Confirmation, Tunnel Vision and Mind-Set: Why Intelligence Failed In Pearl Harbor and Iraq - Research Paper Example

Confirmation, Tunnel Vision and Mind-Set: Why Intelligence Failed In Pearl Harbor and Iraq Introduction There is little doubt that the events in Pearl Harbor and Iraq, with regards to whether or not Saddam Hussein had WMD, signified great failures with intelligence. Simply put, the United States was caught unaware with Pearl Harbor, despite the fact that there was ample evidence coming from the Japanese that they were planning to attack. The United States, however, was led by people who thought that the Japanese simply would never do this, therefore the evidence that they were going to do this was disregarded. With regards to Iraq, the United States, and other countries such as Britain and Australia, assumed that Saddam Hussein had weapons of mass destruction. Because this was the assumption, all of the evidence supporting this assumption was used to fit the prevailing wisdom. In both cases, very powerful psychological biases were at work – confirmation bias, and the closely related phenomenon, tunnel vision and mind-set. Simply put, evidence that supports a bias is used, evidence not supporting it is ignored or discarded. There is ample evidence that this was the case in both Iraq and Pearl Harbor. This paper will explain the different biases, and show how both Pearl Harbor and Iraq fit these biases perfectly. Confirmation Bias, Tunnel Vision and Mind-Set There are a number of reasons why, all of them human, that intelligence analysts may have a hard time properly interpreting data. All of the reasons are focused upon data immersion. (Heuer, 1999, pp. 40-41). When information comes into the intelligence analyst, this information comes in as discrete pieces of data. This data must be put into some kind of meaningful context, so the data must be sorted, selected and organized so that it becomes something that is meaningful to the analyst. However, the data can be sorted and analyzed in a way that would already confirm an inherent bias, or so that it confirms a notion that has already been preconceived. There are great many biases that humans might have, and the intelligence is used to support these biases. The following are some of the biases that might occur. For instance, humans might submit to what is known as a confirmation bias. This means that the facts that are coming in are interpreted in such a way that they confirm existing beliefs. At the same time, the information that is coming in that refutes these existing beliefs is discounted (Ask and Granhag, 2005, p. 45). This may also be known as “tunnel vision.” In tunnel vision, just like with the confirmation bias, the intelligence analyst focuses upon a theory – that there are weapons of mass destruction in Iraq, for instance. The information that comes in through the intelligence channels that support this view are focused upon. The evidence that comes in through the intelligence channels that do not support this view is discarded or ignored (Findley & Scott, 2006, p. 292). In the case of intelligence analysis, confirmation bias may work in one of two ways, or a combination of the two ways. One way is that, as the intelligence information is coming in, the intelligence analyst may selectively “weed out” the information that confirms the bias. The other information that does not confirm the bias may be discarded as “chatter” or “noise.” The other way that confirmation bias may have worked in the case of Iraq, is that the information that is culled for review would have been interpreted in such a way to confirm the bias (Ask & Granhag, 2005, p. 46). This is often seen in prosecutions – a suspect is fingered, and the evidence is reviewed. The evidence that is gathered is examined with an eye towards convicting that suspect. The reviewed evidence is skewed towards this. Perhaps this would also be the reason for the website “unskewed polls,” which came out during the 2012 Presidential election. The progenitor of this site decided that he thought that Mitt Romney would win, but the polls, the evidence, never showed this. So, he went through and “unskewed” the polls so that it would look like Mitt Romney was, in fact, ahead in almost all swing state polls. This is a classic case of making the evidence fit the theory. The reason why confirmation bias and tunnel vision exist is because, as humans, we believe what we want to believe. If evidence comes in that contradicts this, we conveniently put this aside (Nickerson, 1998). Using the example of the election again, Mitt Romney, and many right-wing pundits, thought that, right up until the time that Romney gave his concession speech, that Romney was destined to win. This was in the face of the evidence that showed Romney behind in almost every swing-state poll. They didnt believe those polls – they believed what they wanted to believe. They talked only to others who believed like them, so they were in a kind of right-wing echo chamber, where the reality was that Romney was ahead in the swing-state polls, and he was going to win (Sigman, 2012). In addition to confirmation bias and tunnel vision, humans also tend to have a certain mind-set. A mind set develops quickly and changes slowly (Heuer, 1999, p. 10). A mind-set works in conjunction with tunnel vision. The mind-set is developed, then information is filtered in that either