Visit https://ebooknice.

com to download the full version and

explore more ebooks

(Ebook) Dual-Process Theories of the Social Mind by

Jeffrey W. Sherman, Bertram Gawronski, Yaacov Trope
(eds.) ISBN 9781462514397, 1462514391

_____ Click the link below to download _____

https://ebooknice.com/product/dual-process-theories-of-
the-social-mind-5459038

Explore and download more ebooks at ebooknice.com

Here are some recommended products that might interest you.
You can download now and explore!

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason; Viles, James

ISBN 9781459699816, 9781743365571, 9781925268492, 1459699815,
1743365578, 1925268497

https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

ebooknice.com

(Ebook) Matematik 5000+ Kurs 2c Lärobok by Lena Alfredsson, Hans

Heikne, Sanna Bodemyr ISBN 9789127456600, 9127456609

https://ebooknice.com/product/matematik-5000-kurs-2c-larobok-23848312

ebooknice.com

(Ebook) SAT II Success MATH 1C and 2C 2002 (Peterson's SAT II Success)

by Peterson's ISBN 9780768906677, 0768906679

https://ebooknice.com/product/sat-ii-success-
math-1c-and-2c-2002-peterson-s-sat-ii-success-1722018

ebooknice.com

(Ebook) Master SAT II Math 1c and 2c 4th ed (Arco Master the SAT
Subject Test: Math Levels 1 & 2) by Arco ISBN 9780768923049,
0768923042

https://ebooknice.com/product/master-sat-ii-math-1c-and-2c-4th-ed-
arco-master-the-sat-subject-test-math-levels-1-2-2326094

ebooknice.com
(Ebook) Cambridge IGCSE and O Level History Workbook 2C - Depth Study:
the United States, 1919-41 2nd Edition by Benjamin Harrison ISBN
9781398375147, 9781398375048, 1398375144, 1398375047

https://ebooknice.com/product/cambridge-igcse-and-o-level-history-
workbook-2c-depth-study-the-united-states-1919-41-2nd-edition-53538044

ebooknice.com

(Ebook) Self Control in Society, Mind, and Brain (Oxford Series in

Social Cognition and Social Neuroscience) by Ran Hassin, Kevin
Ochsner, Yaacov Trope ISBN 9780195391381, 0195391381

https://ebooknice.com/product/self-control-in-society-mind-and-brain-
oxford-series-in-social-cognition-and-social-neuroscience-1462242

ebooknice.com

(Ebook) Handbook of Implicit Social Cognition: Measurement, Theory,

and Applications by Bertram Gawronski, B. Keith Payne PhD ISBN
9781606236734, 1606236733

https://ebooknice.com/product/handbook-of-implicit-social-cognition-
measurement-theory-and-applications-1900696

ebooknice.com

(Ebook) Vagabond, Vol. 29 (29) by Inoue, Takehiko ISBN 9781421531489,

1421531488

https://ebooknice.com/product/vagabond-vol-29-29-37511002

ebooknice.com

(Ebook) Reflective and Impulsive Determinants of Human Behavior by

Roland Deutsch (editor), Bertram Gawronski (editor), Wilhelm Hofmann
(editor) ISBN 9781138696877, 1138696870

https://ebooknice.com/product/reflective-and-impulsive-determinants-
of-human-behavior-42487328

ebooknice.com
 ebook
THE GUILFORD PRESS
Dual-Process Theories
of the Social Mind
 Also available

Cognitive Consistency:
A Fundamental Principle in Social Cognition
Bertram Gawronski and Fritz Strack, Editors

Handbook of Implicit Social Cognition:

Measurement, Theory, and Applications
Bertram Gawronski and B. Keith Payne,
Editors

Social Cognition: The Official Journal

of the International Social Cognition Network
Jeffrey W. Sherman, Editor
 Dual-Process
Theories of the
Social Mind

Edited by

Jeffrey W. Sherman
Bertram Gawronski
Yaacov Trope

THE GUILFORD PRESS

New York  London
© 2014 The Guilford Press
A Division of Guilford Publications, Inc.
72 Spring Street, New York, NY 10012
www.guilford.com

All rights reserved

No part of this book may be reproduced, translated, stored in a retrieval system,

or transmitted, in any form or by any means, electronic, mechanical,
photocopying, microfilming, recording, or otherwise, without written permission
from the publisher.

Printed in the United States of America

This book is printed on acid-free paper.

Last digit is print number: 9 8 7 6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data is available from

the publisher.

ISBN 978-1-4625-1439-7
 About the Editors

Jeffrey W. Sherman, PhD, is Professor of Psychology at the University of California,

Davis. His research investigates the cognitive processes underlying social psychology
and behavior, with particular interests in attitude formation and change and how stereo-
types and prejudice affect social perception. Dr. Sherman is Editor of the journal Social
Cognition. A Fellow of the Association for Psychological Science (APS), the Society
for Personality and Social Psychology (SPSP), and the Society for Experimental Social
Psychology (SESP), he is a recipient of other honors including the Theoretical Innova-
tion Prize from SPSP and the Anneliese Maier Research Award from the Alexander von
Humboldt Foundation and the German Federal Ministry of Education.

Bertram Gawronski, PhD, is Professor of Psychology at the University of Texas at Aus-

tin. His research investigates the mental underpinnings and behavioral consequences
of spontaneous and deliberate evaluations of objects, individuals, groups, and social
issues. Dr. Gawronski’s work has been recognized with honors including the Theo-
retical Innovation Prize from SPSP, the Career Trajectory Award from SESP, the Early
Career Award from the International Social Cognition Network, the Early Researcher
Award from the Ministry of Research and Innovation of Ontario, and the Charlotte
and Karl Bühler Award from the German Psychological Society. He is a fellow of APS,
SESP, and SPSP.

Yaacov Trope, PhD, is Professor of Psychology at New York University. His research
investigates the interrelations among cognition, motivation, and self-regulation in the
social context. Particular interests include how psychological distance influences the
representations of objects, and, thereby, the predictions, evaluations, and choices indi-
viduals make regarding those objects; self-­control processes; and the role of affective
states and personal desires in social judgment and decision making. Dr. Trope is a Fel-
low of the American Association for the Advancement of Science, APS, SESP, and SPSP.
His research has been funded by the National Institutes of Health and the National
Science Foundation.

v
 Contributors

Henk Aarts, PhD, Department of Social and Organizational Psychology, Utrecht University,
Utrecht, The Netherlands
Nalini Ambady (deceased), PhD, Department of Psychology, Stanford University,
Stanford, California
Elinor Amit, PhD, Department of Psychology, Harvard University, Cambridge, Massachusetts
David M. Amodio, PhD, Department of Psychology, New York University,
New York, New York
John A. Bargh, PhD, Department of Psychology, Yale University, New Haven, Connecticut
Roy F. Baumeister, PhD, Department of Psychology, Florida State University,
Tallahassee, Florida
Galen V. Bodenhausen, PhD, Department of Psychology, Northwestern University,
Evanston, Illinois
Maarten W. Bos, PhD, Negotiation, Organizations and Markets, Harvard Business School,
Boston, Massachusetts
Pablo Briñol, PhD, Faculty of Psychology, Autonomous University of Madrid, Madrid, Spain
Jimmy Calanchini, BA, Department of Psychology, University of California, Davis,
Davis, California
C. Daryl Cameron, PhD, Department of Psychology, University of Iowa, Iowa City, Iowa
William A. Cunningham, PhD, Department of Psychology, The Ohio State University,
Columbus, Ohio
Ruud Custers, PhD, Department of Cognitive, Perceptual, and Brain Sciences, University
College London, London, United Kingdom
Jan De Houwer, PhD, Department of Psychology, University of Ghent, Ghent, Belgium
Roland Deutsch, PhD, Institute of Psychology, Technical University of Dresden,
Dresden, Germany

vii
viii Contributors

Patricia G. Devine, PhD, Department of Psychology, University of Wisconsin–Madison,

Madison, Wisconsin
Ap Dijksterhuis, PhD, Behavioral Science Institute, Radboud University,
Nijmegen, The Netherlands
Ana Rita Farias, PhD, Lisbon University Institute, Lisbon, Portugal, and Utrecht University,
Utrecht, The Netherlands
Russell H. Fazio, PhD, Department of Psychology, The Ohio State University, Columbus, Ohio
Melissa J. Ferguson, PhD, Department of Psychology, Cornell University, Ithaca, New York
Klaus Fiedler, PhD, Department of Psychology, University of Heidelberg, Heidelberg, Germany
Ayelet Fishbach, PhD, Booth School of Business, University of Chicago, Chicago, Illinois
Patrick S. Forscher, BA, Department of Psychology, University of Wisconsin–Madison,
Madison, Wisconsin
Jens Förster, PhD, Department of Social Psychology, University of Amsterdam,
Amsterdam, The Netherlands
Jonathan B. Freeman, PhD, Department of Psychological and Brain Sciences,
Dartmouth College, Hanover, New Hampshire
Kentaro Fujita, PhD, Department of Psychology, The Ohio State University, Columbus, Ohio
Margarida V. Garrido, PhD, Department of Social and Organizational Psychology,
University Institute of Lisbon, Lisbon, Portugal
Bertram Gawronski, PhD, Department of Psychology, University of Texas at Austin,
Austin, Texas
Sara Gottlieb, BA, Department of Psychology, University of California, Berkeley,
Berkeley, California
Joshua D. Greene, PhD, Department of Psychology, Harvard University,
Cambridge, Massachusetts
Rainer Greifeneder, PhD, Department of Psychology, University of Basel, Basel, Switzerland
James J. Gross, PhD, Department of Psychology, Stanford University, Stanford, California
Ran R. Hassin, PhD, Psychology Department, The Hebrew University, Jerusalem, Israel
E. Tory Higgins, PhD, Department of Psychology, Columbia University, New York, New York
Mandy Hütter, PhD, Department of Psychology, University of Heidelberg,
Heidelberg, Germany
Karl Christoph Klauer, PhD, Institute for Psychology, University of Freiburg,
Freiburg, Germany
Kristen M. Klein, PhD, Analytic Services, Inc., Falls Church, Virginia
Stanley B. Klein, PhD, Department of Psychology, University of California, Santa Barbara,
Santa Barbara, California
Regina Krieglmeyer, PhD, Department of Psychology, University of Würzburg,
Würzburg, Germany
Arie W. Kruglanski, PhD, Department of Psychology, University of Maryland, College Park,
College Park, Maryland
Jennifer S. Labrecque, MA, Department of Psychology, University of Southern California,
Los Angeles, California
 Contributors ix

Nira Liberman, PhD, Department of Psychology, Tel Aviv University, Tel Aviv, Israel
Matthew D. Lieberman, PhD, Department of Psychology, University of California,
Los Angeles, Los Angeles, California
Pei-Ying Lin, MA, Department of Psychology, University of Southern California,
Los Angeles, California
C. Neil Macrae, PhD, School of Psychology, University of Aberdeen, Aberdeen, Scotland
Thomas C. Mann, BS, Department of Psychology, Cornell University, Ithaca, New York
Lucia Mannetti, PhD, Department of Developmental and Social Psychology,
Sapienza University, Rome, Italy
Randy J. McCarthy, PhD, Center for the Study of Family Violence and Sexual Assault,
Northern Illinois University, DeKalb, Illinois
Allen R. McConnell, PhD, Department of Psychology, Miami University, Oxford, Ohio
Agnes Moors, PhD, Department of Experimental Clinical and Health Psychology,
Ghent University, Ghent, Belgium
Gordon B. Moskowitz, PhD, Department of Psychology, Lehigh University,
Bethlehem, Pennsylvania
Thomas Mussweiler, PhD, Department of Psychology, University of Cologne,
Cologne, Germany
Loran F. Nordgren, PhD, Kellogg School of Management, Northwestern University,
Evanston, Indiana
Michael A. Olson, PhD, Department of Psychology, University of Tennessee,
Knoxville, Tennessee
B. Keith Payne, PhD, Department of Psychology, University of North Carolina at Chapel Hill,
Chapel Hill, North Carolina
Richard E. Petty, PhD, Department of Psychology, The Ohio State University, Columbus, Ohio
Antonio Pierro, PhD, Department of Developmental and Social Psychology,
Sapienza University, Rome, Italy
Kimberly A. Quinn, PhD, Department of Psychology, DePaul University, Chicago, Illinois
Dennis Rünger, PhD, Department of Psychology, University of Southern California,
Los Angeles, California
Robert J. Rydell, PhD, Department of Psychological and Brain Sciences, Indiana University,
Bloomington, Indiana
Norbert Schwarz, PhD, Department of Psychology, University of Southern California,
Los Angeles, California
Gün R. Semin, PhD, Faculty of Social and Behavioral Sciences, Utrecht University,
Utrecht, The Netherlands
Luxi Shen, BS, Chinese University of Hong Kong, Hong Kong, China
Gal Sheppes, PhD, School of Psychological Sciences, Tel Aviv University, Tel Aviv, Israel
Jeffrey W. Sherman, PhD, Department of Psychology, University of California, Davis,
Davis, California
Asael Y. Sklar, MA, Department of Psychology, The Hebrew University, Jerusalem, Israel
x Contributors

John T. Skowronski, PhD, Department of Psychology, Northern Illinois University,

DeKalb, Illinois
Steven Sloman, PhD, Department of Cognitive, Linguistic, and Psychological Sciences,
Brown University, Providence, Rhode Island
Robert P. Spunt, PhD, Division of Humanities and Social Sciences, California Institute
of Technology, Pasadena, California
Keith A. Stanovich, PhD, Department of Human Development and Applied Psychology,
University of Toronto, Toronto, Ontario, Canada
Fritz Strack, PhD, Department of Psychology, University of Würzburg, Würzburg, Germany
Madelijn Strick, PhD, Social and Organizational Psychology, Utrecht University,
Utrecht, The Netherlands
Maggie E. Toplak, PhD, Department of Psychology, York University, Toronto, Ontario, Canada
Yaacov Trope, PhD, Department of Psychology, New York University, New York, New York
Richard F. West, PhD, Department of Graduate Psychology, James Madison University,
Harrisonburg, Virginia
Michael T. Wojnowicz, PhD, Department of Statistics, University of California, Irvine,
Irvine, California
Wendy Wood, PhD, Department of Psychology, University of Southern California,
Los Angeles, California
 Preface

The emergence of dual-process theories has been one of the most significant develop-
ments in the history of scientific psychology. The overarching assumption of these theo-
ries is that psychological processes can be divided into two distinct categories depending
on whether they operate in an automatic or controlled fashion. In recent years, this dis-
tinction between “fast” and “slow” modes of thought has even permeated the popular
press and the lay public’s understanding of psychology.
The publication of Shelly Chaiken and Yaacov Trope’s Dual-Process Theories
in Social Psychology in 1999 played a central role in placing the distinction between
automatic and controlled processes at the center of social psychological theory. In so
doing, the book captured a critical turning point in how the field of social psychology
reconciled competing metatheoretical models of the rational, intuitive scientist versus
the satisficing cognitive miser. It also helped to inspire waves of conceptual, method-
ological, and empirical refinements to dual-process thinking and research that have led
to a fundamental reorganization of the theoretical landscape within the field of social
psychology.
In this volume, we have tried to capture the most important developments in the
15 years since the publication of Chaiken and Trope’s seminal book and to organize
and integrate those developments in a novel and coherent manner. Toward this end, we
have invited leading authorities to address the pivotal role of automatic and controlled
processes across many areas of inquiry. The scope of this volume encompasses the core
assumptions and conceptual foundations of dual-process theories, their applications to
a wide range of social phenomena, and contemporary critiques of the duality approach.
Our goals in this book were to describe the ubiquity of dual-process theories; identify
current directions of research on the topic; present key theoretical, methodological, and
empirical developments; and raise important questions for future research.
The book is divided into nine major parts. The chapters in Part I explore the con-
ceptual foundations of dual-process theories, analyzing the theoretical basis for distin-
guishing between two kinds of psychological processes. Part II documents the emerging
trend of extending dual-process to dual-system theorizing, reviewing the core assump-

xi
xii Preface

tions of theories that propose two dissociable mental systems. The chapters in Part III
address conceptual issues in the measurement of dual processes and illustrate recent
advances in the development of formal process models to investigate automatic and con-
trolled underpinnings of social behavior. The chapters in Parts IV through VIII review
the insights that have been gained by applying dual-process theories to a wide range of
questions in social psychology, including research on attitudes and evaluation; social
perception; thinking and reasoning; the interplay among habits, goals, and motivations;
and the range and limits of self-regulation and cognitive control. Finally, the chapters in
Part IX address conceptual criticism of and theoretical alternatives to the dual-process
approach.
In order to provide some common ground and coherence to the chapters in the
book, we requested that the authors of each chapter address three issues that we believe
are critical in contemporary thinking about dual-process models. First, we asked the
authors to clearly define the central features of their model, including the type of dual-
ity addressed by the model, the specific processes proposed in the model, the defining
characteristics of those processes, and any assumptions about the extent to which the
processes are grounded in distinct mental representations or mental systems. Second,
we asked the authors to describe how the processes in their model relate to the distinc-
tion between automatic and controlled processes, and to specify the particular features
of automaticity (e.g., awareness, intention, efficiency, control) that are central to their
model. Finally, we asked the authors to detail the relationship between the proposed
processes, including whether the processes are thought to be mutually exclusive, to
operate in isolation or in parallel, or to interact with one another. Although the chapters
of this volume emphasize different aspects of these issues, the three questions provide
a common framework for much of the material, helping to clarify points of consensus,
contention, and uncertainty.
Given the ubiquity of the dual-process approach in virtually all areas of social
psychology, we hope that the current volume provides an indispensable resource for
researchers interested in the contribution of automatic and controlled processes to
social judgments and social behavior. The book may serve as a valuable supplement for
graduate-level and advanced undergraduate courses on social cognition, judgment and
decision making, attitudes and persuasion, social perception, and self-regulation. By
providing an up-to-date review of contemporary dual-process theories, the book will be
of great interest for any scholar who is interested in the psychological underpinnings of
social behavior.
 Contents

Part I. The Basics

1. Two of What?: A Conceptual Analysis of Dual‑Process Theories 3

Bertram Gawronski, Jeffrey W. Sherman, and Yaacov Trope

2. Examining the Mapping Problem in Dual‑Process Models 20

Agnes Moors

3. Conscious and Unconscious: Toward an Integrative Understanding 35

of Human Mental Life and Action
Roy F. Baumeister and John A. Bargh

4. What Is Control?: A Conceptual Analysis 50

Kentaro Fujita, Yaacov Trope, William A. Cunningham, and Nira Liberman

Part II. Dual-Systems Models

5. Two Systems of Reasoning: An Update 69

Steven Sloman

6. Rationality, Intelligence, and the Defining Features 80

of Type 1 and Type 2 Processing
Keith E. Stanovich, Richard F. West, and Maggie E. Toplak

7. The Reflective–Impulsive Model 92

Fritz Strack and Roland Deutsch

xiii
xiv Contents

Part III. Measurement and Formal Modeling

8. Dual‑Process Theory from a Process Dissociation Perspective 107

B. Keith Payne and C. Daryl Cameron

9. Process Models Require Process Measures 121

Jeffrey W. Sherman, Regina Krieglmeyer, and Jimmy Calanchini

10. Random‑Walk and Diffusion Models 139

Karl Christoph Klauer

Part IV. Attitudes and Evaluation

11. The MODE Model: Attitude–Behavior Processes 155

as a Function of Motivation and Opportunity
Russell H. Fazio and Michael A. Olson

12. The Elaboration Likelihood and Metacognitive Models of Attitudes: 172

Implications for Prejudice, the Self, and Beyond
Richard E. Petty and Pablo Briñol

