Bertram F. Malle, How the Mind Explains Behavior: Folk Explanations, Meaning, and Social Interaction (Cambridge, MA & London: MIT Press, 2004, $38.00, 328 pages, ISBN 0-262-13445-4)—Craig Hamilton, University of California at Irvine

 

How do we explain the behavior of others? To reflect on someone’s behavior is to enter the world of social cognition, since behavior explanations are a tool for making sense of how or why people behave as they do. This is just one of the many points that Bertram Malle clearly and confidently makes in his fascinating new book, How the Mind Explains Behavior: Folk Explanations, Meaning, and Social Interaction. Even if we are mistaken in our explanations of how others behave, the fact that we search for answers reveals a persistent need to understand our social world. Malle, a psychologist at the University of Oregon’s Institute for Cognitive and Decision Sciences, has certainly made a worthwhile contribution to his field here. Although this book might have been called How People Explain Behavior (given the fact that the explanations Malle often refers to are actual—rather than invented—ones offered by his subjects), by placing “mind” in the title Malle highlights the cognitive and conceptual regularities that support behavior explanations across a range of settings. Because life is short and there is never enough time in a day to read all the books one would like to, the conscientious reviewer can do no more than explain this book’s main ideas and hope that readers will want to learn more by reading the book itself!

Malle begins with a simple yet bold confession in his preface: “The ideas that found expression here were first inspired by Jerry Bruner’s magnificent book Acts of Meaning and a seminar held by Fred Dretske and Michael Bratman at Stanford University on action explanation. Thereafter, nothing in the attribution literature meant quite the same for me again” [vii]. What are the rhetorical effect of these statements? To establish Malle’s ethos as a credible scholar who keeps company with important thinkers in elite universities, and to imply that we are reading a conversion narrative where one way of training in psychology is abandoned for another. In Chapter 1, Malle goes on to suggest that he is only picking up where other eminent psychologists have left off, which is another rhetorical strategy that may seem superficially self-effacing but which nevertheless succeeds in preparing readers for the innovative discussion that will follow. Malle begins with a review of the literature, discusses previous research on behavior explanations and attribution theory (especially Heider’s work), and along the way points out two distinctions that have compelled him to write this book. First, there is the person-situation distinction, which made earlier psychologists attribute behavior either to (internal) causes located within a person or to (external) causes found in situations. But for Malle, when “two explanation types—reason explanations (motive attribution) and enabling factor explanations [outcome attributions]—answer such different questions, it is unfortunate that the attribution literature after Heider collapsed them into one” [11-12]. Malle’s initial aim thus seems to be to keep alive distinctions between motive and outcome attributions. Second, to erase the distinction between unintentional and intentional behavior, and at the same time propose different theories for how people explain two types of behavior, is to create confusion where clarity is needed according to Malle.

Explaining behavior appears to start with an act of categorization: is an action unintentional and caused impersonally (e.g. sneezing), or intentional and caused personally (e.g. calling a friend) [8]? Malle, however, maintains that strict adherence to a person-or-situation causal model oversimplifies attribution theory and “does not tell the whole story of behavior explanations” [30]. Rather than simply resuscitate the reputation of attribution theory for the sake of correcting the errors of textbooks, Malle tells us at the end of Chapter 1 that he wants to do more than this. Specifically, Malle would like to examine “a conceptual framework called the folk theory of mind and behavior” [26; italics in original], a framework that applies to intentional and unintentional behavior. Malle devotes Chapter 2 to the folk theory of mind and behavior, which is the “conceptual framework that guide’s people’s cognition of behavior and the mind” and which he sometimes refers to as “theory of mind” [31; italics in original]. Agents, intentions, reasons, beliefs, and desires are key concepts in this framework, and Malle explains how the theory of mind relates to mental problems like autism and simulation theory (often assumed to be the rival hypothesis to the theory of mind). He also discusses the concept of projection (a building block of simulation theory), recognizing that it usually means “other = self” [34]. This equation would describe why cynics believe nobody: they project their self-image onto others and because they skeptically distrust themselves they then skeptically distrust whatever else they see. Malle’s argument, however, is that those who promote either simulation theory or the theory of mind theory (often called “theory-theory,” for short) overlook the fact that “what the two have in common is that they focus on the psychological mechanisms of mental state ascription more so than on the conceptual framework that underpins it [and is] typically presupposed” [35]. Malle has a talent for uncovering these presuppositions, which is why both sides should listen to him.

