Pennebaker and King (1999), who
demonstrated empirically that “language use is a reliable individual difference”
(p. 1300). In the next section, we explore these possibilities by examining the
group differences in Study 2.

. Application to individual differences

To recapitulate the context for Study 2, the study of Richards et al. (2000)
found that, as measured by decrease in infirmary visits after the writing in-
tervention, the sex-offenders showed more health improvement than the non
sex-offenders in response to trauma writing for 3 consecutive days. Can this be
explained by group differences? To find out how the two groups differed, we
conducted oneway ANOVAs on the SSWC categories. The contrasts compared
the trauma writing condition with the trivial writing, and sex offenders versus
non sex-offenders.

Comparison between conditions

Results for the comparison between writing conditions are presented in Table 7.
As shown in Table 7, the Experimental and Control conditions differed along
the following SSWC dimensions:
Affected self : Both sex-offenders and non sex-offenders had relatively more
output on this variable in the experimental than the control condition.
Affect focal: Overall, the experimental condition generated more output on
this variable than the control condition; this is especially true with sex offenders
who had more output on this variable in the experimental condition .
Violent words: Overall, both groups had more output on this variable in the
experimental than the control condition.
The increase in victim role taking (Affected Self), affect-focal words, and
violent and swear words in the experimental condition, in comparison to
the control condition, is to be expected, especially with the prison inmate
population.

Comparison between groups

Results on group comparison are presented in Table 8.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.34 (1982-1991)

 Louise Sundararajan and Lenhart K. Schubert

Table 7. Difference in writing between experimental and control conditions as mea-

sured by SSWC: Study 2 (n = 65)

Experimental Control
SSWC Categories N Mean SD N Mean SD
Affected Self
Main Effect 36 3.23 1.68 29 1.48 1.55 F(1,63) = 18.56 ****
Non Sex Offender 17 3.30 2.02 18 1.55 1.49 F(1,33) = 8.46**
Sex Offender 19 3.17 1.36 11 1.36 1.72 F(1,28) = 10.14**
Affect Focal
Main Effect 36 1.26 0.68 29 0.85 0.78 F(1,63) = 5.22*
Non Sex Offender 17 1.10 0.72 18 0.90 0.70 F(1,33) = 0.68 ns
Sex Offender 19 1.41 0.63 11 0.77 0.93 F(1,28) = 5.17*
Violent Words
Main effect 36 0.48 0.50 29 0.19 0.41 F(1,63) = 6.62 **
Non Sex Offender 17 0.54 0.54 18 0.21 0.49 F(1,33) = 3.52†
Sex Offender 19 0.43 0.47 11 0.14 0.24 F(1,28) = 3.58 †

Note. † p < .07; *p < .05; **p < .01; ***p < .001; ****p < .0001

Table 8. Difference in writing between sex offenders and non sex offenders as measured
by SSWC: Study 2 (n = 65)

Non Sex Offender Sex Offender

SSWC Categories N Mean SD N Mean SD
Emotion as Event
Main Effect 35 .57 .55 30 .22 .27 F(1,63) = 9.97 **
Experimental 17 .74 .58 19 .23 .20 F(1,34) = 13.29 ***
Control 18 .40 .47 11 .20 .38 F(1,27) = 1.34 ns
External Attribution
Main Effect 35 1.17 .73 30 .93 .51 F(1,63) = 2.30 ns
Experimental 17 1.08 .71 19 1.08 .50 F(1,34) = 0.00 ns
Control 18 1.26 .76 11 .67 .44 F(1,27) = 5.39 *
Valence Focus
Main effect 35 .12 .25 30 .29 .40 F(1,63) = 4.34 *
Experimental 17 .05 .09 19 .31 .35 F(1,34) = 8.79 **
Control 18 .19 .34 11 .26 .50 F(1,27) = 0.24 ns
Note. *p < .05; **p < .01; ***p < .001

As shown in Table 8, the non sex- and sex-offenders differed along the follow-
ing SSWC dimensions:
Emotion as event: the non sex offenders had relatively more output on this
variable in general, and especially in the experimental condition.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.35 (1991-2052)

Verbal expressions of self and emotions 

External attribution: The non sex offenders had relatively more output
on this variable only in the control condition, whereas in the experimental
condition the two groups did not differ significantly.
Valence focus: Sex offenders had relatively more output on this variable in
general, and especially in the experimental condition.
In sum, the non sex-offenders seemed to have a baseline (the control condi-
tion) that was relatively more externally oriented in their referential focus, and,
in the experimental condition, seemed to operate in an immersed mode in pro-
cessing emotionally laden information, as evidenced by their increased output
on Emotion as Event and Violent Words. The sex offenders, in contrast, tended
to be more valence-focused in their articulation of affect. Even though in the
experimental condition, sex-offenders increased verbal expressions of a victim
stance as readily as the non sex-offenders, the former’s baseline was to make rel-
atively less external attributions than the latter. Otherwise put, as evidenced by
their relatively less output on categories that concerned the world of facts and
actions, the sex offenders seemed to be less inclined to process emotional infor-
mation in the immersed mode of Secondness. Instead, they outperformed the
non sex offenders in production of affective lexicon that had the potential for
being sufficiently complete signs (Thirdness), such as Affect Focal and Valence
Focus terms.
So, why did the sex-offenders benefit from the trauma writing more than
the non sex-offenders? Richards et al. (2000) speculated that due to social
stigma the sex-offenders might be more inhibited than the non sex-offenders,
and hence benefited more from the disclosure writing paradigm. Our findings
offer a complementary interpretation in terms of individual differences: the
sex offenders might benefit more from social interventions, because of their
valence focus, which was found to be related to “emotional responsivity to the
social environment” (Barrett 1998: 595). Yet, there is much more information
that can be gleaned from the language analysis.
One important source of information on individual differences is linguistic
profiles. An overview of their respective performance on the 15 SSWC vari-
ables gives us contrasting profiles of the two groups – the sex- versus non
sex-offenders, as presented in Figures 1 and 1a.
As Figures 1 shows graphically, when processing emotional information,
the non sex-offenders tended to use incomplete signs – signs higher on the
hierarchy of inefficiency (to the left of the SSWC category axis in Figure 1),
whereas the sex offenders used sufficiently complete signs – such as Affect Fo-
cal and Valence Focus words – at relatively higher frequency. This leads to the
interesting question as to whether sufficiently complete and incomplete signs
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.36 (2052-2103)

 Louise Sundararajan and Lenhart K. Schubert

1.50
1.25
1.00
Percent

0.75
0.50
0.25
0.00
1 2 3 4 5 6 7 8 9 10 11
Category

Non Sex-Offenders (n=35) Sex Offenders (n=30)

Note. SSWC categories: 1. Denial; 2. Violent Words; 3. Emotion as Event; 4. Somatic; 5. External Attribution;
6. Suffering; 7. Affect Non-Focal; 8. Low Activation; 9. High Activation; 10. Valence Focus; 11. Affect-Focal.

Figure 1. Sex- and non sex-offenders’ mean scores on SSWC categories (affective lexi-
con) (Study 2)

are differentially promoted or inhibited by different individuals or groups.

In Study 2 the difference between health improvement after trauma writing
seemed to coincide with this differential use of signs: sex offenders who were
more responsive to the writing intervention were also the ones who promoted
sufficiently complete signs, as evidenced by relatively higher frequency, and in-
hibited incomplete signs, as evidenced by relatively lower frequency, in their
trauma writings, whereas non sex-offenders who were relatively less responsive
to the intervention showed the reverse pattern of sign use.
Another approach to the question concerning individual differences in sign
use is to examine the variability of the text from day to day, instead of averag-
ing the means across days as we have done so far. Indeed, from the Peircean
perspective, the single most important indicator of the quality of a sign is
its variability. Variability has to do with the fact that “it is of the very nature
of thought to grow” (Colapietro 1989: 67). Or in the words of Peirce, “The
highest kind of symbol is one which signifies a growth, or self-development”
(1931–1958, Vol. 4, paragraph 10). One way to approach the variability ques-
tion is to understand it as an index of cognitive flexibility and to measure
accordingly the change in content of texts produced sequentially (Campbell &
Pennebaker 2003). Another alternative is structural instead of content analysis.
From the structural perspective, variability can be indexed by the fluctuation
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.37 (2103-2137)

Verbal expressions of self and emotions 

8.00

6.00
Percent

4.00

2.00

0.00
12 13 14 15
Category
Non Sex-Offenders (n=35) Sex Offenders (n=30)

Note. SSWC categories: 12. Detached Self; 13. Affected Self; 14. Focal Self; 15. Reflexive Self.

Figure 1a. Sex- and non sex-offenders’ mean scores on SSWC categories (personal
pronoun) (Study 2).

in frequency of a sign. This variability in frequency has to do with the reflex-

ive nature of thought: since thought is “dialogical” (thought talking to itself),
sign use involves a dynamic process of thought inhibiting and/or promoting
thought, resulting in fluctuation in the frequency of a sign. Figure 2 shows how
this structural approach to variability can be illuminating.
Figure 2 shows a comparison of the day to day fluctuation of the frequency
of Affect Focal words between sex- and non sex-offenders in the experimental
condition. Consistent with our prediction of the connection between reflexive
thought and variability in verbal expressions of emotions, relatively more fluc-
tuation was exhibited by the sex offenders than the non sex offenders. The non
sex-offenders’ relative lack of fluctuation in sign use is indicative of a stagnant
thought process characteristic of closed information search strategies (Fogas
2001). This is consistent with our speculation of the possibility of turning po-
tentially complete signs (symbols), such as Affect Focal words, into incomplete
ones, such as icons (dictionary entries), as a result of direct access to seman-
tic memory (Fogas 2001). Thus the question concerning alexithymia related
phenomena is not necessarily what – can the individual produce affect focal
words, so much as how – the way an affective term is used, by direct access
to semantic memory or substantive processing of emotional information (For-
gas 2001)? Due to limitation of space, fluctuations on other SSWC variables
will be examined elsewhere (Sundararajan 2004, July). Suffice it here simply
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.38 (2137-2171)

 Louise Sundararajan and Lenhart K. Schubert

1.5
Percent

0.5
1 2 3
Day

Non Sex-Offenders (n=17) Sex Offenders (n=19)

Figure 2. Sex- and non sex-offenders’ mean scores on the Affect Focal category of
SSWC for each day of the three-day traumata writing by psychiatric prison inmates
(Study 2).

to offer for future research the testable hypothesis that the variables of SSWC
change as a function of the individual’s cognitive style in processing emotional
information.

. Summary and conclusion

In this paper we proposed and tested empirically a taxonomy of verbal ex-

pressions of self and emotions. Based on the semiotics of Charles Peirce, our
proposition states that emotion and self expressions are signs with varying de-
grees of liability to fail to traverse the triadic path of their development, and
thus can be categorized accordingly. We further developed a computerized lan-
guage analysis program, SSWC, that translated this taxonomy into 15 or so
scales of pronouns and affective lexicon. Then we conducted two empirical
studies of SSWC to test the construct validity of our taxonomy and explore its
potential contributions to research on individual differences. Preliminary find-
ings suggest that the categories of SSWC mapped out the affective domain in
semantic space in a logically consistent and intelligible manner. Demonstrating
good internal consistency for these categories, factor analysis revealed 5 factors
with a factor pattern that had simple structure and conceptual meaningful-
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.39 (2171-2209)

Verbal expressions of self and emotions 

ness, and that was replicable across independent samples. Reasonable external
validity was also demonstrated with moderate correlations with TAS-20 and
its subscales, and robust correlations with the standard language analysis pro-
gram, LIWC. To explore possibilities of its application to research in individual
differences, we used SSWC to create linguistic profiles of the writing samples,
and demonstrated that our taxonomy rendered visible subtle differences in in-
formation processing strategies between comparison groups. In conclusion,
our semiotics based taxonomy has the potential to extend our understanding
of certain key concepts in the alexithymia literature – such as “lack of intro-
spection” or “externally-oriented thinking” – beyond the three factor structure
of the alexithymia construct, by offering a coherent theoretical framework
and a more fine grained analysis of language use that can give us more nu-
anced information of individual differences in their verbal expressions of self
and emotions.

Acknowledgements

We thank the following individuals for their help with Study 1:

Deborah Hall, Ph.D., Chair, IRB; Chief Psychologist, Rochester Psychiatric
Center; Professors K. R. Sundararajan and Craig Zuckerman of St. Bonaventure
University; Professor Susan F. Roxin of Nazareth University; Emily Krenichyn,
M. S., Nazareth University; Jeffrey Richards, MA., University of Colorado;
Mary M. Fox, Ph.D., University of Rochester; Mary Swift, Ph.D., University
of Rochester.
We thank Professor James W. Pennebaker for providing data for our anal-
ysis in Study 2.

