1 2

Dealing with contradiction in a communicative context: A cross-cultural study

3 4 5

JEAN-BAPTISTE VAN DER HENST, HUGO MERCIER, HIROSHI YAMA, YAYAOI KAWASAKI, and KUNIKO ADACHI

6 7 8 9 10 11 12

Abstract

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

In this paper, we investigate the impact of two factors on the way people address conflicting information. One relates to culture and the other to the communicative context in which contradiction occurs. We compared two theoretical approaches, one that focuses on the former factor—the culturalist approach—and one that focuses on the latter factor—the evolutionary approach. According to the culturalist approach, the way we deal with contradiction can be markedly a¤ected by culture, so that people from cultural environments with di¤erent social practices are more or less inclined to accept contradictions. In particular, this approach predicts that Easterners are more likely to search for a compromise between two conflicting viewpoints than Westerners, who tend to follow a logical principle of noncontradiction. In contrast, the evolutionary approach considers that when contradiction occurs in a communicative context, universal mechanisms designed to deal with the problem of managing deceptive information go into e¤ect and lead to the tendency of giving more weight to one’s own belief than to the other’s conflicting view. We tested these two approaches with Japanese and French participants. Our data supports the evolutionary approach, since both groups showed the same bias of favoring one’s own position when it was challenged by another’s.

33 34 35

1. Introduction

36 37 38 39 40 41 42

Imagine you have acquired some expertise in performing a specific action, like fishing. This implies that you have some beliefs about the best way to fish, including what types of baits and hooks are best to catch certain species of fish, or the rivers or lakes that are stocked with fish. But to what extent are these beliefs entrenched in your mind? Consider, for instance, a case in which someone who is also an expert in the same thing states a Intercultural Pragmatics 3-4 (2006), 487–502 DOI 10.1515/IP.2006.029

1612-295X/06/0003–0487 6 Walter de Gruyter

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 487)

488 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Jean-Baptiste van der Henst et al.

view that is opposite to your own. Would you abandon your beliefs and adopt the new view? Would you, on the contrary, stay your ground? Or, would you entertain the idea that the two opposite views are true, according to the context? There are a number of cues that could influence the stance you take: you might consider the information on which the speaker relies in order to advocate her position and judge whether or not it is true; you might also analyze how she uses such information and how it relates to her thesis (that is, Is the speaker’s thesis justified by her premises?). You might also consider the speaker’s personal characteristics and decide whether she has ever provided you with erroneous information or whether she has deliberately tried to mislead you in the past; you might also consider whether she now has a reason to get you to act in a certain way. All these thoughts would occur intuitively and they would be important in your decision to adopt a position that conflicts with your own. It is unlikely that you would follow the speaker’s view if you discover that the information on which it depends is untrue. Alternatively, if the speaker had always provided you with accurate information that had even led you to revise your previous beliefs, you might be more likely to consider her current view as carrying some truth, even though it contradicts your own position. Although these aspects are likely to play a role in the way that we process contradiction, they are relatively dependent on the specific circumstances in which the contradiction occurs. In this paper, we investigate the general processing factors that influence the way we deal with contradiction in the context of communication. We consider two general factors: one relates to the influence of culture on the processing of contradiction and the other relates to the adaptive problem that communication poses. We then present two cross-cultural experiments that aim to identify which factor predominates. In a recent research study, Richard Nisbett et al. (2001) questioned the assumption that cognitive processes are universal and thus basically the same for all human beings. On the one hand, Nisbett and his colleagues agree that some aspects of cognition are constrained by universal cognitive structures so that no matter what environment human beings grow up in, they tend to have the same expectations about the physical and the social world. On the other hand, they advocate the view that culture can markedly a¤ect cognitive processes. This means that people from di¤erent cultures may not only think about di¤erent things but may also think in radically di¤erent ways. Which aspects of culture most influence our cognitive processes? One important candidate is language. This is the heart of the linguistic relativity hypothesis (Whorf 1956), according to which language largely determines thought processes. While this hypothesis has

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 488)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