The followers of cognitive science will also recognize Malle’s section in Chapter 2 on the evolution of the theory of mind. While discussions of this type may border on speculation, after Darwin it seems that all scientists must be naturalists. By that I mean many scientists may have come to assume that to explain why something exists or how it functions entails explaining its evolution (i.e. where it comes from and how it got that way) even when the entity is a form of human behavior. Malle follows a path here familiar to many cognitive scientists in the age of evolutionary psychology, and they probably will not disagree with Malle’s view of the “five candidate precursors” to the theory of mind in the cognitive evolution of homo sapiens [51]. Those candidates were [i] “the capacity for imitation” and a “noninferential form of empathy;” [ii] “a primitive form of introspection (far from full-blown self-consciousness)”; [iii] understanding another’s “directedness to an object”; [iv] the “ability to appropriately parse the behavior stream into intention-relevant units”; and [v] joint attention, or realizing “that self and another person are both directed at the same object” [52-53]. To his credit, Malle ends the chapter by relating what he has said about the theory of mind to explanations of behavior. He does so by noting that models of behavior explanation that confuse so-called mechanical causation with intentional causation obscure the fact that our “folk concept of intentionality” [61] is rooted in these conceptualizations of causation. Even though it is only in Chapter 4 that Malle presents us with a line drawing to help us visualize the folk concept of intentionality [89], his point here is that our conceptualizations of causation enable our explanations of behavior.

In Chapter 3, Malle turns his attention to why and when people explain behavior. Malle follows Bruner in seeing life as a quest for meaning, and he correctly refers to our drive for seeking clarifications in the face of cognitive dissonance (e.g. unmet expectations in what Bruner once called “canonical” situations). But regarding the “human tendency to find significance” [67], Malle goes one step further. He argues that how we explain and find meaningful the behavior of other people is also how we explain and find meaningful most things in this world. “Finding meaning” is vital to Malle’s model for it defines one of only two motivations for explanations of behavior (managing social interactions is the other motivation). In communication, for example, subjects voluntarily offered behavior explanations to their interlocutors without being asked to do so explicitly in 97% of the cases studied. That means explanations were offered in response to wh-questions in only 3% of cases (i.e. 15 of 451), a finding that reveals how producing explanations involves perceptions of an interlocutor’s implicit anticipations. The pragmatic and social nature of behavior explanations is addressed later in the book although Malle writes in detail in Chapter 3 about four event types: [i] intentional/observable actions; [ii] unintentional/observable behaviors; [iii] intentional/observable thoughts; and [iv] unintentional/unobservable experiences. Data from interpersonal transactions suggest that actions and experiences occur as themes in conversations more frequently than is the case with either behaviors or thoughts. The fact that we talk more often about actions and experiences, than we do about behaviors or thoughts, suggests to Malle that we spend more time explaining certain “behavioral events” at the expense of others. Malle is even more correct here than he realizes. Linguists analyzing conversations in CANCODE (the Cambridge and Nottingham Corpus of Discourse in English) could show Malle that one of our favorite topics of conversation is other people’s business, including why they behave how they behave (especially regarding the speaker).