References

Averill, J. R. (1975). A semantic atlas of emotion concepts. JSAS Catalogue of Selected

Documents in Psychology, 5, 330 (Ms. No. 421).
Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994). The twenty-item Toronto Alexithymia
Scale: I. Item selection and cross-validation of the factor structure. Journal of
Psychosomatic Research, 38, 23–32.
Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994). The twenty-item Toronto Alexithymia
Scale: II. Convergent, discriminant, and concurrent validity. Journal of Psychosomatic
Research, 38, 33–40.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.40 (2209-2313)

 Louise Sundararajan and Lenhart K. Schubert

Barrett, L. F. (1998). Discrete emotions or dimensions? The role of valence focus and arousal
focus. Cognition and emotion, 12 (4), 579–599.
Barrett, L. F., & Russell, J. A. (1998). Independence and bipolarity in the structure of current
affect. Journal of Personality and Social Psychology, 74 (4), 967–984.
Bucci, W. (1995). The power of the narrative: A multiple code account. In J. W.
Pennebaker (Ed.), Emotion, disclosure, & health (pp. 93–124). Washington, DC:
American Psychological Association.
Campbell, R. S., & Pennebaker, J. W. (2003). The secret life of pronouns: Flexibility in
writing style and physical health. Psychological Science, 14 (1), 60–65.
Clore, G. L., Ortony, A., & Foss, M. A. (1987). The psychological foundations of the affective
lexicon. Journal of Personality and Social Psychology, 53 (4), 751–766.
Colapietro, V. M. (1989). Peirce’s approach to the self/A semiotic perspective on human
subjectivity. Albany, NY: State University of New York.
Conte, H. R., Plutchik, R., Jung, B. B., Picard, S., Karasu, T. B., & Lotterman, A. (1990).
Psychological mindedness as a predictor of psychotherapy outcome: A preliminary
report. Comprehensive psychiatry, 31 (5), 426–431.
Dorsey, J. M. (1971). Psychology of emotion: Self discipline by conscious emotional continence.
Detroit, MI: Wayne State University.
Feldman, L. A. (1995). Valence focus and arousal focus: Individual differences in the
structure of affective experience. Journal of Personality and Social Psychology, 69 (1),
153–166.
Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private self-consciousness:
Assessment and theory. Journal of consulting and clinical psychology, 43 (4), 522–527.
Fisch, M. H. (Ed.). (1982). Writings of Charles S. Peirce: A chronological edition (5 Vols).
Bloomington: Indiana University.
Fonagy, P., Steele, M., Steele, H., Moran, G. S., & Higgitt, A. C. (1991). The capacity for
understanding mental states: the reflective self in parent and child and its significance
for security of attachment. Infant Mental Health Journal, 12 (3), 201–218.
Forgas, J. P. (2001). Affect and information processing strategies: An interactive relationship.
In J. P. Forgas (Ed.), Feeling and thinking/The role of affect in social cognition (pp. 253–
282). Cambridge: Cambridge University.
Frijda, N. H. (in press). Emotion experience. Forthcoming in Cognition and Emotion.
Gallagher, S., & Marcel, A. J. (1999). The self in contextualized action. Journal of Con-
sciousness Studies, 6 (4), 4–30.
Gendlin, E. T. (1997). Experiencing and the creation of meaning/A philosophical and
psychological approach to the subjective. Evanston, IL: Northwestern University.
Greco, M. (1998). Illness as a work of thought. London: Routledge.
Greco, M. (2001). Inconspicuous anomalies: Alexithymia and ethical relations to the self.
Health, 5 (4), 471–492.
Gusnard, D. A., Akbudak, E., Shulman, G. L., & Raichle, M. E. (2001). Medial prefrontal
cortex and self-referential mental activity: Relation to a default mode of brain function.
PNAS (Proceedings of the National Academy of Sciences), 98 (7), 4259–4264.
Herpertz, S. C., & Sass, H. (2000). Emotional deficiency and psychopathy. Behavioral
Sciences and the Law, 18, 567–580.
Hoopes, J. (Ed.). (1991). Peirce on signs. Chapel Hill, NC: The University of North Carolina.
JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.41 (2313-2412)

Verbal expressions of self and emotions 

Johnson-Laird, P. N., & Oatley, K. (1989). The language of emotions: An analysis of a

semantic field. Cognition and Emotion, 3 (2), 81–123.
Kitayama, S., & Howard, S. (1994). Affective regulation of perception and comprehension:
Amplification and semantic priming. In S. Kitayama & S. Howard (Eds.), The heart’s
eye: Emotional influences in perception and attention (pp. 41–65). New York: Academic.
Krystal, H. (1988). Integration and self-healing: Affect, trauma, alexithymia. Hillsdale, NJ:
Analytic Press.
Krystal, J. H., Giller, E. L. Jr., & Cicchetti, D. V. (1986). Assessment of alexithymia in
posttraumatic stress disorder and somatic illness: Introduction of a reliable measure.
Psychosomatic Medicine, 48 (1/2), 84–94.
Lambie, J. A., & Marcel, A. J. (2002). Consciousness and the varieties of emotion experience:
A theoretical framework. Psychological Review, 109 (2), 219–259.
Lane, R. D. (1991). LEAS scoring manual & glossary. Department of Psychiatry, University
of Arizona Health Sciences Center.
Lane, R. D., Quinlan, D., Schwartz, G.E., Walker, P., & Zeitlin, S. (1990). The levels of
emotional awareness scale: A cognitive-developmental measure of emotion. Journal of
Personality Assessment, 55, 124–134.
Lane, R. D., Fink, G. R., Chau, P. M. L., & Dolan, R. J. (1997). Neural activation during
selective attention to subjective emotional responses. NeuroReport, 8, 3969–3972.
Lundh, L.-G, Johnsson, A, Sundqvist, K, Olsson, H. (2002). Alexithymia, memory of
emotion, emotional awareness, and perfectionism. Emotion, 2 (4), 361–379.
Marty, P., & de M’Uzan, M. (1963). La ‘Pensée opératoire’. Revue Francaise de Psychanalyse,
27 (suppl.), 1345–1356.
Masley, C., Lumley, M. A., D’Souza, P., & Roberson, T. (March, 2002). Alexithymia and
language use in written disclosure: The importance of subscale analysis. Presented at the
meeting of the American Psychosomatic Society, Barcelona, Spain. Abstract published
in Psychosomatic Medicine, 64, 102.
Ortony, A., Clore, G., & Foss, M. A. (1987). The referential structure of the affective lexicon.
Cognitive Science, 11, 341–364.
Ortony, A., Clore, G., & Collins, A. (1988). The cognitive structure of emotions. Cambridge:
Cambridge University.
Peirce, C. S. (1931–1958). Collected papers of Charles Sanders Peirce (8 Vols, edited by C.
Hartshorne & P. Weiss). Cambridge: Harvard University.
Pennebaker, J. W. (1989). Stream of consciousness and stress: Levels of thinking. In J. S.
Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 327–350). New York: Guilford.
Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an individual
difference. Journal of Personality and Social Psychology, 77 (6), 1296–1312.
Pennebaker, J. W., Francis, M. E., & Booth, R. J. (2001). Linguistic inquiry and word
count/LIWC2001. Mahwah, NJ: Erlbaum.
Richards, J. M., Beal, W. E., Seagal, J. D., & Pennebaker, J. W. (2000). Effects of disclosure
of traumatic events on illness behavior among psychiatric prison inmates. Journal of
Abnormal Psychology, 109 (1), 156–160.
Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: An accessibility model of
emotional self-report. Psychological Bulletin, 128, 934–960.
 JB[v.20020404] Prn:23/11/2004; 10:26 F: Z13111.tex / p.42 (2412-2486)

 Louise Sundararajan and Lenhart K. Schubert

Robinson, M. D., Solberg, E. C., Vargas, P., & Tamir, M. (2003). Trait as default: Extraversion,
subjective well-being, and the distinction between neutral and positive events. Journal
of Personality and Social Psychology, 85, 517–527.
Rosen, S. M. (2004). Dimensions of apeiron/A topological phenomenology of space, time, and
individuation. Amsterdam: Rodopi.
Silvia, P. J., & Gendolla, G. H. E. (2001). On introspection and self-perception: Does self-
focused attention enable accurate self-knowledge? Review of General Psychology, 5 (3),
241–269.
Sundararajan, L. (2001). Alexithymia and the reflexive self: Implications of congruence
theory for treatment of the emotionally impaired. The Humanistic Psychologist, 29
(1–3), 223–248.
Sundararajan, L., & Schubert, L. K. (2002). The Externalizing Scale: A pattern matching
word count program. Computer Science Department, University of Rochester.
Sundararajan, L., & Schubert, L. K. (2003, August). “Lack of introspection” in alexithymia:
Psycholinguistic markers. In L. Sundararajan (chair), Alexithymia – new developments
in theory and research. Symposium conducted at the 111th Annual Convention of the
American Psychological Association, Toronto.
Sundararajan, L. (2004, July). The anatomy of change in writings about trauma: A
Psycholinguistic analysis. Poster to be presented at the Annual Conference of the
International Society for Research on Emotions, New York.
Taylor, G. J., Bagby, R. M., Parker, & J. D. A. (1997). Disorders of affect regulation/Alexithymia
in medical and psychiatric illness. Cambridge: Cambridge University.
Taylor, G. J., & Bagby, R. M. (2000). An overview of the alexithymia construct. In R. Bar-
On & J. Parker (Eds.), Handbook of Emotional Intelligence (pp. 40–67). San Francisco:
Jossey-Bass.
Taylor, G. J. (2000). Recent developments in alexithymia theory and research. Canadian
Journal of Psychiatry, 45, (2), 134–142.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS Scales. Journal of Personality and
Social Psychology, 54 (6), 1063–1070.
Wiley, N. (1994). The semiotic self. Chicago: University of Chicago.
Williamson, S., Harpur, T. J., & Hare, R. D. (1991). Abnormal processing of affective words
by psychopaths. Psychophysiology, 28 (3), 260–273.
JB[v.20020404] Prn:5/11/2004; 14:19 F: Z131P3.tex / p.1 (42-68)

P III

Agency and moral value

JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.1 (41-107)

Apt affect
Moral concept mastery and the phenomenology
of emotions

Elisa A. Hurley
Georgetown University

It is now widely agreed that being a good moral agent requires being
emotionally attuned to the world. The claim that emotions play such a role
typically centers around their cognitive components. The phenomenal
elements of emotions, their distinct feelings, are either relegated to
supporting roles, on the one hand, or explained away, on the other. It is my
suggestion that these standard options for understanding the epistemic role
of emotional feelings are unjustifiably impoverished, and that a proper grasp
of the emotions’ phenomenal elements suggests that the latter play a critical
and independent epistemic role in our practices of moral evaluation.
Following some strands of post-Wittgensteinian philosophy of mind and
epistemology, I identify the single misunderstanding at the root of these two
impoverished views, and suggest an alternative in which the phenomenal
elements are understood as concept-involving and therefore norm-governed.
I argue that if the phenomenal constituents of emotions are so understood,
then it becomes plausible to claim that there is an epistemically privileged
connection between an agent’s capacity for phenomenal emotional responses
and her mastery of moral concepts.

Keywords: Emotions, feelings, affect, norms, holism, cognitivism, moral

concepts, moral epistemology, Wittgenstein, evaluative judgments

It is by now commonplace to suggest that the emotions play a significant role

in moral epistemology. These days, only the most orthodox (and perhaps mis-
guided) student of Kant would deny that at least part of what it is to be a good
moral agent is to be emotionally attuned to the world in the right sorts of ways.
The claim that emotions play such a role typically centers around their cogni-
tive components, which are identified as either evaluative judgments about, or
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.2 (107-149)

 Elisa A. Hurley

evaluatively loaded construals of, the world. (Calhoun 1984, 2003; Elgin 1996;
Greenspan 1989; Gordon 1987; Nussbaum 2001; Solomon 1976/1983; Sorabji
2000). The phenomenal elements of emotions, their distinct feelings, are either
relegated to supporting roles as contingent, causal capacities which often hap-
pen to “signal” circumstances of moral salience, or explained away by reducing
them to the endorsement of distinct propositional contents. It is my suggestion
that these standard options for understanding the epistemic role of emotions
are unjustifiably impoverished, and that a proper grasp of the emotions’ phe-
nomenal dimensions suggests that they might play a critical and independent
epistemic role in our practices of moral evaluation.1
In what follows I will briefly develop these two impoverished views, and
suggest that both ultimately stem from a single source: a certain (mis)under-
standing of the emotions’ phenomenal elements which entails that they can
play only a contingent, causal role in our interactions with the world. Bor-
rowing familiar arguments from other areas of philosophy of mind and epis-
temology, I will diagnose the problem with this understanding and suggest
an alternative in which the phenomenal elements are properly understood as
situated in the space of reasons, that is, as conceptual and therefore norm-
governed. I will argue that if the phenomenal constituents of emotions are
so understood, then it is plausible to suggest that there is an epistemically
privileged connection between an agent’s capacity for phenomenal emotional
responses and her mastery of moral concepts.2 I will conclude by proposing
that emotional affect – i.e., the phenomenal dimension of an emotion – so
understood, might function epistemically in two distinct ways: as noninferen-
tial entitlement to certain value judgments and as a uniquely placed capacity
for discerning instances in which moral concepts apply. What I hope to show,
then, is that being a good moral agent requires being not just emotionally, but,
more precisely, affectively engaged in the world in the right sorts of ways.
Colloquially, we might say that the phenomenal elements in question are
the “oomphings” that give our emotions a certain affective “flavoring.” They
are what it feels like to be angry, sad, proud, ashamed, scared and the like. Now,
before moving on with the argument, it is important to pause and address the
question of whether all emotions carry such a phenomenal “feel,” or whether
we think they must in order to be identified as emotions in the first place. While
it is not my task in this paper to give the necessary and sufficient conditions for
the presence of an emotion state, I do want to offer the following by way of
answer to this question. The first is that the phenomenology I am speaking of
is not a physiological or bodily disturbance, although such disturbances may
in many cases of emotion be part of that emotion’s phenomenology. So the
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.3 (149-198)

Apt affect 

suggestion that emotions have a phenomenological feel all their own is not
the demand that each and every emotion be accompanied by some identifi-
able bodily state that is introspectively of otherwise available to the agent. My
suggestion is rather that it is typical of emotions that there is something that
it feels like to have them, in the sense that the agent is conscious of feeling a
certain way.
Secondly, I certainly want to leave open the possibility that, on occasion,
we might accurately be said to be having an emotion without experiencing it
phenomenologically even in this latter sense, that is, without registering the
presence of an emotion state at all. In other words, I think we ought to say
that, ceteris paribus, emotions have a phenomenal feel or affective experiential
quality, with the understanding that those emotions states which do not involve
phenomenal feelings or experiences are conceptually parasitic upon those cases
in which there is something that it feels like to be in that state. It is the latter
which are paradigmatic emotion states; there is, therefore, some privileged or
essential connection between being in an emotion state and that state’s having
a particular affective feeling or flavor, even if not every case of emotion involves
having such feelings.
Returning to the main argument, then: on one standard philosophical view
of the phenomenal dimensions of emotions, originating with Descartes, they
are brute inner sensations to which we have infallible, introspectible access,
and which are only contingently and merely causally connected with both the
objects at which they are directed and the behavior through which they are
manifested. I think it is clear that so understood, these elements can have no
independent epistemic status. As Wilfrid Sellars’ arguments against the Myth
of the Given have shown, those entities which are only causal are not the right
kinds of things to play a role in epistemic justification (Sellars 1956/1997). Justi-
ficatory practice is concerned with our claims to knowledge. Knowledge claims
must be assessable for correctness – that is, they must be subject to conceptual
norms which are intersubjectively sustained. Causes, however, are not concep-
tual, and are therefore not norm-governed in this way. Things that stand in only
causal relations, then, can have no bearing on our knowledge claims’ status as
justified or warranted. To suggest that phenomenal emotion states, understood
as brutely causal, could play such an epistemic role is to fall into the affective
equivalent of the Myth of the Given.
Now, of course, noncognitivists wholeheartedly embrace the view that
emotional affect is brutely causal in this way (Locke 1689/1978: 229–233; Hume
1740/2000: 181–290; James 1890/1950: 442–485.). In fact, they claim that emo-
tions are entirely constituted by such affect. According to such views, emotions
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.4 (198-252)

