Thinking, Fast and Slow

The first chapter begins with an image of an angry woman’s face, which Kahneman uses to explain System 1. He notes that the reader instantly knows, without effort or choice, that the woman is angry and can make several other intuitive predictions. In contrast, the chapter’s second page contains a multiplication problem that most people could not solve quickly in their heads (17 x 24), which Kahneman uses to illustrate System 2 insofar as it would take some calculation to arrive at a correct answer. By noting several numbers the reader is likely to intuitively know are not the solution to the equation, Kahneman illustrates the rough boundaries between System 1 and System 2.

Kahneman summarizes System 1 as operating automatically, quickly, and without a sense of voluntary effort. He explains that System 2 “allocates attention to the effortful mental activities that demand it,” which makes its operations “often associated with the subjective expression of agency, choice, and concentration” (21).

Unlike other works using the terms “system 1” and “system 2,” this book develops them into full-fledged characters. The latter is identified with the conscious self that makes intentional choices, but System 1 generates the impressions and feelings that prompt and underlie System 2’s operations in a manner that seems effortless. Accordingly, System 1 is responsible for automatic assessments like recognizing that one object is closer than another and understanding a simple sentence.

There are two ways System 1 provides the information that seemingly just appears in the conscious mind. One method depends upon innate skills that all humans and some other animals share, while the other is developed through learned associations over time.

The two systems operate in tandem when an event processed by System 1 automatically activates System 2, such as when a loud noise catches your attention and you immediately wonder about its source.

System 2 contains a multitude of diverse processes, each of which requires attention and can be disrupted if attention is drawn elsewhere. One implication of this feature relates to the exhaustibility of attention. That is, System 2 operations can only occur when the needed attention is available from the limited supply a person has on hand. The attention devoted to intense focus on a task, for example, can effectively prevent a person from giving any attention to other events that would otherwise attract it.

Having introduced his characters, Kahneman then previews the types of interactions between them. It is, on the whole, a very efficient division of labor. System 1 quickly produces correct understandings in familiar situations. However, System 1 has biases and cannot be turned off. Much of the book discusses these problems and the need for effortful System 2 intervention if one wants to correct them.

The conflict between System 1’s intuitive response and System 2’s deliberative approach illustrates one of the latter’s main tasks in our lives: System 2 conducts intentional control over the intuitive reaction that would otherwise flow from System 1. At times, however, we experience System 1’s response very strongly, and we tend to believe it even when System 2 can (and perhaps does) prove it false. Visual illusions are one example of this phenomenon. Cognitive illusions, such as when someone is misled to feel sympathy by a liar, can be much more difficult to overcome.

Finally, Kahneman reemphasizes that System 1 and System 2 are fictional characters. He discusses them as if they are little people inside of someone’s head—but of course they are not. Instead, these “characters” approximate what is known about processes of the mind that appear to represent two different approaches within one person, which have the general properties that Kahneman assigns to his characters.

Kahneman observes that System 2 would be “a supporting character who believes himself to be the hero” if the book were made into a film (32). This arises from the fact that System 2’s operations require effort, but System 2 is lazy in that it is reluctant to expend effort. The upshot of this combination is that System 2 often believes that its efforts are the main event, but it does not realize that those operations are very often (but not always) guided by System 1.

Drawing on prior research that noticed effortful thinking seemed to cause dilation of the pupils, Kahneman designed an experiment with a graduate student that led him to a key insight about how the mind works. Observing the pupils of subjects during various tasks, and especially the effects of adding additional effortful tasks, he observed that subjects generally operated at a relaxed mental pace, but they would devote substantial effort in bursts as necessary to complete tasks. Further, he noted that subjects were far less effective at a given task if their mental effort was already at its peak—bursts of effort were (and are) dependent on the availability of effort to expend.

Electrical circuits in one’s home, measured by the electricity meter, provide an apt analogy for these findings regarding System 2’s capacity to exert effort (as measured by pupil dilation). Kahneman maintains that “[t]he analogy goes deep” before explaining several key processes through it (34). Both the meter and the mind have limited capacity. Rather than breaking a circuit that causes all devices to lose power, however, System 2 protects the most important task demanding attention. This reflects a long evolutionary development that promotes survival.