13. The Associative–Propositional Evaluation Model: 188

Operating Principles and Operating Conditions of Evaluation
Bertram Gawronski and Galen V. Bodenhausen

14. The Systems of Evaluation Model: A Dual‑Systems Approach to Attitudes 204

Allen R. McConnell and Robert J. Rydell

Part V. Social Perception

15. Controlled Processing and Automatic Processing in the Formation 221

of Spontaneous Trait Inferences
Randy J. McCarthy and John J. Skowronski

16. The Dynamic Interactive Model of Person Construal: 235

Coordinating Sensory and Social Processes
Jonathan B. Freeman and Nalini Ambady

17. Person Perception: Integrating Category‑ and Individual‑Level Information 249

in Face Construal
Kimberly A. Quinn and C. Neil Macrae

18. Dual‑Process Models of Trait Judgments of Self and Others: 264

An Overview and Critique
Stanley B. Klein

19. Automaticity, Control, and the Social Brain 279

Robert P. Spunt and Matthew D. Lieberman
 Contents xv

Part VI. Thinking and Reasoning

20. The Human Unconscious: A Functional Perspective 299

Ran R. Hassin and Asael Y. Sklar

21. Metacognitive Processes and Subjective Experiences 314

Rainer Greifeneder and Norbert Schwarz

22. Same or Different?: How Similarity versus Dissimilarity Focus 328

Shapes Social Information Processing
Thomas Mussweiler

23. Visual versus Verbal Thinking and Dual‑Process Moral Cognition 340

Elinor Amit, Sara Gottlieb, and Joshua D. Greene

24. Prolonged Thought: Proposing Type 3 Processing 355

Ap Dijksterhuis, Madelijn Strick, Maarten W. Bos, and Loran F. Nordgren

Part VII. Habits, Goals, and Motivation

25. Habits in Dual‑Process Models 371

Wendy Wood, Jennifer S. Labrecque, Pei‑Ying Lin, and Dennis Rünger

26. Conscious and Unconscious Goal Pursuit: 386

Similar Functions, Different Processes?
Ruud Custers and Henk Aarts

27. The Implicit Volition Model: The Unconscious Nature of Goal Pursuit 400
Gordon B. Moskowitz

28. Promotion and Prevention: How “0” Can Create Dual Motivational Forces 423
E. Tory Higgins

Part VIII. Self-Regulation and Control

29. Beyond Control versus Automaticity: Psychological Processes 439

Driving Postsuppressional Rebound
Jens Förster and Nira Liberman

30. The Explicit and Implicit Ways of Overcoming Temptation 454

Ayelet Fishbach and Luxi Shen

31. Breaking the Prejudice Habit: Automaticity and Control 468

in the Context of a Long‑Term Goal
Patrick S. Forscher and Patricia G. Devine

32. Emotion Generation and Emotion Regulation: 483

Moving Beyond Traditional Dual‑Process Accounts
Gal Sheppes and James J. Gross
xvi Contents

Part IX. Criticism and Alternatives

33. The Limits of Automaticity 497

Klaus Fiedler and Mandy Hütter

34. The Unimodel Unfolding 514

Arie W. Kruglanski, Kristen M. Klein, Antonio Pierro, and Lucia Mannetti

35. Why a Propositional Single‑Process Model of Associative Learning 530

Deserves to Be Defended
Jan De Houwer

36. How Many Processes Does It Take to Ground a Concept? 542

Gün R. Semin, Margarida V. Garrido, and Ana Rita Farias

37. Dual Experiences, Multiple Processes: 560

Looking Beyond Dualities for Mechanisms of the Mind
David M. Amodio

38. Rethinking Duality: Criticisms and Ways Forward 578

Melissa J. Ferguson, Thomas C. Mann, and Michael T. Wojnowicz

Author Index 595

Subject Index 611

 Pa rt I

The Basics
 Chapter 1

Two of What?
A Conceptual Analysis of Dual‑Process Theories

Bertram Gawronski, Jeffrey W. Sherman, and Yaacov Trope

The emergence of dual-­ process theories is mental underpinnings of social phenomena.

probably one of the most significant theo- However, their explanatory and predic-
retical developments in the history of social tive value depends on (1) a clear distinction
psychology. The overarching assumption between operating principles and operating
of dual-­process theorizing is that the men- conditions, (2) conceptual rigor in the defi-
tal processes underlying social phenomena nition of the proposed dualities, (3) precise
can be divided into two distinct categories formulations of empirical hypotheses about
depending on whether they operate in an covariations between processing duali-
automatic or nonautomatic fashion.1 Over ties, and (4) clearly specified links between
the past decades, dual-­process theories have the hypothesized mental processes and the
made important contributions in virtu- causal relations between stimuli and behav-
ally all areas of social psychology, and this ior they are supposed to explain. Expanding
volume showcases the most recent devel- on this analysis, we conclude with a brief
opments since Chaiken and Trope’s (1999) outlook on emerging themes and future
seminal volume, Dual-­Process Theories in directions in dual-­process theorizing.
Social Psychology.
Despite the ubiquity of dual-­process theo-
ries in social psychology, their conceptual Explanatory Function
foundations have also been the subject of
ongoing debates. In this chapter, we pro- A useful framework to clarify the explana-
vide a metatheoretical analysis that aims at tory function of dual-­ process theories is
clarifying the explanatory function of dual-­ Marr’s (1982) distinction among three lev-
process theories, the conceptual nature of els of analysis in psychological research: the
their underlying dualities, and structural computational level, the algorithmic level,
features that characterize different types and the implementational level. Accord-
of dual-­ process theories. Expanding on ing to Marr, research at the computational
this analysis, we discuss the criticism that level is concerned with identifying rela-
dual-­process theories are unfalsifiable and tions between inputs (i.e., stimuli and their
whether the realm of mental processes can broader contexts) and outputs (i.e., judg-
indeed be divided into two distinct catego- ments and behavior). The overarching goal
ries. Our conclusion is that dual-­ process of research at the computational level is
theories have provided vital insights into the to identify which types of inputs produce
3
4 THE B A SIC S

which kinds of outputs under which con- these theories have their theoretical core at
textual conditions. For example, a large the algorithmic level, in that the identified
body of research on behavioral priming substrates are directly linked to two (sets of)
can be described as computational, in that mechanisms that are claimed to translate
it focuses on the particular behaviors that inputs into outputs.
are elicited by exposure to various kinds of Although the proposed positioning of
prime stimuli (for a review, see Bargh, 2006). dual-­process theories at Marr’s (1982) algo-
Research of this kind differs from research rithmic level may seem rather trivial, it helps
at the algorithmic level, which is concerned to clarify the explanatory function of dual-­
with the mechanisms that translate inputs process theories by specifying the empirical
into outputs. This level of analysis resonates phenomena that dual-­process theories aim
with the goal of social-­cognitive research, to explain (explanandum) and the theoreti-
which aims at identifying the mental pro- cal assumptions that are proposed to explain
cesses and representations underlying social these phenomena (explanans). From an epis-
behavior. For example, expanding on the temological point of view, one could argue
identification of input–­ output relations in that research at the computational level aims
studies on behavioral priming, a consider- to explain observed outputs by relating them
able body of research investigated the men- to inputs that cause these outputs. Using the
tal mechanisms that mediate the effects of previous example of behavioral priming,
prime exposure on overt behavior, including exposure to a particular stimulus may serve
motivational (e.g., Cesario, Plaks, & Hig- as an explanation for an observed behav-
gins, 2006), misattribution (e.g., Loersch & ioral response to the extent that the stimulus
Payne, 2011) and self-­related (e.g., Wheeler, can be said to cause the behavioral response.
DeMarree, & Petty, 2007) processes. In other words, the observed behavior rep-
Finally, research at the implementational resents the phenomenon that needs to be
level is concerned with the physical systems explained, and exposure to the prime stimu-
that implement the mechanisms identified lus serves as the event that is supposed to
at the algorithmic level. In social psychol- explain the behavior (causal explanation).
ogy, this approach is prominently reflected However, stating that exposure to the prime
in the emerging field of social neuroscience, explains the behavioral response does not
which is concerned with the neural under- say anything about how the prime caused the
pinnings of social judgments and social observed behavior. This question is central
behavior (Cacioppo, Berntson, Sheridan, & in research at the algorithmic level, in which
McClintock, 2000; Ochsner & Lieberman, the causal relation between prime exposure
2001). For example, expanding on mental and behavior represents a phenomenon that
process theories of prime-to-­behavior effects is in need of further explanation (De Hou-
(e.g., Cesario et al., 2006; Loersch & Payne, wer, 2011). Research at the algorithmic level
2011; Wheeler et al., 2007), research at the provides an answer to this question by iden-
implementational level may investigate the tifying the mental mechanisms that mediate
neural underpinnings of the mechanisms the link between prime exposure and overt
that mediate observed relations between behavior (mechanistic explanation). In this
certain kinds of primes and overt behavior. sense, dual-­process theories offer explana-
Dual-­process theories are located at tions of observed input–­ output relations
Marr’s (1982) algorithmic level of analy- by specifying the mental mechanisms that
sis, in the sense that they identify men- translate inputs into outputs. As we outline
tal mechanisms that translate inputs into in the following sections, this conceptual-
outputs. A central feature of dual-­ process ization has important implications for dual-­
theories is that they postulate two quali- process theorizing in social psychology.
tatively distinct (sets of) mental processes
that mediate between inputs and outputs.
Some dual-­process theories go beyond the Operating Principles
algorithmic level by including assumptions versus Operating Conditions
about the neural substrates that implement
the hypothesized processes (e.g., Lieberman, The first important insight that can be
Gaunt, Gilbert, & Trope, 2002). Yet even gained from relating dual-­process theories
 A Conceptual Analysis of Dual-Process Theories 5

to Marr’s (1982) algorithmic level of analy- ations that translate input X into output
sis is that it resolves the common conflation Y?). From the perspective of Marr’s algo-
of operating principles and operating condi- rithmic level, characterizations of a given
tions. Whereas the concept of operating prin- process as automatic versus nonautomatic
ciples refers to the mental mechanisms that are not sufficient as a conceptual foundation
translate inputs into outputs, the concept of dual-­process theories, because they fail
of operating conditions refers to the condi- to specify the nature of the processes that
tions under which a given process operates translate inputs into outputs. An illustra-
(Gawronski & Bodenhausen, 2009). A cen- tive example is the conscious–­unconscious
tral characteristic of dual-­ process theories duality, which has been used in a manner
is that they are concerned with the question suggesting that it refers to two qualitatively
of whether the mental processes underlying distinct mental processes (e.g., Baumeister,
social behavior operate in an automatic or Masicampo, & Vohs, 2011; Dijksterhuis
nonautomatic fashion. This emphasis has & Nordgren, 2006). Of course, it is pos-
sometimes led to the misunderstanding that sible that conscious and unconscious pro-
labeling a process as automatic or nonauto- cessing of input stimuli produces different
matic is sufficient to characterize the intrin- behavioral outputs via two qualitatively
sic nature of that process. However, stat- distinct mechanisms. However, it is also
ing that a process operates in an automatic possible that conscious and unconscious
or nonautomatic fashion simply specifies processes operate on the basis of the same
when the process is assumed to operate; it mental structures involving the same men-
does not specify how the process translates tal operations (Huang & Bargh, in press).
inputs into outputs. That is, characterizing Simply stating that a mental process is con-
a mental process as automatic versus non- scious or unconscious does not specify how
automatic specifies whether the process this process translates inputs into outputs,
does or does not operate (1) when there is nor does it specify whether conscious and
no conscious awareness, (2) when there is no unconscious processing involve the same or
goal to start the process, (3) when cognitive different mental structures and operations.
resources are reduced, and (4) when there is Thus, although dual-­ process theorizing is
a goal to alter or stop the process (Bargh, often equated with research on awareness,
1994). Thus, although research investi- intentionality, efficiency, and controllability
gating the unawareness, unintentionality, (i.e., operating conditions), its explanatory
efficiency, and uncontrollability of a given goal at the algorithmic level requires clear
process is essential to understand the bound- specifications of the mental mechanisms
ary conditions of observed input–­ output that translate inputs into outputs (i.e., oper-
relations (e.g., does a given input lead to a ating principles). 2
particular output when cognitive resources Another important caveat in this context
are reduced?), such research by itself does is that different features of automatic pro-
not address the question of how the human cessing do not necessarily co-occur (Bargh,
mind translates certain inputs into particu- 1994; Moors & De Houwer, 2006). Coun-
lar outputs. ter to early dual-mode conceptualizations
In terms of Marr’s (1982) framework, assuming an all-or-none relation between
research on operating conditions is located different features of automaticity, the avail-
at the computational level, in that it aims at able evidence indicates that there is virtu-
identifying which types of inputs produce ally no process that is characterized by all
which kinds of outputs under which contex- four features of automaticity. Instead, most
tual conditions (e.g., does input X produce processes studied within social psychology
output Y when participants simultaneously involve combinations of selected features,
perform a secondary task?). Although such making them automatic in one sense and
research is essential to the concern with nonautomatic in another (Bargh, 1992). For
automaticity, the explanatory goal of dual-­ example, a process may be unintentional
process theories goes beyond the computa- and controllable, intentional and efficient,
tional level, in that they aim at specifying unintentional and resource-­ dependent,
the mental mechanisms that translate inputs conscious and uncontrollable, unconscious
into outputs (e.g., what are the mental oper- and resource-­ dependent, controllable and
6 THE B A SIC S

resource-­ independent, and so forth (e.g., coherent categories depending on whether

Fujita, 2011; Gawronski & Bodenhausen, they operate in an automatic or nonauto-
2011; Hassin, Bargh, Engell, & McCulloch, matic fashion should be treated with cau-
2009). This insight has inspired disjunctive tion. Of course, dual-­process theories share
conceptualizations of automaticity, accord- the explanatory goal to identify the mental
ing to which a process can be character- mechanisms that translate inputs into out-
ized as automatic if it meets at least one of puts and the assumption that input–­output
the four criteria of automaticity. According relations are mediated by two qualitatively
to this view, a process can be described as distinct (sets of) processes that operate
automatic if it is (1) unconscious, (2) unin- under different processing conditions. How-
tentional, (3) efficient, or (4) uncontrollable. ever, different theories emphasize different
Although disjunctive treatments of auto- features of automaticity, and the nature of
maticity are rather common in social psy- the proposed mechanisms is not necessarily
chology, they involve a number of problems equivalent.
(Gawronski & Creighton, 2013). First, if the
presence of a single feature is sufficient to
call a process automatic, it is possible that Types of Dual-­Process Theories
a given process has to be described as auto-
matic and nonautomatic at the same time. Despite the concerns about generic treat-
Needless to say, such a description can cause ments, the term dual-­process theory is
considerable confusion if it does not specify sometimes used in the singular to refer to
in which particular sense the process is auto- an overarching theoretical idea that could
matic and in which sense it is nonautomatic. be boiled down to a single theory (see Evans
Second, generic use of the term automatic & Frankish, 2009). However, as the variety
to describe any of the four operating condi- of contributions to this volume illustrates,
tions can lead to confusion about conceptu- there is not really a unifying “essence” that
ally distinct findings that are described with captures what might be regarded as the pro-
the same term. For example, a given Process totype of dual-­ process theories. Instead,
A may be described as automatic because it there are important nuances that are easy
is elicited unintentionally, whereas another to miss if different theories are treated as
Process B may be described as automatic minor variations of the same prototype
because it does not require a large amount theory. Nevertheless, it seems possible to
of cognitive resources. Yet despite their com- classify dual-­process theories in terms of a
mon description as automatic, the two pro- few general characteristics, such as the phe-
cesses may be fundamentally different, for nomena they aim to explain (explanandum),
example, if Process A is resource dependent the theoretical constructs they propose to
and Process B is intentional. explain these phenomena (explanans), and
Based on these considerations, several the- their mathematical formalization.
orists recommended that researchers should
be more precise in their use of terminology
Explanandum
by describing each feature of automatic-
ity with its proper label (i.e., unconscious, When dual-­ process theories started to
unintentional, efficient, or uncontrollable; emerge in the 1980s, their focus was
Bargh, 1994; Moors & De Houwer, 2006). mainly domain-­specific, in that they aimed
Importantly, such decompositional concep- at explaining phenomena in particular
tualizations prohibit simple binary catego- areas of inquiry. Although some of these
rizations of mental processes as automatic theories were based on general processing
or nonautomatic. Because the four features principles from cognitive psychology (e.g.,
of automaticity do not necessarily co-occur, Chaiken, 1987; Trope, 1986), their appli-
a decompositional classification schema cations were specific to particular content
involves 16 rather than two categories of domains within social psychology. Promi-
potential operating conditions. On the basis nent examples include dual-­process theories
of this conclusion, the assumption that the of persuasion (e.g., Chaiken, 1987; Petty &
mental processes underlying social phe- Cacioppo, 1986), attitude–­ behavior rela-
nomena can be divided into two internally tions (e.g., Fazio, 1990), dispositional attri-
 A Conceptual Analysis of Dual-Process Theories 7

bution (e.g., Gilbert, 1989; Trope, 1986), the common denominator of various gener-
prejudice and stereotyping (e.g., Devine, alized dual-­process theories, including theo-
1989), and impression formation (e.g., ries that distinguish between reflective and
Brewer, 1988; Fiske & Neuberg, 1990). A impulsive processing (Strack & Deutsch,
shared feature of these theories is that they 2004), reflective and reflexive processing
explain particular instances of input–­output (Lieberman et al., 2002), and System 1 ver-
relations (explanandum) by postulating two sus System 2 processing (Kahneman, 2003;
(sets of) mechanisms by which the human Stanovich & West, 2000). Deviating from
mind translates inputs into outputs (explan- the initial focus on domain-­ specific phe-
ans). For example, dual-­ process theories nomena, generalized dual-­ process theories
of persuasion aim at explaining the effects are concerned with the basic architecture of
of different features of persuasive mes- information processing (Carruthers, 2009;
sages on attitudes. Their overarching goal Samuels, 2009). Thus, generalized dual-­
is to understand the mental mechanisms process theories differ from domain-­specific
by which different kinds of inputs (e.g., dual-­process theories in terms of the breadth
argument strength, source characteristics) of their explanandum, such that the latter
are translated into outputs (i.e., attitude aim at explaining input–­output relations in
change). Similarly, dual-­process theories of particular content areas, whereas the former
dispositional attribution aim at explaining aim at providing a general account of the
the effects of different kinds of information workings of the human mind.
about potential causes of a person’s behavior
on perceivers’ trait impressions. Again, the
Explanans
overarching goal is to understand the men-
tal mechanisms by which different kinds Beyond differences in terms of the phenom-
of inputs (e.g., behavioral information, ena that dual-­process theories aim to explain,
situational information) are translated into there are nontrivial differences in the mental
outputs (i.e., dispositional attributions). In constructs they propose to explain the phe-
general, domain-­specific dual-­process theo- nomena of interest. Whereas some theories
ries differ in terms of the phenomena they emphasize functionally distinct mental pro-
aim to explain, in that they focus on input–­ cesses (dual-­process theories), other theories
output relations in different content areas attribute different behavioral outcomes to
(explanandum). Although the proposed functionally distinct mental representations
explanations are specific to the phenomena (dual-­representation theories). Moreover,
of interest, their shared assumption is that some theories include assumptions about
different kinds of input–­output relations are both process and representation, assuming
mediated by two (sets of) qualitatively dis- that different outcomes are the product of
tinct processes that operate under different two functionally distinct processing systems
conditions (explanans). (dual-­system theories).
With the beginning of the new millen- A defining feature of dual-­process theo-
nium, the focus of dual-­process theorizing ries is that they explain different kinds of
shifted toward the development of integra- input–­output relations on the basis of two
tive theories that aim at identifying general (sets of) mental mechanisms. For example,
principles that are independent of particular Fiske and Neuberg’s (1990) continuum
content domains. A seminal contribution model of impression formation distin-
in this regard was Smith and DeCoster’s guishes between categorization processes
(2000) conceptual integration of various and piecemeal integration of individuating
domain-­ specific theories within a single attributes. Similarly, Trope’s (1986) two-
dual-­process framework. The central argu- stage model of dispositional attribution
ment of their integrative account is that the distinguishes between perceptual identifi-
multiple dualisms proposed by domain-­ cation of trait-­relevant cues and subsequent
specific theories reflect the operation of two inference of dispositions. Many of these
basic processes that characterize any kind of theories include empirical hypotheses about
human thought irrespective of its content: systematic covariations between operat-
associative versus rule-based processes (cf. ing principles and operating conditions, in
Sloman, 1996). This distinction has become that the proposed mechanisms are assumed
8 THE B A SIC S

