EAESP Small Group Meeting on Counterfactual Thinking

La Baume, Aix-en-Provence, France
May 16-18, 2001


David R. Mandel (University of Victoria) dmandel@uvic.ca
Denis Hilton (Université de Toulouse II - le Mirail) hilton@univ-tlse2.fr
Patrizia Catellani (Catholic University of Milan) catellan@mi.unicatt.it



Since the publication in 1982 of Kahnehman and Tversky's generative article on the simulation heuristic, an increasing number of social psychologists worldwide have sought to understand the psychology of counterfactual thinking. Counterfactual thinking involves bringing to mind ways in which past events might have happened differently. It is a ubiquitous form of thought, and one that is central to the process of mental undoing, which refers to thinking that focuses specifically on the negation of past outcomes and which is part of people's psychological response system to negative and disconfirming events.

The burgeoning literature on counterfactual thinking cuts across several traditional topics in social psychology. One topic concerns how counterfactual thinking relates to a variety of forms of attributional thinking (e.g., attributions of causality, preventability, blame, and responsibility). A closely related topic concerns the influence of counterfactual thinking on judgment (e.g., assessments of culpability and compensation in social and legal contexts) and decision making (e.g., effects on future strategy selection and on choice). Another line of research concerns the effects of counterfactual thinking on emotion, particularly the ways in which such thinking can amplify people's affective responses to the negative and/or disconfirming outcomes that befall them. For instance, counterfactual thinking is implicated in negative feelings of regret, disappointment, shame, guilt, and distress, and in positive feelings of elation, fortunateness, and satisfaction. Work in this area has thus been central to our understanding of the cognition-emotion interface. Research on counterfactual thinking is by now also an important part of the literature on social comparison and comparative thinking.

Given the important consequences of counterfactual thinking for attribution, judgment and decision making, and emotional forecasting and response, the research has branched out into new areas of interdisciplinary concern, such as world politics, economics, marketing, legal philosophy, and cognitive science.

The proposed conference brought together a group of international experts with a shared interest in the area of counterfactual thinking. Given the rapid growth of this intra- and inter-disciplinary research area over the last decade, smaller meetings such as symposia at larger conferences cannot are unable to provide the space and time for presentations and discussion this thriving area deserves. This conference, which spanned three days, thus provided a unique forum for researchers to meet and disseminate their most recent findings in a well-structured environment.



May 16


Barbara A. Spellman
Wine, women, and Wells: Why thinking about more (consequent-changing) counterfactuals leads to greater attributions of causality.

Denis Hilton, John McClure, & Robbie Sutton
Selection of causes from event chains : A comparison of abnormality, intentionality and statistical principles.

John McClure & Denis Hilton
Counterfactual and causal judgments of intentional and physical causes in chains.

Clare R. Walsh & Ruth M.J. Byrne
A computational model of counterfactual thinking: The temporal order effect.

David R. Mandel
Across three Cs : Counterfactuals, causality, and contingency.

Ben R. Slugoski & M. Horne
Counterfactual thinking and implicit causality.

Chuck Tate & Fred B. Bryant
Knowledge structures and counterfactuals: Situating counterfactual thinking in a broader cognitive context.


Donatella Ferrante, Vittorio Girotto, Paolo Legrenzi, & Antonio Rizzo
Controllability effects in counterfactual thoughts.

Sergio Moreno-Rios & Juan Garcia-Madruga
Development of reasoning with counterfactual conditionals.

Mirela Bogdana Saupe
Counterfactual thinking as a coping strategy.

May 17


Dale T. Miller
The counterfactual correspondence bias.

Adam Galinsky, Vanessa L. Seiden, & Victoria Husted Medvec
The dissatisfaction of having your first offer accepted: The role of counterfactual thinking in negotiations.


Susana Segura & Michael W. Morris
Types of counterfactual conjectures and their relation to causal necessity and sufficiency.

Martin Goerke & Jens Möller
"It's not my fault ... but it is me who could change it": Counterfactual and prefactual thoughts of industrial managers.