A difficult task will often come to demand less effort as one develops the relevant skills over time. Further, different tasks require different levels of effort for other reasons, which Kahneman concisely explains. For example, System 2 is able to use “task sets” (36). This relates to the finding that effort is required not only to complete tasks but also to switch between types of tasks. Thus, similar tasks can be grouped, thereby reducing the effort required overall. Another determinate is time pressure. Its addition increases effort required.

Chapter 2 concludes with the observation that humans frequently live their lives by the law of least (cognitive) effort. In general, most of us do not expend mental effort unless it seems necessary or important.

System 2 has a background activity level that is apparent in the aimless thoughts and observations people make on a daily basis. This does not drain its capacity for effort. Thus, Kahneman observes, we can normally take a walk and think at the same time, and the experience is often pleasurable. If, however, we add time pressure, Kahneman suggests that the increased effort becomes a drain, spoiling the pleasure and reducing the efficacy of thinking.

In a rare substantive statement that is not directly backed by research, Kahneman suggests that (based on his observations over a long career) a “law of least effort” applies to most people most of the time with regard to the cognitive effort needed to complete tasks. Therefore, multitasking and the increased pace of activity brought about by technological developments are not intrinsically pleasurable for most people. Quite the opposite.

Yet Kahneman highlights Mihaly Csikszentmihalyi’s work concerning “flow.” This is a state of high effort over long time periods that does not require willpower. In fact, the flow state is often described as joyful, and Csikszentmihalyi refers to it as an “optimal experience” (40). Kahneman easily incorporates flow states into his explanation by pointing out that there is a meaningful difference between effort expended toward concentrating on a task (which can include flow) and that devoted toward focusing one’s attention (the need for which a flow state eliminates). Both forms of effort are relevant, partly because both consume the store of mental energy available.

Kahneman then discusses the work on “ego depletion” that shows people tend to make more errors, be more selfish, and otherwise exhibit less self-control when they are drained of energy in one way or another (physical, emotional, or cognitive).

System 2 serves as a supervisor over whether a person acts on the things suggested by System 1, such as desires. Kahneman discusses a series of mini-tests used in experiments that revealed a very high percentage of intelligent people gave the wrong answers where an intuitive (but obviously wrong, on a moment’s reflection) answer was suggested by the puzzle itself. This reflects a tendency to be mentally “lazy” when there is insufficient motivation.

In concluding the chapter, Kahneman writes that traditional “intelligence” is not the same as the ability to focus attention or maintain willpower to complete tasks well. He cites work by Keith Stanovich and Richard West—who were the first to use the terms “system 1” and “system 2”—that shows the distinction to be real and to have significant effects on individual performance.

Kahneman turns our attention to System 1, which operates primarily through association. To illustrate, the chapter starts with two words that are not necessarily or obviously connected (“bananas” and “vomit”). Then, Kahneman explains the large number of possibilities evaluated and ranked by System 1 to create a causal chain—a story, essentially—regarding the words. All this occurs within approximately one second. Unlike System 2’s processes, this type of “thinking fast” is effortless and is experienced as a sudden self-reinforcing pattern of cognitive, emotional, and physical responses that are both varied and integrated. In other words, the automatic responses generated by System 1 are “associatively coherent” (51).

In addition to the apparently effortless associations brought to mind by System 1 from the mere presence of two words, the reader will also experience a physiological response. That is, the presentation of the ideas represented by the words are almost instantly processed by association, and then the body reacts as if the things represented by those ideas were physically present in reality.

Kahneman then explains “priming effects” (52) to expand our understanding of association. A priming effect is the effect that a prior idea has on the next idea introduced in terms of the associations generated. This means that the order in which ideas are presented to an individual can greatly affect how those ideas are processed by that person.

Psychological research has shown that priming effects can occur due to events of which the conscious mind is unaware. Kahneman illustrates this by explaining that researchers have primed subjects with words associated with old age—but never the word “old” or its close synonyms—which caused those subjects to walk more slowly (apparently reflecting knowledge that old age causes people to walk more slowly) shortly thereafter. The subjects were unaware of the priming and that their gait speed changed in response to it. The effect is self-reinforcing, such that being primed for a concept like “old age” causes subjects to act more like they are old, which then makes them more likely to pick up on subtle priming related to old age.

Kahneman acknowledges that studies of priming effects may threaten our self-image as autonomous agents. For example, research has shown that, at least in the election studied, support for a measure to increase school funding was significantly greater in districts where the polling station was a school, even when other considerations were accounted for.