to operate under different conditions. For Fazio, 2007). Whereas dual-­representation

example, Fiske and Neuberg’s (1990) con- accounts are based on presumed parallels to
tinuum model assumes that the initial cat- the distinction between explicit and implicit
egorization of a target person requires little memory, dual-­ process accounts emphasize
amounts of cognitive resources, whereas the the unintentional activation of attitudes
piecemeal integration of individuating attri- and their reduced impact on overt responses
butes is assumed to be resource-­dependent. when people have the motivation and oppor-
Similarly, Trope’s (1986) two-stage model tunity to deliberate about specific attributes
assumes that the perceptual identification of the attitude object (for a more detailed
of trait-­relevant cues occurs unintentionally, discussion, see Payne & Gawronski, 2010).
whereas dispositional inference is assumed A third category of theories explains dif-
to be an intentional process (for a review, ferent behavioral outcomes by the opera-
see Gawronski & Creighton, 2013). Thus, tion of two distinct processing systems. The
in addition to differences in the particu- shared assumption of such dual-­system the-
lar phenomena they aim to explain, dual-­ ories is that multiple psychological dualities
process theories can be characterized by are systematically correlated, thereby consti-
their assumptions about the nature of the tuting two functionally distinct mental sys-
two processes they propose to explain these tems. Although dual-­system theories differ
phenomena (i.e., operating principles) and in their assumptions about which dualities
the particular conditions under which each represent core features of the proposed sys-
of the two processes is assumed to operate tems, the hypothesized correlations between
(i.e., operating conditions). dichotomous characteristics are often
Deviating from the emphasis on processes, depicted in lists of features that describe one
dual-­representation theories attribute dif- of the two systems as associative, automatic,
ferent behavioral outcomes to distinct slow-­learning, experiential, affective, paral-
mental representations. A useful example lel, and holistic, and the other one as rule-
to illustrate the difference between dual-­ based, nonautomatic, fast-­learning, ratio-
process and dual-­ representation theories nal, cognitive, sequential, and analytic (e.g.,
is research comparing evaluative responses Epstein, 1994; Kahneman, 2003; Sloman,
on traditional self-­ report measures to 1996; Smith & DeCoster, 2000). Although
responses on performance-­based tasks, such some dual-­system theories limit their focus
as the Implicit Association Test (Greenwald, to particular content domains (e.g., Rydell
McGhee, & Schwartz, 1998) or the evalua- & McConnell, 2006), the majority falls
tive priming task (Fazio, Jackson, Dunton, into the category of generalized theories
& Williams, 1995). A common finding in that aim at providing a domain-­independent
this area is that evaluative responses on the account of the workings of the human mind
two kinds of measures show various dis- (e.g., Smith & DeCoster, 2000; Strack &
sociations, including different antecedents, Deutsch, 2004).
different consequences, and discrepant eval-
uations of the same object (for reviews, see
Formalization
Friese, Hofmann, & Schmitt, 2008; Gaw-
ronski & Bodenhausen, 2006). Such dissoci- The different types of dual-­process theories
ations have led some researchers to conclude discussed so far share the feature that they
that self-­report measures reflect conscious, rely only on the informal logic of verbally
“explicit” attitudes, whereas performance-­ formulated propositions rather than math-
based tasks reflect unconscious, “implicit” ematical formalization (for a notable excep-
attitudes (e.g., Greenwald & Banaji, 1995; tion, see Trope, 1986). Over the past decade,
Rydell & McConnell, 2006; see also Wilson, some researchers have started to develop
Lindsey, & Schooler, 2000). Yet counter to formalized theories in which the mecha-
such dual-­ representation accounts, dual-­ nisms that are assumed to mediate input–­
process accounts assume that the two kinds output relations are specified in mathemati-
of measures capture the same attitudinal cal terms. The most prominent example is
representation, the primary difference being Jacoby’s (1991) process dissociation (PD)
the reduced opportunity to control responses model (for a review, see Payne & Bishara,
on performance-­ based measures (e.g., 2009). Other, more complex theories rely on
 A Conceptual Analysis of Dual-Process Theories 9

multinomial modeling (e.g., Conrey, Sher- inconsistent assumptions at various stages

man, Gawronski, Hugenberg, & Groom, of the research process, including the initial
2005; Krieglmeyer & Sherman, 2012; specification of a theory, the generation of
Meissner & Rothermund, 2013; Nadarevic predicted data patterns, the estimation of
& Erdfelder, 2010; Payne, Hall, Cameron, model parameters from observed data, and
& Bishara, 2010; Stahl & Degner, 2007) the assessment of fit between predicted and
and applications of random-­walk and diffu- observed data (Klauer, in press).
sion modeling (e.g., Klauer, Voss, Schmitz, Third, formalized theories have the
& Teige-­Mocigemba, 2007). Regardless of advantage that they directly link the pro-
the specifics of their mathematical under- posed processes to relations between inputs
pinnings, formalized dual-­process theories and outputs. In a strict sense, psychological
have at least three advantages over nonfor- measures do not assess mental processes or
malized theories. mental representations, but their behavioral
First, formalized theories allow research- outcomes (De Houwer, 2011). Although it
ers to quantify the contribution of multiple is rather common to treat behavioral out-
distinct processes to a given behavioral out- comes as “proxies” for mental constructs,
come, which is not possible on the basis of such treatments involve the logical fallacy
verbally formulated theories. The signifi- of affirming the consequent, also known as
cance of this characteristic is reflected in reverse inference (Gawronski & Bodenhau-
the principle of equifinality, which refers sen, in press-a). One possibility to avoid this
to cases in which combinations of differ- problem is to distinguish clearly between the
ent processes produce the same behavioral behavioral outputs captured by psychologi-
outcome. For example, in research on self-­ cal measures and the mental constructs that
regulation, two people may show the same are proposed to explain input–­output rela-
behavioral response when (1) the initial tions. The validity of dual-­process theories
impulse and inhibitory control are weak or (like any other mental process theory) can
(2) the initial impulse and inhibitory control then be tested by deriving predictions about
are strong (Sherman et al., 2008). Formal- input–­ output relations and the particular
ized theories are able to capture such com- conditions under which they should emerge
plex interplays by providing quantitative (Gawronski & Bodenhausen, in press-b). Yet
estimates for each of the proposed processes. the derivation of such predictions requires
Second, formalized theories have the “bridging” assumptions that link the pro-
advantage that their logical coherence can be posed mental constructs to inputs and out-
tested by means of their mathematical con- puts (McGrath, 1981). Domain-­ specific
straints (Klauer, in press). For many verbally dual-­process theories typically include such
formulated theories it can be very difficult to assumptions, but these theories have been
identify logical inconsistencies between their criticized for referring to particular contents
core assumptions, particularly when they (e.g., argument strength vs. source charac-
involve a large number of theoretical claims. teristics) in linking phenomenon-­ relevant
To illustrate this problem, consider a theory inputs to the proposed processes (e.g., cen-
that includes a set of N propositions. Logi- tral/systematic vs. peripheral/heuristic pro-
cally, any subset of N – 1 of these proposi- cessing). This strategy seems problematic
tions could be internally consistent even if to the extent that the processing of differ-
the entire set N is inconsistent. This possi- ent information involves the same mental
bility implies that an exhaustive consistency operations irrespective of its content (see
assessment of N propositions requires scru- Kruglanski, Erb, Pierro, Mannetti, & Chun,
tiny of 2N cases. The resulting capacity prob- 2006; Sherman, 2006). Generalized dual-­
lem is illustrated by the fact that, even if each process theories avoid this problem by pos-
case could be examined in a millionth of a tulating two (sets of) content-­ independent
second, an exhaustive consistency check for processes. However, many of these theories
a theory including 100 propositions would focus primarily on the internal architecture
take longer than the universe has existed (see of the human mind, without specifying links
Johnson-­Laird, 2012). Formalized theories between the proposed processes and the rel-
avoid this problem, because their underlying evant inputs. Formalized dual-­process theo-
mathematical constraints help to identify ries avoid both problems by (1) proposing
10 THE B A SIC S

general processes that are independent of tions under which these processes oper-
particular contents and (2) directly linking ate. For example, a common assumption
the proposed processes to observed input–­ of domain-­specific dual-­process theories is
output relations in their mathematical for- that the processes underlying the effects of
mulations. certain kinds of inputs are intentional and
An important aspect of formalized dual-­ resource-­dependent, whereas those underly-
process theories is the relation between their ing the effects of other kinds of inputs are
processing parameters and the concept of unintentional and resource-­ independent.
automaticity. For example, the processing Examples include processing constraints on
parameters of Jacoby’s (1991) PD model the effects of central versus peripheral cues
have sometimes been interpreted as captur- in dual-­process theories of persuasion (e.g.,
ing automatic and nonautomatic processes Petty & Cacioppo, 1986), behavioral ver-
by virtue of its underlying mathematical sus situational information in dual-­process
structure. This interpretation is reflected in theories of dispositional attribution (e.g.,
the common depiction of the two parameter Gilbert, 1989), and category versus individ-
estimates as automatic (using the acronym uating information in dual-­process theories
A) and controlled (using the acronym C). of impression formation (e.g., Fiske & Neu-
However, such depictions conflate the dif- berg, 1990). Similarly, dual-­ representation
ference between operating principles and theories often attribute a particular feature
operating conditions. Although formal- of automatic processing (e.g., unconscious)
ized theories provide clear specifications to one representation and the opposite fea-
of how inputs are translated into outputs, ture (e.g., conscious) to the other representa-
their mathematical underpinnings do not tion (e.g., Greenwald & Banaji, 1995). Yet
have any implications for when the pro- the idea of covariation between dualities is
posed processes operate (i.e., when there most explicit in dual-­system theories, which
is no conscious awareness; when cognitive propose systematic relations between mul-
resources are reduced; when there is no goal tiple distinct dualities (e.g., Epstein, 1994;
to start the process; when there is a goal to Kahneman, 2003; Lieberman et al., 2002;
alter or stop the process). Questions about Sloman, 1996; Smith & DeCoster, 2000;
the conditions under which a given process Stanovich & West, 2000; Strack & Deutsch,
operates cannot be addressed by mathemati- 2004).
cal formalizations but have to be answered Although assumptions about covariations
on the basis of empirical data. For example, between psychological dualities are very
whether the particular processes captured common, their epistemic value depends on
by the two PD parameters depend on the two conceptual issues (see Moors, Chapter
availability of cognitive resources has to be 2, this volume). First, it is important that the
investigated by testing effects of cognitive two categories underlying a given duality
load or time pressure on the two parameters, are characterized by a clear demarcation, so
and there is nothing in the underlying PD that they constitute nonoverlapping catego-
formulas that would guarantee one or the ries. For example, the distinction between
other outcome. The same is true for all other categorical and individuating information in
features of automaticity, including unaware- dual-­process theories of impression forma-
ness, unintentionality, and uncontrollability tion has been criticized for being ambiguous
(for a more detailed discussion, see Gawron- as to whether a given target characteristic
ski & Creighton, 2013). should be regarded as categorical informa-
tion (e.g., man vs. woman; adolescent vs.
elderly) or as an individuating attribute (e.g.,
Covariation of Dualities male vs. female; young vs. old) (see Kunda
& Thagard, 1996). Second, it is important
Many dual-­process theories hypothesize sys- to distinguish between conceptual defini-
tematic covariations between two or more tions and empirical hypotheses when pro-
psychological dualities (Samuels, 2009). posing systematic relations between two or
At the most basic level, these covariations more dualities. For example, to the extent
involve empirical relations between func- that associative processes are defined as
tionally distinct processes and the condi- unconscious and propositional processes
 A Conceptual Analysis of Dual-Process Theories 11

are defined as conscious (e.g., Mitchell, De is the common equation of System 1 process-
Houwer, & Lovibond, 2009), the relation ing with resource-­independence and System
between the two dualities would be purely 2 processing with resource-­dependence (e.g.,
semantic rather than empirical (cf. Smed- Dhar & Gorlin, 2013). The problem is that
slund, 2000; Wallach & Wallach, 1994). such an equation does not offer anything
In this case, it would make no sense to test beyond post hoc classifications of observed
empirically whether propositional processes effects (Gawronski, 2013). To the extent
require conscious awareness, because that that a given effect is resource-­independent it
is how propositional processes are defined will be categorized as being due to System
in the first place. Any process that is con- 1, but it will be attributed to System 2 if it
scious would be propositional by defini- is resource dependent. Moreover, if an effect
tion. Yet, if the associative–­ propositional that was initially attributed to System 1 turns
dualism is defined without reference to the out to depend on cognitive resources, this
unconscious–­conscious dualism (e.g., Gaw- effect would simply be recategorized as the
ronski & Bodenhausen, 2011), assumptions product of System 2, and vice versa. Without
about systematic covariations between the a clear specification of the operating prin-
two dualities could be subject to empirical ciples of System 1 and System 2 processing,
tests to the extent that (1) the associative–­ the theory does not impose any constraints
propositional distinction specifies particular on the interpretation of a given result. Thus,
input–­output relations that can be expected criticism of dual-­process theories as being
on the basis of the two processes (operat- unfalsifiable can be avoided by (1) conceptu-
ing principles), and (2) the unconscious–­ ally precise definitions of the proposed dual-
conscious distinction specifies the conditions ities and (2) clear formulations of empirical
under which these input–­ output relations hypotheses about their covariations.
should emerge (operating conditions). Thus, To the extent that the two requirements are
a basic requirement for conceptually sound met, dual-­process theories are in fact highly
dual-­process theorizing is that the categories falsifiable, counter to the common criti-
of conceptually distinct dualities are defined cism that they are unfalsifiable. According
in a manner that avoids semantic overlap to Popper (1934), the falsifiability of a given
between these dualities (Moors, Chapter 2, theory increases with the number of events
this volume). Moreover, although concep- that are prohibited by the theory (see Gaw-
tual definitions and empirical hypotheses ronski & Bodenhausen, in press-a). Thus, by
are often conflated in claims about covaria- assuming systematic overlap between two
tions between multiple dualities, it is impor- conceptually distinct dualities, dual-­process
tant that the categories constituting a given theories can be said to prohibit two out of
duality are defined in a manner that allows four possible events. For example, dual-­
unambiguous demarcations between these process theories of dispositional attribution
categories. claiming that the impact of situational infor-
mation is mediated by a resource-­dependent
process, whereas the impact of behavioral
Are Dual-­Process information is mediated by a resource-­
Theories Unfalsifiable? independent process, would be discon-
firmed by any finding showing that either
The quest to distinguish clearly between con- (1) the impact of behavioral information is
ceptual definitions and empirical hypoth- resource-­dependent or (2) the impact of situ-
eses is particularly important in the context ational information is resource-­independent
of criticism that dual-­process theories are (e.g., Krull, 1993; Trope & Gaunt, 2000).
unfalsifiable (e.g., Keren & Schul, 2009). Importantly, the number of events prohib-
If there is no demarcation between the con- ited by dual-­process theories—­and thus their
ceptual definitions of the proposed dualities falsifiability—­increases with the number of
and empirical hypotheses about covariations proposed covariations between dualities.
between dualities, there is a considerable For example, a dual-­systems theory assum-
risk that the value of dual-­process theories ing systematic overlap between four con-
will be reduced to post hoc categorizations ceptually distinct dualities would prohibit
of empirical effects. An illustrative example 14 out of 16 possible combinations. Such a
12 THE B A SIC S

theory would be highly falsifiable, because it attributes.3 Thus, enhanced effects of any
would be disconfirmed by any one of the 14 given attribute may be explained by either
cases that are prohibited by the theory. To be (1) dominance of System 1 processing involv-
sure, certain assumptions about systematic ing the replacement of a target attribute by
overlap between multiple dualities may be another heuristic attribute or (2) dominance
rejected as empirically implausible or incon- of System 2 processing involving the replace-
sistent with the available evidence (Keren ment of the output of System 1 processing.
& Schul, 2009; Kruglanski & Gigerenzer, In the absence of “bridging” assumptions
2011). For example, the assumption that that link the hypothesized mental constructs
the associative–­ propositional, affective–­ to external inputs and overt outputs, it can
cognitive, and unconscious–­conscious duali- be rather difficult to determine whether a
ties have systematic overlap may be rejected given theory is consistent or inconsistent
on the grounds that there are cognitive asso- with a particular finding (McGrath, 1981).
ciative processes, unconscious cognitive pro- This difficulty can lead to the impression
cesses, conscious affective processes, and so that dual-­process theories are unfalsifiable,
forth. However, because such rejections are although their assumptions about covaria-
based on empirical arguments, they invali- tions between two or more dualities clearly
date rather than support the common con- prohibit a large number of possible events.
tention that dual-­process theories are unfal- In other words, although dual-­process the-
sifiable. ories are falsifiable in principle if they are
Despite our rejection of generalized claims evaluated in terms of Popper’s (1934) struc-
of unfalsifiability, we believe that there is tural criterion, some of them are practically
another, less obvious feature that can make unfalsifiable, because they are formulated
dual-­process theories unfalsifiable. Many in a manner that makes it difficult to deter-
domain-­specific dual-­process theories were mine which relations between external stim-
formulated in a manner such that the rel- uli and overt behavioral responses would be
evant inputs were specified at the level of inconsistent with these theories. This fea-
external stimuli (e.g., source characteristics, ture allows researchers to apply some dual-­
situational information, category member- process theories to virtually every empirical
ship). The same was true for the proposed outcome in a post hoc fashion (e.g., post hoc
outputs, which were clearly specified at categorizations of a given effect as being
the level of overt responses (e.g., evaluative driven by System 1 vs. System 2). However,
judgments, trait judgments). These specifi- it is rather difficult to derive a priori predic-
cations made it relatively easy to determine tions that could pose a strong challenge to
whether a theoretically derived prediction these theories.
about input–­ output relations is consistent
or inconsistent with a given finding. With
the emergence of generalized dual-­ process How Many Processes Are There?
theories, the focus shifted toward properties
of the mind’s internal architecture. In these A common question about dual-­ process
theories, many of the core assumptions are theories is whether the realm of mental pro-
specified at the mental level, without refer- cesses can indeed be meaningfully divided
ence to external stimuli and overt responses. into two distinct categories. Whereas some
For example, in Kahneman’s (2003) dual-­ theorists have argued that dual-­ process
system theory of judgment and choice, Sys- theories clearly demonstrated their explana-
tem 1 processing is claimed to involve a tory and predictive value (Deutsch & Strack,
process of attribute substitution, in which a 2006), others have argued for the superior-
target attribute is substituted by a heuristic ity of single-­process (e.g., Kruglanski et
attribute that comes more readily to mind. al., 2006) or multiple process alternatives
Moreover, System 2 processing is assumed (e.g., Sherman, 2006). Yet when discuss-
to monitor and, if necessary, modify or over- ing the question of how many processes
ride the outputs of System 1. However, the there “really” are, it is important to note
theory itself does not specify which stimu- that existence claims—­ including claims
lus attributes count as target attributes and about the existence of one, two, or multiple
which ones should be regarded as heuristic processes—­are ontological in nature. In the
 A Conceptual Analysis of Dual-Process Theories 13