489

been widely debated, it has received some empirical support from di¤erent fields of cognition. For instance, Levinson (1997) presented evidence indicating that spatial language has an influence on the way people locate objects. In some languages, speakers tend to describe objects in space in a relative manner and use spatial expressions such as ‘‘to the left of ’’ or ‘‘in front of,’’ while in other languages, speakers tend to focus more on absolute directions such as north, south, east, and west. Several empirical studies involving non-linguistic tasks revealed that speakers of the latter kind of languages were more likely to conceive space in an absolute way (Levinson 1997). Other studies have revealed that reading and writing habits (from left to right versus from right to left) are responsible for directional biases in various cognitive tasks such as line bisection tasks (Jewell & McCourt 2000), line extension, and line estimation tasks (Chokron, Bernard & Imbert 1997; Singh, Vaid & Sakhuja 2000), recall and naming tasks (Nachshon, Shefler & Samocha 1977; Padakannaya, Devi, Zaveria, Chengappa & Vaid 2002), and drawing tasks (Vaid, Singh, Sakhuja & Gupta 2002). Moreover, in the field of numerical cognition, di¤erences in number naming systems seem to result in di¤erences in the way children count (Miller, Smith, Zhu & Zhang 1995). There is also some work showing that color terms have an influence on the perception on color (Roberson, Davies & Davido¤ 2000). Nisbett et al. (2001) consider di¤erent cultural factors, apart from language, that are likely to influence the way we think; in particular, social practices. According to them, di¤erent organizations of the social world focus on di¤erent aspects of reality, which lead to di¤erent conceptions of the world and of its functioning. This, in turn, results in di¤erent systems of thought that influence the way cognitive processes operate. To illustrate their view, they compare the Western world, influenced by the ancient Greek civilization, and the Eastern world (in particular, China, Japan, and Korea), influenced by the ancient Chinese thought. Social organization in ancient Greece was based on the individual: a person was thought to have the ability to control some aspects of the world, to have a sense of personal agency, and to act on the basis of individual decisions by moderating the influence of social constraints. In contrast, the social system of ancient China was much more complex, and involved a high degree of social obligation towards the family, the landlord, the emperor’s representatives, and so forth. An individual was viewed as a part of the group and was expected to act in relation to the group in order to preserve its cohesion. It follows that maintaining in-group harmony became a principal goal of the individual. In contrast, confrontation, conflict, debate, and competition were more likely to occur within Western social systems that promoted the individual as an entity independent of the group.

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 489)

490 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Jean-Baptiste van der Henst et al.

Nisbett et al. (2001: 293) made a key distinction between the holistic and dialectical way of thinking in the East Asia and the analytic and logical way of thinking in the West: ‘‘we define holistic thought as an orientation to the context or field as a whole, including attention to relationships between a focal object and the field, and a preference for explaining and predicting events on the basis of such relationships. Holistic approaches rely on experience-based knowledge rather than on abstract logic and are dialectical, meaning that there is emphasis on change, a recognition of contradiction and of need for multiple perspectives. We define analytic thought as involving detachment of the object from its context, a tendency to focus on attributes of the object to assign it to categories, and a preference for using rules about the categories to explain and predict the object’s behavior. Inferences rest in part on the practice of decontextualizing structure from content, the use of formal logic and avoidance of contradiction.’’ Indeed, laws of formal logic involve a law of noncontradiction so that a proposition has to be either true or false but cannot be both. If it is the case that P is true, it follows that non-P is false. The application of the rule of non-contradiction implies that when an individual is confronted with two opposite viewpoints, he will think that one is right and the other is wrong. Dealing with contradiction consists in eliminating one of the two viewpoints in order to rule out contradiction. In contrast, according to dialecticism and holistic thought, reality is not static but is in constant change. One cannot give a fixed and durable description of the world. A statement can convey some truth about some aspect of the world at a given time but a contradictory statement can also convey some truth about another aspect of the world or at another time. It follows that contradiction is inherent to reality and ought not to be eliminated. When two opposing viewpoints are encountered, dialectic thinking consists in finding a compromise or a middle way in order to recognize and a‰rm that they both contain some truth. Peng and Nisbett (1999) provided some empirical support for these differences by analyzing how people from Western and Eastern cultures deal with contradiction. For instance, they found that Chinese students exhibited a greater preference for proverbs containing contradictions (e.g., ‘‘beware of your friends not your enemies’’) than non-dialectical proverbs, whereas Americans showed an opposite pattern of preference. They also analyzed social contradictions, that is, contradictions that involve conflicts between people. In their third study, Chinese and American participants were given two real-life conflicts (a mother-daughter conflict over values and a playground conflict) and then had to describe how such conflicts occurred and how they could be resolved. The results indicated a clear cultural di¤erence: Chinese participants provided more dialectical