Priming for the idea of money produces an increase in independence, meaning that subjects are less concerned about others and less willing to sacrifice for or cooperate with others. Kahneman notes the potentially troubling impacts of persistent cultural reminders of money. He also notes other priming effects of potential concern. For example, priming for mortality appears to increase people’s support for authoritarian rule.

Kahneman focuses on priming because the conclusions of this research must be accepted despite the lack of any subjective experience that confirms them. Kahneman points out that our subjective sense of who we are is largely a product of System 2. Priming effects, which have a demonstrable effect on behavior whether or not we believe they affect us personally, relate to System 1.

In daily life, we are consistently assessing the status of our well-being by a variety of measures. One aspect of this, which accounts for many others, is a spectrum from “cognitive ease” to “cognitive strain.” The former suggests things are well, while the latter reflects the level of effort required and the extent of unmet needs.

Using Figure 5 on Page 60, Kahneman discusses cognitive ease as being fostered by repeated experience, clear display, primed idea, and good mood. In turn, the figure shows ease feeding feelings of familiarity, truth, good, and effortlessness. However, cognitive ease is also likely to produce more superficial and casual thinking. In contrast, cognitive strain produces vigilance, suspicion, greater effort into activities, and less likelihood to err. Strain also tends to reduce creativity and intuition.

The sense of familiarity that we feel with things we have seen before creates an illusion of remembering—a sense of prior experience even where there is only a prior encounter of a word. That illusion of remembering and the underlying familiarity relate to cognitive ease, if only because we can more quickly recognize that which we have seen before.

Familiarity can also foster illusions of truth. Kahneman observes that a “reliable way to make people believe in falsehoods is frequent repetition, because familiarity is not easily distinguished from truth” (62). Beyond that tenant of propaganda and advertising, repetition of part of a phrase tends to lead people to believe the entire phrase if it is heard later (regardless of falsity).

Building on these and related precepts, Kahneman explains that one can increase the persuasive power of a message by drawing on the studies of truth illusions, which suggest that legibility should be maximized (to reduce strain). Further, he suggests that people regard ideas stated in simpler language as more credible, based partly on studies showing that pretentious language is regarded as an effort to mask a lack of intelligence. Further, memorable messages (such as those in verse or those that rhyme) are more likely to be regarded as true. Finally, Kahneman suggests that cited sources are weighted more heavily if the authors’ names are relatively easy to pronounce. This is because System 2 tends to avoid anything that seems like it involves effort. Despite all these tricks to enhance apparent credibility by appealing to System 1, Kahneman reminds readers that System 2 can and does overcome such tricks to sort out the truth when a person has sufficient motivation to do so.

Cognitive strain’s relationship to the activation of System 2 is similar to the relationship of smiling or frowning to the related emotion. They are two-way associations. Smiling or frowning tends to create the associated emotion just as the emotion creates the facial expression. Likewise, strain tends to activate System 2 and, in the same way, the activation of System 2 tends to create greater strain. They are self-reinforcing. Once System 2 is activated, people are more likely to reject the intuitive answers suggested by System 1.

In fact, cognitive ease is apparently associated with good feelings. Further, mere exposure to something, which is connected with cognitive ease and a feeling of familiarity, has been repeatedly proven to positively influence people’s opinions toward the thing. This “exposure effect” applies to items that are never consciously registered and is actually stronger if the exposure is not consciously registered. This demonstrates System 1’s operation outside of System 2’s conscious awareness.

Kahneman then surveys the extensive research showing that good mood is linked with creativity and intuition but also gullibility and increased logical errors. The reason for this, in bio-evolutionary terms, is not hard to understand. When a person experiences pleasure, it generally means that individual is doing relatively well. Cognitive ease is a cause of pleasant feelings, but it is also a result of them. Specifically, with more pleasant feelings, people are more likely to experience a sense of logical cohesion. Here, too, the level of cognitive effort and the quality of the individual’s emotional state are mutually reinforcing.

Kahneman focuses on System 1’s abilities and limitations in Chapter 6. He states that “the main function of System 1 is to maintain and update a model of your personal world, which represents what is normal in it” (71). That model consists of the associations between “ideas of circumstances, events, actions, and outcomes that co-occur with some regularity” in a contemporaneous or simultaneous manner (71). This model grows, and certain associations gain strength, over time as they recur, thereby reflecting the events of your life and creating the lens through which you experience the present and hold expectations about the future.