philosophy of science, ontological claims fall assumptions over and above the hypothesis
into the realm of metaphysics, which means that information processing is guided by a
that they cannot be tested empirically (e.g., single process (e.g., modulation of this pro-
Popper, 1934; Quine, 1960). From this per- cess by five conceptually distinct parameters;
spective, it is not possible to test whether see Kruglanski et al., 2006). Thus, when
there are one, two, or multiple processes. evaluating theories on the basis of their par-
However, researchers can make decisions simony, it does not suffice to count the num-
about the usefulness of ontological claims ber of processes they propose. What matters
by empirically testing assumptions about for the criterion of parsimony is the total
the proposed processes. To the extent that number of assumptions that is required to
dual-­process theories help us to understand explain a given finding.
past observations of input–­ output rela- A final issue concerns the most appropri-
tions (explanation) and correctly forecast ate level of abstraction in theorizing about
future observations of input–­ output rela- mental processes. In some sense, every men-
tions (prediction), there is reason to believe tal process can be described by the shared
that their assumptions are correct, includ- feature that they follow if–then rules. Some
ing their ontological assumptions about the researchers have interpreted this possibility
existence of certain mental processes. How- as an argument for the superiority of single-­
ever, if the predictions of a given theory are process theories that attribute all kinds of
continuously disconfirmed, it seems likely input–­output relations to the same process
that researchers will at some point reject its of rule-based inference (e.g., Kruglanski et
underlying ontological claims (cf. Peters & al., 2006). However, descriptions of mental
Gawronski, 2011). Note, however, that in processes at such a high level of abstraction
such cases it is not the existence-­claim itself do not go far beyond claiming that all men-
that is confirmed or disconfirmed, but the tal processes follow some kind of regular-
assumptions that are made about the pro- ity instead of being random (Gawronski &
posed entities. In this sense, the number of Creighton, 2013). Moreover, to the extent
mental processes is not an objective fact that that such if–then conditionals directly refer
can be studied independently of the assump- to inputs as the antecedent and outputs as
tions that are made about them. Instead, the the consequent, their theoretical claims
number of mental processes that we assume would be located at Marr’s (1982) compu-
to exist inherently depends on our theories, tational level of analysis. However, compu-
in that we accept the ontology of those theo- tational analysis fails to specify the mental
ries that help us to understand past obser- processes and representations that translate
vations (explanation) and correctly forecast inputs into outputs at the algorithmic level,
future observations (prediction). which is a central concern of dual-­process
An important issue in this context is the theorizing. In this sense, the debate between
quest for parsimony, which stipulates that advocates of single-­process and dual-­process
researchers should favor theories that involve theories seems at least partly spurious, in
fewer assumptions to explain a particular that the conflicting claims refer to different
empirical finding (Gawronski & Boden- levels of analysis (cf. De Houwer & Moors,
hausen, in press-a). Single-­process theorists in press; see also Sherman, 2006).
sometimes appeal to the quest for parsimony, Similar considerations apply to the debate
arguing that dual-­process theories are less between dual-­process and multiple-­process
parsimonious than single-­ process theories, theories. In some sense, every process can
because they postulate two qualitatively be further analyzed for subcomponents
distinct processes rather than a single one. that constitute this process. For example,
However, an often-­overlooked aspect of par- although the distinction between associative
simony is that it refers to the total number and propositional processes may serve as
of theoretical assumptions that are required the conceptual foundation of a dual-­process
to explain a given finding rather than the theory of evaluation (Gawronski & Boden-
number of assumptions of what might be hausen, 2006), the functional principles of
considered the core of a given theory. To the two processes may be further divided
explain a particular finding, single-­process depending on whether they operate during
theories have to rely on a host of additional the formation or the expression of evaluative
14 THE B A SIC S

representations. Moreover, one could argue assumptions about their operating condi-
that the process of propositional validation tions reflect empirical hypotheses about
involves a number of subprocesses, including whether they operate (1) when there is no
the default affirmation of validity, the moni- conscious awareness, (2) when there is no
toring of consistency, and the resolution of goal to start the process, (3) when cognitive
inconsistency (Gawronski & Bodenhau- resources are reduced, and (4) when there is
sen, 2011). In this sense, every dual-­process a goal to alter or stop the process. Operat-
theory may be regarded as an oversimpli- ing conditions are typically investigated by
fication, in that its proposed dualities can means of appropriate manipulations, includ-
always be divided into multiple subordinate ing supraliminal versus subliminal presenta-
processes. In fact, for many theories that tion times (awareness), the presence versus
have been categorized as dual-­process theo- absence of instructions to perform a par-
ries, the term is actually a misnomer given ticular mental operation (intentionality),
that the majority of these theories propose the relative difficulty of a simultaneously
more than two processes that are involved in performed secondary task (efficiency), and
the translation of inputs into outputs. instructions not to perform a particular
We argue that the optimal level of theo- mental operation (controllability). Any of
retical abstraction cannot be determined a these manipulations can be conceptualized
priori but has to be decided on the basis of as involving contextual inputs (or input
the research question (Sherman, 2006). For characteristics) that may moderate the rela-
many social psychological questions, the tion between a primary input and its output
conceptual distinctions proposed by dual-­ (De Houwer & Moors, 2012). In line with
process theories have clearly demonstrated this conceptualization, Marr (1982) located
their value in explaining and predicting the operating conditions at the computational
phenomena of interest. However, for other rather than the algorithmic level, which is
questions, more fine-­ grained theories may consistent with our emphasis on the distinc-
be needed to account fully for the available tion between operating principles and oper-
evidence. Yet regardless of the chosen level ating conditions.
of abstraction, it is important that the pro- Nevertheless, it is certainly possible to
posed process dimensions be clearly defined investigate the mental processes by which
and their conceptual definitions be distin- the inputs of operating conditions influ-
guished from empirical hypotheses about ence the input–­output relations produced by
covariations between different dimensions. another process. In this case, the moderat-
ing effect of a given input on other input–­
output relations represents the phenomenon
Emerging Themes that needs to be explained (explanandum),
and Future Directions and researchers may develop mental process
theories that explain how this input moder-
An interesting line of research that is start- ates input–­output relations (explanans). For
ing to emerge from dual-­ process theoriz- example, researchers may investigate the
ing concerns the mental processes underly- mental processes by which the affordances
ing operating conditions. This statement of secondary tasks influence the operation of
may seem a little puzzling given our strong another mental process. Similarly, research-
emphasis on the distinction between oper- ers may be interested in the mental processes
ating principles and operating conditions. by which instructions to inhibit a particular
Yet it is entirely possible to stipulate a clear process influence the operation of that pro-
distinction between operating principles and cess. Such theories differ from traditional
operating conditions and, at the same time, dual-­process theories, in that they focus on
ask questions about the processes underly- how secondary processes influence the oper-
ing operating conditions. ation of the primary processes postulated by
Traditionally, dual-­process theories aimed traditional dual-­process theories. Applied to
at identifying the mental processes underly- the four features of automaticity, potential
ing particular kinds of input–­ output rela- themes of such theories may include (1) the
tions. These processes are conceptually processes by which subliminal versus supra-
defined by their operating principles, and liminal stimulus presentations moderate the
 A Conceptual Analysis of Dual-Process Theories 15

execution of a primary process, (2) the pro- process theories to be generative and appli-
cesses by which the presence versus absence cable to a broad spectrum of social psycho-
of instructions to perform a particular men- logical phenomena. Since the publication
tal operation moderates the execution the of Chaiken and Trope’s (1999) seminal vol-
primary process, (3) the processes by which ume, Dual-­Process Theories in Social Psy-
secondary task performance moderates the chology, these theories have continued to be
execution of a primary process, and (4) the highly generative and have further expanded
processes by which instructions not to per- the range of application. At the same time,
form a particular mental operation moder- attempts at integration have shifted the focus
ate the execution of the primary process. to formulating unitary distinctions between
What may already be clear from this con- different mental processes that cut across
ceptualization is that research on the mental multiple domains. Our hope is that future
processes underlying operating conditions advances will help dual-­process theories to
requires a sufficient understanding of the further enhance the basic science desiderata
input–­output relations that are produced by of conceptual coherence, precision, general-
the relevant primary processes. Although ity, and testability without sacrificing their
such an assumption may seem questionable applicability to the problems humans face as
to critics of dual-­process theorizing, cogni- members of dyads, groups, and societies.
tive science has a long history of research In this spirit, our main goal in this chapter
on similar questions, including the nature of has been to analyze the explanatory func-
consciousness (Baars, 2002), working mem- tion of dual-­process theories, the conceptual
ory (Baddeley, 2010), and executive control nature of their underlying dualities, and the
(Braver, 2012). Of course, the paradigms structural features that characterize differ-
in this research tend to be much less com- ent types of dual-­process theories. Drawing
plex than the ones typically used by social on Marr’s (1982) distinction among compu-
psychologists. Nevertheless, we believe that tational, algorithmic, and implementational
research under the guidance of dual-­process levels of analysis, we have argued that a cen-
theories has accumulated a sufficiently large tral explanatory function of dual-­ process
body of knowledge to move to the next level theories is to specify the mental mechanisms
of inquiry by studying the interplay between that translate inputs into outputs. From this
the proposed primary processes and the sec- perspective, descriptions of a given process
ondary processes that determine the condi- as automatic versus nonautomatic simply
tions of their operation. To the extent that specify when the process is assumed to oper-
the theories inspired by this research adhere ate (operating conditions). However, they
to the distinction between operating prin- do not specify how the proposed process
ciples and operating conditions for both translates inputs into outputs (operating
primary and secondary processes, they may principles). The latter question requires clear
also provide valuable insights into when specifications of the underlying mental oper-
and why certain features of automaticity ations, which are essential for mechanistic
do or do not co-occur (Moors & De Hou- explanations at Marr’s algorithmic level of
wer, 2006). Examples of emerging themes analysis. To the extent that dual-­ process
in this regard include recent claims about theories include precise and nonoverlap-
the resource-­ independence of inhibitory ping definitions of the proposed dualities,
control (e.g., Fujita, 2011) and the resource-­ their hypotheses about covariations between
dependence of unconscious processing (e.g., processing dualities are highly falsifiable,
Hassin et al., 2009). in that they prohibit a considerable num-
ber of possible events. Yet the explanatory
and predictive value of dual-­process theories
Conclusion essentially depends on (1) a clear distinction
between operating principles and operating
Emerging in the 1980s, the first generation conditions, (2) conceptual rigor in the defi-
of dual-­process theories has yielded a multi- nition of the proposed dualities, (3) precise
plicity of dualities in theorizing about social formulations of empirical hypotheses about
information processing. Cumulatively, this covariations between processing duali-
work has demonstrated the ability of dual-­ ties, and (4) clearly specified links between
16 THE B A SIC S

the hypothesized mental processes and the Bargh, J. A. (1994). The four horsemen of auto-
causal relations between stimuli and behav- maticity: Awareness, intention, efficiency, and
ior they are supposed to explain. control in social cognition. In R. S. Wyer & T.
K. Srull (Eds.), Handbook of social cognition
(pp. 1–40). Hillsdale, NJ: Erlbaum.
Notes Bargh, J. A. (2006). What have we been priming
all these years?: On the development, mecha-
1. Automatic processes are often contrasted nisms, and ecology of nonconscious social
with controlled processes. Yet the term con- behavior. European Journal of Social Psychol-
trol has been used to refer to either (1) a par- ogy, 36, 147–168.
ticular feature of nonautomatic processing Baumeister, R. F., Masicampo, E. J., & Vohs,
(i.e., controllability) or (2) an umbrella con- K. D. (2011). Do conscious thoughts cause
cept subsuming multiple different features behavior? Annual Review of Psychology, 62,
of nonautomatic procesing (i.e., awareness, 331–361.
intentionality, resource dependence, control- Braver, T. S. (2012). The variable nature of cog-
lability). To avoid conceptual confusion, we nitive control: A dual mechanisms framework.
use the term nonautomatic as the semantic Trends in Cognitive Sciences, 16, 106–113.
antonym of the term automatic instead of the Brewer, M. B. (1988). A dual process model of
more common term controlled (see Moors & impression formation. In T. K. Srull & R. S.
De Houwer, 2006). Wyer (Eds.), Advances in social cognition
2. Some dual-­ process theories derive empiri- (Vol. 1, pp. 1–36). Hillsdale, NJ: Erlbaum.
cal assumptions about operating conditions Cacioppo, J. T., Berntson, G. G., Sheridan, J.
from their conceptual definitions of operat- F., & McClintock, M. K. (2000). Multilevel
ing principles (e.g., inferences about resource integrative analyses of human behavior: Social
dependence from the conceptual distinction neuroscience and the complementing nature of
between sequential and parallel processing). social and biological approaches. Psychologi-
Although such inferences are theoreticaly cal Bulletin, 126, 829–843.
valuable to clarify the conceptual basis of the Carruthers, P. (2009). An architecture for dual
derived hypotheses about operating condi- reasoning. In J. St. B. T. Evans & K. Frank-
tions, it is important to note that operating ish (Eds.), In two minds: Dual processes and
principles and operating conditions are nev- beyond (pp. 109–127). New York: Oxford
ertheless conceptually distinct in these theo- University Press.
ries, in that the former describe the nature of Cesario, J., Plaks, J. E., & Higgins, E. T. (2006).
the mental processes that translate inputs into Automatic social behavior as motivated prepa-
outputs, whereas the latter refer to the partic- ration to interact. Journal of Personality and
ular conditions under which these processes Social Psychology, 90, 893–910.
are assumed to operate. Chaiken, S. (1987). The heuristic model of per-
3. Note that the specification of heuristic attri- suasion. In M. P. Zanna, J. M. Olson, & C. P.
butes as coming more readily to mind does Herman (Eds.), Social influence: The Ontario
not provide a clear conceptual demarcation, Symposium (Vol. 5, pp. 3–39). Hillsdale, NJ:
because the accessibility of any given attribute Erlbaum.
can vary as a function of its salience. Chaiken, S., & Trope, Y. (Eds.). (1999). Dual-­
process theories in social psychology. New
York: Guilford Press.
References Conrey, F. R., Sherman, J. W., Gawronski, B.,
Hugenberg, K., & Groom, C. (2005). Separat-
Baars, B. J. (2002). The conscious access hypoth- ing multiple processes in implicit social cog-
esis: Origins and recent evidence. Trends in nition: The Quad-Model of implicit task per-
Cognitive Sciences, 6, 47–52. formance. Journal of Personality and Social
Baddeley, A. (2010). Working memory. Current Psychology, 89, 469–487.
Biology, 20, 136–140. De Houwer, J. (2011). Why the cognitive
Bargh, J. A. (1992). The ecology of automaticity: approach in psychology would profit from a
Toward establishing the conditions needed to functional approach and vice versa. Perspec-
produce automatic processing effects. Ameri- tives on Psychological Science, 6, 202–209.
can Journal of Psychology, 105, 181–199. De Houwer, J., & Moors, A. (2012). How to
 A Conceptual Analysis of Dual-Process Theories 17

define and examine implicit processes? In control as more than the effortful inhibition of
R. Proctor & J. Capaldi (Eds.), Implicit and impulses. Personality and Social Psychology
explicit processes in the psychology of science Review, 15, 352–366.
(pp. 183–198). New York: Oxford University Gawronski, B. (2013). What should we expect
Press. from a dual-­process theory of preference con-
De Houwer, J., & Moors, A. (in press). Levels of struction in choice? Journal of Consumer Psy-
analysis. In B. Gawronski & G. V. Bodenhau- chology, 23, 556–560.
sen (Eds.), Theory and explanation in social Gawronski, B., & Bodenhausen, G. V. (2006).
psychology. New York: Guilford Press. Associative and propositional processes in
Deutsch, R., & Strack, F. (2006). Duality models evaluation: An integrative review of implicit
in social psychology: From dual processes to and explicit attitude change. Psychological
interacting systems. Psychological Inquiry, 17, Bulletin, 132, 692–731.
166–172. Gawronski, B., & Bodenhausen, G. V. (2009).
Devine, P. G. (1989). Stereotypes and prejudice: Operating principles versus operating condi-
Their automatic and controlled components. tions in the distinction between associative
Journal of Personality and Social Psychology, and propositional processes. Behavioral and
56, 5–18. Brain Sciences, 32, 207–208.
Dhar, R., & Gorlin, M. (2013). A dual-­system Gawronski, B., & Bodenhausen, G. V. (2011).
framework to understand preference construc- The associative–­ propositional evaluation
tion processes in choice. Journal of Consumer model: Theory, evidence, and open questions.
Psychology, 23, 528–542. Advances in Experimental Social Psychology,
Dijksterhuis, A., & Nordgren, L. F. (2006). A 44, 59–127.
theory of unconscious thought. Perspectives Gawronski, B., & Bodenhausen, G. V. (in press-
on Psychological Science, 1, 95–109. a). Theory evaluation. In B. Gawronski & G.
Epstein, S. (1994). Integration of the cognitive V. Bodenhausen (Eds.), Theory and explana-
and the psychodynamic unconscious. Ameri- tion in social psychology. New York: Guilford
can Psychologist, 49, 709–724. Press.
Evans, J. St. B. T., & Frankish, K. (Eds.). (2009). Gawronski, B., & Bodenhausen, G. V. (in press-
In two minds: Dual processes and beyond. b). Social-­cognitive theories. In B. Gawron-
New York: Oxford University Press. ski & G. V. Bodenhausen (Eds.), Theory and
Fazio, R. H. (1990). Multiple processes by which explanation in social psychology. New York:
attitudes guide behavior: The MODE model as Guilford Press.
an integrative framework. Advances in Exper- Gawronski, B., & Creighton, L. A. (2013).
imental Social Psychology, 23, 75–109. Dual-­process theories. In D. E. Carlston (Ed.),
Fazio, R. H. (2007). Attitudes as object-­ The Oxford handbook of social cognition
evaluation associations of varying strength. (pp. 282–312). New York: Oxford University
Social Cognition, 25, 603–637. Press.
Fazio, R. H., Jackson, J. R., Dunton, B. C., & Gilbert, D. T. (1989). Thinking lightly about oth-
Williams, C. J. (1995). Variability in automatic ers: Automatic components of the social infer-
activation as an unobtrusive measure of racial ence process. In J. S. Uleman & J. A. Bargh
attitudes: A bona fide pipeline? Journal of Per- (Eds.), Unintended thought (pp. 189–211).
sonality and Social Psychology, 69, 1013–1027. New York: Guilford Press.
Fiske, S. T., & Neuberg, S. L. (1990). A contin- Greenwald, A. G., & Banaji, M. R. (1995).
uum of impression formation, from category-­ Implicit social cognition: Attitudes, self-­
based to individuating processes: Influences esteem, and stereotypes. Psychological
of information and motivation on attention Review, 102, 4–27.
and interpretation. Advances in Experimental Greenwald, A. G., McGhee, D. E., & Schwartz,
Social Psychology, 23, 1–74. J. L. K. (1998). Measuring individual differ-
Friese, M., Hofmann, W., & Schmitt, M. (2008). ences in implicit cognition: The Implicit Asso-
When and why do implicit measures predict ciation Test. Journal of Personality and Social
behaviour? Empirical evidence for the mod- Psychology, 74, 1464–1480.
erating role of opportunity, motivation, and Hassin, R. R., Bargh, J. A., Engell, A. D., &
process reliance. European Review of Social McCulloch, K. C. (2009). Implicit working
Psychology, 19, 285–338. memory. Consciousness and Cognition, 18,
Fujita, K. (2011). On conceptualizing self-­ 665–678.
18 THE B A SIC S