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 490)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

491

responses (i.e., responses that took both views into account and that adopted a middle way) than non-dialectical responses (i.e., responses that favored only one side; 72% vs. 28%). In contrast, American participants gave fewer dialectical responses than non-dialectical ones (26% vs. 74%). In another study on argumentation, in which participants had to compare two arguments (one dialectical and the other not) leading to the same conclusion, Peng and Nisbett observed that American participants preferred logical arguments following the law of non-contradiction, whereas Chinese participants preferred dialectical arguments in line with the principle of holism. Finally, in their last study, participants had to judge the plausibility of arguments concerning scientific topics. Participants had to judge a single argument or the same argument with another conflicting argument that could be more or less plausible. When they had to assess the two opposing arguments, American participants rated the plausibility of the most plausible arguments as greater than when they had to evaluate it alone (i.e., without contradiction). This may be explained by the application of the law of non-contradiction that leads one to polarizing judgments. In contrast, when they had to assess the two contradictory arguments, Chinese participants rated them as equally plausible (in other words, the less plausible arguments were considered to be more plausible than when they were evaluated alone). This result is in line with the view that Chinese follow a compromise approach by accepting the notion that truth can be found in two opposite positions. Although the data reported by Peng and Nisbett revealed di¤erent ways of dealing with contradiction, one may question the importance of these di¤erences in everyday life. It seems unlikely that Easterners would be willing to accept any kind of contradiction. In the experiments Peng and Nisbett describe, participants are provided with two contradictory positions but are not strongly committed to either of these positions. They are relatively external to the debate they have to consider. Take, for instance, the study on social contradiction; participants did not have to impose their own views on the subject and did not have to advocate them against an opponent. Their role was that of a judge, not of a contender taking part in a debate. A similar remark can be made about this last study: although participants had to evaluate the plausibility of arguments, they were not put in a position in which a contradictory viewpoint was posited in order to challenge their own perspective. In the present study, we attempt to analyze how people react when a speaker communicates to them a position that conflicts with a view they already hold. This raises the question of the general stance we take towards communicated information. On the one hand, most of our

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 491)

492 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Jean-Baptiste van der Henst et al.

knowledge is socially transmitted by means of communication. Schoolteachers, journalists, friends, neighbors, and so on, provide us with information. Through social communication, one may surpass the knowledge individually acquired on one’s own. The fact that a large part of our knowledge has a social origin has some implications for our cognitive processes. On the one hand, people must be able to assimilate this socially transmitted knowledge and to accept it as having merit. On the other hand, those processes can be manipulated. We can reasonably trust that individually acquired knowledge does not result from processes susceptible to deception (at least in our natural environment; see Sperber 2001). However, this is not so true of knowledge obtained through social communication. Our interlocutor may want us to believe something so that we will behave in a certain way. If she believes that a piece of information would lead us to act in a way that is advantageous to her, and that we are likely to believe this information, she might choose to communicate something false on purpose. Thus, social communication carries with it not only the risk that our interlocutors may be mistaken, but that they may want to manipulate us (Dawkins & Krebs 1978; Sperber 2001). It follows that a significant amount of socially transmitted beliefs are likely to be false. This is a general problem posed by communication. How can this problem be solved? The ‘‘Machiavellian Intelligence hypothesis’’ endorsed by numerous primatologists and researchers specializing in evolution postulates that for primates, ‘‘high level’’ cognitive abilities have evolved by natural selection in order to deal with the complexity of their social environment (Byrne & Whiten 1988; Whiten & Byrne 1997). Among these abilities, some may be dedicated to manipulation—either for manipulating others or for warding o¤ manipulation attempts by others (Cosmides & Tooby 1992; Sperber 2001). If communicated information was given the same weight as individually acquired information, people would be not be able to thwart manipulation attempts and they would often be misled (Sperber 2001). The avoidance of manipulation relies on a person’s ability to give more or less weight to a piece of information, that is, on the ability to modulate the e¤ect it may have on the context in which it is processed. All things being equal, one may predict that in order to deal with the adaptive problem posed by the social transmission of knowledge, greater weight will be given to individually acquired information than to socially transmitted information. It follows that when these two types of knowledge come into conflict people will tend to favor their own position. This approach has received empirical support from the literature in social psychology. Yaniv (2004; see also Yaniv & Kleinberg 2000) has

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 492)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