Philip E. Tetlock
Poking counterfactual holes in covering laws: The tension between theory-driven and imagination-driven cognition in historical reasoning

R. Ned Lebow
What's so different about a counterfactual?

Patrizia Catellani & Patrizia Milesi
"We wouldn't have lost the election, if only…": Counterfactual thinking and collective action in political activists.

Neal J. Roese & Adam Galinsky
Counterfactuals as entertainment: Psychological underpinnings of pleasurable supposition in cognition and literature.

May 18


Marcel Zeelenberg, Kees van den Bos, Eric van Dijk, & Rik Pieters
The inaction effect in the psychology of regret.

Karl Halvor Teigen
Counterfactual thinking and luck.

Christopher O. Fraser
The role of emotional responses in counterfactual thinking and self-blame for negative outcomes involving the self.

Sharif El Leithy, Gary P. Brown, & Ian Robbins
Counterfactual thinking and post-traumatic stress reactions.


(listed alphabetically)

1. Patrizia Catellani and Patrizia Milesi (Catholic University of Milan)

“We wouldn’t have lost the election, if only…”. Counterfactual thinking and collective action in political activists

Previous studies have shown that counterfactuals influence preparation to action at an individual level. The two studies presented here investigated whether counterfactuals may influence the intention of getting involved in collective action. Activists of a political party that had been defeated in a recent political election were asked to think “how differently things might have gone, if only…”. Counterfactuals were classified according to their direction (upward vs. downward), their target (own party vs. other parties), and their being general vs. specific. Results showed that, in the first place, upward counterfactuals were largely prevailing over downward counterfactuals. Secondly, the more counterfactuals were general and focused on the subject’s own party, the higher was the intention of getting involved in collective action. Finally, counterfactual thinking was shown to predict action intention even when other predictors of collective action were controlled for. These findings are consistent with a functional view of counterfactual thinking.

Correspondence to:
Patrizia Catellani
Department of  Psychology
Catholic University of Milan
L.go Gemelli 1, I-20123 Milano
E-mail: catellan@mi.unicatt.it

2. Sharif El Leithy (Canterbury Christchurch University College), Gary P. Brown (Canterbury Christchurch University College) and Ian Robbins (Southwest London and St. George’s Mental Health NHS Trust)

Counterfactual thinking and post-traumatic stress reactions

The aim of the present study was to develop a phenomenological account of counterfactual thinking (CT) following trauma and to examine the associations between CT, traumatic stress reactions and adaptive/maladaptive cognitions and behaviors. A further aim was to examine how people manage CT, and how this may be associated with traumatic stress reactions. A sample of individuals who had been victims of violent, non-sexual assaults were recruited. Participants completed a series of questionnaires assessing posttraumatic stress and anxiety and depression symptoms, PSTD-related cognitions, and assumptions about the world. They also participated in a procedure designed to elicit trauma-related counterfactual thinking patterned after the procedure of Roese and colleagues (e.g. Roese & Olson, 1992). The responses were also content analyzed to identify factors related to predictions drawn from Kahneman and Tversky’s simulation heuristic, including the allocation of blame and causation. Initial findings bearing on the study hypotheses are presented.

Correspondence to:
Gary Brown
CASPD, Salomons
Broomhill Road
Tunbridge Wells, Kent
E-mail: g.brown22@salomons.org.uk

3. Donatella Ferrante (University of Trieste), Vittorio Girotto (University of Trieste  & LPC, Aix-en-Provence), and Antonio Rizzo (University of Siena)

Controllability effects in counterfactual thought

In their counterfactual thoughts about how events could have turned out differently, people tend to mutate controllable actions rather than uncontrollable circumstances. McCloy & Byrne (in press) showed that only inappropriate controllable actions are more mutated than uncontrollable events. These results seem to suggest that the controllability effect is a special case of the exceptionality effect, given that socially inappropriate actions represent exceptions with respect to prevailing social norms. In the present study we revisited the controllability factor. Specifically, we conducted two experiments to test the hypotheses that controllable actions are more mutable than uncontrollable events, regardless of their social appropriateness; that actions differ in their mutability, as a function of their different level of controllability; and that the more mutable actions are those that activate alternatives in which the protagonist of the scenario did not perform them. The obtained results corroborated our hypotheses.