Huang, J. Y., & Bargh, J. A. (in press). The Self- attributional inference. Advances in Experi-
ish Goal: Autonomously operating motiva- mental Social Psychology, 34, 199–249.
tional structures as the proximate cause of Loersch, C. & Payne, B. K. (2011). The situated
human judgment and behavior. Behavioral inference model: An integrative account of the
and Brain Sciences. effects of primes on perception, behavior, and
Jacoby, L. L. (1991). A process-­ dissociation motivation. Perspectives on Psychological Sci-
framework: Separating automatic from inten- ence, 6, 234–252.
tional uses of memory. Journal of Memory Marr, D. (1982). Vision: A computational inves-
and Language, 30, 513–541. tigation into the human representation and
Johnson-­Laird, P. N. (2012). Mental models processing of visual information. New York:
and consistency. In B. Gawronski & F. Strack Freeman.
(Eds.), Cognitive consistency: A fundamental McGrath, J. E. (1981). Dilemmatics: The study
principle in social cognition (pp. 225–244). of research choices and dilemmas. American
New York: Guilford Press. Behavioral Scientist, 25, 179–210.
Kahneman, D. (2003). A perspective on judg- Meissner, F., & Rothermund, K. (2013). Esti-
ment and choice: Mapping bounded rational- mating the contributions of associations and
ity. American Psychologist, 58, 697–720. recoding in the Implicit Association Test: The
Keren, G., & Schul, Y. (2009). Two is not always ReAL model for the IAT. Journal of Personal-
better than one: A critical evaluation of two-­ ity and Social Psychology, 104, 45–69.
systems theories. Perspectives on Psychologi- Mitchell, C. J., De Houwer, J., & Lovibond, P.
cal Science, 4, 533–550. F. (2009). The propositional nature of human
Klauer, K. C. (in press). Mathematical modeling. associative learning. Behavioral and Brain Sci-
In B. Gawronski & G. V. Bodenhausen (Eds.), ences, 32, 183–198.
Theory and explanation in social psychology. Moors, A., & De Houwer, J. (2006). Automa-
New York: Guilford Press. ticity: A conceptual and theoretical analysis.
Klauer, K. C., Voss, A., Schmitz, F., & Teige-­ Psychological Bulletin, 132, 297–326.
Mocigemba, S. (2007). Process components Nadarevic, L., & Erdfelder, E. (2011). Cognitive
of the Implicit Association Test: A diffusion-­ processes in implicit attitude tasks: An experi-
model analysis. Journal of Personality and mental validation of the Trip Model. European
Social Psychology, 93, 353–368. Journal of Social Psychology, 41, 254–268.
Krieglmeyer, R., & Sherman, J. W. (2012). Dis- Ochsner, K. N., & Lieberman, M. D. (2001). The
entangling stereotype activation and stereo- emergence of social-­ cognitive neuroscience.
type application in the Stereotype Mispercep- American Psychologist, 56, 717–734.
tion Task. Journal of Personality and Social Payne, B. K., & Bishara, A. J. (2009). An integra-
Psychology, 103, 205–224. tive review of process dissociation and related
Kruglanski, A. W., Erb, H.-P., Pierro, A., Man- models in social cognition. European Review
netti, L., & Chun, W. Y. (2006). On paramet- of Social Psychology, 20, 272–314.
ric continuities in the world of binary either Payne, B. K., & Gawronski, B. (2010). A history
ors. Psychological Inquiry, 17, 153–165. of implicit social cognition: Where is it coming
Kruglanski, A. W., & Gigerenzer, G. (2011). from? Where is it now? Where is it going? In B.
Intuitive and deliberative judgments are based Gawronski, & B. K. Payne (Eds.), Handbook
on common principles. Psychological Review, of implicit social cognition: Measurement,
118, 97–109. theory, and applications (pp. 1–15). New
Krull, D. S. (1993). Does the grist change the York: Guilford Press.
mill?: The effect of the perceiver’s inferential Payne, B. K., Hall, D. L., Cameron, C. D., & Bis-
goal on the process of social inference. Per- hara, A. J. (2010). A process model of affect
sonality and Social Psychology Bulletin, 19, misattribution. Personality and Social Psy-
340–348. chology Bulletin, 36, 1397–1408.
Kunda, Z., & Thagard, P. (1996). Forming Peters, K. R., & Gawronski, B. (2011). Are we
impressions from stereotypes, traits, and puppets on a string?: Comparing the effects
behaviors: A parallel-­ constraint satisfaction of contingency and validity on implicit and
theory. Psychological Review, 103, 284–308. explicit evaluations. Personality and Social
Lieberman, M. D., Gaunt, R., Gilbert, D. T., & Psychology Bulletin, 37, 557–569.
Trope, Y. (2002). Reflexion and reflection: Petty, R. E., & Cacioppo, J. T. (1986). The
A social cognitive neuroscience approach to Elaboration Likelihood Model of persuasion.
 A Conceptual Analysis of Dual-Process Theories 19

Advances in Experimental Social Psychology, chology: Conceptual integration and links to

19, 123–205. underlying memory systems. Personality and
Popper, K. R. (1934). Logic of scientific discov- Social Psychology Review, 4, 108–131.
ery. New York: Basic Books. Stahl, C., & Degner, J. (2007). Assessing auto-
Quine, W. V. O. (1960). Word and object. Cam- matic activation of valence: A multinomial
bridge, MA: MIT Press. model of EAST performance. Experimental
Rydell, R. J., & McConnell, A. R. (2006). Under- Psychology, 54, 99–112.
standing implicit and explicit attitude change: Stanovich, K. E., & West, R. F. (2000). Individ-
A systems of reasoning analysis. Journal of ual differences in reasoning: Implications for
Personality and Social Psychology, 91, 995– the rationality debate? Behavioral and Brain
1008. Sciences, 23, 645–726.
Samuels, R. (2009). The magical number two, Strack, F., & Deutsch, R. (2004). Reflective and
plus or minus: Dual-­process theory as a theory impulsive determinants of social behavior.
of cognitive kinds. In J. St. B. T. Evans & K. Personality and Social Psychology Review, 8,
Frankish (Eds.), In two minds: Dual processes 220–247.
and beyond (pp. 129–146). New York: Oxford Trope, Y. (1986). Identification and inferential
University Press. processes in dispositional attribution. Psycho-
Sherman, J. W. (2006). On building a better logical Review, 93, 239–257.
process model: It’s not only how many, but Trope, Y., & Gaunt, R. (2000). Processing alter-
which ones and by which means. Psychologi- native explanations of behavior: Correction or
cal Inquiry, 17, 173–184. integration? Journal of Personality and Social
Sherman, J. W., Gawronski, B., Gonsalkorale, Psychology, 79, 344–354.
K., Hugenberg, K., Allen, T. J., & Groom, Wallach, L., & Wallach, M. A. (1994). Gergen
C. J. (2008). The self-­regulation of automatic versus the mainstream: Are hypotheses in
associations and behavioral impulses. Psycho- social psychology subject to empirical test?
logical Review, 115, 314–335. Journal of Personality and Social Psychology,
Sloman, S. A. (1996). The empirical case for two 67, 233–242.
systems of reasoning. Psychological Bulletin, Wheeler, S. C., DeMarree, K. G., & Petty, R.
119, 3–22. E. (2007). Understanding the role of the self
Smedslund, G. (2000). A pragmatic basis for in prime-to-­behavior effects: The active-­ self
judging models and theories in health psychol- account. Personality and Social Psychology
ogy: The axiomatic method. Journal of Health Review, 11, 234–261.
Psychology, 5, 133–149. Wilson, T. D., Lindsey, S., & Schooler, T. Y.
Smith, E. R., & DeCoster, J. (2000). Dual-­ (2000). A model of dual attitudes. Psychologi-
process models in social and cognitive psy- cal Review, 107, 101–126.
 Chapter 2

Examining the Mapping Problem

in Dual‑Process Models
Agnes Moors

Multiprocess models divide the realm of automatic influence on decision making,

mental processes into two (e.g., Gawronski and systematic information (e.g., persuasive
& Bodenhausen, 2006, 2011; Lieberman, arguments of the speaker) a nonautomatic
2003; Sloman, 1996; Smith & DeCoster, influence (e.g., Chaiken, Liberman, & Eagly,
2000; Strack & Deutsch, 2004), three (Lev- 1989). By mapping various dichotomies onto
enthal & Scherer, 1987), or four sets of each other, dual-­process models may grow
processes (e.g., Conrey, Sherman, Gawron- into dual systems models (e.g., Deutsch &
ski, Hugenberg, & Groom, 2005; Sherman, Strack, 2006).
2006). The sets of processes in dual-­process The question one may ask is whether it
models often correspond to dichotomies is justified to map dichotomies onto each
such as affective versus nonaffective, heu- other. I propose investigating the mapping
ristic versus systematic, automatic versus problem in a stepwise manner. A first step is
nonautomatic, associative versus rule-based, to engage in the conceptual analysis of vari-
perceptual versus conceptual, associative ous dichotomies to see whether they can be
versus propositional, and subcortical ver- defined independent of each other. If it turns
sus neocortical. It is perfectly legitimate to out that they can be so defined, a second step
create sets on the basis of one or another is to investigate empirically whether there
dichotomy. Most dual-­ process models, is overlap among dichotomies in the real
however, take things one step further. They world. To be able to go from the first to the
choose two or more dichotomies and make second step, however, an intermediate step is
a priori assumptions of overlap among them required, in which the definitions proposed
(Keren & Schul, 2009; Moors & De Hou- in the first step should be translated into
wer, 2006b). A few examples: Associative methods for diagnosing occurrent processes
processes are often said to be automatic, as belonging to one or the other term of the
and rule-based processes nonautomatic dichotomy. In this chapter, I engage in a
(e.g., Smith & DeCoster, 2000). Associative conceptual analysis of the following dichot-
processes are often said to operate on per- omies: automatic versus nonautomatic, asso-
ceptual codes, and rule-based processes on ciative versus rule-based, perceptual versus
conceptual codes (e.g., Leventhal & Scherer, conceptual, and associative versus propo-
1987). Heuristic information (e.g., attrac- sitional (cf. the first step discussed earlier).
tiveness of the speaker) is said to exert an After that, I examine whether the definitions

20
 Examining the Mapping Problem in Dual-Process Models 21

obtained in the first step lead to criteria for ical research, we might miss out on relevant
the diagnosis of processes (cf. the intermedi- findings. Thus, my aim in this section is to
ate step discussed earlier). examine whether there is a satisfactory way
to define each dichotomy without having to
make appeal to another dichotomy.
Conceptual Analysis My proposal is to situate the various
dichotomies within a levels-­ of-­
analysis
Definitions of sets are often of the intensional framework. This framework proposes that
kind, specifying the necessary and sufficient any process can be described at (minimally)
conditions for something to be an exemplar three levels of analysis (e.g., Anderson, 1987;
of a set (i.e., one necessary and sufficient Marr, 1982; Pylyshyn, 1980). At the func-
condition, or a collection of necessary condi- tional level, a process is described as the rela-
tions that are jointly sufficient). Intensional tion between an input and an output. This
definitions can be descriptive or prescriptive. level includes the content of inputs and out-
Descriptive definitions describe the rules puts, as well as the conditions under which
that laypersons (implicitly or explicitly) fol- the process operates. For example, affective
low (or seem to follow) when using concepts. processing or evaluation can be described as
Prescriptive definitions stipulate the rules the relation between a stimulus and affective
that scientists prescribe for using concepts. properties (e.g., positive or negative valence).
The development of prescriptive definitions The processing of heuristic information can
is often guided by two principles: A first be described as the relation between a stimu-
principle is to stay close to natural language lus and heuristic properties (e.g., the attrac-
or the way in which laypersons use the term tiveness of the speaker). Conditions include
(descriptive definitions). As such, it should those related to automaticity versus nonau-
cover a sufficient number of intuitively clear tomaticity (e.g., uncontrolled–­ controlled;
exemplars (Green, 1992). A second prin- unconscious–­conscious, efficient–nonef-
ciple stems from the desire to cut nature at ficient, fast–slow). The algorithmic level
its joints or to provide a useful heuristic for articulates the mechanisms involved in
future research. These joints are not abso- transforming input into output, as well as
lute but context dependent and in this sense the formats of the representations or codes
involve an element of choice (i.e., they are that the mechanisms operate on or produce.
stipulative). My choice of definitions is For example, evaluation can proceed with
explicitly guided by the desire to avoid over- an associative mechanism (retrieving stored
lap among dichotomies. The main argument valence) or a rule-based one (e.g., weigh-
for this is that empirical research can only ing the implications for various concerns),
be carried out with concepts that are logi- and both might operate on various types of
cally independent (e.g., Smedslund, 1992). codes (e.g., conceptual versus perceptual,
For example, when the dichotomies of auto- associative versus propositional). The hard-
matic versus nonautomatic and associative ware level is concerned with the physical
versus rule-based are defined independently, implementation of processes in the brain. At
four combinations are open to investiga- this level, a process is described in terms of
tion: associative and automatic, associative its neurophysiological substrates (e.g., neo-
and nonautomatic, rule-based and auto- cortical versus subcortical areas or circuits).
matic, and rule-based and nonautomatic. For example, evaluation is often linked to
When, by contrast, automatic processes are circuits involving the amygdala.
defined as associative ones, and rule-based The levels-­of-­analysis framework makes it
processes are defined in terms of nonauto- easy to conceive of the various dichotomies
matic features (e.g., conscious), there is no as nonoverlapping. This is because the lev-
point in empirically investigating any com- els in the framework are only loosely related
bination. Investigating whether associative to each other. For instance, one functional
processes are automatic, for instance, would process (e.g., evaluation) can be accounted
be equally absurd as investigating whether for by more than one mechanism (e.g., by
bachelors are indeed single and male. By a rule-based or associative process). Also,
excluding certain combinations from empir- presenting automaticity as a matter of con-
22 THE B A SIC S

ditions provides a natural way to decouple it conditions that is sufficient for the process
from processes (described at any level). The to operate. A process is uncontrolled in the
next sections include detailed definitions of counteracting sense when it operates regard-
dichotomies and compare them with alter- less of the presence or absence of a coun-
native definitions, specifically with regard to teracting goal within the set of conditions.
how successful they are in avoiding overlap A process is unconscious when it operates
with other dichotomies. under the condition of a lack of awareness.
A process is efficient when it operates under
the condition of minimal available atten-
Automatic versus Nonautomatic
tional capacity. A process is fast when it can
Contemporary feature-­based views of auto- be completed under the condition of mini-
maticity define it as an umbrella term for a mal available time. In summary, a process
number of features, such as uncontrolled, is automatic when it operates under subop-
unintentional, unconscious, efficient, and timal conditions; a process is nonautomatic
fast (Bargh, 1992; Moors & De Houwer, when it operates only under optimal condi-
2006a). These features roughly can be tions.
defined as follows. An uncontrolled process Logan (1988) proposed to replace feature-­
is one that is not influenced by a person’s based definitions of automaticity with a
goals about the process. Goals about pro- mechanism-­based definition. Specifically, he
cesses can be either of the promoting kind proposed to define automatic processes as
(e.g., the goal to engage in the process, which ones that are based on single-­step memory
can be split into goals to start, continue, and retrieval. Single-step memory retrieval is
complete the process) or of the counteracting very much akin to what others have called
kind (e.g., the goals to avoid, alter, or stop an associative mechanism, that is, the acti-
the process). A process is uncontrolled in the vation of a memory trace leading to previ-
promoting sense when it is not caused by the ously computed outputs. Such a definition
goal to engage in it (or to start, continue, or of automaticity is obviously not successful in
complete it). Another word for uncontrolled avoiding conceptual overlap with the asso-
in the promoting sense is unintentional. A ciative versus rule-based dichotomy.
process is uncontrolled in the counteracting
sense when it is not counteracted (avoided,
Associative versus Rule-Based
altered, or stopped) by the goal to do so
(avoid, alter, or stop). A process is uncon- The associative versus rule-based dichot-
scious when the person has no awareness of omy has a central role in research domains
it. I think awareness is possible for processes such as problem solving (Anderson, 2005;
described at the functional level (i.e., the Mayer, 1992), categorization (Rouder &
relation between input and output) but not Ratcliff, 2006; Smith & Sloman, 1994),
likely for processes described at the algorith- reasoning (Sloman, 1996), memory (Smith
mic or hardware level (Moors & De Hou- & DeCoster, 2000), and grammar learning
wer, 2006a). People may be aware that they (Reber, 1967). It refers to the mechanisms
evaluate a person as good or bad, but they situated at the algorithmic level of analy-
probably lack insight into the mechanisms, sis. A rule-based mechanism consists of the
codes, and brain circuits involved. An effi- application of a rule to an input (or a rep-
cient process is one that makes minimal use resentation of it). Computation of the rule
of attentional capacity. A fast process is one produces an output. An associative mecha-
that is completed within a short time inter- nism consists of the activation of stored
val. representations of similar past inputs. This
The definitions of automaticity features activation, in turn, spreads to associated
just given can be reformulated in terms of stored representations that determine the
operating conditions (Bargh, 1992; Moors output. As an illustration from the domain
& De Houwer, 2007; Moors, Spruyt, & De of problem solving, consider the behavior of
Houwer, 2010). For example, a process is two persons paying 4 euros at the counter
uncontrolled in the promoting sense (unin- after having ordered two beers that cost 2
tentional) when the goal to engage in it is not euros each. One person’s behavior is caused
part (or only a redundant part) of the set of by the calculation of the rule “N × M = 4,” in
 Examining the Mapping Problem in Dual-Process Models 23

which N is the number of beers ordered and ciations can be described with nonabstract
M is the price of one beer. The other person’s rules, in which the premise consists of only
behavior is caused by the activation of a constants. Constants are representations of
memory that she paid 4 euros last time when concrete instances. The nonabstract rule of
she ordered the same. As an example from the beer example (if two beers, then 4 euros)
categorization, consider that dolphins are cannot be applied to novel situations (e.g.,
(correctly) categorized as mamals based on ordering five coffees) unless they resemble
the rule that mamals nurse their offspring. the situation in the premise (e.g., ordering
Based on similarity with fish exemplars, they two soft drinks).
can also be (incorrectly) classified as fish. It A third criterion is that rule-based mecha-
may be noted that some researchers speak nisms can operate only if there is a perfect
of the similarity-­based mechanism instead match between the input and the premise,
of the associative mechanism, because the whereas associative mechanisms can also
initial activation of stored representations is operate if there is a partial match between
based on similarity (Hahn & Chater, 1998; the input and the premise (or stored repre-
Pothos, 2005). sentation; Hahn & Chater, 1998). In the
To arrive ultimately at a way to diagnose beer example, the abstract rule N × M can-
processes as either rule-based or associa- not be computed unless values for N and M
tive, the previous definitions need to be are available. The nonabstract rule to pay
formalized. What is computation? And in 4 euros when ordering two beers can be
what sense is it different from activation of applied to any situation that is similar to this
stored knowledge?2 I discuss several criteria one (the more similarity, the stronger will be
that have been proposed as capturing the the activation of the memory trace). Because
essence of the rule-based versus associative of the complementary forces of abstrac-
dichotomy. First, it has been suggested that tion and partial matching, both rule-based
rule-based mechanisms are ones that can be and associative mechanisms can account
described by an if–then clause. For example, for generalization toward new exemplars.
“if N × M = p, then pay p euros.” However, Rule-based mechanisms can produce gener-
several authors have correctly noted that the alization by virtue of abstract variables; the
associative mechanism can be described by associative mechanism can produce general-
an if–then clause as well. For example, “if ization by virtue of partial matching (partial
two beers, then pay 4 euros.” In fact, any matching compensates for the lack of vari-
mechanism that connects an input to an out- ables). Nonabstract rules can thus be applied
put can be expressed in an if–then format. to new input, but only by virtue of similarity
The if-part specifies the input, and the then- among the input and the constants specified
part specifies the output. This has led some in the premise.
authors to conclude that associative and A fourth criterion spelled out by some
rule-based mechanisms are not qualitatively theorists is that rule-based mechanisms are
different, and that an associative mechanism ones that follow rules, whereas the associa-
is just a simple type of rule-based mecha- tive mechanism merely conforms to rules.
nism (Kruglanski, Erb, Pierro, Mannetti, & Rule following means that a mental rule sits
Chun, 2006). Other authors have searched between the input and the output of a pro-
for other criteria to distinguish both types of cess and causes the output; rule conforming
mechanisms. means that the relation between input and
A second criterion is that rule-based, output merely can be described or summa-
but not associative, mechanisms can be rized according to a rule (Hahn & Chater,
described by abstract rules (e.g., Sloman, 1998; Pylyshyn, 1980; Searle, 1980; Sloman,
1996; Smith, Langston, & Nisbett, 1992) 1996; Smith et al., 1992). Numerous phe-
in which the premise (the if-part) contains nomena can be described with rules without
variables. Variables are abstract represen- there being actual rule following. Examples
tations that can be instantiated with more are the planetary motions and the dances
than one constant. In the abstract rule of the of honeybees. Planets and bees are unlikely
beer example (if N × M, then p), the vari- to have internal representations of the rules
ables N and M can be instantiated with an that describe their behavior. Something
infinite range of numbers. By contrast, asso- similar is assumed to occur with the asso-
24 THE B A SIC S