 Elisa A. Hurley

have no intentional contents of their own. Emotions as a class are “inner sensa-
tions” only contingently related to a conceptual content which is itself embod-
ied in beliefs and other cognitive states. As such, emotional responses might in
fact be “triggered” in a causal sense by the presence of certain moral features,
but it is our other “properly” cognitive or rational faculties which actually eval-
uate those features as morally significant. That is, the emotions themselves are
wholly outside the space of reasons.
Cognitivists about emotions, however, hold that they properly belong
within the space of reasons. The cognitivist approach – which currently domi-
nates the philosophical and psychological literature on the emotions – is char-
acterized most generally by the conviction that the concept of each emotion-
type restricts what can logically be the object of any token emotion of that
type, i.e., it circumscribes the range of things any particular instance of that
emotion can be directed toward or about. The object of fear, for instance, must
be something that either does, or is seen as, threatening harm. Part of what it
means to be the emotion of fear, to be identified and individuated as such, is
to be about or directed at an object of (perceived) danger or harm. And simi-
larly with other emotions – they are conceptually linked to the sorts of things
that are or can be their objects, and it is by means of these links that we iden-
tify them as the emotions that they are (Kenny 1963: 52–75). The cognitivist
picture, then, entails that emotions have intentional contents, that these con-
tents are conceptual and that it is in virtue of these contents that emotions are
identified and individuated. It is, furthermore, this concept-involvement that
in turn entails that emotions, like other intentional states,3 are norm-governed,
and therefore publicly available.
The insistence, then, on an ineliminable phenomenal component, under-
stood along the lines of the standard view presented above, seems to undermine
the emotions’ claim to these very features. Phenomenal elements, understood
as brutely causal, could accompany the intentional elements only in some con-
tingent way. Cognitivists who want to countenance such elements, then, are
committed to a “hybrid” view of emotions in which they are made up of two
distinct components – one properly cognitive, the other noncognitive – with
the latter component equipped to play only a supporting epistemic role along
the lines sketched above. Such a picture seems to create an irresolvable tension
between the solipsistic nature of such phenomenological experiences and the
epistemic authority that the deliverances of concept-involving emotions would
be thought to enjoy.
Of course, some cognitivists have attempted to revalue the emotions in
their entirety by reducing their phenomenal components to the endorsement of
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.5 (252-295)

Apt affect 

the particular value judgments which are the emotions’ intentional contents.4
So, for instance, Martha Nussbaum, proposing a sophisticated neo-Stoic ac-
count of emotions, holds that the particular “feel” of an emotion is nothing
more than the vivacity or detail with which we grasp its particular evalua-
tive content. Because, according to her view, that evaluative content usually
involves those things that are central to my well-being, embracing such con-
tent just is the upheaval of an emotion (Nussbaum 2001: 45). In this way, then,
the threatened solipsism in avoided – the phenomenal elements are reduced
to intentional elements that are uncontroversially in the space of reasons, that
is, with whose concept-involvement, and therefore epistemic import, we are
already familiar. The phenomenal, then, falls out of the picture altogether.
It seems, then, that the cognitivist’s two options for understanding the
phenomenal dimensions of emotions – i.e., hybridization on the one hand,
reductionism on the other – are motivated by the same basic assumption that
drives the noncognitivist to deny that the emotions carry any epistemic import
whatsoever: the assumption that the phenomenal elements of emotions can be
plausibly understood only along a causal picture of brute, private sensations.
While according to the first line of thinking, this entails that emotions, prop-
erly construed as wholly phenomenal, can play no independent epistemic role,
according to the second, it entails that the phenomenal elements must be re-
ducible to some properly conceptual components in order for the emotions to
play such a role.
Now, I do not want to suggest that some cognitivist accounts are not quite
successful on their own terms. I am entirely sympathetic with the cognitivist’s
desire to situate the emotions squarely in the space of reasons, and there is
clearly something right about Nussbaum’s analysis of their intentional compo-
nents. Furthermore, if the affective elements are private and causal in the way
suggested above, then of course they can only be situated in the space of rea-
sons if they are reduced to something properly concept-involving. This is, of
course, just the lesson learned from Wittgenstein’s Private Language Argument
(hereafter PLA).
But I think this is to get the PLA only partially right. The PLA does not sug-
gest that phenomenal experiences to which we have, in some sense, privileged
access are thereby banished wholesale from the space of reasons. The point of
the PLA, as I understand it, is instead that the very having of such experiences
depends, in a holistic sense, upon the existence of a norm-governed, public
practice of using emotion concepts. Recall, the PLA claims that our “phe-
nomenal experience” concepts could not acquire their meaning from ostensive
definitions, from introspecting our own personal experiences and simply slap-
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.6 (295-341)

 Elisa A. Hurley

ping a name on them (Wittgenstein 1953, §293, §243–258). To use a concept

is to make ourselves responsible to the norms that settle the correctness or
incorrectness of doing so, and such norms must be administered intersubjec-
tively. But once we have learned such concepts and become competent users
of them, there is no reason to think that we couldn’t then apply them to our
own phenomenal experiences, even when we keep individual instances of those
experiences to ourselves.
This point should be familiar from cases of color perception. It is by now
commonplace to suggest that there is something that it is like to see red. But al-
though this experience is properly understood as phenomenal, what it is to
count as an episode of seeing red is, we might say, beholden to the norms
governing ‘redness’ as a color concept. As Sellars suggests, in order to have
the actual experience of seeing red, I must be competent with a whole host
of concepts including not just ‘red’, but also those concerning color distinc-
tions, standard viewing conditions, lighting circumstances and the like. This
phenomenal episode would not count as a seeing of red (or perhaps “count” as
anything epistemically relevant at all) were I not already to have competency
with these and other concepts, all of which are, of course, norm-governed. This
is not to deny, then, that when I am in the situation in which I report seeing red
and exhibit red-seeing behaviors, there is something that I am feeling, some-
thing that it is like to be having such an experience, and that it is in this sense
mine and mine alone (Wittgenstein 1953, §304). We deny only that the mean-
ing of redness-seeing could be private, or that one could know that what one
was seeing was red without first being a competent user of a family of norm-
governed concepts. Our experiences of redness-seeing remain in our intuitive
sense private.
As the perceptual story goes, then, once I do have the relevant concepts, the
very phenomenal appearance of a red object usually gives me good reason to
judge that the object in front of me is actually red, assuming that I am in stan-
dard viewing conditions. If I am, then I am noninferentially entitled to such
claims about the presence of redness. I may legitimately make an ‘observation
report’ about it. But, importantly, I might not in fact make such a report. That
is, this picture does not require that for each and every experience of redness, it
must be the case that there is some overt behavior (verbal or otherwise) off of
which others will “read” my experience and therefore be able to assess it – the
experience is not norm-governed in this sense. The point is not just that having
the redness-seeing experience is conceptually dependent on publicly assessable
manifestations of redness-seeings (i.e. verbal reports or other overt behaviors),
but that such dependence is holistic. It is conceptually impossible for me to
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.7 (341-385)

Apt affect 

have any personal phenomenal experiences of redness before I have the ability
to make public reports about redness, and therefore before I have the family of
concepts governing redness which play a role in intersubjective discourse. But
this is consistent with the claim that on any particular occasion I might expe-
rience seeing red and not in fact manifest any overt behavior that indicates to
others that I am.
Analogously, then, the fact that I do not report or outwardly manifest an
emotion every time I experience it does not mean that I cannot be said to be
“having” the emotion experience. As with color experiences, my affective expe-
rience being part and parcel of the emotion that it is is conceptually parasitic,
in a holistic sense, on a practice of public emotion manifestation in which the
meanings of emotion terms are determined, taught and learned intersubjec-
tively. An example should help to make this clear. Suppose I claim that I am
feeling happy. I, however, show no outward signs of being happy on this par-
ticular day – I act aloof and despondent, I don’t laugh at those things I usually
find funny, I snap at people, etc. On the one hand, we might be inclined to
believe what I say about feeling happy and just chalk up my failure to display
it to some other local circumstance. Certainly, if at the end of the day I am
once again laughing at jokes and seem to be enjoying the company of the peo-
ple I am with, then perhaps I really was “feeling” happy all along. The account
presented here need not deny the intuitive plausibility of such an explanation,
since it holds that we may legitimately ignore the occasional “gap” between my
phenomenal experience of, and the public manifestations of, an emotion.
But suppose that my behavior and dispositions continue in this way, and
I seem in fact to be manifesting signs of depression – I sleep 15 hours per
day, I stop caring about the way I look, I become more and more reclusive,
I stop eating, all the while reporting that I am in fact happy. Suppose when you
ask why I am happy I am at a loss for a reason. And suppose such behavior
continues. It seems that one is now in a position to legitimately challenge my
continued report that what I feel is in fact ‘happiness’. That is, if the words, “I
feel happy” occur to me regularly divorced from all contexts in which I could
justify them as being appropriate, they would gradually lose their meaning al-
together – whether I am convinced that what I am feeling is ‘happiness’ or not.
It is in this sense, then, that my phenomenal experience of happiness is holis-
tically dependent on there being publicly available evidence of something that
is appropriately happiness-meriting. Now, clearly there is more to say here. A
complete account of such holistic dependence would obviously require some
explanation of this relation such that we could tell when it is legitimate to
charge an emotion experience claim with violating it, that is, with altogether
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.8 (385-463)

 Elisa A. Hurley

failing to fall within the scope of the what is (holistically defined as) happiness.
While such an account will have to wait for another time, we should still be
able to take away the relevant lesson about the conceptual norm-governedness
of phenomenal experiences.
So, the proposal on the table is that the phenomenal dimensions of emo-
tions, properly understood, are indeed something over and above the endorse-
ment of particular propositional content, something which in one sense can be
considered private. It is, furthermore, this something extra which is concept-
involving and therefore epistemically valuable in its own right. So how, one
might now ask, do such affective elements do their independent epistemologi-
cal work? It seems plausible to suggest that phenomenal emotional responses so
conceived are (perhaps uniquely) useful tools for identifying the moral features
of the world: appropriately interpreted within a particular context, emotional
affect might be, in a sense, a shorthand way of making moral value judgments.
However, there is nothing so far to suggest that I might not be perfectly com-
petent at making moral judgments in the absence of such capacities. That is,
none of this shows that the capacity for the appropriate phenomenal response
is in fact necessary for moral concept mastery.
I, however, am inclined to think that such capacities might in fact be neces-
sary. I can only begin to suggest why here. To return again to the analogy with
color, it is of course the case that not all judgments of redness involve my seeing
red. Suppose a trustworthy friend describes to me in detail her new cherry-red
Porche, and I, having no reason to doubt her at her word, judge that her new
car is in fact red. I might go on to make other inferences involving the redness
of her car – that it would look purple in a blue-lit room, that it is quite different
from the faded red of my own 1982 Honda Civic, that, given its color, I would
look decidedly better behind the wheel than she does, etc.
But of course, this is critically different from experiencing the redness first
hand. Many would argue that the person who does not have the phenomeno-
logical experience of seeing red in the presence of red things, whether or not she
does in fact make reliable judgments about such things, does not have mastery
of the concept red. Part of what it is to possess such mastery is to have the req-
uisite phenomenological experiences of seeing red in the cases in which redness
is in fact present, all against my background competencies with the concepts of
color distinctions, lighting circumstances, standard viewing conditions and the
like. That is, it is to be able to make noninferential observation reports, in the
sense identified earlier, about redness.
Perhaps we can tell an analogous story in the case of emotional responses
and moral features. That is, we might say that the good person in the pres-
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.9 (463-479)

Apt affect 

ence of cruelty is analogous to the normally sighted and standardly situated

individual in the presence of redness: both have a phenomenal response which,
because it is placed against a normative background, and assuming it is “appro-
priate,” entitles them to make other conceptual moves concerning the features
to which they are responding. As in the case of the person skilled with redness,
the skilled moral agent’s emotional responses give her noninferential entitle-
ment to the claim that cruelty is in fact present here. There is something that
it is like to be in the presence of cruelty that is distinct from making value
judgments about cruelty through a process of inference from certain natu-
ral features of a situation – that is, of knowing that something is worthy of a
“cruelty response”. In the skilled moral agent, that something is, we might say,
equivalent to an observation report about the cruelty here present. Of course,
whereas in the standard viewing conditions, there is generally only one accept-
able observation report to be made – that which accords with the features that
the object actually possesses – in the moral case, there might be a bounded
range of appropriate responses to the presence of cruelty. We might say that
in ‘standard morality conditions’, these responses share a certain “negative”
affective flavoring, in that these responses seem intimately tied to my being
motivated to intervene and change the situation, to recoil in horror, to being
disposed to protest, and the like; that is, to bringing about the motivational, be-
havioral or verbal tendencies that we think are the justified moral responses to
the presence of cruelty. Were I not to experience one of the range of negatively
flavored affective responses in the presence of truly cruel features, then I would
be only partially competent with the concept itself – there would be something
deficient in me qua moral agent. That is, if we agree that part of what it is to
be competent with a moral concept is to be able to arrive at a judgment about
it noninferentially, to make observation reports about it, then we might have
an argument for saying that phenomenally flavored emotional responses are
required for moral concept mastery.
Two caveats are in order here. First, the claim that in order to qualify as
a good moral agent I must experience one of a range of negatively flavored
affective responses in the presence of cruelty should, once again, be interpreted
holistically. It would, of course, be unnecessarily demanding to insist that such
cruelty responses are required of me in every cruelty-instantiating situation
in order for me to count as competent with the concept. Instead, I want to
suggest that there is a privileged connection between having the response and
being competent with the concept, such that we might say, “ceteris paribus, the
good moral agent experiences a merited phenomenal cruelty response in the
presence of cruelty.” While we will once again have to leave for another time
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.10 (479-524)