Correspondence to:
Donatella Ferrante
University of Trieste
Via S. Anastasio 12
I-34134 Trieste
E-mail: ferrante@univ.trieste.it

4. Christopher O. Fraser (Auckland University of Technology)

The role of emotional responses in counterfactual thinking and self-blame for negative outcomes involving the self.

Previous research has shown that counterfactual thinking is not always associated with an increase in blame for the mutated event. The first of a series of four studies presented here showed that, for serious outcomes where actions of the self are antecedents and different self-actions could have prevented the outcome, there was an associated increase in self-blame. The second study showed that this self-blame was further influenced by the self-presentation strategy chosen, such as making either excuses or concessions about one’s actions. In two further studies, individual differences in guilt-proneness were shown to influence self-presentation strategy following self-mutations, but had no effect on the mutation focus itself. However, individuals with higher levels of empathy were more likely to mutate their own actions. These results suggest that blame attributions are the outcome of both spill-over effects from the initial counterfactual thinking and modifications occurring in subsequent processing, such as empathic or self-serving explanations.

Correspondence to:
Christopher Fraser
School of Education and Social Sciences
Auckland University of Technology
Private Bag 92006, Auckland 1020
New Zealand
E-mail: chris.fraser@aut.ac.nz

5. Adam Galinsky (University of Utah), Vanessa L. Seiden, and Victoria H. Medvec (Northwestern University)

The dissatisfaction of having your first offer accepted: The role of counterfactual thinking in negotiations

In this paper we explore the role of counterfactual thoughts in determining satisfaction with negotiated outcomes. In a scenario experiment and a simulated negotiation, immediate acceptance of a first offer increased counterfactual activation and decreased satisfaction. This reduction in satisfaction emerged even when the objective outcomes of negotiators whose first offers were immediately accepted were equal to or better than the outcomes of negotiators whose first offers were not immediately accepted.
The simulated negotiation experiment explored the functional and dysfunctional consequences of counterfactual activation on preparation for a subsequent negotiation. Upward counterfactual thoughts were positively related to the amount of preparation for a subsequent negotiation (a functional effect); on the other hand, upward counterfactual thoughts were negatively correlated with the likelihood of making future first offers (a dysfunctional effect, see Galinsky & Mussweiler, 2001). Our results find that counterfactual thoughts, rather than satisfaction, drive the effects on future behavior.

Correspondence to:
Adam Galinsky
Management Department
David Eccles School of Business
University of Utah
1645 E. Campus Center Dr. #106
Salt Lake City, UT 84112-9304
E-mail: galinsky@business.utah.edu

6. Martin Goerke and Jens Möller (University of Bielefeld)

“It’s not my fault ... but it is I who could change it”: Counterfactual and prefactual thoughts of industrial managers

We explored the contents of  N=120 managers’ counterfactual and prefactual thinking. Participants were instructed to focus on one subordinate (strong in performance, average, or weak), to generate either counterfactual (antecedents that would have led to a different performance) or prefactual thoughts (changes necessary to produce a different future performance), and to simulate both worse and better performance. Answers in the resulting 2 (time: counterfactual vs. prefactual) x 3 (subordinate: strong vs. average vs. weak) x 2 (direction: upward vs. downward) design were coded regarding reference (self vs. subordinate vs. outside). A significant four-way interaction (time x target x direction x reference) revealed that in concentrating on performances of weak subordinates, managers generated few self-referent upward counterfactuals. In contrast, self-referent upward prefactuals with regard to weak subordinates were particularly frequent: Managers saw themselves as not responsible for the weak past performance, but claimed numerous own options of improving future performance.