ciative mechanism. However, unlike with sterhuis, & Galinsky, 2008). Given that con-
planets or bees, associations are internal to sciousness is a feature of nonautomaticity,
the organism (mental) and they do mediate such a definition is not successful in avoiding
between input and output. Moreover, it has overlap with the automatic–­ nonautomatic
been argued that rules can be implemented dichotomy. As such, it precludes empirical
by patterns of associations (like the ones research concerned with whether rule-based
proposed in connectionist models; Fodor & mechanisms can be automatic (e.g., Hélie,
Phylyshyn, 1988), or that rules are approxi- Roeder, & Asch, 2010; Hélie, Waldschmidt,
mations of such patterns (Smolensky, 1986). & Asch, 2010; Van Opstal, Gevers, Osman,
These views again suggest no fundamental & Verguts, 2010).
distinction between associative and rule- The next dichotomies to be discussed,
based mechanisms: They are the same thing perceptual versus conceptual and associa-
described on different levels of analysis. tive versus propositional, characterize the
A fifth criterion proposed by some theo- format of the representations or codes on
rists is that the rules in rule-based processes which mechanisms operate. The terms con-
must be not only internally represented and ceptual and propositional indicate a resem-
causal but also symbolic (e.g., Hahn & blance, respectively, with words and sen-
Chater, 1998; Smith & DeCoster, 2000). tences in natural language. The resemblance
A symbolic representation of a rule has the is not physical, but structural or functional
rule as its content or object (but see Searle, (what can be done with it), and various
1980). Because patterns of associations do bases for this resemblance have been high-
not represent rules symbolically (they are lighted.
subsymbolic representations of rules at best),
activation of them does not count as rule-
Perceptual versus Conceptual
based processing. Other theorists (Clark,
1990; Fodor & Pylyshyn, 1988; Pinker & The dichotomy of perceptual versus con-
Prince, 1988; Smith et al., 1992) have argued ceptual codes is central in the domain of
against the view that the rules in rule-based imagery and has often been referred to as
processes must be symbolically represented. picture-­like versus verbal-­ like (Barsalou,
They allow rules to be hardwired in the 1999).3 Conceptual codes are thought to be
system. Fodor and Pylyshyn (1988, p. 61) abstract4 (e.g., the word dog applies to many
maintained that the only things that must be instances of dogs), context-­ independent
symbolically represented in a computational (e.g., the same code for dog is activated in all
system are the data on which the rules oper- contexts), and amodal (i.e., the same code is
ate. Patterns of associations are typically activated by a picture of a dog, and the word
not built in but are shaped by the environ- dog in writing or pronounced out loud; Bar-
ment. Yet one may wonder why the distinc- salou, 1999). Perceptual (or sensory) codes
tion between innate and learned would be have perceptual (or sensory) features and
crucial in determining whether something are therefore concrete (e.g., an image of a
counts as a real rule (i.e., a generator of rule- dog cannot be applied to all dogs), context-­
based processing). In summary, several cri- dependent (e.g., a different dog code is acti-
teria have been proposed to demarcate rule- vated in the context of a dark alley than in
based from associative mechanisms: abstract the context of a dog house), and modal (an
versus nonabstract, perfect versus partial image of a dog activates a code other than
matching of the premise, and rule follow- that produced by the word dog pronounced
ing versus rule conforming. With regard to out loud). Some authors add motor features
the latter criterion, some authors consider to these perceptual codes and call them
internal representations and causal influence embodied (Barsalou, 1999). Activation of
as sufficient ingredients for rule following, embodied codes leads to partial reenactment
whereas others also include the ingredient or mental simulation of the perceptions and
symbolic representation. actions that typically occur in the presence
An alternative proposal is to define rule- of the represented object, which makes them
based processes as ones that are conscious, or dynamic. This is a further difference with
at least conscious during application (Cleer- conceptual codes, which are considered
emans & Destrebecqz, 2004; Zhong, Dijk- static.
 Examining the Mapping Problem in Dual-Process Models 25

Associative versus Propositional these definitions comprise two parts: one

part is a verb (activation, validation), and
Propositional codes are often said to be com-
the other is a type of representation (associa-
positional, which means that they comprise
tion, proposition). Thus, in these definitions,
meaningful parts that can be recombined
mechanisms are partly defined in terms of
with other parts to form new representa-
formats of representations. Other authors
tions (i.e., productivity; Fodor & Phylyshyn, (e.g., De Houwer, Chapter 35, this volume)
1988).5 In that sense, they are just like sen- also define the associative mechanism as the
tences in natural language, which also com- activation of an association, but the proposi-
prise words that can be recombined with tional mechanism as anything that happens
other words to form different sentences. (validation or activation) with a proposition.
An organism that is capable of responding According to this view, the pure activation
differently to the same stimulus in differ- of a proposition is also called a proposi-
ent contexts must have the capacity to form tional mechanism. Here too, mechanisms
propositions in the sense that it can isolate are defined in terms of formats of represen-
parts and reuse them in new combinations. tations.
In addition to being compositional, sev- One might ask whether it is possible to
eral authors have proposed that the content avoid such overlap between mechanisms and
of propositional codes also must be truth representations. Can associative and propo-
evaluable (Anderson, 1978). Thus, in addi- sitional mechanisms be defined in such a way
tion to comprising (recombinable) parts that they are allowed to operate on any type
(e.g., cat–roof), propositional codes also of representation? I think this is indeed pos-
must specify a relation between these parts sible for the associative mechanism but not
that can be evaluated as true or false (e.g., for the propositional mechanism. The asso-
the cat is on the roof). In order for a relation ciative mechanism can be minimally defined
to be truth evaluable, it must have a certain as the activation of a representation in mem-
quality. These so-­called “qualified” relations ory, without specifying the format of the
(e.g., A is B, A is bigger than B, A is under representation that is activated. The propo-
B) are contrasted with unqualified relations, sitional mechanism, on the other hand, if it
also termed (pure) associations, in which the is defined as validation or the evaluation of
nature of the relation between the parts is something as true or false, can only oper-
unspecified (e.g., A–B; De Houwer, Chapter ate on representations that are truth evalu-
35, this volume). This aligns with the idea able. As discussed earlier, only propositions,
that to hold a proposition (e.g., “the apple is defined as complex representations compris-
red”) is not to hold a list of concepts (apple, ing a qualified relation between two elemen-
is, and red) but to predicate something (red- tary representations, can be evaluated as true
ness) onto something (the apple; Soames, or false. This is not possible for elementary
2012). To sum up, propositional codes are representations (e.g., the concept cat), or for
defined here as complex representations associations, defined as complex representa-
that comprise a qualified relation between tions of unqualified relations between two
elementary representations. Associations elementary representations (e.g., cat–roof).
are defined as complex representations that This being said, some theorists (e.g., Gaw-
comprise unqualified relations between ele- ronski & Bodenhausen, 2011) seem to sug-
mentary representations. gest that associations give rise to the spon-
Theorists in the domain of learning (e.g., taneous formation of propositions, and that
Gawronski & Bodenhausen, 2011) have the latter can be validated. For instance, the
used the contrast between associations and association “banana–­good” gives rise to the
propositions as the basis for distinguishing proposition “banana is good,” which can be
between two types of mechanisms: associa- evaluated as true or false. But how does the
tive ones and propositional ones.6 The asso- organism know which quality to attribute
ciative mechanism is often defined as the to an unqualified relation? The banana–­
activation of an association in memory, and good example suggests that the organism
the propositional mechanism, as the valida- follows a default rule, according to which
tion of a proposition, that is, the evaluation associations suggest a positive or even iden-
of it as true or false. It may be noted that tity relation (i.e., an “is a” relation). But
26 THE B A SIC S

such a default rule cannot be applied to the tion between the two terms of a dichotomy.
cat–roof example. A better solution might This has implications for the diagnosis of
be to argue that associations tend to be the mental processes underlying some per-
interpreted in a way that is consistent with formance as belonging to either term of a
existing propositions. But even then, given dichotomy.
that people may endorse multiple proposi-
tions about the same pair of concepts (the
Automatic versus Nonautomatic
banana is good when I am in a fruit mood,
otherwise it is bad; the cat may be under In the former section I defined automaticity
the roof), the question of which proposition as an umbrella term for a collection of fea-
is most likely to be chosen as a frame for tures. In addition, I favor a gradual approach
a given association remains unanswered. In to automaticity (Logan, 1985; Moors & De
any case, the transition from associations to Houwer, 2006a; Shiffrin, 1988). This grad-
propositional codes is something that is left ualness is manifested in two ways. First, I
largely unspecified. suggest that a process can be automatic with
Theorists have defined the distinction regard to some but not other features. For
between associative and propositional example, it seems that some processes are
mechanisms not only in terms of types of fast and efficient but can be counteracted
representations but also in terms of features (Uleman & Moskowitz, 1994). Second,
of nonautomaticity, in particular the fea- each automaticity feature can itself be con-
ture conscious (e.g., De Houwer, Chapter sidered as gradual. A process more or less
35, this volume; Mitchell, De Houwer, & can be uncontrolled, unconscious, efficient
Lovibond, 2009). For example, it has been and/or fast. Although I believe the gradual
proposed that the formation of a proposition approach to automaticity is theoretically the
must be conscious, or at least that the output most cautious approach, it does not provide
of this formation process, the proposition an objective criterion for the diagnosis of a
itself, must be conscious at some point. In process as automatic or nonautomatic. This
addition, it has been proposed that evalua- problem can be dealt with by specifying
tion of the proposition as true (i.e., belief of the sense in which (i.e., the feature accord-
the proposition) must be conscious at some ing to which) one considers a process to be
point. These theorists accept that once a automatic or nonautomatic, and by making
proposition is stored in memory,7 it may be relative conclusions or choosing an intersub-
retrieved in an unconscious way, and that jective criterion (i.e., convention) for calling
conscious and unconscious propositions some process controlled or uncontrolled,
may exert an unconscious (and otherwise conscious or unconscious, efficient or nonef-
automatic) influence on subsequent behav- ficient, and fast or slow (Moors & De Hou-
ior and thought. Even so, the fact that con- wer, 2006a).
sciousness of the proposition is put forward Researchers have proposed to examine
as a necessary element in the definition of (1) the uncontrolled nature of a process by
the propositional mechanism creates overlap removing aspects of the procedure (e.g.,
with the automatic–­ nonautomatic dichot- instructions or other cues) that might induce
omy. Such overlap can be avoided by adopt- a processing goal (to engage in or counter-
ing a minimal definition of the propositional act a process), or by blocking implementa-
mechanism as validation or truth evaluation tion of such a goal (e.g., Klauer & Teige-­
irrespective of consciousness. Mocigemba, 2007; Moors, De Houwer,
Hermans, & Eelen, 2005; Spruyt, De Hou-
wer, & Hermans, 2009); (2) the unconscious
Diagnosis nature of a process by presenting stimuli
subliminally (Greenwald, Klinger, & Schuh,
The conceptual exercise in the previous 1995); (3) the efficient nature of a process
section shows that it is possible to define by increasing the load of a secondary task
most dichotomies in nonoverlapping ways. in dual-task studies (e.g., Hermans, Crom-
A closer look at these nonoverlapping defi- bez, & Eelen, 2000); and (4) the fast nature
nitions reveals that several of them do not of a process by using short stimulus onset
allow one to draw an all-or-none distinc- asynchronies combined with short response
 Examining the Mapping Problem in Dual-Process Models 27

deadlines (Draine & Greenwald, 1998; for is indistinguishable from applying the rule
a detailed review, see Moors, Spruyt, & De that mammals are animals that nurse their
Houwer, 2010). offspring. Thus, evidence for generaliza-
tion toward stimuli that share only abstract
features with previously encountered ones
Associative versus Rule-Based
(Reber, 1989; Marcus, Vijayan, Bandi Rao,
Assessment of rule-based mechanisms sepa- & Vishton, 1999) is equally compatible with
rate from or in addition to the associative rule-based and with associative mechanisms
mechanism also suffers from the problem (Redington & Chater, 1996; but see Sloman
that the criteria adduced to distinguish & Rips, 1998; Smith et al., 1992).
between the two mechanisms not always Third, advocates of the idea that rules
lead to unequivocal ways to diagnose perfor- in rule-based mechanisms must be sym-
mance as based on one or the other mecha- bolically represented have proposed to use
nism. First, as explained earlier, both rule- verbal reports to investigate the content of
based and associative mechanisms are able representations. If a person has a symbolic
to account for generalization toward new representation of a rule, there is a possibility
stimuli (abstract rules by virtue of variables, that he or she can verbally report the rule.
associations by virtue of partial matching). Unfortunately, the criterion of verbal report-
Second, abstraction is a gradual notion ability is unsuitable for research concerned
(Hahn & Chater, 1998). The variables in with automatic (in the sense of unconscious)
abstract rules and the constants in nonab- rule-based processing. It is also not suited
stract rules occupy two points on a con- for studying nonsymbolic rules, such as sub-
tinuum. Variables can be instantiated with symbolic patterns of associations or built-in
a larger set of things than can constants, rules.
but the variables in abstract rules often can- The research literature contains several
not be instatiated by just anything (e.g., in other proposals for how to register the oper-
the abstract rule, “if N × M = p, then pay ation of rule-based mechanisms independent
p euros,” N and M must be instantiated by of the operation of the associative mecha-
the number of beers ordered and the price nism, but none of the proposed methods
of one beer, not by the number of words seems unequivocal (cf. reviews by Hahn &
spoken and the color of the curtains). At the Chater, 1998; Rouder & Ratcliff, 2006; Slo-
extreme abstract end of the continuum are man, 1996; Smith et al., 1992). One option
logical rules in which the variables can be is to continue searching for a distinction
substituted by anything (e.g., if [X and Y], between rule-based and assocative mecha-
then X). Conversely, constants often hold nisms that does lead to an undisputable
some level of abstraction (e.g., in the nonab- empirical test. Another option is to abide
stract rule “if two beers, then pay 4 euros” by a gradual view of this distinction (e.g.,
the representation of two beers can itself Pothos, 2005) or to specify an intersubjec-
be instantiated by more than two unique tive criterion (i.e., convention) for calling a
instances of beer). At the extreme nonab- mechanism rule-based or associative.
stract end of the continuum are constants
that represent a unique instance. Given the
Perceptual versus Conceptual
relative nature of abstraction, no objective
line can be drawn between variables and Diagnosing representations as perceptual
constants. This is reflected in the idea that (picture-­like, concrete, modal, dynamic) ver-
activation of stored knowledge can be based sus conceptual (verbal-­like, abstract, amo-
on concrete as well as abstract similarities dal, static) has met with similar difficulties.
(e.g., similar function). Similarity even may Various authors (Anderson, 1978; McDow-
pertain to abstract relations among variables ell, 1994; but see Dretske, 1981; Peacocke,
(cf. Goldstone, 1994; Redington & Chater, 1992) have suggested that the two are
1996; Hummel & Holyoak, 2003). For informationally equivalent. Any detail of a
instance, dolphins are perceptually similar picture-­like representation can be described
to fish but they are similar to other mammals verbally. Furthermore, activation can spread
in the abstract sense that they nurse their from amodal and disembodied representa-
offspring, and the latter type of similarity tions to the sensory and motor systems, and
Random documents with unrelated
content Scribd suggests to you:
The Project Gutenberg eBook of Ugor Ágnes
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: Ugor Ágnes

Author: György Szemere

Release date: May 18, 2020 [eBook #62166]

Language: Hungarian

Credits: Produced by Albert László from page images generously

made
available by the HathiTrust Digital Library

*** START OF THE PROJECT GUTENBERG EBOOK UGOR ÁGNES ***

 Megjegyzések:

Az eredeti képek elérhetők innen:

https://hdl.handle.net/2027/uc1.aa0006787360.
Facebook oldalunk: http://www.facebook.com/PGHungarianTeam.
 MAGYAR IRÓK
ARANYKÖNYVTÁRA

XXII. KÖTET.

SZEMERE GYÖRGY:

UGOR ÁGNES

BUDAPEST, 1909.
GRILL KÁROLY KÖNYVKIADÓVÁLLALATA
IV., VERES PÁLNÉ-UTCA 16.

UGOR ÁGNES
*
IRTA

SZEMERE GYÖRGY
 BUDAPEST, 1909.
GRILL KÁROLY KÖNYVKIADÓVÁLLALATA
IV., VERES PÁLNÉ-UTCA 16.
A NAP nyomdája, Budapest, VIII., Rökk Szilárd-utca 9
 ELSŐ FEJEZET.

A könnyelmü csók.