493

analyzed the weight that people give to advice in a set of empirical studies. In the first phase of the task, participants had to answer fifteen questions about historical events (e.g., ‘‘In what year was the Suez Canal first opened?’’) by indicating their best estimate, as well as lower and upper boundaries. In the second phase, participants were again asked the same questions along with estimates provided by an advisor. Participants thus had to give a second answer that would or would not take the advisor’s estimate into account. Yaniv defined a ‘‘weight of advice’’ that would take the value of 0 if a participant completely discounted a piece of advice and the value of 1 if she completely endorsed the advice. His findings indicated a clear tendency to discount other’s opinions with a mean weight of advice of 0.27, which was significantly smaller than the averaging value of 0.5 (Yaniv 2004, study 1; see also Yaniv & Kleinberg 2000). According to Yaniv, these data can be accounted for by an ‘‘informational asymmetry’’ between the access of evidence supporting one’s own opinion and the access of evidence supporting the advisor’s opinion. While an individual can analyze, to some extent, the reasons and facts on which his own beliefs rely, he cannot evaluate as easily why an advisor adopts a particular view. More precisely, since an individual can not access why an opponent makes a particular claim, this may cause him to favor his own view in order to ward o¤ manipulation. The bias favoring one’s own opinion has also been observed by Mercier and Van der Henst (2005) with a di¤erent empirical procedure that involved a more direct conflicting situation of communication. In the first experiment, participants were provided with short scenarios describing a real-life situation in which two pieces of information coming from two di¤erent sources conflicted with each other. In the ‘‘self-other’’ condition, the participant was first placed in a position to hold an initial viewpoint that was then challenged by another person that advocated an opposite opinion. In the ‘‘other-self ’’ condition, the initial viewpoint was taken by someone and the participant was put in a position of challenging it by advocating the opposite view. Participants then had to make a decision that reflected the choice between one and the other opinion. The data showed a clear bias towards the position participants were assigned to. In the ‘‘self-other’’ condition, 65% of the answers were in line with the initial view, which was the view participants hold (meaning that 35% were against); in the ‘‘other-self ’’ condition, 79% of the answers were in line with the challenging view, which again was the view that participants hold (21% were against). Although these studies report a clear self-other di¤erence e¤ect, it is noteworthy that they have been carried out with populations taken from the Western world. This raises the question of the universality of the

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 493)

494 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

Jean-Baptiste van der Henst et al.

processes underlying such an e¤ect. As indicated above, cultural psychologists have argued that cognitive processes generating contradiction are dependent upon cultural factors: Easterners are more likely to accept contradictions and are thus more likely to endorse diverging viewpoints than Westerners. The culturalist hypothesis predicts that the bias favoring one’s own opinion should be less pronounced, or even absent, for Easterners. Indeed, if Easterners follow a ‘‘middle way’’ approach, they would be more willing to accept a conflicting viewpoint than Westerners that aim at eliminating contradictions. In contrast, according to the evolutionary hypothesis, the tendency to favor one’s own opinion is a result of a set of adaptive mechanisms designed to prevent the acquisition of false beliefs through social interaction. Such mechanisms are considered universal and should be e¤ective in both cultures. The present study tests the cultural and evolutionary approaches by comparing how Japanese and French participants deal with contradictions in the context of communication. We report two experiments that involve scenarios similar to those used by Mercier and Van der Henst (2005). The two experiments mainly di¤er according to the answer format provided to participants. In Experiment 1, participants were forced to make a choice between two opposite opinions, whereas in Experiment 2 participants were allowed to make a more fine-grained evaluation of opinions through the use of a scale. If the self-other di¤erence e¤ect can be observed in both cultures with no cross-cultural di¤erence, this would corroborate the evolutionary prediction and would challenge the cultural prediction. If the self-other di¤erence e¤ect is observed in both cultures but is less pronounced for Japanese than for French participants, this would be compatible with the two hypotheses: the prediction of universality made by the evolutionary hypothesis and the cultural modulation prediction made by the cultural hypothesis would both be supported. If the self-other e¤ect is observed for French but not for Japanese participants, this would corroborate the cultural prediction and would challenge the evolutionary prediction. Finally, a puzzling result would be a higher self-other e¤ect for Japanese than for French participants.

35 36 37

2.

Experiment 1

38 39

2.1.