Correspondence to:
Martin Goerke
Universität Bielefeld
Abteilung für Psychologie
Postfach 100131
33501 Bielefeld
E-mail: martin.goerke@uni-bielefeld.de

7. Denis J. Hilton (Université Toulouse 2 – le Mirail), John L. McClure, and Robbie Sutton (Victoria University of Wellington)

Causal selection from event chains :A comparison of abnormality, intentionality and statistical principles

In this study we investigate selection of causes from chains of events that we call unfolding causal chains, which are characterised by the seemingly predictable and inevitable production of an end-result through a chain of intervening events, given the occurrence of a triggering event. We find that people are more likely to select distal causes as explanations of accidents if they involve human actions (e.g. an act of sabotage) rather than natural (e.g. heavy storms). We then test four statistical models to explain these causal preferences: two based on covariation analysis, and two others based on sufficiency analysis. Our results indicate that causal preferences are more strongly correlated with perceived sufficiency than perceived covariation, and that the two kinds of sufficiency we studied (non-conditionalised and conditionalised sufficiency) contributed independent variance to the prediction of causal preferences. However, further analyses suggested that our measures of perceived sufficiency did not completely explain the tendency to prefer intentional explanations, suggesting that intentions are not preferred as explanations simply because they are more "sufficient in the circumstances".

Correspondence to:
Denis Hilton
UFR de Psychologie
Université Toulouse 2 - le Mirail
5, allées Antonio Machado
31058 Toulouse Cedex 1
E-mail: hilton@univ-tlse2.fr

8. Richard Ned Lebow (Mershon Center, Columbus, OH)

What’s so different about a counterfactual?

I contend that the difference between so-called “factual” and counterfactual arguments is greatly exaggerated; it is one of degree, not of kind.  Both arguments ultimately rest on the quality of their assumptions, the chain of logic linking causes to outcomes and their consistency with available evidence. I review the criteria for counterfactual experimentation proposed by social scientists who have addressed this problem and find many of their criteria unrealistic and overly restrictive.  The methods of counterfactual experimentation need to be commensurate with the purposes for which they are used.  I discuss three generic uses for counterfactual arguments and thought experiments, propose eight criteria appropriate to plausible world counterfactuals and conclude by looking at the special problems associated with historical counterfactual experiments.

Correspondence to:
R. Ned Lebow
Mershon Center
1501 Neil Avenue
Columbus, OH 43201
E-mail: lebow.1@osu.edu

9. David R. Mandel (University of Victoria)

Across three Cs: Counterfactuals, causality, and contingency.

The present paper discussed the relations between counterfactual conditionals, causal judgments, and contingency representations. First, it was demonstrated by way of examples that, contrary to past claims, counterfactual conditional do not necessarily convey causal assertions. Second, a general framework outlining the relations among (a) counterfactual, semifactual, and isofactual propositions, (b) features of implicit hypotheses, (c) states of the world, and (d) tests of necessity or sufficiency was described. Finally, an experiment was reported that demonstrated a three-way dissociation among judgments of counterfactual undoabilty, causality, and conditional probability in a hypothetical case of multiple sufficient causes. Whereas causal judgment focused on factors that directly contributed to the isofactual outcome, probability and counterfactual judgments were more sensitive to antecedents that made a similar outcome inevitable from a prefactual stance. The difference between isofactuals and prefactual inevitabilities is considered in terms of time-uncertainty interactions.

Correspondence to:
David R. Mandel
Department of Psychology
University of Victoria
P.O. Box 3050
Victoria, British Columbia
Canada  V8W 3P5
E-mail: dmandel@uvic.ca

10. John McClure (Victoria University of Wellington) and Denis Hilton (Université Toulouse 2 – le Mirail)

Counterfactual and causal judgements of intentional and physical causes in chains

The present studies compare causal and counterfactual judgements of intentional and physical causes, when those causes are placed in causal sequences. Hart and Honoré (1985) claimed that in causal sequences, intentional causes and physical causes are seen as equally necessary for the outcome, but that intentional causes are seen as better explanations than physical causes, and that the most recent intentional cause is judged the primary cause of events. Spellman (1997) claimed that causal judgements in relation to causal sequences reflect how much each cause changes the probability of the outcome. To evaluate these two models, scenarios described intentional and physical versions of distal and proximal causes preceding events. Intentional causes were judged better explanations than physical causes, but these judgements were not based on counterfactual necessity.  Judgments of necessity were affected more by the position of the cause in the sequence than by the nature of the cause (intentional or physical). The findings are discussed in relation to Hart and Honoré’s and Spellman’s theories of counterfactual and causal judgements in causal chains.