Ez a történet egy csókon fordul meg, amelyet egy férfi egy

leánynak adott…
Hiszen ilyen csókon fordul meg okos emberek állitása szerint is,
minden a nap alatt: a háboru, a müvészet, a civilizáció, sőt bizonyos
agyafurt tudományos praktikával meg is lehet állapitani, teszem a
szerelmi csók háboruvá való fejlődési evoluciójának minden egyes
mozzanatát, összes jelenségeivel együtt: csakhogy az a csók,
amelyből a következő igaz mesének szálai kifutnak, erre a célra nem
szolgálhatna bizonyitási alapul. Nem rendes szerelmi csók volt. Egy
házas ember adta, aki szerette a feleségét, nyegleségből s egy
urileány kapta váratlanul, minden bevezetés nélkül, mert
szobaleánynak hitték. Tehát a természetes előföltételek, amelyek a
kis dolgokból nagyokat alkotnak, hiányzottak az elemeiből. Nem volt
benne a szerelem. Ennélfogva következményei sem gyökereztek a
természet megszokott rendjében: egészen más, a rendestől teljesen
eltérő, fiziológiai és pszihikai jelenségeket eredményeztek, mint
amilyeneket csóktól józan észszel várni lehetett volna.
Ime:
Államférfit csinált egy jobbfajta patópál-szerü magyar nemesből,
megnemesitett egy krudélis természetü vénkisasszonyt,
férjhez adott egy gyönyörü leányt,
kemény kozákká változtatott egy gyámoltalan gavallért s magyar
urrá egy angol baronetet.
 Mindeme tények nem voltak ugyan közvetlen következményei a
csóknak, de annyi okozati összefüggésben legalább állottak vele,
mint teszem a trójai háboru annak egyik klasszikus példányával.
Mert annyi bizonyos, hogy a nélkül nem történhettek volna meg…
A szenvedő fél, (aki a csókot kapta), Ágnes kisasszony volt, a
felvidéki hires-nevezetes Ugor-nemzetség sarja; a cselekvő: báró
Bálvándy Antal, erdélyrészi földbirtokos, állitólag rabonbán ivadék…
Az eset ugy történt, hogy Bálvándy fajkosokat akart venni az
Ugorék országhirü juhászatából. Nem sajnálta a fáradtságot, maga
utazott föl Dormándra, (az Ugorék törzsfészkébe) a kosokért. A ház
urát nem találta otthon. Egy öreg inas bevezette a nappali szobába,
ahol egy orrig bekötött fejü leány, rövidre szabott perkálszoknyában
éppen port törölt.
A vendég hangos jókedvvel üdvözölte a szobacicát (mert annak
hitte), az, (nem lehet tudni minek nézte a poros csizmás urat),
dünnyögött valamit, különben alig hederitett reá, több kérdésére
nem is felelt, ami mód felett boszantotta a derék férfiut, mert igen
szeretett csevegni, mint a vénasszonyok.
De volt Bálvándynak egy más nagy hibája is. Szobaleánytól lefelé
nem igen tisztelte a nőt. Csak bolondozni szeretett vele. Ha nem
látták, ugy nekiszelesedett a szoknya közelében, mint egy
sárospataki diák.
Ágnesnek nem is láthatta az arcát az orrára lehuzott kendő miatt,
de megpillantotta vörhenyeges dus haját és ez elég volt neki. Midőn
a leány éppen egy antik csészével babrált, háta mögé lopódzott és
orvul elkapván a derekát, egy csókot cuppantott a nyaka közé. Csak
ugy, csupa taktikából, hogy megoldja a szemérmeteskedő szobacicus
nyelvét. Praktizálta ő ezt a módszert nem egyszer és mindig bevált.
Most azonban bizony felsült vele egy kissé.
Mindenekelőtt kiesett az antik csésze a szobacica kezéből és ezer
darabra tört. Aztán – a helyett, hogy még mélyebbre huzta volna, –
lekapta a leány a fejkendőjét, mire a merénylő legott megbánta
merényletét, kivilágosodván, hogy a szobacica arcvonásai nem
állottak harmóniában a szép hajával.
De ezt még csak elviselte volna a báró, hanem a leány
magatartása már valósággal megijesztette.
– Ki ön? – kérdezte ez elfulva, halálsápadtan.
– Én – felelt a báró megszeppenve, – báró Bálvándy Antal
vagyok, kosokért jöttem.
– És – folytatta a leány (s most meg minden vére az arcába
szaladt), – hogy mert engem megcsókolni?
– Ejnye – feszengett a megszeppent férfiu rosszat sejtve, – hisz
nem kell ahhoz valami nagy kurázsi… Ne haragudjon, lelkem, – tette
hozzá, – jóvá teszem a hibát.
Gondolta magában: egy tizessel majd csak kibékiti az erényes
szobaleányt.
Ágnes egy pillanatig merőn nézte a férfit, – (volt mit látnia rajta:
pompás, daliás alak volt), – nézte, aztán lesütötte szemét.
– Az más, báró Bálvándy – jegyezte meg halkan, – azt hittem,
hogy nős ön.
– Hiszen az vagyok.
A leánynak villámot szórt a szeme.
– Akkor hogy akarja jóvá tenni? – kérdezte fenyegető hangon.
A báró szeme nem szórt villámot, csak nyilladozni kezdett. Egy
szobaleány, vélte, nem merne vele ilyen hangon beszélni.
– Tulajdonképpen ki maga, lelkem? – hebegte riadtan.
– Ugor Ágnes vagyok.
S gőgösen fölvetette szemét a leány.
 Bálvándy legjobb szeretett volna a föld alá sülyedni, de minthogy
ezt nem tehette, kénytelen volt más kibuvót keresni.
– Akkor persze – dadogta, – kissé nehezen fog menni a dolog.
Mert mit tehetek egyebet, minthogy bocsánatot kérjek
nagyságodtól… Elvégre is – magyarázta kinjában, – szóra sem
érdemes az eset, tévedés forgott fenn… s mit árt kegyednek az a pici
csók!
– Az nem ugy van – jelentette ki Ágnes, megnyomva minden
egyes szótagot, – velem nem lehet ugy bánni, mint egy
szobaleánynyal.
– Hát hogy van? Hogy legyen? – kérdezte Bálvándy kimeresztve
szemét.
– Ugy, hogy én várni fogok magára, mindaddig, amig abban a
helyzetben nem lesz, hogy elégtételt adhasson nekem.
– Milyen elégtételt, az Istenért?
– Olyat, amilyet egy urileány egy ilyen inzultus fejében
megkövetelhet egy uriembertől.
Hát ez elég érthető beszéd volt. Bálvándy összekulcsolta kezét a
feje fölött.
– Az ég szerelmére, mondtam, hogy nős vagyok!
– Én várok…
– De hiszen husz esztendős a feleségem és imádom…
– Azt előbb kellett volna meggondolnia…
S hiába esküdözött a férfi ezután is: hogy a világért sem válik el
szép és fiatal párjától. Ágnes makacsul megmaradt a mellett, hogy
vár, vár és vár.
 Elhatározását kivételesen közölte szüleivel is. Sőt azoknak már
azt is bevallotta, hogy mire számit.
– Valami belső érzés, több a sejtelemnél, – mondta fanatikusan,
– azt sugja nekem, hogy meghal az az asszony, hiába szép és fiatal…
Szülői elképedtek. Szinte azt hitték, hogy meghibbant a leányuk
agya…
Hát ezt az eszelősséget bizony nehezen fogja megérteni és
valószinünek itélni még az is, aki ismerte Ugor Ágnest, de talán
eligazodik rajta, ha megismeri a történetét.
Ki volt ő?!…
Szülői a család szégyenének mondották egymás között, rokonai a
család eszének. Ismerősei hol skrimpiciának, hol sárkány-kigyónak. A
cselédek és a köznép földöntuli hatalmat tulajdonitottak neki. Sőt
Matyis, a pankuhár egyenesen azt állitotta, hogy nem is igazi ember,
hanem boszorka, hites felesége Asmódiás ördögnek és a pokol
küldetésében jár a földön… Valóságos legendák kerengtek róla,
amire bőven okot szolgáltatott titokzatos, erőszakos lénye,
zárkózottsága – még a szülőivel szemben is – s fölényes, mindenkit
lekicsinylő modora.
Szobaleánya azt állitotta róla, hogy átvirrasztja az éjszakákat,
ilyenkor (megleste a kulcslyukon) hajmeresztő dolgokat müvel. Hogy
mik lettek légyen ezek a hajmeresztő dolgok, azt nem merte (?) a
komorna elárulni, csak annyit mesélt el a konyha népének, hogy
csudatevő szerei vannak, (alighanem az emberek megrontására),
egyebek között egy pókvérrel megtöltött kék üvegcse, egy
selyemmadzagon lógó kutyafark stb.
Az igazság azonban az volt, hogy azért virrasztotta át Ágnes az
éjszaka egy részét, mert olvasott. Kivételes mértékben kifejlődött
tudásvágya nem hagyta nyugodni, folyton képezte magát.
Huszonötéves korára annyi ismeretet gyüjtött össze, hogy kiállotta
volna a versenyt akárki polihisztorral. Betéve tudta Schoppenhauert
és a Korpusz Juriszt, Kantot, Shakespearét és a bibliát. Amellett
olyan választékosan tudott beszélni, mint Chatam lordja s ugy
zongorázott, mint a záporeső. De legjobban szerette Nietzsét. Mert
magára ismert a szertelen tudós tulemberében.1)
És valóban, csodálatos kvalitásai voltak a vénleánynak. A többi
emberétől teljesen elütő, minden nőiség nélkül szükölködő volt a
jelleme. Fanatikusa volt az akaratnak és az igazságnak. Energiája
nem ismert akadályokat, talán a büntől sem riadt volna vissza, ha az
eszközöknél többre becsüli vala a célt. Afféle tulzott,
természetellenes mértékben kialakult példánya volt fajának. Szinte
ugy tetszett, mintha egy hatodik érzéke is lett volna, amivel belelát a
jövőbe s áthidalja s távolságokat. Csudálatos előérzetei és sejtései
voltak. Akárcsak valami láthatatlan erő hipnózisa igazgatta volna a
lelkét s cselekedeteit… Egyszer megjósolta, sőt percre megmondta
egy rokona halálának idejét, máskor kitalálta egy felbontatlan levél
tartalmát. Igaz, hogy sokat foglalkozott a spiritizmussal és hitt a
telepátiában. A hit pedig csudákat sejtet és csudákat müvel…
Nyilván eme rejtélyes kvalitások táplálták a cselédség azt a hitét,
amely emberfeletti ranggal ruházta fel a ház kisasszonyát. De talán
magát Ágnest is azok vezérelték, midőn valamelyes rendkivüli tettre
szánta el magát. Különben, hogy lett volna érthető, hogy férjévé
szándékozott tenni egy házasembert, aki nem szerette őt s még csak
nem is válás, de egy fiatal szép asszony halála révén remélt célhoz
jutni?
Pedig nem is volt ez az elhatározása ötletszerü. Nem bánta meg,
nem józanodott ki belőle, nem mondott le róla.
Várt!…
Várt, várt! Nap-nap után, hónapról-hónapra. Évről-évre. Öt
esztendeig, tizig, tizenötig. Már harminchatéves volt és egy
árnyalattal se gyöngült a bizakodása.
És harminchatodik életévében végre csakugyan valóra vált
csudával határos előérzete. A még mindig viruló és fiatal Bálvándyné
kitért az utjából: meghalt.
Ágnes alig birta, vagy talán nem is akarta leplezni diadalmas
örömét.
– Most már rajtam a sor!…
Szülei már nem is igen törődtek az ő „lelketlen“ leányukkal. Csak
éppen hogy megtürték maguknál kötelességből. Ágnest ez egy
csöppet sem bántotta. Nem vágyott szeretetre, sőt megvetette azt.
Talán alig is volt lelke, inkább csak ösztöne, mint az ősembernek, aki
a gyengébbek leigázásából élt…
Hát ezuttal kiszemelt párját kellett leigáznia… Munkához látott.
Akaratzsenije minden energiáját szabadjára eresztette, hogy
biztositsa a célját. Érezte, hogy rendkivüli eszközökhöz kell
folyamodnia, ha le akarja győzni azt a távolságot, amely őt
Bálvándytól elválasztotta. Olyan tervet eszelt ki, ami szinte láthatóan
magán viselte a téboly bélyegét… Választottja Erdélyben lakott. (Az
idevágó adatokat régen beszerezte volt), nyáron Bálvándon, télen
Kolozsvárott. Gazdag ember volt, független és még mindig fiatalos.
Ha ujból megakart volna házasodni, ott virágzott körülötte a bérces
kis haza egész előkelő leányflórája: bizvást válogathatott volna
benne, aligha kapott volna kosarat. Nagy parti volt, mint mondani
szokás, méltó akármelyik mágnásleány kezére…
Ágnes teljesen tisztában volt azzal, hogy terve két főakadály
kiméletlen feldöntése nélkül nem érhet célt. Ez a két főakadályának:
áldozata vagyoni függetlensége és a geográfiai távolság, mely őt tőle
elválasztotta… Őrült gondolatok! Ám Ágnes nem ismert lehetetlent,
sőt azt tartotta annak, hogy bárki akaratával szemben ne az ő
erősebb, szivósabb akarata érvényesüljön.
Azzal kezdte a vénleány, hogy levelet irt egy Budapesten züllő s
örökös pénzzavarral küszködő unokatestvérének: bizonyos Hervay
Muki bárónak, látogatná meg őt a dormándi kastélyben azonnal:
beszélni akar vele az adóságairól.
 MÁSODIK FEJEZET.

Muki báró.

Muki báró a kaszinóban kapta meg Ágnes levelét, ebédközben,

éppen a liba előtt. Egy darab mellet még kiszedett a tányérjára,
aztán felbontotta a levelet. Evett és olvasott. Nyelte a libát és a
sorokat, de rosszul nyelte, mindkettővel egyszerre nem tudott
megbirkózni. A liba torkán akadt s midőn lenyelte végre, az esze
állott meg a levéltől.
– Amicskó, – sugta szomszédjának, – a kitüzött terminusra nem
adhatom meg az ezer forintodat családi gyász miatt.
– Hogy-hogy?
– Él egy kuzinkám Hontban, aki vagy megbolondult, vagy a
halálán van.
– Várok, csak ne mókázzál, – intette le a megbizhatatlanságáról
hires bárócskát, barátja, Zsirkay gróf, unottan, fölényesen.
– Nem mókázom, amicskó, – állitotta ez sértődötten, de tekintve
adósi mivoltát, meghunyászkodó önérzettel, – itt a levél… Ki akarja
fizetni az adósságaimat.
– Akkor csakugyan megbolondult, – mondta a gróf szárazon.
Muki báró erre már komolyan megorrolt. Alig várta, hogy lenyelje
az ebédet: hazahajtatott és még az éjjeli vonattal elrobogott
Dormánd felé.
 Utközben, mint megannyi mérges darázs, mindenféle cifra
gondolat röpült ki a fejéből s elkezdte a fülét rágni s orrát
boszantani.
… Nini, ötlött az eszébe, hátha nem bolondult meg, hanem épen
belőle akar az a fránya Ágnes bolondot csinálni?… Vagy talán arra
aspirál, hogy elvegye?… Huh, akkor inkább felakasztja magát!…
Hohó, csipte meg egy ujabb, még gonoszabb gondolat az orrát,
hátha a vagyonomból akar kiboritani!… Eszébe jutott a nagybátyja
végrendelete, különösen a klauzulája: Abban az esetben, ha kedves
öcsém, Hervay János Mária negyvenötéves koráig nem házasodnék
még, az öröksége tárgyát képező ingatlan terhére további ötvenezer
forintot tartozik Zsófia hugom, Ugor Menyhértné leányának,
Ágnesnek kifizetni…
… Világos, – sóhajtott Muki báró, – ezt az ötvenezer forintot
akarja tőle Ágnes kicsikarni. Azt ugyan – konstatálta, – hiába
csikargatja, mert immár sikerült neki birtokát annyira eladósitania,
hogy árverés esetén ötvenezer krajcárja sem maradna meg belőle,
ám mit ér ilyen élvezet, mikor, miként a hasisélvezőnek, neki is meg
kell az árát adnia. Egy licitáció és fuccs a hitelnek, amiből immár
évek óta él!…
Egyelőre csak a Muki álmának lett fuccs. Egy ideig még
elkinlódott fekvőhelyén a báró, hátha sikerül elcsipnie. De nem
sikerült, mert még jobban megdagadt a feje ujabb rajokat eresztő
rémképeitől. Hát felkelt és kinyitotta a kupé ablakát, hogy
kirepüljenek rajta a gonosz szellemek.
A szellemek nem repültek ki, de a friss levegő, az betódult a
kupéba. Muki kidugta fájós fejét a hideg nedves éjbe és belebámult
a holdba. Tetszett neki a nyájas fényü gömbölyüség az égen, sokért
nem adta volna, ha odaviszi vala a gyorsvonat… Ott aligha van leány,
kártya, pezsgő, amik itt lent, ezen a barna földön, megejtik az
emberfiát…
 Mély sóhaj szakadt föl az inci finci gavallér kebeléből. Csupa ideg
volt, csupa jóság. Nem bántott soha senkit, csak egy ellensége volt:
önmaga. És mégis elzüllött. A szerencse elhagyta, az élet agyba-főbe
vágta. Pedig nem is volt nagyon könnyelmü, csak ugy élt, mint a
többi hozzá hasonló gavallér. Az volt a baja, hogy nagyon érzékeny
volt és kissé hiu. Az emberek meghizelegték, kihasználták a
gyengeségét, kifosztogatták…
– Most végképp hoppon vagyok, – morfondirozott a jó Muki, –
nem tudom, sziven lőjjem-e magamat, vagy fejbe? Mi a biztosabb?…
Majd Dormándon kieszelem, – vetett véget tépelődésinek, aztán nem
gondolt többé a jövőjére, régi emlékeit szedegette elő… Azok szépek
voltak nagyon… A dormándi pázsit, – odavaló birtokos volt az ő atyja
is, – ahol a parasztgyerekekkel birkózott, az első pónija, az érettségi
vizsgája, amit kitüntetéssel tett le s amiért nagymamájától száz
forintot, szüleitől aranyórát kapott… Aztán első szerelme: Ilonka,
akinek a keze egyszer ott remegett az övében…
Csak ennyi volt. S mégis, ez a pici édes emlék több
gyönyörüséget okozott a bárócska szivének, mint a temérdek csók
összevéve, mit asszonyajakról valaha szakitott… És mégsem tette
párjává Ilonkát. Gyáva volt, félt szüleitől, akik partihajhásznak
nevelték, félt az életől, a küzdelemtől, a szegénység apró, piszkos
tragédiáitól, amelyekkel – előtte lebegett a példa, – át meg
átszövődött szülei élete is… Mert akkor még szegény volt Muki,
mikor megszerette Ilonkát: atyja nyakig adósságban, ő meg
fizetésnélküli minisztériumi segédfogalmazó. Nagybátyját később
örökölte meg: két héttel az Ilonka esküvője után. Mily tragikum!…
Muki báró behunyta szemét, s szabad szárnyra eresztette
képzeletét, hogy hozná neki vissza az ő Ilonkáját. És a jóságos
tündér egy pillanat alatt eleget tett a kivánságának… A leányka
hozzásimult önfeledten, bugva s vágyakozó szerelmesének
odakinálta forró, remegő ajakát. Az beleszédült a menyei
gyönyörüségbe, elolvadt a végtelenség étherében: elaludt…