Participants

40 41 42

Ninety-three Japanese business students from the University of Kinki (45 females and 48 males) and 82 French business students from the

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 494)

Dealing with contradiction in a communicative context 1 2

495

University of Lyon 3 took part in this experiment (52 females and 30 males).

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

2.2.

Method

In this experiment, participants had to read a small scenario. They had to imagine themselves as having a particular opinion on how to carry out an activity in which they were hypothetically quite experienced. More specifically, they had to hold the view that a first option (O1) is better than a second option (O2) for carrying out this activity. One out of four similar daily life scenarios were provided to each participant. In the ‘‘fishing scenario,’’ they had to hold the view that one river was more suitable for fishing than another because it contained more fish. In the ‘‘salesperson scenario,’’ they had to hold the view that one particular route was faster than another to go from one place to another. In the ‘‘motorbike scenario,’’ they had to hold the view that one motorbike had a greater acceleration than another. Finally, in the ‘‘pasta scenario,’’ they had to hold the view that a specific flour ensured a quicker cooking time of noodles than another. Then participants had to imagine that they encountered another person who was as experienced as themselves in this activity. However, this second person claimed that the second option O2 was preferable, thus holding a view that was at odds with the first participant’s position. Participants were then asked whether they would adopt O1 (their preferred position at the beginning of the scenario) or O2 (the position advocated by the other person) in the future. Here is how the ‘‘salesperson scenario’’ was framed: You are a salesperson working in Brittany [In the Japanese version, an area of Japan]. You make the trip between Brest and Quimper [two Japanese cities in this area] very often. There are two possible routes, one that passes through Landrieu and the other through Guersac. You think that the route that passes through Landrieu is much faster than the one that passes through Guersac. In a roadside restaurant, you meet another person who is also a salesperson. She/he works in the same district and also makes the trip between Brest and Quimper very often. You are coming to discuss the two possible routes and she/he is telling you: ‘‘I think that the route that passes through Guersac is much faster than the one that passes through Landrieu.’’ In the future which route will you choose to make the trip between Brest and Quimper as fast as possible? The one that passes through: Landrieu or Guersac? Participants of both populations received the same scenarios apart from the fact that proper nouns had a Japanese or a French connotation.

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 495)

496 1

Jean-Baptiste van der Henst et al.

Table 1. Percentages of answers for Japanese and French Participants in Experiment 1

2 3

Japanese group N ¼ 93

French group N ¼ 82

71 29

72 28

4 5 6

% of answers compatible with one’s own perspective % of answers compatible with the other’s perspective

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

2.3.

Results and discussion

Table 1 shows the percentages of answers for the two groups of participants. It clearly indicates that participants favored their own perspective over the perspective of the other person. Seventy-two percent of the French participants gave answers that were in line with the initially provided view. This result significantly di¤ers from a pattern of answers attributing the same weight to the two views (Binomial test, p < 104 , two-tailed). The Japanese participants showed a similar e¤ect and did not follow a middle way approach in dealing with contradictions: 71% provided answers in accordance with their own perspective, a rate that was significantly higher than 50% (Binomial test, p < 104 , two-tailed). A comparison of Japanese versus French participants indicates that both groups showed a similar pattern of answers and did not significantly di¤er (w 2 ð1Þ ¼ :02, n.s.). Finally, the di¤erence between female and male participants was not significant either: in the Japanese group, 68.9% of the female participants and 72.9% of the male participants gave an answer in line with their own perspective (w 2 ð1Þ ¼ :02, n.s.). In the French group, 73.1% of the female participants and 70% of the male participants gave an answer in line with their own perspective (w 2 ð1Þ ¼ :09, n.s.). In Experiment 1, the results obtained with the French participants are consistent with findings obtained in earlier studies: participants did show a bias towards their own position and the data indicate similar values to those obtained by Yaniv (2004; Yaniv & Kleinberg 2000) and Mercier and Van der Henst (2005). Interestingly, Japanese participants showed the same e¤ect and thus did not take the other person’s view into account any more than the French participants. This experiment challenges the culturalist prediction stating a greater tendency for Japanese to follow a middle way between conflicting positions. It corroborates the evolutionary prediction, according to which the reported bias results from adaptive and universal mechanisms designed to prevent oneself against potentially misleading information. One could, however, object that the methodology used in Experiment 1 did not facilitate the production of ‘‘middle-way’’ answers. If Japanese participants had followed a strict middle-way approach, their answers

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 496)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10

497

would have been equally distributed between the two opposite positions. However, with the forced-choice procedure, participants had to make an all-or-nothing judgment and were not given the opportunity to provide fine-grained judgments about each position. This may have exaggerated the tendency to value one’s own position, since there was no way to express the view that both positions can carry some truth. One might wonder if Japanese participants would show a greater tendency to value the opposite position if they had the opportunity to make a more fine-grained judgment about it. Consequently, in Experiment 2, we o¤ered participants the possibility of providing answers on a scale.