Correspondence to:
John McClure
School of Psychology
Victoria University of Wellington
P.O. Box 600, Wellington
New Zealand
E-mail: john.mcclure@vuw.ac.nz

11. Dale T. Miller, Penny S. Visser, and Brian Staub (Princeton University)

The counterfactual correspondence bias

Two studies support the hypothesis that observers’ impressions of actors reflect not only what actors do, but what they can easily be imagined doing. Participants observed 10 year-old boys take a math test in a context in which the temptation to cheat and the constraints against cheating varied. When the temptation to cheat was high, but the likelihood of getting caught was also high, observers perceived a non-cheating actor to be less trustworthy than a control actor whom they did not see. This effect was not found when the temptation to cheat was low, suggesting that its occurrence under high temptation resulted from observers in that condition generating the counterfactual thought that the actor would have cheated had the likelihood of detection been low. The implications of the counterfactual correspondence bias are discussed, as are the more general links between counterfactual thought and the person perception process.

Correspondence to:
Dale T. Miller
Department of Psychology
Princeton University
Princeton, NJ 08544-1010
E-mail: dmiller@princeton.edu

12. Sergio Moreno Rios (Univesidad de Granada) and Juan Garcia-Madruga (UNED, Spain)

Development of reasoning with counterfactual conditionals

117 pupils between 7 and 14 years old were tested in a reasoning task with factual and semifactual conditionals. Results showed that, like adults, children accept less inferences with semifactual than with factual conditionals. Also like adults, children were able to notice that the negation of the consequent was not possible, and also they noticed that negating the antecedent did not lead to a negation of the consequent but to an affirmation of it. However, unlike adults, they could not connect information from both situations: the affirmation of the consequent led to opposite antecedents, and therefore «nothing can be concluded». That is, they did not appreciate that from a semifactual. Results were explained on the basis of the mental model theory, assuming the existence of a developmental sequence in the ability of carrying out mental operations.

Correspondence to:
Sergio Moreno Rios
Departamento de Psicologia Evolutiva y de la Educaciòn
Universidad de Granada
Campus Universitario de Cartuja
18071 Granada
E-mail: semoreno@ugr.es

13. Neal J. Roese (Simon Fraser University) and Adam Galinsky (University of Utah)

Counterfactuals as entertainment: Psychological underpinnings of pleasurable supposition in cognition and literature

Counterfactuals are fun. Imagining what might have been had one's parents never met, or how life might have turned out differently had some romantic flame been followed, can be self-indulgently pleasurable. Moreover, the popularity of recent scholarly and fictional treatments centering on historical alternatives (what if Hitler had triumphed in WWII?) attests to the notion that there are some counterfactuals that are particularly compelling, haunting, and intriguingly entertaining. We discuss five explanations that explain which counterfactuals are particularly entertaining: 1) expectancy-value framework, 2) artistic conceptions of theme and variation, 3) propensity and trajectory, 4) family resemblance, and 5) invitations to further elaborative thought.

Correspondence to:
Neal J. Roese
Department of Psychology
Simon Fraser University
Burnaby, BC V5A 1S6
E-mail: roese@sfu.ca

14. Mirela B. Saupe (Babes-Bolyai University)

Counterfactual thinking as a coping strategy

This paper presents the results of 2 pilot studies regarding the affective consequences of counterfactual thinking and the possibility of using a specific pattern of counterfactuals as an affective coping response to negative life events. The first study showed that depressed subjects made more upward counterfactuals than did the non-depressed subjects. They also made more additive counterfactuals but these counterfactuals were focused on aspects of the event that were not under personal control. The second study showed that high self-esteem subjects reported more downward counterfactuals after a negative event than low self-esteem subjects and had also more mood improvement as a consequence. The second study focused on the mediator effect of self-esteem counterfactual production. Data from these studies are discussed in the context of clinical psychology. Specifically, the paper is concerned with the practical implication of these preliminary findings in the cognitive therapy of depression and posttraumatic stress disorder.