 Elisa A. Hurley

the question of how to fill in this ceteris paribus clause, we should at the very
least say that were I never to experience one of these responses, I certainly could
not be considered a master of the moral concept.
Second, the account I have just provided means that I will have to allow
that the person who feels what we would intuitively call “glee” at the prospect
of torture, but associates this feeling with being motivated to stop the cruel
act, of recoiling in horror, of exclaiming “how horrible!”, is not criticizable on
account of her affective feelings in such a case. My suggestion has been that it
seems plausible to think that only a certain range of affective experiences will be
able to do the right epistemic work, will be able to move and motivate us in the
right sorts of ways. But since my point is just that our “oomphings” must allow
us to track the moral features of the world, if your tickle happens to morally
track what my pain does, then I’ll simply have to say, “so much the worse for
me and my pain.”
I think this picture of the role of phenomenal emotional responses in track-
ing the moral features of the world is an attractive one, but I will not pursue it
further here. Instead, I want to suggest that there is a second, independent rea-
son for thinking that the capacity for phenomenal responses might be, albeit
in a different way, necessary for moral concept mastery. Given my (perhaps by
this point, obvious) sympathies with the non-naturalist camp championed by
John McDowell and David Wiggins, I think it might in fact be impossible to
identify the class of cruel things – the “shape of cruelty” – without having an
emotional response to those things. One of the oft-recognized epistemic virtues
of our emotional capacities is their tendency to be revelatory of information to
which we might have no other access. It seems plausible to suggest, for instance,
that I might witness the giving of a gift, but would never have recognized this
particular instance of gift-giving as insidiously cruel if I did not first respond to
the situation, in a way that might surprise me, with indignation or outrage. Of
course given the picture presented here, this response would only be merited if
one could, in principle, tell a story about what features of this gift-giving do in
fact make it cruel. Otherwise, we would remain understandably skeptical about
whether this is a case in which my emotional response is to be trusted. But to
require that such a story is in principle available is not to insist that we must be
able to tell that story before being justified in a particular emotional response.
What this suggests is that the class of the ‘cruel’ may be inaccessible to us if we
do not possess the affective capacities to be moved by cruelty in certain ways.
Although we might have all sorts of background commitments about cruelty
and those features and concepts to which it is related, it is possible for us, in any
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.11 (524-574)

Apt affect 

given circumstance, to fail to see the cruelty present here without being “hit”
emotionally with it (Elgin 1996: 146–169).
My central claim here is that the phenomenological deliverances of emo-
tions, as themselves concept-involving, have an independent epistemic role to
play. However, I want to be clear that this account need not insist that the de-
liverances of emotions enjoy decisive epistemic authority. Nothing I have said
denies the reality that emotions have a tendency to be recalcitrant, and that the
force with which they deliver their information can often take us off course or
make us blind to the real features of the situation at hand. It is, I think, this fea-
ture which accounts in large part for the entrenched conviction that emotions
are uniquely problematic, epistemologically speaking. To be sure, it might take
a great act of will not to be carried along immediately to assent by the phe-
nomenal force of my emotional experience. But such an act of will is possible,
and emotions, properly understood, are the appropriate objects of such acts of
will. Just as I am legitimately open to challenge when I am in the nonstandard
viewing conditions, have the phenomenological experience of seeing redness
and then start making inferential moves from such an appearance, if I am in
the presence of cruelty or some other moral feature and have an unmerited
emotional reaction, I might similarly be called upon to justify or withdraw my
emotional response. At the very least, I am open to criticism if I take myself to
be entitled to that response.5
Just as we know the standard conditions for the tenability of the deliver-
ances of our perceptual capacities, we develop standards that specify when and
to what extent emotional reactions, too, are tenable, or when we have reason
to doubt or challenge their deliverances. As with other epistemic states, if there
is reason to question the appropriateness or legitimacy of a certain emotion,
then I have other methods for justifying or correcting it. I might look to other
beliefs and attitudes I hold and see if this judgment fits in with them coher-
ently and/or consistently, I might attempt to verify my supposed evidence, or
I might appeal to some intersubjective standard within an evaluative practice.
There need be nothing sui generis or mysterious about this particular kind of
epistemic fallibility or our capacities for addressing it.
On the other hand, we need not deny that the emotions may be signifi-
cantly more recalcitrant than our perceptual capacities: whereas in the case of
color perception, we typically either have no reason to doubt the deliverances
of our perceptual capacities or are pretty well aware of when we’re in non-
standard viewing conditions and should therefore not trust those deliverances,
in the case of emotions, the phenomenal “appearings” may have more sway
over us, even when we know that they are distorted, inaccurate or unjustified.
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.12 (574-621)

 Elisa A. Hurley

There might in fact be fewer cases, then, in which we are noninferentially en-
titled to the evaluations embodied in our emotional responses. But this seems
to me to suggest only that the emotions are more complex than some of our
other epistemic capacities, not that they ultimately function in epistemically
disanalogous ways.
Catherine Elgin, in fact, has suggested that emotion states and belief states
have very similar epistemic functions, whatever their respective conceptual
contents turn out to be. For example, just as my anger might structure the
ways in which I see the world, beliefs, too, supply an orientation, or struc-
turing perspective. Whether I simply believe that women are subordinated in
the workplace or am angry about the (perceived) fact that they are, I avail
myself of certain patterns of salience, expectations and states of perceptual
awareness. If emotions and beliefs function similarly in their structuring and
salience-generating capacities, then, Elgin suggests, from an epistemic stand-
point there seems to be little at stake in making the distinction between the
two (Elgin 1996: 153–159). While I think that there will be some important
differences between emotion and belief states – I think, for instance, that their
phenomenology will be markedly different – I agree with Elgin that from an
epistemic standpoint, i.e., in terms of beholdenness to justificatory norms, the
distinction seems largely overstated. I think we do better to regard differences
in epistemic function between beliefs and emotions as differences primarily of
degree, not differences in kind.
My suggestion, then, has been that the impoverished options for treat-
ing the phenomenal elements of emotions result from a failure, or perhaps a
refusal, to countenance the ways in which a familiar argument about our phe-
nomenal experiences’ holistic dependence on conceptual norms applies equally
to emotions. It is a mystery to me why moral psychologists and philosophers of
mind who work on the emotions continue to seem satisfied with an outmoded
Cartesianism which was long ago recognized as a legitimate target for critique
in other areas of philosophical investigation. As far as I can see, there is no prin-
cipled reason to think that the emotions represent a radically different kind of
epistemic phenomenon, one that could not be accounted for by this familiar
picture in the ways suggested above. It seems to me, then, that the burden of
proof is now on those who would insist that emotions are, from an epistemic
standpoint, intrinsically problematic. I myself find it hard to imagine why they
should be so.6
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.13 (621-682)

Apt affect 

Notes

. I am concerned here only with the epistemic contributions of emotion states – the ways
in which they help us come to moral verdicts, or deliver moral knowledge, in the first place.
In this paper, I therefore leave aside another critical role that emotions arguably play, that
is, as appropriate responses to situations which have already been morally evaluated. Clearly
these two roles may be closely related, and perhaps even temporally inseparable, but they
are, I think, conceptually quite distinct.
. To be clear, I do not think such mastery is appropriately understood as an all or nothing
affair – i.e. you either have it or you don’t – but instead as existing along a continuum. I
think that we can be better and worse users of moral concepts, that there are degrees of
competency as a moral agent. This is supported by the picture of emotions presented here,
because if emotions are properly understood as residing within the space of reasons and
not in the space of causes, then our emotional capacities are norm-responsive and therefore
the kind of things which can be trained, honed, and educated such that they might become
more discriminate and appropriate over time.
. I purposely use the rather vague ‘intentional states’ here, leaving open the question of how
conceptually robust the intentional contents of such states must be. The degree or nature of
conceptualization required by a cognitivist picture of the emotions is far from settled within
the philosophical literature. While some suggest that the evaluative element of an emotion
must be a belief, others hold that emotions might require only weaker forms of conceptual-
ization such as construals, adverbial attitudes, imaginings or entertainings. It is beyond the
scope of this paper, to take a stand on the nature of the conceptual contents of emotions. My
intuition is that it is too strong, and not phenomenologically accurate, to require that the
cognitive content be fully propositional in the manner required by belief. Emotions likely
embody a wider variety of cognitive states, all of which involve conceptualization, but many
of which will not be propositional in form.
. To be sure, there has emerged recently a strain of neo-cognitivists (e.g., Michael Stoker,
Claire Armon-Jones, and Robert Roberts) who seek to give non-reductionist accounts of the
affective elements of emotions states. These accounts, however, remain to my mind unsatis-
factory. Stocker (1996) speaks of “psychic feelings” which exhibit “conceptual complexity,”
but he does not explain how such conceptual complexity is related to the irreducible “feel-
ings” of an emotion state. Claire Armon-Jones (1991) gives an account which understands
the felt aspects of emotions as adverbial construals of certain thought-contents. But given
that Armon-Jones just stipulates that adverbial construals have a particular, technical sense
that just is the felt aspect in question, it is unclear that her account really tells us anything
useful about irreducible affectivity.
. I actually think we are criticizable, at least to some degree, for having the emotional
response in the first place. I am compelled by recent arguments suggesting that we are
responsible for cultivating our emotional capacities in the appropriate ways, and that emo-
tions are to a large extent within our voluntary control (see the work of Nancy Sherman
(1999) for instance ). Here, then, the analogy with color perception might break down. That
is, in the perceptual case, I might be held responsible, or even “criticized,” as an epistemic
agent if I take myself to be noninferentially entitled to certain deliverances in nonstandard
 JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.14 (682-765)

 Elisa A. Hurley

viewing conditions, but I am nevertheless not responsible for having those perceptual ex-
periences, given the existence of those conditions. In the emotion case, however, it seems
like I might also be legitimately criticizable for having the response, because it might be a
sign that I have not cultivated the appropriate kinds of emotional sensitivities. I think this
is an important piece of the overall project of epistemically revaluing the emotions, but it is
conceptually distinct from the claims I make here.
. I would like to thank Carin Ewing, Nathaniel Goldberg, Mark Lance, Daniel Levine, and
John Haldane for engaging me in thought-provoking discussions on these topics. I am also
grateful to audiences at the University of Southampton and Brown University where earlier
drafts of this paper were presented. Finally, I owe a debt of gratitude to Matthew Burstein,
Margaret Little and Thane Naberhaus for invaluable comments on earlier versions of this
paper.

References

Armon-Jones, Claire (1991). Varieties of affect. Toronto: University of Toronto Press.

Calhoun, Cheshire (1984/2003). Cognitive emotions? In Robert Solomon (Ed.), What is an
emotion? Classic and contemporary readings (pp. 236–247). Oxford: Oxford University
Press.
Elgin, Catherine Z. (1996). Considered judgment. Princeton: Princeton University Press.
Greenspan, Patricia S. (1989). Emotions and reasons: An inquiry into emotional justification.
London: Routledge.
Gordon, Robert M. (1987). The structure of emotions: investigations in cognitive philosophy.
Cambridge: Cambridge University Press.
Hume, David, (1740/2000). A treatise of human nature. Oxford: Oxford University Press.
James, William (1890/1950). The principles of psychology, Vol. II. New York: Henry Holt and
Company.
Kenny, Anthony (1963). Action, emotion and will. London: Routledge and Kegan Paul.
Locke, John (1689/1978). An essay concerning human understanding. Oxford: Clarendon
Press.
McDowell, John (1994). Mind and world. Cambridge, Massachusetts: Harvard University
Press.
Nussbaum, Martha C. (2001). Upheavals of thought: The intelligence of emotions. Cambridge:
Cambridge University Press.
Roberts, Robert C. (1988). What an emotions is: A sketch. The Philosophical Review, 97,
183–209.
Sellars, W. (1956/1997). Empiricism and the philosophy of mind. In Herbert Feigl &
Michael Scriven (Eds.), Minnesota Studies in Philosophy of Science, Vol. 1 (pp. 253–359).
Minneapolis: University of Minnesota Press. Reprinted by Cambridge, Massachusetts:
Harvard University Press.
Sherman, Nancy (1999). Taking responsibility for our emotions. In Ellen Frankel, Fred
Miller, & Jeffrey Paul (Eds.), Responsibility (pp. 309–321). Cambridge: Cambridge
University Press.
JB[v.20020404] Prn:23/11/2004; 10:28 F: Z13112.tex / p.15 (765-785)

Apt affect 

Solomon, Robert C. (1976/1983). The passions. Notre Dame, Indiana: University of Notre
Dame Press.
Sorabji, Richard (2000). Emotion and peace of mind. Oxford: Oxford University Press.
Stocker, Michael with Elizabeth Hegeman (1996). Valuing emotions. Cambridge: Cambridge
University Press.
Wiggins, David (1987/1998). A sensible subjectivism? In his Needs, values, truth (3rd
edition). Oxford: Clarendon Press.
Wittgenstein, Ludwig (1953). Philosophical investigations. G. E. M. Anscombe (Trans.). New
Jersey: Prentice Hall.
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.1 (41-110)

The Varieties of Religious Experience

considered from the perspective of James’s
account of the stream of consciousness

Thomas Natsoulas
Department of Psychology, University of California, Davis

William James produced a second major psychological work, in which he

focussed on the individual’s stream of consciousness. The Varieties of
Religious Experience addresses, among other things, the phenomenology of
the concrete experiences wherein one apprehends oneself to stand in an
immediate relation to whatever one may consider to be the divine. The
present article starts exploring the later volume from the perspective of
James’s detailed conception of the stream of consciousness as presented in
The Principles of Psychology. Special attention is given to the (a) the evidently
nonsensory perceiving of the unperceivable, including the divine, (b) the two
main aspects belonging to every state of consciousness – the feeling aspect
and the cognitive aspect – however “abstract” a state may be, (c) the primacy
of the qualitative over the conceptual, including some of James’s arguments
against the Intellectualists, and (d) the character of mystical experiences,
both the religious and nonreligious kind.