Á
 Ám jóságos hü fantáziája álmában sem hagyta magára Mukit.
Tovább szőtte a szép meséket, mig – egyet zökkent a gőzös…
Csingilingi, krrr! Dormánd öt perc… A vonat megállott. A kalauz
fölébresztette Mukit.
– Méltóságos uram, helyben vagyunk.
– Köszönöm, tessék egy kis borravaló. Segitsen lehurcolkodnom.
A kalauz füttyentett egy pályaőrnek, s leadta neki a báró
disznóbőr bőröndjeit.
Muki leszállott. Alig érintette lába a földet, fülébe nyillalott az
unokahuga hangja:
– Jó napot, Muki.
– Jaj lelkem, pá, szervusz! – üdvözölte a báró hugát. – Hogy
vagy? Puszi.
Öszecsókolóztak, de utána mindketten megtörölték a szájukat.
– Ejnye, hogy kijöttél elém! Mily kedves vagy Ágneske! –
áradozott Muki szokásból, mert olyan volt a természete: nem volt
benne semmi objektivitás.
Annál több volt Ágnesben.
– Ugyan – utasitotta ez vissza unokabátyja édeskés szavait, – ne
komédiázzál. Nem azért kérettelek, te sem azért jöttél, hogy örüljünk
egymásnak. Szükségünk van egymásra, pont.
Muki elröstelte magát.
– Hehe, de tréfás vagy, kuzin… Persze, persze, – mekegte, –
nekem szükségem van a te jóságodra.
– Nekem meg – mondta Ágnes, – az ügyességedre.
 – Ügyességemre? – csudálkozott a kis báró. – Nem vagyok én
ügyes.
– Majd eligazitalak én, csak tégy ugy, ahogy mondani fogom…
Kocsiba szállottak… Az állomást a kastélytól jó egynegyedórai ut
választotta el. Ágnes kihasználta az időt az utolsó másodpercig.
Folyton ő beszélt, Muki csak makogott néha, mint a kecskegida,
mikor az anyját keresi. Muki az eszét kereste.
– Van-e pénzed? Ugy-e nincs? – Kezdte Ágnes. – Hát majd adok
én neked annyit, amennyi kell.
A bárócska roppant megkönnyebbült a biztató kezdettől. Igen sok
felesleges vér szaladt a szive felé, amitől csuklani kezdett: kluk, hluk,
hihilukk…
Ágnes folytatta:
– Persze nem ingyen. Erdélybe fogsz menni és megismerkedsz
Bálvándy Antallal.
– Meeg?
– Ne szólj közbe… Behizelged magadat a barátságába és ki nem
ereszted a körmeid közül.
Muki önkénytelenül a körmére nézett, de nem szólt semmit,
bámulatos képpel leste huga további szavait:
– Azt mondják, igen szolid ember, hát el fogod lumpositani.
– Lu… lu…
– Ne feleselj!… Igenis, megkedvelteted vele a pezsgőt, kártyát és
lovakat…
– A nőket is? – kérdezte vigyorogva Muki, mert most már
elemében érezte magát.
Ágnes gondolkodott egy pillanatig, aztán igy felelt:
 – Azokat nem… Hanem – folytatta, – sok pénzt fogtok költeni az
– övéből.
De ezt már nem értette Muki.
– Én is az övéből?
– Te is az övéből, – világositotta föl Ágnes szigoru szóval. – El
fogod nyerni a pénzét.
– É-én? Nem tudok én nyerni.
– Ne okoskodjál. A kaszinóban nem tudsz nyerni, mert azok
kozákok, de olyantól, aki nem tud kártyázni, nyerhetsz még te is.
– I-ijnye, biz én ezt sem értem.
– Nem is kell. Van ötvenezer forintom, amit Miklós bácsitól
örököltem, mint tudod. Ebből tizezer forintot mindjárt átadok neked.
Mukinak megszünt a csuklása, ahelyett nyeriteni kezdett, mint
egy jókedvü paripa: hihihi, hehehe.
– Ezzel a tizezer forinttal, – folytatta a vénleány, – ha csak egy kis
ügyesség van benned, lefektetheted Bálvándyt.
– Lefektethetem? hogy érted?
– Ne add a kis ártatlant. Elmégy vele Bécsbe és Párisba s addig is
pikettirozni fogtok a kupéban, mig csak egy csöpp pénze marad.
Akkor kölcsön adsz neki…
De ezen már mégis fölháborodott a minden izében korrekt
bárócska.
– Ágnes, az Istenért, hova gondolsz? Nem vagyok én sipista!
– Oda gondolok, – világositotta fel Ágnes, – hogy elmultál
negyvenötéves s a Miklós bácsi végrendelete értelmében további
ötvenezer forinttal tartozol nekem. Igaz, hogy nincs egy vasad se, de
a birtokodat mégis elvehetem… Hát oda gondolok.
 Szegény Muki meglapult a batárban, mint a nyul a füben.
Ágnes ütötte a vasat.
– És arra gondoltam, hogy oda ajándékozom neked az ötvenezer
forintot, ha jól viseled magadat…
– Ehe… De miért pályázol éppen arra a szegény Bálvándyra?
– Ahhoz semmi közöd… Azt akarom, hogy képviselő legyen. Hogy
miért? Hát ahhoz sincs semmi közöd. Te leszel a főkortese, te fogod
megválasztatni. Pont.
A bárócskának végképp széjjel mászott a két nagy leffentyüs
füle.
– És – bátorkodott megjegyezni, – ha nem fog akarni kártyázni,
se fellépni, mit tegyek?
– Akkor – jelentette ki Ágnes megkerülve az egyenes feleletet –
kiperelem belőled a Miklós bácsi hagyományát.
– Jaj, Ágnes!
– Ellenben, ha jól csinálod, adok neked ujabb tizezer forintot.
Többet is adnék, mert nekem nem kell a pénz, de hátha a
választásra kell.
Muki összecsapta kezét.
– Istenem, hát olyan fontos az neked, hogy Bálvándy képviselő
legyen?
– Nem az a fontos, hanem az, hogy Pestre jöjjön lakni. Mert én
nem szaladgálhatok utána Erdélybe.
Nyilladozni kezdett már a bárócska szeme.
– Ahán, hát… ühüm, a férjednek szántad… Igaz is, ideje, hogy
férjhez menj… De miért választottad éppen őt? És honnan
ismered?…
Á
 Ágnes nem válaszolt. A fogat éppen befordult a parkba. Muki a
hársfákat nézte. Az egyikre egy monogrammot vésett ezelőtt
huszonöt esztendővel. Az ő és az Ilonka nevének kezdőbetüiből…
– Vajjon – mélázott, – kiforrta-e a hársfa héjja a régi sebet,
avagy őrzi-e még (miként ő) a nyomát ma is?…
A terrasz előtt megállott a kocsi. Egy öreg férficseléd unottan
kecmergett ki az ódon szürke kastélyból s egy rövid fejbiccentéssel
üdvözölte az érkezetteket. Aztán levette a böröndöket a bakról.
Mukit nem ismerte meg…
Pedig ezelőtt negyven esztendővel, amikor még szép fiatal legény
volt, de sokat hancurozott vele! Hátára vette és elkezdett ügetni,
mint egy sánta ló: vegyenek dohoskását!… A többi gyerek, a Hervay
és Ugor-családok legifjabb nemzedéke vett a dohoskásából:
megcsipdeste Mukit. Az sivalkodott, de soha se sirt a csiklandós
tréfáért, inkább igen megszerette érte az akkor még Miska sorban
lévő Mihályt, amiért huszéves daliás huszár létére leereszkedett
hozzá, a hatéves gyermekhez…
Mukinak majdhogy ki nem csordult a könye. Szivébe sajgott az
öreg cseléd mogorvasága s hogy meg nem ismerte őt. Ám nem tett
neki érte szemrehányást. Ugy adta, mintha ő sem ismerte volna
meg. Némán követte a kastélyba, a huga utasitása szerint:
– Majd Mihály a szobádba vezet. Mosakodjál meg, ha akarsz…
Különben – jegyezte meg, – maradhatsz ugy is, ahogy vagy, a
komédiákat, hálisten, kiirtottam a házból…
Gondolta magában a báró: kiirtottál te egyebet is a házból, a
csöndes békességet, az előkelő szellemet…
Elfoglalván szobáját, megberetválkozott Muki, felrakta a
tyukszemgyürüket és kopaszodó fején összepomádézta széjjeleső
haját. Egyszóval összetákolta, a mennyire lehetett, cingár, romlatag
testét. Délfelé járván az idő, mindjárt felrántotta a ferencjózsefet is s
az ablakba hajló futórózsatőről egy öklömnyi csokrot szerkesztett és
a gomblyukába tüzte. Azzal felkerekedett, hogy megkeresse
nagybátyját és nagynénjét. Végig ődöngött a kongó boltives folyosón
s találomra betolt egy ajtót. Éppen az ebédlőbe nyitott be. Idestova
harminc éve, hogy utoljára látta a kedves termet. Valósággal
meghatotta előkelő komorsága és naiv pompája. Merész, csúcsives
boltozata égnek volt kipingálva, freskóban természetesen. Az
ultramarin-kék levegőben arany csillagok reszkettek és fecskék
röpködtek szerteszét. A mennyország szélét bárányfelhő-fodor
jelezte, a közepén angyalok tanyáztak, kicsinyek és nagyok; a
kicsinyek meztelenül röpködtek, cicáztak, a nagyok görög fehér
lepelben másfélméteres réztrombitákat fujtak.
Muki arra gondolt, hogy milyen kedves jó emberek lehettek az
ősök, akik számára a kastély épült… Csillagos ég, fecskemadár,
angyalkák!…
De nem volt ideje a mélázásra. Kordult az ajtó s összekuszálta
ábrándképét. Nagybátyja lépett a terembe: az öreg Ugor Menyus.
Pompás pátriárka-alak, arca csupa üde pirosság, haja csupa hó. Öles
teste kemény, mint a vas, csak apró tipegő járása árulta el, hogy
lábának izmait mégis elnyütte az idő… Tárt karokkal és harsogó
Istenhozottal fogadta rég nem látott rokonát.
– Ajaj, öcsécském, csakhogy láthatlak még egyszer a halálom
előtt… Te, te rossz, haszontalan rokon!…
Muki protestálni akart a halál ellen, de az öreg ur nem engedte
szóhoz, ahelyett belelóditotta egy hintaszékbe. A hintaszék elkezdett
veszettül kalimpálni az egyensulyát elvesztett Muki alatt. Az öreg
nem törődött vele, nyugodtan folytatta:
– Aj jaj, öcsécském, vége a szép időknek! A percek elrepülnek és
soha sem jönnek többé vissza. S mindig elvisznek magukkal valamit.
Egy darabka agyvelőt, egy kicsinyke lelket, életkedvet, emléket; az
ember vén korára azon veszi magát észre, hogy elfogyott. Kiégett,
mint az öreg pipa. Aztán jön a halál és azt mondja: most már
terhére vagy a földnek, alló a sirba! És az ember belemegy a sirba,
ha akar, ha nem… Hát édes, kedves öcsécském, – fakadt ki az öreg,
– ez igazságtalanság. Mig fiatal az ember, nem tud élni, elpocsékolja
a kincseit, mire meg megtanul spórolni Isten adományával – mars a
föld alá!…
Roppant szeretett az öreg ur a halál témájával babrálni. Bár
görcsösen ragaszkodott az élethez. Egyszer a nevenapján azzal
köszöntötte fel a papja, hogy éljen száz esztendőt. És kikapott a pap.
Menyus ur leszidta minden teketória nélkül.
– Kedves barátom, ha kiván valamit, jót kivánjon, az én
nagyapám százhat évig élt.
Igazat mondott… Általában metuzsálemi kort értek el az Ugorok.
Szivós faj voltak különben is és roppantul vigyáztak az egészségükre.
Nem koptatták munkával a testüket, se gonddal a lelküket. Csak
éltek egyenes vonalban, nem nézvén se jobbra, se balra és igen sok
ételt és pénzt fogyasztottak. Mindaddig, amig az anyaföld (azaz az
ősi birtok) ingadozni nem kezdett a talpuk alatt. Akkor megijedtek és
összehuzták magukat, ahelyett, hogy jövedelmük szaporitásán
jártatták volna eszüket. Kikoplalták – mint mondani szokás, – a
haladó kor veszedelmeit.
Ilyenformán, ha nem is gyarapodtak az Ugorok anyagilag, de
nem is fogyhattak. Életük első felét eldáridózták mindközönségen,
fényes házat és telivér lovakat tartottak, mert gőgösek voltak, mint a
rozmár, másik felét meg a dáridók folytán megbillent pénzügyi
egyensuly szanálására forditották. Megszelidültek, igás lovon jártak,
három évig viseltek egy ruhát, sőt elfelejtettek mosakodni is
mindaddig, amig ki nem fizették összes adósságaikat. S miután
addig éltek az Ugorok, ameddig akartak, örököseikre többnyire
tehermentes birtokot, sőt egy kis pénzt is hagytak. Ezek azután
bátran folytathatták az uri tempót, folytatták is, hogy valahogy el ne
kallódjék az Ugorok minduntalan megszakitott nimbusza.
Ilyenformán nagy gavallér volt valamennyi fiatal Ugor és nagy
réparágó valamennyi öreg. Menyus ur is. Bár ő valamivel kevésbbé
mint vele egykorácsu elődei és oldalági rokonai. Neki nem engedte
meg a leánya, hogy hájjal ápolja a bagaria csizmáját, mert tartott az
orrára. A zsámiskával sem érte be Ágnes, mert tartott a gyomrára.
Sőt uri fogat is kellett a fránya leánynak, inas is, mert a rangjára
tartott. És ki is csikart magának mindent, amit akart. Szülei nem
mertek vele szembehelyezkedni. Mert energia volt benne, azokban
meg semmi se. Nem volt jellemük. Az öreg asszony martirnak
vallotta magát, egész nap jajgatott, az öreg ur meg prézsmitált és
morgott. Mikor nem morgott, hazudott, De az öreg asszony is
hazudott, mikor nem jajgatott. Olyan csuda házaspár volt: nem
tudott az egyenes uton haladni, váltig kerülgette egymást és az
igazságot. Nem is koccantak össze soha, mert a döntő pillanatban
belebujtak a képmutatás és hazugságok utvesztőjébe…
Egy szó mint száz, tótok voltak Ugorék. Ojtott tótok. Immár
hatszáz év óta, hogy nemességet szerzett a családjuk, tagjai
vérzettek is a hazáért becsülettel, de a faji kvalitásaik
leglényegesebb jegyei: a kétszinüség és sunyiság, nem tudott róluk
lekopni. Ágnes talán egymaga volt kivétel a hires fa hajtásai között…
… Menyus bácsi, miután kikeseregte magát a halál ellen, végre
szóhoz engedte jutni az öccsét is.
– No, öcsécském, most már beszélj te is.
Szegény Mukiban kevés volt a szó, különösen, ha kalimpált alatta
a szék.
Kikecmergett hát a himbából és megrázogatván vékony
lábszárait, megragadta a bácsija hosszu és fekete körmös kezét és
elkezdte rázni, mintha kondoleálni akarna neki. Az öreg urnak –
nyilván félreértette – nem tetszett ez a furcsa modor.
– Ej, hát mit sajnálgatsz, azért nem vagyok még a sirban!
Muki kézzel-lábbal protestált.
– Ments Isten, hiszen nem sajnálom én, édes kedves bátyám!…
 – De sajnálsz – makacskodott az öreg. – Akarom tudni, hogy mi
sajnálnivalót találsz rajtam?
– Semmit, – erősködött Muki, – Isten ucscse, nem sajnállak, csak
meg vagyok hatva.
– Mitől te?
Mukinak egyszeribe megeredt a szava.
– Annyit játszottam gyermekkoromban itt: ebben a kertben, a
gyümölcsös padláson, a halastónál és (hozzátámaszkodott a nagy
családi asztalhoz) e körül a jó vén butor körül!… Nem tehetek róla,
most, hogy a hangodat hallom, visszakérezkednek régi kedves
emlékeim a szivem közepébe és – bizonyisten sirnék, ha nem
röstelleném.
– Csak sirj öcsécském, – biztatta az öreg, most már
megnyugodva, – a köny nem szégyen és könnyit a lelken.
És elbeszélte életének egy pár keserves esetét, a mikor neki is
sirnia kellett.
A sirás igaz volt, mert a sirhatnámság is hozzátartozik a felvidéki
urak természetrajzához, de az esetek már költve voltak. Nem volt az
öreg Menyusnak életében egyetlen olyan keserves esete sem, a mi
komoly ok lett volna a sirásra. Ha sirt, minden külső befolyás nélkül
sirt, szabadon, mint a milyen szabadon a fülemüle csattog: a szive
szavára. Csupa passzióból felzaklatta idegeit, elsajnálgatta magát,
minden ok nélkül, különösen ha nagyon unatkozott…
… Muki, hiába biztatta a nagybátyja, nem követte a tanácsot.
Inkább nénje felől kérdezősködött.
– Az most beteg, – állitotta az öreg Menyus, – semmi esetre se
fog kijönni.
És ebben a pillanatban benyitott a néni.
 Muki eléje szaladt és kezet csókolt neki. A néni nagyot sóhajtott s
megölelte öcscsét szótlanul, mint a sorsüldözöttek, a kik nem tudnak
már örülni a legnagyobb örömöknek se.
– Igen beteg szegény Zsófi, – magyarázta ki Menyus bácsi a
felesége sóhaját és némaságát.
Jól tette, hogy igy magyarázta, (ha a meggyőződése ellenére is,
mert egyszeribe elapadt az öreg asszonyka könyforrása, hogy a
legnagyobb méltatlankodásnak adjon helyet.
– Én beteg? – hápogta elfojtott indulattal. – Honnan veszed, édes
Menyuskám, hogy beteg vagyok? Nincs nekem – állitotta, – semmi
bajom, csak, ah, szerencsétlen vagyok.
– Hisz az a legnagyobb betegség, – állitotta Menyus bácsi. –
Inkább odaadnám a kisujjamat, csak szerencsétlen ne legyek soha…
Ilyen modorban folyt a társalgás. Lefetyeltek, lotyogtak az
öregek, egyetlen szavukat sem vették komolyan. Ha feszült az
ellentmondás fonala, eleresztették egyszerre mindaketten, hogy
végképp el ne szakadjon valahogy. Mert komolyan mérgeskedni
egyikük sem akart: az koptat, nyüvi az életerőt…
Muki, a ki rég elszokott a felvidéki levegőtől, csak ámult-bámult a
vén gyerekeken. Sehogy se tudta megérteni, hogy miért
bolondoznak egymással, amikor lehetnének okosak és komolyak is.
Hisz az sokkal könnyebb.
Nem tudta Muki, mennyire ragadós a rossz szokás. S hogy idővel
kedvessé, sőt bájossá válhatik az üres locsogás is, afféle játszi
incselkedés szinét veszi fel annak a szemében, a kinek nem
ambiciója, hogy edzze a jellemét.
Bizonyos: a jellemakrobaták erőlködésében van valami
visszataszitó nagyképüség s viszont demokritoszi bölcseség rejlik
abban, hogy az ember komolyan ne vegye se magát, sem
embertársait. Hanem persze születni kell erre is. Mert a kemény
ember nem fog magából bolondot üzni soha.
Á
 Ám Muki báró minden volt, csak nem kemény. Volt némi jelleme,
de csak olyan puha, mint a kocsonya. Ha megnyomták, rezgett, ha
ráfujtak, akkor is rezgett. Hát Menyus bácsi megnyomta, Zsófi néni
meg-megfujta s harmadnapra már ő is megtanult füllenteni. Talán
maga se tudta, de ugy hazudott és kertelt, mint a záporeső.
A bácsiék igen megszerették érte s már-már terveket kombináltak
a megragasztására. Hisz ugy pászolt volna nekik vőnek! Igaz ugyan,
hogy el volt adósodva, de ők meg már abban a korban voltak, a
melyben az Ugoroknak nem szokott adósságuk lenni. Segithettek
volna Mukin… Aztán Ágnesnek is ott volt az ötvenezer forintja.
Egyszóval, vélték az öregek, alighanem meg lehetne a Muki
vagyonát menteni, különösen ha olyan kezekbe kerül, mint az
Ágneséi…
Igy szőtték az öregek ábrándjaik fonalából a hálót a Muki
számára, vagyis ellene. Inkább a számára, mert hiszen, vallották
bácsiék, ha nem fogatja meg magát, előbb-utóbb az adósok
börtönébe kerül… Tudták, hogy ma már nincs is efféle börtön, de
azért csak igy mondták, mert volt valaha és mert jobban pászolt
volna nekik, ha ma is lenne. Akkor, vélték, egész biztos lenne, hogy
elsüthetik végre Ágnest. Már pedig az volt az ő egyedüli életcéljuk.
Szerettek volna végre-valahára megpihenni, holott a leány
minduntalan tüzes csóvákat hajigált békéjük olajába. Kigunyolta
anyját, mert martirt csinált magából; apját, mert folyton duruzsolt és
örökké élni akart. Egyszóval megzavarta szülői ártatlan kedvteléseit…
Leányuk elsikkasztása kérdésében elejétől fogva összetartottak az
öregek s ez az együttérzés volt az egyedüli kapocs, amely
negyvenkét évi frigyükből erkölcsi köteléket, ethikai értelemben vett
házasságot csinált. Mert egyébként szétfelé huztak minden vonalon…
Az öreg pár reménykedését nagyban táplálta az a körülmény,
hogy Ágnes egész nap sülve-főve volt Mukival. Ebből ki lehetett
kombinálniok, hogy ő is komolyan pályázik reá… Különben miért
ragadta volna minduntalan magával a kertbe?… Ott suttogott,
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com