Correspondence to:
Mirela B. Saupe
Department of Psychology,
Babes-Bolyai University,
E-mail: mbogdana@hotmail.com

15. Susana Segura (University of Malaga) and Michael W. Morris (Columbia University)

Might have been vs. would have been: Perceived causal relationships as determinants of asserted likelihood of counterfactual outcomes

The authors propose a distinction between "might have been" and "would have been" conjectures and predict they differentially depend on the perceived causal relationship between the antecedent and the outcome. "Would have been" claims require a stronger relationship (involving causal necessity).  To test this against opposing claims in the literature, we varied the causal necessity and sufficiency of an antecedent event for an outcome and measured the conjectures that people endorsed about the outcome.  Further, we varied the outcome valence (positive vs. negative) languages (English vs. Spanish) to test the generality of the pattern. Findings showed that the rated plausibility of "would have been" conjectures is generally lower and, more dependent on perceived necessity than that of "might have been" or causal conjectures.  Moreover, "would have been" conjectures were accompanied by stronger causal inferences linking changed antecedents with changed outcomes, suggesting they serve in the preparative functions of counterfactual thinking.

Correspondence to:
Susana Segura
Departamento de Psicologia Basica
Facultad de Psicologia
Campus de Teatinos
29071 Malaga
E-mail: s_segura@uma.es

16. Ben R. Slugoski and M. Horne (James Cook University, Townsville)

Counterfactual thinking and implicit causality

Early work on the relationship between causal and counterfactual reasoning supported an intimate connection between the two. However, more recent researchers have disputed that such an intimate connection exists. Whether pro or con,  virtually all studies in the area have utilized scenario-type formats, which we believe can be faulted for order and other artifacts. We conducted two studies that examined the relationships between participants' causal and counterfactual judgments using more implicit measures modelled on the 'verb effect' paradigm. Using six action and six state verbs, we had participants generate responses to both causal and counterfactual sentence stems, predicting that the counterfactual completions would mirror the causal completions in the proportions of participants providing subject or object completions.Taken together, the results suggest a quite intimate relationship between judgements of causality and counterfactuality, as would be expected from a merging of the abnormal conditions focus model (Hilton & Slugoski, 1986) with Brown and Fish's (1983) causal schema hypothesis.

Correspondence to:
Ben R. Slugoski
School of Psychology
James Cook University
Townsville QLD 4811
E-mail: ben.slugoski@jcu.edu.au

17. Barbara A. Spellman (University of Virginia)

Wine, women, and Wells: Why thinking about more (consequent-changing) counterfactuals leads to greater attributions of causality

Researchers now agree that causal and counterfactual reasoning ask different questions and serve different functions; yet we must acknowledge that they are related.  Evidence for the relation comes from experiments showing that (a) the mere existence of an obvious counterfactual alternative to an outcome may affect causal attributions and (b) considering counterfactual alternatives to an outcome may affect later causal attributions.I present evidence from two groups of experiments.  First, we examine order effects when subjects make both causal and counterfactual judgments. Second, we examine how the number and kind (consequent-changing or not) of counterfactual alternatives presented to subjects affects causal judgments. Then I tie these findings into a theory of the relation between causal and counterfactual reasoning in which: (a) causality judgments are based on how much events increase the probability of the outcome above its baseline probability and (b) counterfactual reasoning is used to create those probability estimates.

Correspondence to:
Barbara A. Spellman
Department of Psychology
University of Virginia
P.O. Box 400400
Charlottesville, VA  22904-4400
E-mail: spellman@virginia.edu

18. Chuck Tate and Fred B. Bryant (University of Oregon)

Knowledge structures and counterfactuals: Situating counterfactual thinking in a broader cognitive context

This study investigates two related questions, both concerning counterfactual thinking and its relationship to social knowledge structures. The first question investigates whether people’s folk theories about agency and determinism in human behavior matter differently for the specific content of mutated events or characteristics, for ascriptions of causality, and for ascriptions of responsibility. The second question investigates whether perspective manipulations are differentially related to the above measures. We present data from 210 participants which suggest that counterfactual content and the ascriptions listed above are influenced by folk theories about human behavior and perspective manipulations, with more consistent support for the former than the latter. This study therefore provides evidence that counterfactual thinking exists within and is informed by the content of social knowledge structures.

