Theoretical and Empirical Evidence of Human Rationality - Essay Example

?Introduction Whether humans are rational or not would depend upon how they reason, for reasoning is at the heart of rationality and is opposed to intuition, which is where most of the biases described below begin. However, much of the research done shows that humans are not really rational, when it comes to measuring rationality against an objective normative standard. Rationality would mean that an individual is efficient towards reaching his or her goals, and that the individual would be able to see a situation and see the most efficient solution. Unfortunately, humans are unable to do this, for the most part, so they rely upon heuristics which leads to biases. These biases, such as confirmation bias, anchoring, base rate neglect and overconfidence, make up the first section of this essay. The biases are present in behavioral economics, which dictates that humans will behave in a way that is efficient for them in an economic sense, and this makes up the second portion of this essay. There is some indication that humans can be rational when it comes to domain specific tasks – such as exposing cheating or enforcing social contracts. Because of this, more study should be done regarding other domain specific tasks to indicate if there are pockets of rationality in other domain specific areas. Moreover, there is considerable thought that the irrational nature that is detected on the tests could be the result of something other than irrationality at work – such as poorly worded tests, computational errors, incorrect norms being applied, or inattention of the participants. Therefore, it is possible that humans are not as irrational as these tests might presume. This analysis makes up the third and last portion of this essay. Discussion Reasoning is a part of cognition, which is virtually wholly made up of using inference. There is some question of whether or not reasoning has a separate cognitive system unto itself, or whether or not it is merely a part of the whole of cognition (Manktelow & Chung, 2004, p. 66). That said, reasoning must be distinguished from intuition. The main difference between reason and intuition is that reasoning is done deliberately and consciously, while intuition springs forth from the mind in a spontaneous fashion, without effort or a conscious search (Kahneman, 2003, p. 1450). Thus, when a person is doing an income tax form, he is using reasoning; when that same person revolts at eating a piece of chocolate that is the shape of a cockroach, that person is working from intuition (Kahneman, 2003, p. 1450). Reasoning is a function that is only in the higher order beings, as there has been no evidence that reasoning is present in animals or children who are pre-verbal (Mercier & Sperber, p. 3). There are three basic types of reasoning, according to Samuels et al. (2004). They are descriptive, which describes how humans actually reason; normative, which describes how humans should reason; and evaluative, which describes the difference between how humans actually reason and how they should reason. In other words, there is an assumed standard that has been established by the normative project, so researchers interested in the evaluative project are interested in finding out how actual reasoning fit the assumed standard (Samuels et al., 2004, p. 1). These are the bases of deciding whether humans are rational or irrational – does their reasoning fit what is normative? If this is the case, then rationality can be presumed, for this would mean that the individual is making decisions that benefit him or herself. There are a series of normative rules that prescribe how humans should behave. One of these is cancellation, which means that a human will eliminate “any state of the world that yields the same results, regardless of one’s choice” (Tverskey & Kahneman, 1986, p. s252). Cancellation is important because only one state will be realized, which makes it easy to evaluate the other options separately for each state. Transivity is another rule, which means that each option in an array of choices can be assigned a value that is independent of all other options. Dominance is a third normative rule, which states that, if one option is obviously better than in one state and at least as good in all states, then that option should be chosen. Invariance is yet another rule, which states that “different representations of the same choice problem should yield the same preference” - so alternative descriptions of the same problem would lead to the same solution (Tverskey & Kahneman, 1986, pp. S252-s253). If an individual operates by these normative rules, then that individual can be said to be rational. Rationality comes from a combination of perception, memory, consciousness and reason (Audi, 2004, p. 19). According to Tversky & Kahneman (1986), the normative standard is bolstered by the fact that humans should be efficient in reaching goals, which would mean that they would be governed by the normative project, as acting rationally would get them closer to this goal; and, as competition favors rational choice, and optimal decisions increase the likelihood that an individual can survive in a competitive environment, humans should gravitate towards normative projects (Tversky & Kahneman, 1986, p. s251). However, most individuals do not operate by the above