Keywords: William James, Varieties of Religious Experience, consciousness,

stream of consciousness, nonsensory perceiving, psychology of religion,
feeling, intellectualism, mysticism

Such is the human ontological imagination, and such is the convincingness of

what it brings to birth. Unpicturable beings are realized, and realized with an
intensity almost like that of an hallucination. (James 1902/1982: 72)
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.2 (110-158)

 Thomas Natsoulas

. Introduction

The Varieties of Religious Experience: A Study in Human Nature consists of the

Gifford lectures William James delivered at Edinburgh in 1901 and 1902. James
suggests a certain conviction is found at every religion’s heart: There exists an
unseen order with which we should harmoniously live. The third lecture, “The
Reality of the Unseen,” treats of the sense of reality people have with refer-
ence to the special objects of their religious experience. By this point, James
already has indicated what he means by the “religious” kind of experience. Re-
ligion should be understood to refer to certain experiences – as distinct from
the institutions, philosophies, and theologies developed from them.
Religion, . . . as I now ask you arbitrarily to take it, shall mean for us the feelings,
acts, and experiences of individual men in their solitude, so far as they apprehend
themselves to stand in relation to whatever they may consider the divine.
(James 1902/1982: 31; original italics)

Absent such concrete experiences, James expresses doubt (p. 431) that any
religious philosophies or theologies would ever have been framed.
James’s lecture on experiencing the unperceivable draws a distinction be-
tween two general kinds of presence to consciousness: (a) the presence of
objects to thought; and (b) their presence to the senses, their sensible pres-
ence. Typically, the concrete objects of religious experience – such as a diety
whom one worships – and the abstract objects of religious experience – such
as the Christian God’s infinity, omniscience, and absoluteness – are present
only to thought. One’s awareness of these concrete and abstract objects is usu-
ally devoid of any distinctive sense-content corresponding to them. Yet one
may strongly believe in their superexperiential existence, and no less so than
in the reality of items for which one can give definite descriptions. Indeed,
feelings of the real presence of objects of religious experience may be “as con-
vincing . . . as any direct sensible experience can be, and they are, as a rule,
much more convincing than results established by mere logic ever are” (James
1902/1982: 72).
Therefore, James’s earlier arguments, in The Principles of Psychology
(1890/1950b: 299–306), regarding “the paramount reality of sensations” must
receive some qualification. As I bring out later, one sometimes has, in religious
experience, a greater sense of the realness of the object of one’s belief than of
other things, not excluding oneself. A diminution in the felt sense of one’s own
realness may also occur at other times, for example, during experiences that are
mystical but not religious – a distinction that requires explication.
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.3 (158-200)

The Varieties of Religious Experience considered 

. Contra the Intellectualists

Objects of religious experience are not unique with respect to the absence of
sense-contents corresponding to them. Experiences unrelated to religion are
also informed by abstract ideas. For example, the exercise of concepts with
abstract reference is, according to James (1902/1982), a very common ingredi-
ent of our everyday perceptual awarenesses of the physical environment. Thus,
James states abstract ideas
form the background for all our facts, the fountain-head of all the possibilities
we conceive of. They give its “nature,” as we call it, to every special thing.
Everything we know is “what” it is by sharing in the nature of one of these
abstractions. We can never look directly at them, for they are bodiless and
featureless and footless, but we grasp all other things by their means, and in
handling the real world we should be stricken with helplessness in just so far
forth as we might lose these mental objects, these adjectives and adverbs and
predicates and heads of classification and conception. (James 1902/1982: 56)

Those who have studied The Principles may find James’s claim in Varieties that
we apprehend all other things by means of concepts surprising. Among other
locations in James’s masterwork, his famous chapter on the stream of con-
sciousness contains sustained criticism of a certain group of thinkers whom
James calls “the Intellectualists.” These thinkers held that the apprehension of,
among much else, relations and universals requires that special acts of relat-
ing intelligence be performed. They proposed that these acts, or actions of an
Ego, are pure thoughts and are in their nature superior to the kind of mun-
dane mental-occurrence instance – which is brought into being automatically
by the total brain process – of which James proposes that the entire stream of
consciousness consists.
According to James, the states of consciousness that are the basic durational
components of streams of consciousness comprise the entire mental side of re-
ality. One after another, it is states of consciousness alone of which our mental
life consists, even doubly so if there proceeds as well within us a second stream
of consciousness. Thus, in explaining any psychological function, there is no
theoretical need to introduce an actus purus intellectûs, which swoops down
from on high in order to do the job. Againsts the Intellectualists, several cases
are cited by James (1890/1950a) wherein we mentally apprehend and cannot
articulate the item that we are apprehending – as we surely could if, instead, our
apprehensions of the item were of a conceptual sort. The list below indicates a
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.4 (200-272)

 Thomas Natsoulas

few arguments of James’s suggesting as much, along with others supportive of

his view (cf. Natsoulas 2000–2001b).
1. Subtle or implicit communications often indicate the presence of objective
relations without their assigning any names to these relations, owing to the
priority of firsthand awareness of these relations over our thinking about them.
2. The relations we know of as existing in nature are more numerous than
any language can do justice to in all their shades and, therefore, we must have
discovered these relations in the environment, rather than having conceptually
constructed them.
3. In any instance in which a state of consciousness has an intrinsic perceptual
aspect or component – whenever a state of consciousness is a “perceptual state
of consciousness” – there is involved as part of it perceptual awareness of more
than one can say, as well as awareness of environmental or bodily features that
one is at a loss to describe.
4. The noticeable difference in how thunder upon silence sounds, as compared
to how more thunder upon thunder sounds, is a matter of one’s undergoing
two qualitatively different auditory experiences, not a matter of one’s merely
thinking it to be the case that two different kinds of sequence took place.
5. Perceiving something and thinking it are two different psychological pro-
cesses that are normally discriminable from each other firsthand.
6. Some of our states of mental expectation, which we commonly undergo, do
not specify their objects conceptually. In such cases, it can be the case that we
expect something to occur, or to be present, that is unidentified by us before
we come to perceive it.
7. Trying with difficulty to recall a name, one often apprehends intead an in-
tensely active phenomenological gap that is qualitatively distinctive and varies
in this respect with the name that one is seeking.
It would be inconsistent to understand James as abandoning in Varieties his
basic thesis of the priority of nonconceptualizing feelings over awarenesses
that conceptualize the objects of their apprehension. I take James to be say-
ing that what, in The Principles, he had called “knowledge-about,” as distinct
from “knowledge of acquaintance,” is what is taking place in the form of any
conceptualizing state of consciousness. That is, not all our awarenesses – only
those that are conceivings of things or states of affairs – involve the exercise
of concepts. Thus, the general fact would remain, as James (1890/1950a) had
previously stated it:
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.5 (272-319)

The Varieties of Religious Experience considered 

What we are only acquainted with is only present to our minds; we have it,
or the idea of it. But when we know about it, we do more than merely have
it; we seem, as we think over its relations, to subject it to a sort of treat-
ment and to operate upon it with our thought . . . . Feelings are the germ and
starting point of cognition, thoughts the developed tree. The minimum of
grammatical subject, of objective presence, of reality known about, the mere
beginning of knowledge, must be named by the word that says the least. Such a
word is the interjection, as lo! there! ecco! voilà! or the article or demonstrative
pronoun introducing the sentence, as the, it, that. (p. 222; original italics)

This passage does not propose the existence of states of consciousness that lack
a cognitive aspect. No state of consciousness occurs that does not instantiate
the property of intentionality. They are all of them, whatever else they may be,
cognitive states. Consult in this connection the first paragraph of James’s fa-
mous chapter on the stream of consciousness. That paragraph says something
different from the well-known statement regarding early infancy that psychol-
ogists like to quote from James. Although states of consciousness vary in their
veridicality and breadth, or how much each of them apprehends, no state of
consciousness is less a cognition than any other state of consciousness.
Nor does the above passage suggest that some states of consciousness are
of such an abstract purity – that their essential nature is of such a kind – that
they are literally missing, per impossibile, an intrinsic “sensitive body,” or feel-
ing aspect (James 1890/1950a: 478; Natsoulas 1998). No state of consciousness
is more abstract, less finite, less of a concrete feeling, than any other state
of consciousness. Even an awareness having something universal or general
for its object is, nevertheless, a “perfectly determinate, singular, and transient
thing . . . . a perishing segment of thought’s stream, consubstantial with other
facts of sensibility” (James 1890/1950a: 474).
In James’s view, every one of the many basic durational components con-
stituting any stream of consciousness is perforce a qualitative mental state. Not
a single state of consciousness is proposed to ever occur that is lacking a quali-
tative or feeling aspect. This negative generalization applies as well to all those
durational components of the stream that have among their objects abstract
items or items that are not themselves possibly perceivable. It does not matter
how abstract or unperceivable the objects of a particular state of consciousness
are. According to James, states of consciousness are all qualitative mental states
and will therefore seem, when they occur, “warm and intimate” to any inner
awareness that one may have of them.
I have discussed James’s feeling aspect of the states of consciousness in two
recent articles (Natsoulas 1998, 2000–2001a). Let me say at this point only the
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.6 (319-370)

 Thomas Natsoulas

following three things. Later in the present article, I shall return to the pri-
ority James assigned to states of consciousness as instances of sensibility or
feeling, over the conceptualizing character that, as he acknowledges, many of
them do possess.
1. In brief, the feeling aspect of every state of consciousness is that essential,
qualitative way, or form by which, it feels the totality of items, properties,
events, relations, and so on, that it apprehends or seems to apprehend – the
latter in cases of objects of consciousness having no existence. A mental state,
however neutral affectively, that is no more or less than a visual experience, say,
would be a state that visual-qualitatively feels its objects.
2. The cognitive aspect of every state of consciousness is an aspect of
its phenomenological structure, no less so than the state’s feeling aspect.
Thus, the cognitive aspect too belongs to the subjective part of what James
(1902/1982: 498) refers to as “the world of our experience.” The objective part
of this world is “the sum total of whatsoever at any given time we may be think-
ing of ” (p. 498). Consistently, James could have expressed the objective part as
being anything at all that we may happen to be apprehending, experiencing,
occurrently aware of, or thinking about. In contrast, the cognitive aspect is an
intrinsic feature of the act of apprehending: the state of consciousness itself of
whatever kind it may be. The cognitive aspect is that “outwardly pointing” fea-
ture of a state that makes it possible for it to be about something beyond itself.
A controversial matter, not to be considered here, is whether a mental state, a
single pulse of mentality, can ever have itself as object. James (1890/1950a: 190)
holds self-intimating states of consciousness are impossible.
3. James comes to discuss mystical experiences late in Varieties. He likens to
sensations as opposed to conceptual thoughts the apprehensions that mystical
experiences characteristically involve (p. 405). The mystical apprehensions are
cognitive, for they possess intentionality, but their objects are not subject to
articulation and the character of the experience is incommunicable. And yet
the objects of these apprehensions are felt firsthand to be as real as anything,
and often as more real than anything mundane. Notwithstanding the fact that
firsthand reports of mystical experiences insist that these experiences are non-
sensory. For example, James’s (1902/1982) long quotation from Saint John of
the Cross includes this sentence:
In [the mystical knowledge of God], since the senses and the imagination are
not employed, we get neither form nor impression, nor can we give any ac-
count or furnish any likeness, although the sweet-tasting wisdom comes home
so clearly to the innermost part of the soul. (p. 407)
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.7 (370-429)

The Varieties of Religious Experience considered 

. Abstract objects and the sense of reality

How should we construe James’s following conclusion drawn from the con-
crete experiential instances that he considered? How does what I am claiming
about his general understanding of states of consciousness fit with his follow-
ing description of many states of consciousness, including those involving, as
part of them, awarenesses directed in the way James defined as religious? What
does the kind of awareness James is referring to amount to intrinsically?
It is as if there were in human consciousness a sense of reality, a feeling of objec-
tive presence, a perception of what we may call “something there,” more deep
and more general than any of the special particular “senses” by which the cur-
rent psychology supposes existent realities to be originally revealed. If this were
so, we might suppose the senses to waken our attitudes and conduct as they so
habitually do, by first exciting this sense of reality; but anything else, any idea,
for example, that might similarly excite it, would have the same prerogative of
appearing real which objects of sense normally possess.
(James 1902/1982: 58; original italics)
We undergo states of consciousness that are accompanied by a strong impres-
sion of their giving us experiential contact with an ultimate reality that lies
beyond the one which our senses reveal to us. Above, James is adverting not
so much to the existence of a hidden, unperceivable order, but to our having
experiences of the same. Sometimes, our stream of consciousness is as though
we are perceiving – other than by the particular senses – a general order that
(a) lies beyond the objects that are capable of projecting photic energy or some
other kind of energy to our point of observation and that (b) is no less external
to us than those objects, belongs no less than they do to external reality and is
present to our consciousness from outside it.
Moreover, James is vaguely suggesting a means whereby this impression
is produced in us. That is, the abstract objects of our states of consciousness
are taken to be really present owing to the fact that an impression of reality is
“excited” in us by abstract ideas. Abstract ideas are supposed to do this in some-
thing like the way the senses excite an impression of reality. But it is unclear how
James believes abstract ideas “similarly excite” this impression.
Our senses – better, our perceptual systems – pick up stimulus information
and resonate to it (Gibson 1966, 1979/1986). But does any such information
specify the existence or properties of that ultimate hidden order in which we
deeply believe? If indeed it is the case that no stimulus information specifies
a hidden order, then how does an abstract idea on its own give us a sense
of reality with respect to its object? Activity in the sense receptors does yield
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.8 (429-480)