Correspondence to:
Chuck Tate
Department of Psychology
1227 University of Oregon
Eugene, OR, 97403-1227
E-mail: ctate1@darkwing.uoregon.edu

19. Karl Halvor Teigen (University of Tromsø)

Counterfactual thinking and luck

Good and bad luck events from daily life contain a strong counterfactual component. People perceive themselves and others as lucky when a worse outcome could “easily” have happened. They are more lucky the closer they are to the downward counterfactual outcome and the worse it could have been. This is typically the case in dangerous situations and in situations in which they admit to have behaved carelessly. Downward counterfactuals are typically suggested by the situation, either in terms of a close call, or by a general, negative trajectory that is changed “in the last minute”. Thus good luck situations are not necessarily positive, whereas bad luck situations are almost always negative, because people tend to think that negative outcomes could have been avoided and thus generate upward counterfactuals when such outcomes occur. We also show that own luck is related to gratitude and other’s luck to envy. Both these emotions presuppose a realization of “it could have been different”.

Correspondence to:
Karl Halvor Teigen
Department of Psychology
University of Tromsø
N-9037 Tromsø
E-mail: karlht@psyk.uit.no

20. Philip E. Tetlock (The Ohio State University)

Poking counterfactual holes in covering laws: The tension between theory-driven and imagination-driven cognition in historical reasoning

This paper reports both correlational and experimental studies of historical reasoning among professional observers of world politics. The correlational studies demonstrate that counterfactual reasoning is largely theory driven. Abstract covering-law beliefs are potent predictors of reactions to close-call counterfactuals that imply history could have been easily re-routed with minimal modifications of antecedent conditions. The experimental studies demonstrate that experts are capable of surprising themselves as they explore the imaginative implications of dissonant what-if scenarios. Indeed, the more detailed experts explorations of the counterfactual worlds, the greater the risk that they will assign too much subjective probability to too many classes of scenarios. The paper closes by defining good judgement as a process of reflective equilibrium, of striking a balance between theory-driven and imagination-driven patterns of thinking and between the causal and moral intuitions primed by complementary factual and counterfactual framings of historical questions.

Correspondence to:
Philip E. Tetlock
Department of Psychology
The Ohio State University
1885 Neil, Townsend Hall
Columbus, Ohio 43210
E-mail: tetlock.1@osu.edu

21. Clare R. Walsh and Ruth M.J. Byrne (University of Dublin, Trinity College)

A computational model of counterfactual thinking: The temporal order effect

We report the results of two experiments that examine what events people focus on in their counterfactual thoughts. Previous research shows that people tend to mentally undo the most recent event in an independent sequence. We demonstrated that the representation of the counterfactual events rather than the representation of the factual events may be the source of this temporal order effect. In our experiments, we varied the framing of the winning conditions that could have led to a counterfactual outcome. We gave participants scenarios about two players picking red or black cards to win a prize of £1000, with the factual outcome that both players picked black cards and lost. We showed that the temporal order effect can be eliminated and even reversed when the counterfactual events are manipulated. We describe a computational model that simulates our theory of the mental representations and cognitive processes underlying the temporal order effect.

Correspondence to:
Clare Walsh
Psychology Department
University of Dublin
Trinity College
Dublin 2
E-mail: http://web.uvic.ca/psyc/dmandel/mailtocwalsh@tcd.ie

22. Marcel Zeelenberg (Tilburg University), Kees van den Bos (Free University Amsterdam), Eric van Dijk (Leiden University), and Rik Pieters (Tilburg University)

The inaction effect in the psychology of regret

Ample research has shown that decision outcomes produce more regret when they are caused by actions as compared to inactions. Such research largely ignored the fact that decisions are often made in response to earlier outcomes, and that sometimes these prior outcomes may promote action and hence make inaction more “abnormal”. Four experiments investigated these effects of prior outcomes on decision regret. It was manipulated whether information about a prior outcome was positive, negative or absent. As hypothesized, positive or absence of prior outcomes caused people to regret action more than inaction. However, as predicted and counter to previous research, in the case of negative prior outcomes, inaction resulted in more regret than action; a finding that we label the ”inaction effect”. The final experiment, focusing on the comparison of regret with the related emotion disappointment, revealed that these effects are typical for regret, and this experiment demonstrates the need to be specific about the emotion under investigation.