normative rules, which would mean that most humans have some degree of irrationality. Tversky & Kahneman, like Samuels, reject the view that humans correspond with normative rational projects, stating that “the logic of choice does not provide an adequate foundation for a descriptive theory of decision making” (Tversky & Kahneman, 1986, p. s252). Likewise, Samuels et al. (2004) do not believe that humans comport with normative rational standards. The basic premise of the Samuels article was that humans lack the basic ability to reason effectively, and, because of this, they must reach for heuristics that lead to an assortment of biases (Samuels et al., 2004, p. 2). There are a number of different biases that are examined by the Samuels article that prove or disprove the notion that humans are basically without rationality. One of these is an examination of tasks that require probabilistic judgment. Probabilistic judgment concerns making inferences about something based upon facts that are already known. Since inferring something new from something old is uncertain, then the laws of probability fill in the gap (Chater et al., 2006, p. 287). That said, probabilistic judgment gives way to many fallacies. One is the conjunction fallacy, which means that, when asked to rank the probabilities of certain scenarios, research participants rank the probability of a compound scenario higher than that of one of its conjunctive elements (Samuels et al., 2004, p. 5). So, for instance, the possibility that a dog is hungry and tired is ranked as more probable then the possibility that the dog is hungry, despite the fact that the dog cannot be hungry and tired without first being hungry. This shows a degree or irrationality, because, obviously, if an individual was thinking rationally, he or she would realize that if a dog is hungry and tired that he is also hungry, and that the dog being hungry would be a more likely event than the dog being hungry and tired. Another bias is known as “base rate neglect “ (Samuels et al., 2004, p. 6). In this, participants were informed that individuals were interviewed, and that 70% of the people interviewed were lawyers and 30% were engineers, or vice versa. Then the individuals were asked to infer from a description whether the individual was a lawyer or an engineer. Some of these descriptions were stereotypes of lawyers, some of engineers, and others were neutral. However, the base rate – which states that the men were either 70% lawyers or 70% engineers – was ignored by the participants (Samuels et al., 2004, p. 6). This is an example of irrationality, as a rational mind would figure out that, if it is known beforehand that there is a 70-30 split, that assigning 90% of the identities as lawyers would violate the rational tenet that was set forth at the beginning of the project. Another bias is confirmation bias, which refers to the fact that human beings “tend to seek and interpret information in ways that are partial towards existing beliefs,” while discounting or avoiding information that contradicts these existing beliefs (Ask and Granhag, 2005, p. 45). A nearly identical phenomenon is “tunnel vision”, in which a suspect is focused upon, and all evidence that builds a case for conviction is selectively filtered, and all evidence that would exculpate is ignored. (Findley & Scott, 2006, p. 292). The psychological basis for confirmation bias is that people believe what they want to believe, and seek out information that bolsters what they want to believe while ignoring evidence that contradicts this. (Mercier & Sperber, p. 24). Confirmation bias also explains why people seek out information that conforms with their beliefs, as opposed to opening their minds to information that does not. (Nickerson, 1998, p. 199). Confirmation bias is a sign of irrationality, simply because a rational mind would work through a problem and not get stuck just because an existing belief dictates a certain outcome. A rational mind would be an open mind, and a mind is not open when it is essentially stuck on one thing and one thing alone, being blinded to all other possibilities. Another bias is the “the anchoring effect.” The anchoring effect is where participants use the first values in a sequence as a stronger influence than subsequent values. Thus, when presented with the problem of computing 1x2x3x4x5x6x7x8, participants typically find the product of this is much less than the product of 8x7x6x5x4x3x2x1, even though these two products will be exactly the same (Samuels et al., 2004, pp. 7-8). These biases are examples of how humans operate irrationally. Because there are certain normative ways that humans are to reason, and humans are not able to reason in these ways, they must rely upon the heuristic biases that are stated above. If humans were operating at a normative rational level, these heuristics would not be used, because the individual would be able to reason in a way that biases would not come into play. Moreover, reasoning and rationality has been studied in the area of behavior economics. In this area, paternalism has grown because of the theory that either a) certain groups of individuals, historically women, minors and idiots, do not have the ability to rationalize in their own best interests, therefore the state must step in to protect these individuals or b) even individuals who are considered intelligent and