For Chapter 4 and Chapter 5, Malle puts forward his theory of behavior explanations by dividing it into two major parts: conceptual structure and psychological construction. The elegance of Malle’s model is in the details. He distinguishes intentional from unintentional behavior, and carefully demonstrates that there are three modes we use to explain intentional behavior: reasons, causal history of reasons explanations (a cumbersome term Malle often shortens to CHRs), and enabling factor explanations [91]. For example, if the question “Why did you go running?” is answered “Um, because I wanted to get in better shape, and [...] I figured that I can do that by going running every day” [93], then a reason is used to explain an intentional behavior. Malle’s brilliant insight is that the difference between reasons and CHRs “has nothing to do with the classic person-situation dichotomy” [103]. For instance, if the question “Why was Nina using drugs?” is answered “She was at a party,” then we have evidence of causal factors outside the agent being evoked to answer the question. The behavior may be intentional, but the explanation of the behavior is not a reason; rather, it is a CHR. This is not to say that there is only one way to explain intentional behavior. On the contrary, if we read that “By choice, Ian worked 14 hours a day last month,” and we say in return that he did so either “To make more money” (reason explanation) or because “He is driven to achieve” (CHR), then we have offered two different kinds of explanation for the very same intentional behavior [104]. While Malle jokes about the chicken crossing the road, he argues at length that enabling factor explanations are commonly used to explain difficult actions. If reasons or CHRs are insufficient for us to explain why a student got a perfect score on a math exam, then we may offer “He’s a stats whiz” as an enabling factor explanation to account for the student’s behavior in a more satisfactory manner [110]. Although Malle has fewer things to say about unintentional behavior in Chapter 4, his summary here persuasively defends his folk-conceptual theory of behavior explanations by arguing that it accounts for data in a way that the standard tools of attribution theory cannot. I suspect this is the point that psychologists will most appreciate it if they have their doubts about attribution theory.

When it comes to psychological construction, Malle quickly introduces a shopping analogy in order to demonstrate, thanks to a flow chart [119], that behavior explanations involve many choices in much the same way that a shopper at supermarket might also make choices. Here Malle discusses specific causes and factors in reasons, CHRs, and enabling factor explanations in order to clarify in three ways of explaining behavior rather than the standard two (internal reasons versus external enabling factors). What Malle is uncovering for us are the cognitive processes or “principles that guide the psychological construction of explanations” [145], processes that include knowledge structures, simulation or projection, covariation, direct recall, and rationalization [143].

These processes are examined in depth again in Chapter 6 when Malle studies the pragmatics of explanations. The social nature of behavior explanations is clear when it comes to explaining socially desirable behavior versus socially undesirable behavior. For instance, based on experimental results such as “I slapped my mom. Why? I was really angry and frustrated” [160], Malle found that people use CHRs like these 42% of the time when explaining undesirable actions, compared to using CHRs 22% of the time to explain socially desirable actions [161]. This is why a CHR is seen as a “mitigating device” [163] used extensively during “impression management,” which is one way to define the behavior explanations that actors and observers offer to others. For example, in the case of Syria's Interior Minister Ghazi Kanaan, who allegedly committed suicide on 12 October 2005, I suspect Malle’s findings would lead us to predict that CHRs will be used extensively to explain a socially undesirable action like suicide. Clearly, Malle’s findings have an incredibly wide range of application. For their part, sociolinguists might also take an interest in Chapter 6 in Malle’s study of interpersonal verbs and discourse analysis based on the fact that mental state markers really matter to the pragmatics of explaining behavior. For instance, suppose “Cliff asks Jerry: ‘Why did your girlfriend refuse dessert?’ Jerry responds by saying: ‘She thinks she’s been gaining weight’ vs. ‘She’s been gaining weight’” [168]. The verb “thinks” in the first reply marks a mental state while in the second reply the mental state is unmarked, a difference with an immediately visible social impact. Malle later adds that whereas actors often begin with unmarked explanations, observers often begin with marked explanations [183], a finding that should interest narratologists who study thought reports in literary texts.