Correspondence to:
Marcel Zeelenberg
Department of Economic and Social Psychology
Tilburg University
PO Box 90153
5000-LE Tilburg
The Netherlands.
E-mail: m.zeelenberg@kub.nl


The past 15 years have witnessed an increasing number of studies on counterfactual thinking, that is thinking of how past events might have happened differently. Counterfactuals are often generated in everyday life, especially to undo negative outcomes, but they are also frequent in more specialised fields, such as legal reasoning. Research on this issue has contributed to several traditional topics of social psychology, such as attributional thinking, judgment, and decision making. Recently, it has been extended to various areas of interdisciplinary concern, such as politics, history and economics.

The aim of the Small Group Meeting at Aix-en-Provence was to integrate the advances in recent research on counterfactual thinking, by bringing together a significant representative of researchers interested in the issue. Some 25 researchers from Europe, United States and Australia attended the meeting, and worked together for three days in the inspiring framework of La Baume, a former Jesuit seminary on the outskirts of Aix.

Day One was devoted to basic theoretical and empirical issues in the study of counterfactual thinking (Barbara Spellman, Denis Hilton, John McClure, Clare Walsh, David Mandel, Ben Slugoski, and Chuck Tate). Relationships with causal thinking were especially considered, showing on the one hand how generating or being presented with a counterfactual alternative to an outcome may affect causal attributions and, on the other hand, how counterfactual and causal judgment do not always overlap. The various talks differed as regards their theoretical approach, spanning from "pure" cognition to wide consideration of intentional and motivational factors. Thanks to this, and to the large time space allotted to each talk, discussion turned out to be especially alive and fruitful. A further opportunity of discussion was offered by the poster session held in the afternoon, including Donatella Ferrante, Sergio Moreno and Mirela Bogdana.

Day Two featured two sets of presentations. The first set dealt with counterfactuals in interpersonal contexts, with Dale Miller presenting interesting data on how availability of counterfactual scenarios may affect person perception judgments, Adam Galinsky highlighting the role of counterfactuals in negotiations, and Susana Segura and Michael Morris focusing on linguistic formulation of counterfactual statements. The second set of presentations focused on counterfactuals in business, history, politics, and literature (Martin Goerke, Phil Tetlock, Ned Lebow, Patrizia Catellani, and Neal Roese). Consideration of possible alternatives is a common reasoning strategy in all these knowledge domains and may serve different functions: preparing future individual and collective action in organizational and political contexts, explaining and argumenting in relation to historical events, raising involvement and fun in readers of narratives. The lively atmosphere of Aix, with its pleasant open-air bars in horsechestnut-lined squares, concluded the day in the best possible way.

Day Three included a set of talks on the affective correlates and consequences of counterfactual thinking (Marcel Zeelenberg, Karl Teigen, Christopher Fraser, and Sharif El Leithy). The conditions under which counterfactuals may generate more or less strong emotional reactions were examined in detail, considering both negative emotions following simulation of better outcomes (regret, disappointment, worry), and positive emotions and perceptions following simulation of worse outcomes (gratefulness, perception of luck). The meeting concluded with a round-table and a sight-seeing trip to Cassis, a delightful seaside port where the group stayed over for dinner in a restaurant with a magnificent view of the harbour.

In brief, the meeting attained its main goal of offering an articulate and updated image of current research in the field of counterfactual thinking. The peaceful and green environment of the ancient La Baume seminary favoured conversation and exchanges among all the participants who, in many cases, were meeting each other for the first time. Comparing different perspectives, making comments on each other’s research, and suggesting developments for future research were activities in which all participants took part, offering a confirmation of the liveliness of this research field and laying the foundations for future meetings of a similar kind to be held in the future.

This page maintained by David R. Mandel