 Thomas Natsoulas

activity in the brain. But abstract ideas cannot themselves affect the brain pro-
cess. From among the mental, only the states of consciousness can affect the
brain process – and they can do so only to a certain degree, according to James
(1890/1950a), only by furthering or hindering the brain process as it is already
proceeding.
In support of the thesis that a sense of reality can be excited by receipt
of relevant stimulus information, James (1902/1982: 59) brings up the kind of
awareness we call a “presentiment.” In having a presentiment, the person has
awareness of something there but has no inner awareness of states of conscious-
ness involving any kind of perceptual awareness of that item. The person’s
pertinent states of consciousness may be perceptual, they may involve per-
ceptual awareness, but these states do not include, according to the testimony
of the person, any kind of awareness of the presentimental item that would
subjectively qualify as perceptual.
James quotes a report of this evidently nonperceptual kind of awareness
from an intimate friend of keen intellect. The friend stated that he “felt” some-
thing come into his room and stay close to his bed for a minute or two. He
had a powerful sense of this entity’s real presence near him, and he was deeply
affected emotionally by his awareness of the entity. Although he had no kind
of perceptual awareness of the entity, the friend stated that he was aware of
its presence “far more surely” than he had ever been aware of the presence of
“any fleshly living thing.” He also reported apprehending the entity’s “instan-
taneously swift” departure through the door.
James’s friend described as well two other similar happenings and, with the
following words, summarized what took place in his stream of consciousness
on all three occasions:
The certainty that there in outward space there stood something was indescrib-
ably stronger than the ordinary certainty of companionship when we are in the
close presence of ordinary living people. The something seemed close to me,
and intensely more real than any ordinary perception. Although I felt it to be
like unto myself, so to speak, or finite, small, and distressful, as it were, I didn’t
recognize it as any individual being or person.
(James 1902/1982: 60; original italics)

The friend also reported having, more than once, an equally intense and abrupt
sense of presence “of a sort of mighty person” comprising “some ineffable
good.” This presence filled him with joy instead of, as in the above three cases,
abhorrence.
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.9 (480-525)

The Varieties of Religious Experience considered 

James quotes a few further, similar examples from other informants. These
people too were aware, by what seemed to them nonsensory means, of some-
thing present right there in their immediate environment. And James com-
ments on the “seat” of such awarenesses, briefly proposing an hypothesis
regarding their source (see later).
It is natural at this point to recall James’s (1902/1982) rejection of “med-
ical materialism” in his first lecture, titled “Religion and Neurology.” Medical
materialism refers to any doctrine that would account for religious experience
exclusively in terms of effects of brain or bodily function on mental processes.
In comparing with his own view of religion, James first expresses a belief con-
sistent with his account of the stream of consciousness in The Principles. He
agrees with the “general postulate of psychology” according to which every
state of consciousness has “some organic process as its condition . . . Every one
of them without exception flows from the state of the possessor’s body at the
time” (James 1902/1982: 14).
However, James then points out (a) that the thoughts and feelings mak-
ing scientific activity itself possible are no less organically conditioned than
religious experiences are and (b) that the significance of both categories of ex-
perience is not decidable from the dependence of the states of consciousness
involved on bodily conditions. Therefore, the Jamesian fact that, for example,
the awarenesses of nonsensory presence are produced, as all states of conscious-
ness are, by the total brain process does not mean the awarenesses should be
discredited, that they do not have significance. Certainly, they possess contents,
which make no reference to their most proximal cause, and these contents
should be considered in relation to the rest of what we feel and know. James
(1902/1982) summed up his general view of religious opinions: “Their value
can only be ascertained by spiritual judgments based on our own familiar feel-
ing primarily; and secondarily on what we can ascertain of their experiential
relations to our moral needs and to the rest of what we hold true” (p. 18).

. The nature of the sense of reality in the evident absence

of sensory presence

What, more specifically, is the sense of reality we experience even when what
seems to us to be really before us has no sensory presence so far as inner
awareness can tell? The impression of real presence nevertheless is more like
a sensation, James suggests, than like an intellectual operation. James might
have suggested the sense of reality is more analogous to the conclusion of an
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.10 (525-581)

 Thomas Natsoulas

inferential process – analogous not to the whole process, of course, but to the
consequent occurrent belief that something is really there.
Instead, James is suggesting that the sense of reality under discussion is like
a sensational content (visual, auditory, etc.) that corresponds to an external ob-
ject of awareness, although when we have nonsensory awareness of something
as being there, ordinary sensations that correspond to that of which we have
awareness seem absent to us. And the sense of this object’s real presence can
be introspectively distinguished from the object’s presence to thought. James
quotes the self-report of a person who continued to think about the same un-
perceived and unperceivable object over the years but the “electric current”
of its felt presence no longer flowed into him, and so the object, now only
an object of thought, no longer provided him with the moral support that
it used to.
As I mentioned, all states of consciousness have a feeling aspect accord-
ing to James. This would apply equally to those states of consciousness whose
objects have, one or more of them, a sensible presence and to those states of
consciousness all of whose objects are present only to thought. However, none
of the latter category of state may actually exist in James’s view. He claimed
in The Principles that “our own bodily position, attitude, condition, is one of
the things of which some awareness, however inattentive, invariably accompa-
nies the knowledge of whatever else we know” (James 1890/1950a: 241). An
object felt to be really present in the immediate environment without having
any sensible presence
(a) is apprehended as part of a larger sensory context,
(b) is apprehended by means of a state that has the body among its objects,
and
(c) is apprehended by means of a state that is no less qualitative than any other
state of consciousness.

James’s (1902/1982: 63) tentative hypothesis regarding the seat of the general
sense of present reality associates this sense with feelings arising from the mus-
cles, specifically from their occurrent state of readiness to respond. A state of
consciousness that includes feelings of incipient behavior and connects the ac-
tion readied with an object – even an object of awareness that lacks all sensible
presence – will include as well a sense of the respective object’s real presence.
A major problem with James’s latter conjecture is, as I see it, simply this: We
are often prepared to act with respect to environmental items that have not yet
entered our immediate environment. Yet we do not experience these items as
being present until they have a sensible presence for us. That is, although we are
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.11 (581-627)

The Varieties of Religious Experience considered 

ready for them, we distinguish their actual arrival from their not having arrived
as yet. They do not seem to be there before us in nonsensory form simply as
a result of our being muscularly prepared to respond to them. Therefore, two
other hypotheses concerning the sense of real presence of unperceived items
are worth mentioning.

1. James may be on the right track in looking toward what else, besides the
nonsensory phenomenological presence of the special item of interest, is part
of the content of the same state of consciousness. Part of the surrounding en-
vironment does have sensory presence in the state of consciousness and the
stimulus information picked up and resonated to by the perceptual system
may include informational features that specify properties of the “unperceived”
object. Thus, Gibson (1966) writes:
Organs of perception are sometimes stimulated in such a way that they are not
specified in consciousness. Perception cannot be “extrasensory,” if that means
without any input; it can only be so if that means without awareness of the
visual, auditory, or other quality of the input. An example of this is the “ob-
stacle sense” of the blind, which is felt as “facial vision” but is actually auditory
echo detection. The blind man “senses” the wall in front of him without re-
alizing what sense has been stimulated. In short, there can be sensationless
perception, but not informationless perception. (p. 2)

So too, within a single modality. Of all that which one is having awareness of at
any moment by means of the visual perceptual system, only some of it is qual-
itatively present. The surfaces seen usually obstruct photic energy that would
otherwise project to the point of observation from parts of other surfaces. Yet
those partially occluded surfaces are perceived not as fragmented but as ex-
tending behind the surfaces that are partially obstructing the light reflected
from them (Gibson 1979/1986; Michotte, Thinès, and Crabbé, 1964/1991).

2. The awareness of unperceivable objects with a strong sense of their real pres-
ence may result from the functioning of internal “loops” within the respective
perceptual system. For example, the visual system “can also operate without
the constraints of the stimulus flux . . . . The visual system visualizes” (Gibson
1979/1986: 256). And it can do so independently of the stimulus information
currently being picked up by the perceptual system. Since, in such cases, none
of the current stimulus information specifies properties of the object being ex-
perienced, its visualization (whether qualitative or not) with a sense of reality
will depend on whether tests of reality are performed. For whatever cause, if
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.12 (627-676)

 Thomas Natsoulas

such tests are suspended, then the object will seem to be really present before
one. As Gibson (1979/1986), whose account this is, states,
The dreamer is asleep and cannot make the ordinary tests for reality. The drug-
taker is hoping for a vision and does not want to make tests of reality. There
are many possible reasons why the hallucinasting patient does not scrutinize
what he says he sees, does not walk around it or take another look at it or test
it. (p. 257)

Gibson is speaking here of cases in which the visual system gives awareness that
is qualitative and he is discussing the fact that this sometimes occurs in the ab-
sence of corresponding stimulus information yet with a sense of the object’s
real presence. However, a perceptual system also sometimes gives nonqualita-
tive awareness of objects. Presumably, such awareness can occur too when the
perceptual system operates without the constraints of the stimulus flux. An ac-
tivity of the visual or another perceptual system can give awareness of objects
without the perceptual system’s also producing sense-contents corresponding
to them yet there can be a sense of their real presence.

. In the distinctively religious sphere of experience

Many people “possess the objects of their [religious] belief, not in the form of
mere conceptions which the intellect accepts as true, but rather in the form of
quasi-sensible realities directly apprehended” (p. 64). In his lecture on the felt
real presence of the unperceivable, James (1902/1982) considers religious expe-
riences, too, that involve a sense of real presence. There is, as well, an important
category of human experience wherein the individual (a) is aware of himself or
herself as standing in a relation of directly apprehending what he or she con-
siders to be the divine or some feature or aspect of the divine and (b) takes this
relation of direct apprehension to be one in which the object of religious belief
has real presence albeit of a nonqualitative kind.
I mentioned in the preceding section a person who continued to think
about the same unperceived and unperceivable “It” over the years but the “elec-
tric current” of its presence had somehow, unfortunately, gotten turned off. His
description of “It” was as an “Absolute Reality” existing beyond the perceived
and perceivable world. But he distinguished “It” from a wholly unknowable re-
ality. For he could enter into a firsthand relation with it. There could be a kind
of nonsensory communication between them by which the man drew strength
and vitality at difficult times. He had a very strong sense of the real presence
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.13 (676-734)

The Varieties of Religious Experience considered 

to him of a higher mind – before he lost the ability to make contact, perhaps
because he no longer believed in its existence as strongly as before.
Similarly, as James’s examples show, people report communicating with
God, usually in the form, it would seem, of their sharing thoughts with God.
That is, there occurs what seems to be a “joint or mutual consciousness” (to
quote the OED). See my article on the original use of the English word con-
sciousness to refer to the occurrent sharing of knowledge of something with
one or more other individuals (Natsoulas 1991). To be conscious would be, as
it were, to be “scious con,” that is, to have an apprehension of some actual or ap-
parent fact along with someone else. Relevantly to my present topic, the OED
quotes from literature as follows to illustrate the use of the word consciousness
in its earliest meaning. In 1681, someone wrote of “consciousness, or mutual
knowledge of persons in their worship.” James (1902/1982) mentions a report
that suggests the OED’s first, interpersonal sense of consciousness may apply,
and will surely seem to the person to apply, to his or her experience of God’s
presence. A boy of seventeen stated,
Sometimes as I go to church, I sit down, join in the service, and before I go out
I feel as if God was with me, right side of me, singing and reading the Psalms
with me . . . . And then again I feel as if I could sit beside him, and put my arms
around him, kiss him, etc. When I am taking Holy Communion at the alter, I
try to get with him and generally feel his presence. (p. 71)

People sometimes report they hear the voice of God speaking to them. But,
more often, communications from God take the apparent form of states of con-
sciousness produced by God in one, God having awareness of states referring
to him and occurring in one’s stream along with the states he produces there.
Thus, in some religious experiences, it is by means of one’s stream of conscious-
ness itself that God answers one’s thoughts addressed to him. A young man
stated, “God is quite real to me. I talk to him and often get answers. Thoughts
sudden and distinct from any I have been entertaining come to my mind after
asking God for his direction” (James 1902/1982: 71).
Indeed, James suggests that a self-report containing the following sentence
epitomizes the religious experience of “thousands of unpretending Christians”:
“He answers me again and again, often in words so clearly spoken that it seems
my outer ear must have carried the tone, but generally in strong mental im-
pressions” (James 1902/1982: 70). This person adds that these inner communi-
cations from God usually have for content a text of Scripture directly relevant
to the practical concerns with which the person is currently preoccupied.
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.14 (734-778)

 Thomas Natsoulas

James (1902/1982) comments that, among those people who experience

the objects of their religious belief as “quasi-sensible realities directly appre-
hended,” their sense of the real presence of the objects to them may fluctuate
and, concomitantly, their faith alternates between warmth and coldness. He
adds that every religious person has likely had, in a time of crisis, his or her
faith strengthened by “a directer vision of the truth, a direct perception, per-
haps, of a living God’s existence” (p. 68). A case of what James is here referring
to can be found in the correspondence of James Russell Lowell:
As I was speaking, the whole system rose up before me like a vague destiny
looming from the Abyss. I never before so clearly felt the Spirit of God in me
and around me. The whole room seemed to me full of God. The air seemed
to waver to and fro with the presence of Something I knew not what. I spoke
with the calmness and the clearness of a prophet. (James 1902/1982: 66)

In similar self-reports from other people, James finds mention of (a) the non-
perceptual character of how the divine “appears” to them at such moments
and (b) the immediate sense of the reality of what is “appearing.” Indeed,
when a certain clergyman described the evident reality of what he experienced,
he stated God was there no less than he himself was and he, the clergyman,
was, if anything, the less real of the two. The clergyman had a greater immedi-
ate sense of God’s realness than of his own. Here are two relevant comments
corresponding to the above two points.

1. An extremely explicit statement of God’s real but nonsensory presence ap-

pears in a self-report of an experience lasting four or five minutes that James
classifies as “mystical”:
In this ecstasy of mine God had neither form, color, odor, nor taste; more-
over . . . the feeling of his presence was accompanied with no determinate
localization. It was rather as if my personality had been transformed by the
presence of a spiritual spirit. But the more I seek words to express this intimate
intercourse, the more I feel the impossibility of describing the thing by any of
our usual images. At bottom the expression most apt to render what I felt is
this: God was present, though invisible; he fell under no one of my senses, yet
my consciousness perceived him. (James 1902/1982: 68)

This man also described his experience in terms of being “penetrated . . . al-
together” with God’s goodness and power. This statement may suggest some
spatial localization for the real presence the man felt. That is, God was “here,”
in me, rather than “there” in the environment. However, the man might have
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.15 (778-828)

The Varieties of Religious Experience considered 

felt God to be everywhere – which is how, frequently, mystical experiences have

been described.