The capacity for surprise emerges from these expectations regarding the future. Kahneman explains that two types of surprise exist. The first is based on known, conscious expectations and arises when such events do or do not occur as expected. These expectations constitute what we understand and perceive as normal.

To illustrate how norms affect our System 1 operations, Kahneman notes that when things exist in the same context in our view of the normal, most people do not detect a substitution of the name “Moses” for “Noah” in a sentence about animals on an ark. This is primarily because both are biblical characters, but it also helps that they are the same number of syllables. Yet if the name “George W. Bush” was substituted into the same sentence, System 1 would register it as unusual because it does not fit the context of the ideas generated by the other words.

Kahneman then explains that we treat causation similarly in that we are apparently wired from birth to infer causation. Infants as young as six months old, for example, have been shown to assume causation in watching visual events that may have no causal connection but nonetheless appear to present cause and effect.

Finally, Kahneman notes that he discusses System 1 as either an agent or an associative machine without regard to consistency. He presents System 1 this way because of the human mind’s strong tendency to see causation in events.

Continuing the discussion of System 1, Kahneman borrows a line from comedian Danny Kaye: “Her favorite position was beside herself, and her favorite sport was jumping to conclusions” (79). System 1 jumps to conclusions, making a choice between several possible interpretations without a person’s awareness, thus leading to one seemingly automatic interpretation that enters consciousness.

System 1 does not deal in uncertainty or doubt. Those are in the domain of System 2. Instead, System 1 contributes to the phenomenon of confirmation bias because it is inclined to believe what it perceives. If a false statement is made, System 1 initially attempts to find a way to believe it by making it fit with what is previously known. That is, System 1 undertakes a “positive test strategy” in attempting to confirm the statement. This contributes to confirmation bias in that what is earlier known is likely to shape perception of what is later perceived in a manner that tends to confirm the earlier knowledge. Determining whether something is false, on the other hand, requires System 2.

In a related manner, System 1 creates the “halo effect,” in which a person or thing is understood as all good or all bad based on the initial characteristics known about the person or thing. That is, System 1 strives for—and thus tends to exaggerate—emotional coherence.

One of the more important features of System 1 is introduced by Kahneman at this point. Kahneman describes this phenomenon as “what you see is all there is” and uses the acronym WYSIATI as shorthand from this point forward.

WYSIATI means exactly what it sounds like: For System 1, there is no deliberative process of filling in blanks. Its conclusions are formed solely on the basis of readily available information. Armed only with such available information, System 1 creates the best possible story that it supports. Information that is not currently known simply does not exist to System 1. This concept is crucial to matters discussed later in the book, such as overconfidence and framing effects.

Chapter 8 explains the core mechanism that underlies biases and heuristics: System 1’s continual “basic assessments” of the world and the mind, which occur without intent or effort because hard questions are easily substituted for easier ones. Among other features driving such basic assessments is the fact that System 1 deals well with averages but poorly with sums.

System 1 can also translate values across dimensions. It is not difficult to process the rather strange question of how intelligent someone would be if he were as intelligent as he his tall, for example. This is called intensity matching.

System 1 also takes a mental shotgun approach, partly in conjunction with System 2. That is, when the mind attempts to address one question, it tends to assess many factors (relevant and irrelevant) and answer much wider questions than necessary.

Kahneman suggests that we rarely feel “stumped” because if faced a question that we cannot immediately answer, System 1 substitutes an easier question and provides an immediate answer. These easier questions are what Kahneman calls “heuristic questions.” The word “heuristic” refers to “a simple procedure that helps find adequate, though often imperfect, answers to difficult questions” (98).

It is not practical for an individual to fully and rationally grapple with all the difficult questions that arise. Instead, heuristic questions immediately arise and provide an answer that seems to address the more difficult question. Sometimes these answers work fairly well, and sometimes they produce serious mistakes.

When people are asked about their general happiness and later asked about their dating lives (or another emotionally evocative aspect of life), the two show no correlation. When the order is reversed, however, the correlation is extremely high. This is an example of substitution. It reflects the laziness of System 2, which would struggle to answer the question about overall happiness, and WYSIATI, through which System 1 draws on the emotions evoked by the earlier question.

The “affect heuristic” describes the tendency to form judgments on the basis of emotions about a person or thing rather than through an evaluative process. In this instance, System 2 acts as an apologist for the emotion-driven snap judgments of System 1.