rational might act against their best interest if they are put into a desperate situation (Camerer et al., 2003, p. 1214). Hence, there was a need for regulations to basically save these individuals from themselves. Behavioral economics sought to explain these situations where individuals do not act in their own best interest, and cognitive theory was used to explain the theories about economic choice (Camerer et al., 2003, p. 1215). Departures from rationality that were detected in the literature regarding behavioral economics were used as justifications for paternalistic regulations. Some of these biases are as follows. Default means that humans are much more likely to stick with an existing policy or legislators, even when the costs of switching would be relatively low. This resistance to change is related to the concept of loss aversion, which means that individuals put great emphasis on losses than gains, and, since switching to something new denotes that one might lose, then the person is averse to switching. One might also gain from the switch, but, since this is given less credence then the possibility one might lose, loss aversion means that individuals are less likely to switch (Camerer et al., 2003, p. 1224). Thus, they want to go with the devil they know rather than the devil they don’t, so to speak. Related to this is the anchoring effect in behavioral economics, which means that, for instance, an initial price, though arbitrary, once established in one’s mind, will shape present and future prices (Ariely, 2008, p. 26). Another bias is related to framing and information disclosure. This is where individuals either ignore fundamentals that economists would assume would be relevant, or putting too much emphasis on things that economists would assume would be irrelevant (Camerer et al., 2003, p. 1230). Yet another bias is that individuals will respond to emotion rather than reason (Camerer et al., 2003, p. 1239). However, not all theories are as pessimistic as the above, which states that, no matter what, humans are not rational and do not have the capabilities to be rational, therefore they must rely upon heuristics to get by in the world. There is a contradiction in the literature, that was exposed by Gigerenzer & Hug (1992), who hypothesize that reasoning and the ability to reason depends upon content, and is domain specific as opposed to domain general. Gigerenzer & Hug were influenced by Cosmides (1989), who introduced the social contract theory, the premises of which can be stated “if you take a benefit, then you pay a cost” (Mantkelow, 1999, p. 106). Gigerenzer & Hug hold up social contract theory as an example of human reasoning that is domain specific and highly developed to solve an important adaptive problem (Gigerenzer & Hug, 1992, p. 2). Gigerenzer & Hug contrast the theory about social contracts and its relation to human reasoning with the above biases, such as the confirmation bias, anchoring, etc., by stating that these other biases were domain general reasoning, or content general, in that these biases can apply to any set of facts or circumstances, and what is studied in social contracts are domain specific, or content specific (Gigerenzer & Hug, 1992, p. 3). Their study found that human biases are more pronounced in the content general areas than in content specific areas, which means that biases do not reach through all the domains of human reasoning (Gigerenzer & Hug, 1992, p. 9). Moreover, Stanovich and West (2000) posit that there could be other reasons why individual behave irrationally on the standardized tests. These tests could be skewed by performance error such as inattention; may be skewed by computational errors; and may be skewed by applying the wrong normative model to the task at hand; and the subject may be construing the problem differently, thus applying the appropriate answer to a different problem (Stanovich & West, 2000, p. 421). An example of this is the “Linda problem,” which is an example of conjunctive errors, as described above. Linda is described as educated, bright, and active in anti-nuclear demonstrations. Subjects ranked the possibility that Linda was a bank teller and feminist as higher than the possibility that Linda was a bank teller. Although this seemed to be a conjunctive fallacy - if Linda is a bank teller and feminist, then she is a bank teller, therefore being a bank teller must be ranked higher than the possibility of being a bank teller and feminist – Stanovich & West posited that there were tacit assumptions at work that prevented participants from reading this problem as intended, and it would be reasonable for participants to assume that, if Linda is a bank teller alone that she would not be a feminist. Therefore, this could be the reason for the oft-cited conjunctive fallacy in this problem, not that there is some other kind of irrational bias going on (Stanovich & West, 2000, p. 433). As for the computational errors, Stanovich & West found correlations between cognitive ability and heuristics; and cognitive ability and other tasks (such as sunk cost effects, outcome bias, “if only” thinking, searching for unconfounded variables). This suggests that there might be individual differences in rational thought (Stanovich & West, 2000, p. 423). For the possibility that the wrong normative model has been applied, Stanovich & West point out that quanitifying