2. The sense of God’s reality as being greater than one’s own suggests a possibly
fruitful line of investigation. Does “perceiving” the unperceivable in instances
of religious experience tend to be accompanied by “depersonalization?” The
latter term refers to a kind of experience in which one remains aware of who
one is yet feels oneself altered in some way and somehow less real than normal.
In describing depersonalization in The Psychology of Anomalous Experience,
Graham Reed (1972) writes of it as an “uncanny sense of unreality” (p. 127);
the individual “no longer feels himself to be real by comparison with his en-
vironment” (p. 129). Reed brings out that many people occasionally undergo
experiences of depersonalization. A survey of college students Reed mentions
found depersonalization associated with stressful situations, especially social
situations involving embarrassment. However, Reed does not speak of the suf-
fering of one’s own apparent realness from the heightened realness of another
object of one’s current awareness. Yet this is not an unreasonable expectation
given Reed’s interpretation of depersonalization as owed to an “anomalous em-
phasis”: an unusual change in how attention is distributed among the objects
of one’s current states of consciousness.
Reed also comments on, among other matters, certain instances in which
the experience of depersonalization is recurrent and part of a depressive mental
disorder. Such patients interpret their disturbing depersonalization as an indi-
cation of their being “cut off from God’s grace” (p. 128). This is likely also a
feature of certain states of consciousness wherein the “perceived” presence of a
diety evokes strong negative emotion.
Another report provided by James, concerning “intimate communion with
the divine”, has some comparative pertinence to the sense of one’s seeming to
be less real than the divine. A man described a number of his religious experi-
ences as involving “the temporary obliteration of the conventionalities which
usually surround and cover my life” and “a temporary loss of my identity.” He
may or may not have experienced himself as less real than the divine, but he did
speak of an alteration in self-awareness: a change for the duration of the expe-
rience in what he felt himself to be. Reed would not consider such experiences
to be instances of depersonalization; for, Reed insists, we retain an awareness
of our personal identity in depersonalization. Rather, the man may have had
mystical experiences in which he temporarily lost the sense of himself as a sep-
arate being. His account of the experiences has him taking in great expanses
of the natural world while standing on a high mountain summit – which is
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.16 (828-879)

 Thomas Natsoulas

not an unusual kind of setting mentioned in the self-reports and writings of

nature mystics.
Indeed, in James’s later lectures in the series, the ones concerning mysti-
cism, he quotes (p. 394) reports in which one loses not only the sense of one’s
own separateness but also the sense of the presence of God as a separate person-
ality. Instead, one experiences himself and God, both, as being a unity with the
totality of nature. In effect, God is nature and so is the person himself. Com-
pared to the “ordinary” experiences of God’s immediate real presence, God
and oneself are “lost” yet they continue to have existence in the form of a total
reality of supreme power and love with which one has spontaneously merged.
Following James’s definition of religious experience, perhaps we should say
this: As the experience just described proceeds, it varies or changes from the
religious kind to another kind. Recall that James’s definition requires one to
apprehend oneself standing in relation to the divine. This does occur in some
mystical experiences that involve a sense of one’s having merged with the di-
vine. As James (1902/1982) states, “In mystic states we both become one with
the Absolute and we become aware of our oneness” (p. 419). We continue to
have awareness of ourselves, now as part of a divine whole. However, if all that
one is aware of is nature, and so neither of God nor of self, then one is not ap-
prehending oneself in the way James’s definition of religious experience would
require, no matter how much awe the apprehension of the totality of nature
may elicit.
The objection may be forthcoming that my point depends on how nature is
experienced; that is, nature may be apprehended as itself the divine. However,
in such a case, the second term in the definitionally required relationship would
still be missing. In the component states of consciousness that would qualify
the experience as mystical yet not religious, there would be awareness of the
divine as such but there would be no awareness of oneself as apprehending it,
because one does not distinguish oneself from that whole which one takes as
nature and divine.
Admittedly, as this mystical experience proceeds, it will include – see below
James’s fifth property of mystical experience – the occurrence of some states
of consciousness that are inner awarenesses of other states of consciousness.
However, inner awareness need not involve the appropriation of its objects,
that is, the appropriation of the states of consciousness that it apprehends. In-
stead, inner awareness may be, so to speak, “anonymous” – as David Woodruff
Smith (1989: 94) points out occurs in some types of meditation: whose aims are
to eliminate all awareness of self. In such cases, states of consciousness occur
and inner awareness of them is not, for example, an awareness of one’s see-
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.17 (879-932)

The Varieties of Religious Experience considered 

ing a chair, although the latter is what is taking place. In that kind of mystical
experience, just the seeing, and not anyone who is seeing, is apprehended.
Similarly, James (1902/1982: 400) quotes from a Hindu work on yoga that
describes mystical experiences in which “no feeling of I” exists. However, to
know later on of this absence, there must have been at the time of the mystical
experience, I should think, some inner awareness of what was taking place in
the stream. Moreover, given what is remembered, this inner awareness must
have been anonymous; that is, there must have been no appropriation of states
of consciousness to oneself – which is commonly a part of inner awareness. If
there had been such appropriation at the time of the mystical experience, the
experience would not be remembered later on as lacking an “I-feeling”.

. Mystical experience

Earlier, here and elsewhere, I observed that the qualitative – the feeling aspect
belonging to all states of consciousness – has an important priority for James
over the conceptual or articulable. Having examined numerous self-reports on
the felt real presence of the unperceivable, James proffers in his third lecture a
brief defense of mysticism over rationalism that rests upon a feature of human
nature which favors feeling over word. With reference to the metaphysical and
religious sphere, James argues that what we accept as articulate reasons for what
we believe or disbelieve in depends on our inarticulate feelings as to what is real
and what is unreal.
Our impulsive belief is here always what sets up the original body of truth,
and our articulately verbalized philosophy is but its showy translation into
formulas. The unreasoned and immediate assurance is the deep thing in us,
the reasoned argument is but a surface exhibition. Instinct leads, intelligence
does but follow. If a person feels the presence of the living God after the fashion
shown by my quotations, your critical arguments, be they never so superior,
will vainly set themselves to change his faith. (James 1902/1982: 74)

James adds that the primacy of the intuitive sense of what is real – over the
other deliverances from the total brain process that take the form of reasons
and arguments – is simply a fact. That is, James has not yet said it is better
that human nature is as it is in this regard. However, James explicitly claims
that the part of mental life of which rationalism can give an account is “rel-
atively superficial.” And he explains the high prestige of rationalism as owed
to its articulateness, argumentativeness, and logicality. Also, James describes
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.18 (932-979)

 Thomas Natsoulas

the rationalist approach to mental life as loquacious, clever, belittling, logic-

chopping, and as setting the rules in accordance with its own purposes. Those
adopting a rationalist approach have their own primary feelings concerning
what reality consists of, and they construct arguments showing that the con-
trary feelings about reality are false. Or they simply demand that one rationally
demonstrate to them, to their satisfaction, that those contrary feelings are true.
James (1902/1982) begins his two lectures on mysticism, late in the series,
with the promise that they will wound up broken threads from previous lec-
tures. Thus, the religious experiences under discussion have their “root and
centre in mystical states of consciousness” (p. 379). Such states are to be dis-
tinguished by their instantiating the first two of the following five properties,
and usually the third and fourth property as well. The fifth property, listed here
separately, is mentioned by James in his comments on the fourth property.

1. Ineffability. A “mystical” state of consciousness possesses inarticulable con-

tent; an adequate report of its content cannot be given in words. James is
proposing, I take it, that an important part of the total content of the state
is ineffable, whereas some parts may be expressible. People are able to re-
port on their mystical states up to a point after which words fail them. James
(1902/1982) states, “This incommunicableness of the transport is the keynote
of all mysticism” (p. 405). I believe understanding the altered states of con-
sciousness would deepen from systematic investigation into what exactly is or is
not communicable about mystical experiences. Among James’s most interest-
ing pages begin with this sentence: “In spite of the repudiation of articulate self-
description, mystical states in general assert a pretty distinct theoretic drift”
(pp. 415–416). Toward the end of these pages, this intriguing statement ap-
pears: “Not conceptual speech, but music rather, is the element through which
we are best spoken to by mystical truth” (pp. 420–421). Are the contents of
mystical states fully communicable, then, with the proper form of music, the
kind of music that, sans words, succeeds in conveying “ontological messages”?

2. Noetic quality. Its ineffability does not mean that a mystical state of con-
sciousness is noncognitive. It is inaccurate to characterize any such state as
lacking intentionality. James’s statement that mystical states are “more like
states of feeling than like states of intellect” (p. 380) does not imply that mys-
tical states of consciousness are not mental apprehensions. Any person who
witttingly undergoes a mystical state of consciousness would testify to having
been aware of something, at least something found to be completely indescrib-
able. Not only are mystical states cognitive; also, they usually seem, to inner
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.19 (979-1025)

The Varieties of Religious Experience considered 

awareness, authoritative apprehensions of deeper truths than the articulable.

Indeed, James goes further still in a footnote: “Consciousness of illumination is
for us the essential mark of ‘mystical’ states” (p. 408). Absent an impression of
new depths of truth being revealed, however incommunicable, the experience
would not count as mystical.

3. Transiency. Using the term a mystical state of consciousness, James has in

mind not just a single one of the successive pulses of mentality constituting a
stream of consciousness one after another. In effect, he distinguishes between
mystical experiences and the individual durational components of the stream
constitutive of them. A mystical state is a fairly long temporal section of the
stream of consciousness. James speaks of mystical experiences that last as much
as an hour or two. Mystical experiences consist of a sequence of mystical states of
consciousness. The latter sentence is consistent with James’s thinking although
the verbal distinction is mine. With this distinction, one can raise regard-
ing the composition of a mystical experience such questions as: Are all states
of consciousness constitutive of a mystical experience mystical states of con-
sciousness? During a mystical experience, are nonmystical states interspersed
among the mystical states in some pattern?

4. Passivity. James writes mystical experiences are usually felt by inner aware-
ness as conditions of passivity; one’s will seems in abeyance while they last. It
may even seem one’s stream is under the control of someone or something else.

5. Memorabilty. According to James, some memory of the content of a mys-

tical experience always remains, together with a conviction of the experience’s
importance. This fact, if it is one, would entail, as I see it, that at least some of
the states of consciousness that constitute a mystical experience must be ob-
jects of inner awareness. If no awareness of a mystical experience took place
when it occurred, the experience would have no later memorability, in my
view. Assuming James’s conception of inner awareness, requiring a separate
state of consciousness in every case, it would follow that, among the states
defining a mystical experience, states of consciousness are interspersed that
are inner awarenesses of them. These states of consciousness, which are di-
rected on other states in the same stream, would be of the same general kind as
states that are inner awarenesses of mental states such as anger or fatigue. James
(1890/1950a: 190) suggested that there is, when these inner awarenesses are oc-
curring, a considerable modification of the anger or fatigue of which they are
the awarenesses. That is, these inner awarenesses are themselves weaker states
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.20 (1025-1075)

 Thomas Natsoulas

of anger or fatigue because, James presumably holds, each has for an additional
focus the one or more states of consciousness that are its objects. Analogously, a
mystical experience would be expected to wax and wane in some respects with
the frequency of inner awareness.
As mentioned, not all mystical experiences are considered religious ex-
periences. James estimates only half of mysticism is religious mysticism. To
be religious experiences, mystical experiences must be consistent with James’s
definition of religion quoted in the introduction to the present article. An expe-
rience wherein one feels with awe the surroundings to be full of undecipherable
meaning is a mystical experience but not, ipso facto, a religious experience.
Here is how James expresses the paticular metaphysical insight common to
all his mystical experiences upon inhaling nitrous oxide:
It is as if the opposites of the world, whose contradictoriness and conflict make
all our difficulties and troubles, were melted into a unity. Not only do they, as
contrasted species, belong to one and the same genus, but one of the species,
the nobler and better one, is itself the genus, and so soaks up and absorbs its
opposite into itself. (James 1902/1982: 388; original italics)

These experiences fell short of being the religious kind. However, James allows
drug-produced mystical experiences can be religious experiences of the kind he
quoted reports of in his third lecture. That is, a sense of the divine’s real pres-
ence may occur as well in states of consciousness produced by drugs, although
it never happened in James’s case.
I stated above, “If there had not been awareness of the state when it oc-
curred, it would have no later memorability.” My statement may seem contra-
dicted by Saint Teresa’s reports on her mystical experiences. She writes of an
extreme form of passivity wherein the natural action of all her faculties is sus-
pended. Although “the soul is fully awake as regards God” during her union
with God, she adds that “she is as it were deprived of every feeling, and even if
she would, she could not think of any single thing” (p. 408).
But, I ask, how could she tell us all these things about what happend dur-
ing her mystical experiences unless she had inner awareness at the time? James
makes no comment on this score, but he clearly thinks of her mystical experi-
ences of God as religious, which implies – as indeed her own report does –
that she apprehended herself during the experiences to stand in a relation
to the divine and was aware of taking up a certain personal attitude toward
God. She could not have been so self-aware, I should think, unless she appre-
hended during a mystical experience some of the states of consciousness that
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.21 (1075-1126)

The Varieties of Religious Experience considered 

occurred therein. Compare the following statement of hers. It is certainly not a

description of someone blind to her mystical states as they occurred.
It was granted me to perceive in one instant how all things are seen and con-
tained in God. I did not perceive them in their proper form, and nevertheless
the view I had of them was of a sovereign clearness, and has remained vividly
impressed upon my soul . . . . The view was so subtile and delicate that the
understanding cannot grasp it. (James 1902/1982: 411)