11 12 13

3. Experiment 2

14 15 16 17 18 19

3.1.

Participants

Eighty-nine Japanese business students from the University of Kinki and 82 French business students from the University of Lyon 3 took part in this experiment.

20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

3.2.

Method

Experiment 2 used the same scenarios as Experiment 1, but participants did not have to choose between one of two opposite opinions. Instead, they had to give weight to each opinion on a 10-point scale (from 0 to 9). Each opinion was expressed by a statement that participants had to evaluate on the scale after reading the scenario. For instance, the two opinions and the two scales of the ‘‘salesperson scenario’’ presented above were the following: The route that passes through Landrieu is the fastest Never Always 0 1 2 3 4 5 6 7 8 9 The route that passes through Guersac is the fastest Never Always 0 1 2 3 4 5 6 7 8 9 As in Experiment 1, participants had to address only one scenario.

38 39 40 41 42

3.3.

Results and discussion

Participants were classified into three categories. First, there were participants favoring their own perspective: these gave a higher value to their

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 497)

498 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Jean-Baptiste van der Henst et al.

initial position than to the position of the other person. The second category consisted of participants assigning the same value to the two opinions. These participants can be considered to follow a strict middle-way approach since they judged both opinions to be equally true. Finally, participants of the third category were those favoring the position of the other person: they assigned a lower value to their own position than to that of the other person. Table 2 presents the percentages of participants in these three categories. As in Experiment 1, there were more participants favoring their own perspective than participants favoring the perspective of the other person. This e¤ect was significant in the French group (Binomial test, p < :005, two-tailed), as well as in the Japanese group (Binomial test, p < :103 , two-tailed). If we discard the participants who assigned the same value to the two opinions, we can see that the data are remarkably consistent with those of Experiment 1: 73.2% of the Japanese participants who provided a polarized answer favored their own perspective compared to 71.2% of the French group. The distribution of the three categories of answers did not significantly di¤er across cultures (w 2 ð2Þ ¼ 1:64, n.s.). Let us note, however, that there were more participants following a strict middle-way approach in the Japanese group than in the French group, but this di¤erence was not statistically significant (37.1% vs. 28%, w 2 ð1Þ ¼ 1:58. n.s.). We now turn to the analysis of the rating scales. Table 3 presents the mean values that participants gave to the two conflicting statements. It appears that the results are highly similar for the two groups. First, both

25 26 27

Table 2. Percentages of Japanese and French participants for the three categories of answer

28 29 30 31 32

Participants favouring their own perspective Participants favouring neither one or the other perspective Participants favouring the other’s perspective

Japanese participants

French participants

46,1 37,1 16,8

51,2 28,1 20,7

33 34 35 36

Table 3. Means (and standard deviations) of ratings of the two statements for Japanese and French participants

37 38

Japanese participants

French participants

5,73 (2,20) 4,53 (2,32) 1,20 (3,41)

5,79 (1,96) 4,55 (2,08) 1,24 (3,44)

39 40 41 42

Statements in line with the participant’s perspective Statements in line with the opponent’s perspective Di¤erence between the two types of statements

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 498)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