Malle turns his attention to explaining the behavior of self and others in Chapter 7, and the behavior of individual and groups in Chapter 8. Just as we learned in previous chapters that there were differences in how we explain behavior that is good or bad, intentional or unintentional, in Chapter 7 Malle argues that people as “observers are biased toward assuming intentionality for harmful behaviors” [177]. In other words, when a bad thing happens we usually explain it by assuming someone intentionally caused it to occur. “Accidents don’t just happen,” as a popular public service announcement for the reduction of car accidents in America used to put it, and our intentionality bias compels us to accept this statement as true. But Malle defends our bias as follows: “Consider the enormous costs of falsely assuming that an explosion was an industrial accident when in fact it was a deliberate terrorist attack” [177]. While the intentionality bias might explain the psychological roots of conspiracy theories, a desire to avoid the costs of making the wrong assumption was apparent, for example, among French citizens who refused to believe that the AFZ factory explosion in Toulouse on 21 September 2001 was an accident. But for Malle the bias is only one part of the picture; the other part involves the “actor-observer asymmetry.” An analysis of 4000 explanations shows that actors offer more reasons than observers do to explain intentional behaviors, while observers offer more CHRs than actors do to explain intentional behaviors [178]. This asymmetry, however, should not obscure the fact that actors and observers still have something in common: they offer more reasons than CHRs for explaining intentional behavior. Ideas like these help Malle argue that psychologists can no longer use one distinction (e.g. the person-situation dichotomy) to account accurately for another (e.g. the actor-observer asymmetry) when the two distinctions are not related. If Malle is consistent in arguing that attribution theory has its limits, he is equally consistent in arguing that the person-dichotomy also has its explanatory limits.

After discussing self and other, in Chapter 8 Malle moves onto individuals and groups in order to test his “working hypothesis”: “People use their folk theory of mind to make sense of groups just as they use this folk theory to make sense of individuals” [194]. One of the problems here, of course, is that “people interact less with groups than they talk about them,” and this is especially true of aggregate groups [208]. Whereas aggregate groups include individuals who act in similar ways without coordinating their actions (e.g. “High school seniors in the US vandalize school property”), jointly acting groups may function more coherently (e.g. “Seniors at Irvine High School vandalized school property this year”). Just as we distinguish intentional from unintentional behavior, so too do we distinguish aggregate group behavior from jointly acting group behavior. We do so by offering more CHRs (about 45%) to explain intentional aggregate group actions than for explaining individual (about 30%) or jointly acting group actions (about 20%) [206-208]. Despite these differences, however, Malle’s point is that we analogically reason about groups in much the same way we reason about individuals: we ascribe intentionality and agency to groups just as we ascribe intentionality and agency to individuals [194-197]. In other words, just as we attribute mental states to separate people, so too do we attribute them to groups in the belief that groups have minds too. Political theorists familiar with the work of Charles Beitz will recognize the implications of Malle’s findings for the analysis of international relations, especially the Hobbesian question of whether or not the state of nature that applies for individuals can also apply (analogically) to groups like nation-states. Finally, although Malle admits the data from psychological experiments regarding stereotypes are still inconclusive, he nevertheless ends the chapter by mentioning the fact that research on stereotypes in the future might produce findings that can be useful to psychologists interested in explanations of group behavior and the generalizations we make about stereotyped groups.

For his conclusion, in Chapter 9, Malle reminds us that behavior explanations are cognitive tools that provide meaning and social tools that manage interactions. Malle also reminds us that his theory contains three levels: the conceptual, the psychological or cognitive, and the linguistic. These levels complicate the views offered by the dichotomies of attribution theory, but Malle argues it is necessary to change’s one view when confronted with data that attribution theory simply cannot account for in any satisfactory manner. However, as Malle realistically and honestly concedes, “Revolutions rarely happen in social psychology” [225], so he is aware of just how difficult it will be for the folk-conceptual theory of behavior explanation to supplant the more popular and better established attribution theory of behavior explanation. Historians of science, of course, are familiar with such moments in time where competing scientific theories result in incommensurability and practitioners find themselves forced with a choice: to stick with the old way of doing things or to adopt an altogether new approach to their object of study. For his part, Malle thinks that social psychologists can no longer ignore current developments in cognitive science, especially when those developments have major implications for theories of behavior explanations. Only time will tell if cognitive science is indeed the future of social psychology.