James is impressed by the exactness of the distinctions that Saint Teresa uses
in her descriptions of her experiences and takes the quality of her reports to
indicate the reality of the experiences, as opposed to her having come later to
believe that they had occurred in the form that she describes. That is, she gives
to James indication of having been a good witness to her mystical experiences.
Among James’s conclusions from his discussion of mystical experience is
one of direct relevance to the discussion in the present article. He suggests that
mystical experiences constitute another, second “kind” of consciousness. I be-
lieve that James thinks of them in this way because he believes that they may be
a way of apprehending reality more broadly and completely than we ordinarily
do. However, from his consistent perspective, there are only streams of con-
sciousness as he has defined these. Sometimes – much more often according
to Varieties than according to The Principle – there may flow simultaneously a
second stream of consciousness in the one person. Otherwise, there exists no
“other” consciousness. Rather, the identical states of consciousness make up a
stream that make up a mystical experience. A mystical experience is a temporal
section of someone’s stream. This section may markedly differ in its contents
from other sections of the same stream, but it is no less made up exclusively
of states of consciousness that posssess both a cognitive and feeling (or quali-
tative) aspect. And it is no less produced pulse by successive pulse by the total
brain process.
This is not to deny, neither with nor against James, that states of conscious-
ness vary in the breadth of their apprehension and in their veridicality, their
conformance with reality. James is suggesting that it is not pre-decided that a
mystical experience will be less veridical than a nonmystical experience, nor
that it will not be more veridical in some cases. He states,
It must always remain an open question whether mystical states may not pos-
sibly be such superior points of view, windows through which the mind looks
out upon a more extensive and inclusive world . . . The counting in of that
wider world of meanings, and the serious dealing with it, might, in spite of all
 JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.22 (1126-1185)

 Thomas Natsoulas

the perplexity, be indispensible stages in our approach to the final fullness of

the truth. (James 1902/1982: 428)

. Final comment

There is at least one large, highly relevant topic that requires the attention of
a separate article: the role that James assigns in Varieties to what he calls “the
subconscious self.” He states that “the ‘more’ with which in religious experience
we feel ourselves connected is on its hither side the subconscious continuation
of our conscious life” (p. 512). There is in all of us – James now wants to say – a
kind of secondary consciousness that has important effects on the main stream
of consciousness. He states, “There is actually and literally more life in our total
soul than we are at any time aware of ” (p. 511).
Thus, a part of the total brain process that is held to be producing, pulse by
successive pulse, our primary consciousness produces a second stream that has
effects on the first stream. Or is it the neurophysiological causes of the second
stream that also affect the primary stream? The latter notion would seem to be
more consistent with The Principles. James speaks in Varieties of a second world
of experiences that sometimes impacts on the world of ordinary consciousness.
And he is now able to say that some of our states of consciousness that make
up our primary stream “come” from a certain mental source external to that
stream. More consistently, however, I should think, states of consciousness do
not literally incur, or run in, from outside a stream, but the part of the total
brain process that produces the secondary stream is a determinant as well of
which states of consciousness are produced in the primary stream.
Moreover, the secondary effects on the primary stream may well appear
firsthand to have a source that is exerting control over it. In religious experi-
ences, this control is felt to be divine and, according to James, may indeed have
its ultimate source in the divine. Thus, the focal point of the divine’s contact
with us would be our total brain process or the secondary consciousness that
this process produces to flow along with the primary stream.

References

Gibson, J. J. (1966). The Senses Considered as Perceptual Systems. Boston: Houghton Mifflin.
Gibson, J. J. (1979/1986). The Ecological Approach to Visual Perception. Hillsdale: Erlbaum.
James, W. (1890/1950a). The Principles of Psychology, Volume 1. New York: Holt.
JB[v.20020404] Prn:23/11/2004; 10:32 F: Z13113.tex / p.23 (1185-1230)

The Varieties of Religious Experience considered 

James, W. (1890/1950b). The Principles of Psychology, Volume 2. New York: Holt.

James, W. (1902/1982) The Varieties of Religious Experience. Harmondsworth: Penguin.
Michotte, A., Thinès, G., & Crabbé, G. (1964/1991). Amodal completion of perceptual
structures. In G. Thinès, A. Costall, & G. Butterworth (Eds.), Michotte’s Experimental
Phenomenology of Perception. Hillsdale: Erlbaum.
Natsoulas, T. (1991). The concept of consciousness1: The interpersonal meaning. Journal for
the Theory of Social Behaviour, 21, 63–89.
Natsoulas, T. (1998). On the intrinsic nature of states of consciousness: James’s ubiquitous
feeling aspect. Review of General Psychology, 2, 123–152.
Natsoulas, T. (2000–2001a). The stream of consciousness: XXII. Apprehension and the
feeling aspect. Imagination, Cognition and Personality, 20, 275–295.
Natsoulas, T. (2000–2001b). The stream of consciousness: XXIII. Contra the Intellectualists
(first part). Imagination, Cognition and Personality, 20, 383–403.
Reed, G. (1972). The Psychology of Anomalous Experience. Boston: Houghton Mifflin.
Woodruff Smith, D. (1989). The Circle of Acquaintance. Dordrecht: Kluwer.
JB[v.20020404] Prn:23/11/2004; 13:45
Index
A 197–213, 219, 226, 232–4,
abstract objects 309
adaptation 5, 7, 11, 38, 87
adaptivity, emotional 3
adrenal 94
affect logic 23ff.
affect-congruent effects 58–9
affected self 251ff.
affect-incongruent effects 58–9
affective neuroscience 23ff., 81ff.
affective stimuli 56–74
affordances x, 160, 171, 203–5
aftereffect 5, 6, 18, 19
agency 50, 55, 70, 157–79, 182, 184,
188–92, 247, 252, 267, 287–89,
295, 299
aggression 38–41, 53, 55, 76, 78, 79,
86, 141
agitation 60
Alexithymia Provoked Response
Questionnaire 259ff.
alexithymia xii, 243–284
alpha motoneurons 160ff.
amygdala 29, 86, 90, 94, 101, 108,
115, 131, 167
anger 86, 92, 95, 106, 115, 125–6,
141, 150, 237, 255, 298, 321–2
anxiety 3, 12, 13, 16–18, 31, 38, 49,
51, 85, 138, 234, 237–9,
appetitive phase 166–7
approach/avoidance 60
arousal 27–30, 33, 35, 39, 45, 54, 57,
62–69, 76, 80, 85, 167, 169, 173,
249–51, 256, 259, 264, 272–4,
attention 28–29, 33, 37–42, 48,
60–61, 65, 70, 74, 100, 114, 122,
126, 131–2, 159, 175, 189,
F: Z131IND.tex / p.1 (61-151)
249–56, 259, 263–4, 269–70, 274,
303, 317, 324
attributional triggers 28ff.
autonomic activation (during
imagined action) 161
autonomic nervous system 94, 161,
166, 174, 250
B
basal ganglia 210ff.
basic emotions 36, 243–4, 247, 251,
260, 268, 274–5, 277, 280–83,
291, 306
behavioral facilitation system 167
belonging 95, 102,
body gamma loop 160ff.
butterfly-effect 41
C
CARE system 32ff.
cerebellum 89, 172, 179, 211, 213
channel functions 26ff.
chaos 12, 49, 50, 55
cholinergic neurons 84
cingulate 34, 86, 91, 170, 211
clonal selection 84
cognitivism (in ethics and emotion
studies) 287ff.
corticospinal excitability 172
covert action 157ff.
creative functioning test 8, 14, 21st
creativity 4, 8, 9, 14, 16–19, 99
curiosity 24, 31, 38, 102, 167
cutaneous image 169
 JB[v.20020404] Prn:23/11/2004; 13:45
 Index
D frustration 86
defenses 3, 9, 12, 14, 30
denial of emotion 267
denial 11, 220, 253–4, 261–8, 271,
273–4, 278
depression see depressive
depressive 13–17, 32, 41, 45. 317
detached self 251ff.
disgust 36, 95, 96, 115, 119
disinhibition 211
dominance 56–74
dopamine 26, 30, 41, 55, 86, 115,
167, 210–11
double dissociation disorders 202ff.
E
egocentric frame of reference 159ff.
embodiment ix, xii, 27, 106, 166,
174, 176, 181–95, 270, 290, 298
emotional context 56ff.
enaction 181ff.
enactive theories 181ff.
estrogen 163
evaluation 46, 57–62, 66–74, 126,
174, 189, 255–6, 287–8, 298
event-related potential 121–135
explanatory gap 137ff., 193ff.
external attribution 255
extraversion 99, 284
F immune system 81ff.
face x, 10–18, 52, 61–2, 76, 80ff.,
121–135
FEAR system 31ff.
fear 95ff., 125ff.
feedback 56ff.
feeling, in William James’s sense
308ff.
field dynamics 25ff.
first-person perspective 220ff.
focal self 251ff.
free will xiii, 197ff.
Freud 217ff.
frontal cortex 27, 29, 33–4, 54, 167,
170–1, 194, 211–12, 213–14,
F: Z131IND.tex / p.2 (151-233)
fusimotor system 172ff.
G
GABAergic interneurons 173ff.
gamma motoneurons 157, 160–1,
167, 172–7
gating 211
glutamate 86
glutamatergic computations 25–7
goals 187ff.
grouping of responses 200ff.
guilt 36, 92, 95–6, 115, 232–7
H
happiness 30, 95–6, 125ff., 293–4
hard problem 137ff., 193ff.
higher-order perception 140ff.
histaminergic neurons 84
histrionic 13
homeostasis 85, 87, 92, 167, 190
hope 167
Huntington’s disease 210ff.
hypothalamus 84, 86, 90, 94, 118,
163–9, 201
I
imprint-hypothesis 47
inhibition of the alpha motoneurons
172
innate motor programs 162
inner awareness 224ff.
insult 57
intellectualism 303
intentionality 27ff., 55–7, 87,
157–71, 178, 181, 193, 226,
307–8, 320
International Affective Photo System
57, 62ff.
introspection 243ff.
isolation 12, 14, 17, 108
JB[v.20020404] Prn:23/11/2004; 13:45
K O
knowledge argument 137ff.
L opioids 86
language x, 81, 101–118, 139, 142,
145–51, 177, 183, 194
Libet’s veto 197ff.
limbic system 49, 102–3, 165
Linguistic Inquiry and Word Count
260ff.
locus coeruleus 83, 96
lordosis 163ff.
M perception 3, 4, 6, 10, 11, 17, 19, 23,
memory 23, 27, 29, 37–42, 46–8,
56ff., 72, 102, 105, 107, 122, 133,
151, 169, 184, 190, 194, 200, 204,
211–15, 240, 279, 283, 321
meta-contrast technique 9
microgenesis 4, 21
misconstrued emotions 217ff.
monoaminergic neurons 84
moral agency 287ff.
moral concept mastery 287ff.
motor imagery 160ff.
motor intentionality 171
motor system 157ff., 174–5, 214
motor-evoked potentials 172
motor-tone setting circuits 162
muscle spindles 157, 160ff.
muscle stretch reflex 160ff.
mysticism 303, 318–22
N 220–22, 225, 227–30, 233, 238–9
neural Darwinism 81ff.
neurotensin 86
neurotransmitter pathways 162ff.
non-linear dynamics 23ff., 40ff.
norepinephrine 26, 83, 86
norms (social and moral) 287ff.
nucleus accumbens 86
nurturance 86
F: Z131IND.tex / p.3 (233-309)
Index 
object mode (of self) 250ff.
obsessive-compulsive 12, 212
olfactory image 169
oxytocin 86
P
PAG 86, 170
palmar flexion (induced) 171
PANIC system 32ff., 85ff.
Parkinson’s disease 210ff.
perceptgenesis 3, 4, 5, 9–11, 13–14,
17–21
25, 27, 37, 42, 57, 59–65, 67, 69,
71–5, 122–7, 131, 140, 183–6,
193, 198, 200–09, 218, 221,
224–40, 251–2, 166–7, 292,
297–9, 109–13, 316
personality 4, 13, 20, 34, 66, 75–9,
91, 97–9, 103, 116–19, 188, 213,
223, 249, 258, 282–4, 316, 318,
325–6
phenomenal consciousness 137ff.
phenomenology xi, xii, 48, 159, 209,
217ff., 284, 287–8, 298, 303, 325
phobia 18–19, 31
pituitary 94, 118
planned action schema 165
plasticity of emotional systems 29ff.
PLAY system 32–33, 85ff.
positive approbation 66–74
post-traumatic stress disorder 30ff.
preconscious 5, 13, 17, 168, 217,
prefrontal cortex 211 see also frontal
cortex
pre-supplemental motor area 210
problem-space theory 189
projection 13–14, 17
prolactin 86
proprioception 157
proprioceptive image 169ff.
 JB[v.20020404] Prn:23/11/2004; 13:45 F: Z131IND.tex / p.4 (309-385)

 Index

psychoanalysis 12, 21, 37, 54, 228 self-image 12, 61

psychology of religion 303ff. semiotics 243ff.
sense impressions 137ff.
serotonin 26, 86
Q
sex 30, 39, 66–72, 85, 92, 108, 163,
qualia 137ff.
165–75, 253, 255, 258, 267–80
shame 36
R signs 244ff.
RAGE system 31–2, 85ff. somatization 264ff.
readiness potential 199ff. somatosensory systems 157, 171,
reflective consciousness 159, 170 173, 176, 214
reflexivity 248ff. spiral after-effect test 5, 6, 16, 19
regression 14, 17 state functions 26ff.
religion 303ff. steroids 86
religions experience 303ff. Stroop color word test 5
representation 246ff. subjectivity 3, 7, 11, 17, 18, 249–50
repression 13–14, 17, 217, 229–31, subliminal event related potentials
236–40 124ff.
rod-and-frame test 6, 18 substance P 86
routinization of action 201ff. Sundararajan-Schubert Word Count
243ff.
supervisory attentional system 211
S surprise 95ff., 125ff.
sadness 24, 30, 36–8, 54, 59, 76,
95–6, 126, 273
sciousness, in William James’s sense T
315 task-switching 189
secondary emotions 94ff. temporal lobe 28
second-order awareness 244ff. temporal thickness 169–70
SEEKING system 30, 33, 85, 96, 99, Toronto Alexithymia Scale-20 259ff.
101–2, 166–8
selective attention 202ff.
U
SELF (as distinct brain system) 162
unconscious 217ff.
self vi, x, xi, 5–7, 10–14, 19–20, 33,
utilization behaviour 204
42–3, 47, 57, 61–4, 73–4, 90, 110,
112, 141, 157–9, 162, 165–70,
174–6, 188–90, 207, 219, 223, V
243–280, 308, 312, 315–24 valence 66–74
self-assessment manikin 62ff. vasopressin 86
Self-Consciousness Scale 259ff. ventral tegmental area 86, 167