499

Japanese and French participants assigned a higher weight to their own view than to the conflicting view (for Japanese participants, t test ð88Þ ¼ 3:5, p < :001; for French participants, t test ð81Þ ¼ 3:3, p < :005). Second, the values given to the two statements were very similar for the two groups and the mean di¤erence between the two groups did not di¤er significantly (t test ¼ 0:081, n.s.). Finally, if participants were sensitive to the additivity property of the two scales, their rating should add up to 9 (i.e., if they answered 9 on one scale, they should answer 0 on the other). The results indicated that they were not, since the mean value of the two ratings provided by participants was significantly higher than 9 (mean ¼ 10,3, standard deviation ¼ 2:72. z ¼ 6,26, p < :001). No di¤erence was observed between the two groups (for the Japanese group, mean ¼ 10:34, standard deviation ¼ 2:1, for the French group, mean ¼ 10:26, standard deviation ¼ 3:16, t test ð169Þ ¼ :19. n.s.). By showing a clear tendency to favor one’s own perspective in both groups, Experiment 2 replicated the results of Experiment 1. Although we used a methodology that o¤ered participants the opportunity to provide more fine-grained answers than in Experiment 1, Japanese participants did not show a significantly greater tendency to follow a middleway approach: the rate of participants assigning the same value to both conflicting parties was not significantly higher in the Japanese group than in the French group (although it was in the direction predicted by the culturalist approach) and the mean di¤erence between the two opinions was not significantly less important in the Japanese group than in the French group. These results again confirm the prediction of the evolutionary approach, that is, that both Japanese and French participants would exhibit the same bias towards their own views. Although there was no di¤erence between the two groups, Experiment 2 revealed that participants did not entirely reject the opposite viewpoint and did not assign very high values to their own view. Indeed, about a third of the participants did assign the same value to the two statements, and the mean di¤erence between them was not very large (i.e., slightly more than one point on the scale). This indicates that participants did not apply an all-or-nothing approach and considered both opinions to be capable of containing some valuable information. This may of course result from the experimental procedure. Participants did not have to defend a belief they personally worked out, but were committed to a fictitious position and had to pretend that they were advocating it. Additionally, the interpersonal conflict was also purely fictitious. The aspect of pretense involved in our experiments (and in many other experiments in psychology) can weaken the polarization towards one’s own position. We may assume that if participants had experienced a genuine conflict in which

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 499)

500 1 2 3 4

Jean-Baptiste van der Henst et al.

their actual beliefs were challenged they would have been more committed to them. However, our experimental procedure was successful in eliciting a bias towards the participants’ own perspectives and, more importantly, in showing that both groups tend to react in the same way.

5 6 7

4.

Conclusion

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

In this paper, we investigated the impact of two factors on the way people address conflicting information. One relates to culture and the other to the communicative context in which contradiction occurs. We compared two theoretical approaches, one that focuses on the culturalist approach and one that focuses on the evolutionary approach. According to the culturalist approach advocated by Nisbett et al., the way we deal with contradiction can be markedly a¤ected by culture, so that people from cultural environments with di¤erent social practices are more or less inclined to accept contradictions. In particular, this approach predicts that Easterners are more likely to search for a compromise between two conflicting viewpoints than Westerners, who tend to follow a logical principle of non-contradiction. In contrast, the evolutionary approach considers that when contradiction occurs in a communicative context, universal mechanisms designed to deal with the adaptive problem of managing deceptive information operate. This would lead to the universal tendency of giving more weight to one’s own belief than to the other’s conflicting view. We tested these two approaches with Japanese and French participants and our data supports the evolutionary approach. Both groups showed the same bias of favoring one’s own position when challenged by another person’s position. The results obtained in the two experiments reported here are similar to those previously observed with populations from the Western world but the absence of di¤erence between Japanese and French participants depart from the findings obtained by Peng and Nisbett (1999). One could account for this di¤erence by the fact that in Peng and Nisbett’s study, the cross-cultural comparison included Chinese and American students. One could thus argue that Japanese and French manage contradiction in a di¤erent way than did Chinese and Americans. This would suppose that the Japanese are less inclined to adopt the ‘‘Eastern’’ holistic view or that the French are less prone to confrontation and debate than their counterparts in the Western world. We think that these two possibilities are unlikely. First, because studies with Japanese participants have shown that they are more prone to attend to contextual information and to relations among objects than Americans (Masuda & Nisbett 2001; Miyamoto,

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 500)

Dealing with contradiction in a communicative context 1 2 3 4 5 6 7 8 9 10 11

501

Nisbett & Masuda 2006); such a result is a strong cue in favor of the thesis that Japanese participants do have a holistic way of thinking. Second, given the feature of popular protest that characterizes social and political practices in France (Puyo & Van Eersel 1978), it is di‰cult to believe that French people are adverse to debate. In conclusion, we do not think that the absence of di¤erence between Japanese and French participants is due to specific characteristics of these two populations but rather that it is a consequence of universal mechanisms designed to face the adaptive problem involved in our experiments. In contrast, Peng and Nisbett’s studies did not include any situation raising the adaptive problem posed by communication.