responses is a difficult endeavor that relies upon different disciplines, such as statistics and logic, and psychologists have invoked inappropriate norms to such biases as base rate use and overconfidence (Stanovich & West, 2000, p. 423). Jones (2004) states that the available norms are incomplete or are accorded an importance that have been substantially inflated, and that norms and ideals are fault, which contributes to the erroneous belief that females are especially not rational (Jones, 2004, p. 103). Hence, Panglossians, who dispute the irrationality of subjects, tend to state that the problem lies with the researchers, not the subjects (Stanovich & West, 2000, p. 423). For Panglossians, the norm is rejected, and normative standard is substitute that fits perfectly with the descriptive model of the subjects’ performance, which restores human rationality to ideal proportions (Stanovich & West, 2000, p. 424). Conclusion Much of the evidence that is present about rationality shows that humans are not really rational due to an enormous number of biases that occur at the intuitive level. Hence, while the ability to reason would imply that humans should act rationally, thus comporting with normative prescriptions for human behavior, they do not, due to underlying psychological phenomena that undermine humans’ ability to formalistically reason. This is seen not only in the realm of psychology, but also in the realm of economic behavior. That said, Gigerenzer & Hug posit that such biases do not always work in every domain, and that some types of reasoning are domain specific, such as reasoning related to social contracts and the detection of cheating. This would imply that, from an evolutionary standpoint, the need to regulate social behavior and expose cheating is somehow essential to the survival of the species, as this is a domain that is seemingly untouched by human biases. This represents a kind of gap in the literature, in that not enough literature has touched upon reasoning in a domain specific way, such as Gigerenzer & Hug, and the majority of the literature examines biases in a domain general context. This is one area that needs to be better examined, in that other domain specific reasoning should be examined to determine if Gigerenzer & Hug have discovered an isolated phenomenon, or if it is a phenomenon that can be extrapolated to other domain specific areas of human cognition. Sources Used Ariely, D. (2008). Predictably irrational: The hidden forces that shape our decisions. New York: HarperCollins Publishers. Ask, K. and Anders, P.G. (2005). Motivational sources of confirmation bias in criminal investigations: The need for cognitive closure. Journal of Investigative Psychology and Offender Profiling, 2, 43-63. Audi, R. (2004). Theoretical rationality: Its sources, structure and scope. In Mele, A. & Rawling, P. (2004) The Oxford Handbook of Rationality. Oxford: Oxford University Press. Camerer, C., Issacharoff, S., Loewenstein, G., O’Donoghue, T. & Rabin, M. (2003). Regulation for conservatives: Behavioral economics and the case for “Assymetrical Paternalism.” University of Pennsylvania Law Review, 151(3), 1211-1254. Chater, N. & Oaksford, M. (2000). The rational analysis of mind and behavior. Synthese, 122(1/2), 93-131. Chater, N., Tenenbaum, J. & Yuille, A. (2006). Probabilistic models of cognition: Conceptual foundations. Trends in Cognitive Sciences,10(7), 287-291. Findley, K.A. and Scott, M.S. (2006). The multiple dimensions of tunnel vision in criminal cases. Wisconsin Law Review, 291-397. Gigerenzer, G., Hoffrage, U. & Kleinbolting, H. (1991). Probabilistic mental models: A Brunswikian theory of confidence. Psychological Review, 98(4), 506-528. Gigerenzer, G. & Hug, K. (1992). Domain-specific reasoning: Social contracts, cheating and perspective change. Cognition, 43, 127-171. Gigerenzer, G. & Goldstein, D. (1996). Reasoning the fast and frugal way: Models of bounded rationality. Psychological Review, 103(4), 650-669. Heuer, R.J. Psychology of Intelligence Analysis. History Staff, Center for the Study of Intelligence, Central Intelligence Agency, 1999, http://murd0c.net/reprimander/PsychologyofIntelligenceAnalysis.pdf (accessed March 3, 2011). Heuer, Richards J. (Winter 2005). Limits of intelligence analysis.” Orbis, 75-94. Jones, K. (2004). Rationality and gender. In Mele, A. & Rawling, P. (2004) The Oxford Handbook of Rationality. Oxford: Oxford University Press. Kahneman, D. (2003). Maps of bounded rationality: Psychology of behavioral economics. The American Economic Review, 93(5), 1449-1475. Mantkelow, K. (1999). Reasoning and Thinking. New York: Psychology Press. Mantkelow, K. & Cheung Chung, M. (2004). Psychology of Reasoning. New York: Psychology Press. Mercier, H. & Sperber, D. Why to humans reason? Arguments for argumentative theory. Accessed online. Available at: http://www.dan.sperber.fr/wp-content/uploads/2009/10/MercierSperberWhydohumansreason.pdf Samuels, R., Stich, S., Faucher, L. (2004). Reason and rationality. Handbook of Epistemology, 1-50. Stanovich, K. & West, R. (2000). Individual difference in reasoning: Implications for the rationality debate? In Gilovich, T. & Kannehan, D. Heuristics of Judgment: Extensions and Applications. New York: Cambridge University Press. Tversky, A. & Kahneman, D. (1986). Rational choice and framing of decisions. The Journal of Business, 59(4), S251-S278. Read More