12 13 14

References

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Chokron Sylvie, Jean-Marc Bernard, and Michel Imbert. 1997. Length representation in normal and neglect subjects with opposite reading habits studied through a line extension task. Cortex 33: 47–64. Cosmides, Leda, and John Tooby. 1992. Cognitive adaptations for social exchange. In Jerome H. Barkow, Leda Cosmides, and John Tooby (eds.), The Adapted Mind. Oxford: Oxford University Press. 19–136. Dawkins, Richard, and John Krebs. 1978. Animal signals: Information or manipulation? In John R. Krebs and Nicholas Davies (eds.), Behavioural Ecology. Oxford: Basil Blackwell. 282–309. Jewell, George, and Mark E. McCourt. 2000. Pseudoneglect: A review and meta-analysis of performance factors in line bisection tasks. Neuropsychologia 38: 93–110. Levinson, Stephen C. 1997. Language and cognition: The cognitive consequences of spatial description in Guugu Yimithirr. Journal of Linguistic Anthropology 7: 98–131. Masuda, Takahiko, and Richard E. Nisbett. 2001. Attending holistically versus analytically: Comparing the context sensitivity of Japanese and Americans. Journal of Personality and Social Psychology 81: 992–934. Mercier, Hugo, and Jean-Baptiste van der Henst. 2005. The source of beliefs in conflicting and non-conflicting situations. In Bruno Bara, Lawrence W. Barsalou, and Monica Bucciarelli (eds.), Proceedings of the 27th Cognitive Science Conference. Mahwah, NJ: Lawrence Erlbaum. Miller, Kevin F., Catherine M. Smith, Jianjun Zhu, and Houcan Zhang. 1995. Preschool origins of cross-national di¤erences in mathematical competence: The role of number naming systems. Psychological Science 6: 56–60. Miyamoto, Yuri, Richard E. Nisbett, and Takahiko Masuda. 2006. Culture and the physical environment: Holistic versus analytic perceptual a¤ordances. Psychological Science 17: 113–119. Nachshon, Israel, Gabriel E. Shefler, and Dalia Samocha. 1977. Directional scanning as a function of stimulus characteristics, reading habits, and directional set. Journal of Cross Cultural Psychology 8: 83–99. Nisbett, Richard E., Kaiping Peng, Incheol Choi I, and Ara Norenzayan. 2001. Culture and systems of thought: Holistic vs. analytic cognition. Psychological Review 108: 291– 310.

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 501)

502 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Jean-Baptiste van der Henst et al.

Padakannaya Prakash, M. L. Devi, B. Zaveria, S. K. Chengappa, and Jyotsna Vaid. 2002. Directional scanning e¤ect and strength of reading habit in picture naming and recall. Brain and Cognition 48: 484–490. Peng, Kaiping, and Richard E. Nisbett. 1999. Culture, dialecticism, and reasoning about contradiction. American Psychologist 54: 741–754. Puyo, Jean, and Patrick van Eersel. Sacre´s Franc¸ais, ou Les nouveaux cahiers de dole´ances. Paris: Stock. Roberson, Debi, Ian Davies, and Jules Davido¤. 2000. Color categories are not universal: Replications and new evidence from a stone-age culture. Journal of Experimental Psychology: General 129: 369–398. Singh, Maharaj, Jyotsna Vaid, and Tripti Sakhuja. 2000. Reading/writing v.s. handedness influences on line length estimation. Brain and Cognition 43: 398–402. Sperber, Dan. 2001. An evolutionary perspective on testimony and argumentation. Philosophical Topics 29: 401–413. Vaid, Jyotsna, Maharaj Singh, Tripti Sakhuja, and G. C. Gupta. 2002. Stroke direction asymmetry in figure drawing: Influence of handedness and reading/writing habits. Brain and Cognition 48: 597–602. Whorf, Benjamin L. 1956. Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf, ed., John B. Carroll. Cambridge, MA: MIT Press. Yaniv, Ilan. 2004. Receiving other people’s advice: Influence and benefit. Organizational Behavior and Human Decision Processes 93: 1–13. Yaniv, Ilan, and Eli Kleinberger E. 2000. Advice taking in decision making: Egocentric discounting and reputation formation. Organizational Behavior and Human Decision Processes 83: 260–281.

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

(AutoPDF V7 13/10/06 10:53) WDG (148
225mm) TimesM J-1632 Inter. Prag., 3:4 PMU:I(CKN[W])13/10/2006 pp. 487–502 1632_3-4_06 (p. 502)