supports the mind-set or refutes it. Supportive information is given weight and paid attention to. Non-supportive information is ignored or discarded. This is shown by hypothesis – the amount of information necessary to product a hypothesis is not as much as the information that is necessary to refute an existing hypothesis (Heuer, 1999, p. 11). Heuers example of a mind-set is an experiment that was conducted around a series of drawings. The initial drawing either resembled a man who looks a lot like Ronald Reagan, or a woman who is sobbing into one of her hands. Slowly, image by image, the initial image morphs into the opposite image – the woman becomes the man, or the man becomes the woman. The middle image has an equal chance of being one or the other. The test subjects, when viewing this middle image, either thought that it was the man or the woman, depending upon which image they began with. (Heuer, 2005, p. 81). This shows the power of the mind-set – those who had an initial mind-set that the image was a woman held on to that mind-set even when the stimuli became ambiguous; the opposite was true with the subjects who had the initial mind-set that the image was a man.1 There are also different biases, or what Bar-Joseph and Levy (2009) refer to as unmotivated biases and motivated biases. The motivated biases are those which which have at their root the mindsets and belief systems that are described above. The person doesnt want to be biased, but he or she is, because of the cognitive biases that all of us tend to have. Motivated biases are another category, and these are motivated by more conscious biases. Such as the conscious desire to advance the persons own interest – such as, with the Iraq intelligence failure, the intelligence analysts might have been under pressure to find a certain result, political pressure, so they worked their bias towards this result. Bar-Joseph and Levy (2009) states that, a step above the motivated bias, which is, itself, a step above the unmotivated bias, is the deliberate action to distort intelligence. The way that they distinguish the two is that, even with motivated bias, there are still unconscious processes at work. The intelligence analyst may not really be aware that he or she is distorting the evidence because of political gain. However, when one consciously distorts the evidence, there is a blatant attempt to make the evidence fit the hypothesis. In other words, the person knows that he or she is distorting evidence, and is doing it deliberately in an attempt to influence others, so that his or her own policy preferences look more favorable. Bar-Joseph and Levin (2009) present the extreme examples of this – dictators, such as Saddam Hussein and Joseph Stalin often had people executed if they were telling them information that they did not want to hear. Therefore, there was a motivation to deliberately distort the information that they received, because it was literally a matter of life or death to them. This is an extreme example, of course, but it is one reason why dictators often are caught unaware that a coup was about to occur, or an invasion, such is the case with Saddam Hussein and Operation Desert Storm. Pearl Harbor and Iraq Pearl Harbor, according to Bar-Joseph and Levy (2009), was the product of unmotivated biases. In this case, the pre-existing belief that underlied the unmotivated biases was that Japan couldnt win a war with the United States, therefore there was not a reason to ever believe that they would attack the United States. The belief was that Japan knew this to be true, therefore they would be smart enough to not attack American soil. Therefore, even though there were indications that Japan was, indeed, gearing up for war with the United States, this intelligence was ignored by the intelligence analysts, because it did not fit in with their preconceived notion that Japan would never attack them. Moreover, even though there were some in the intelligence community who did imagine that Japan would attack the United States, did not imagine that it would happen at Pearl Harbor. Bar-Joseph and Levy (2009) refer to this as a “lack of imagination,” which is the same type of lack of imagination that plagued the United States some 60 years after Pearl Harbor, as shown by 9/11. Just like in Pearl Harbor, the United States intelligence officials never imagined that something like 9/11 could happen, so they ignored the evidence that such an attack was underfoot. It was also the case, in the Pearl Harbor disaster, that there was ample information that the Japanese were going to attack America, according to Bar-Joseph and Levy (2009). Therefore, it was never a case that America did not have indications that this would happen. However, there was too much information that was coming in from Japan. It was therefore difficult to distinguish the chatter and noise from the actual information. Much of what was coming into the American intelligence community during this time could have been used to support the thesis that Japan was going to attack the United States. Much of it would not necessarily support this thesis. However, since the mindset was that Japan would not attack, the information that was coming into the community was dismissed as chatter and noise. Another reason for the intelligence failure in Pearl Harbor, and has often been cited by historians who criticize the handling of the Al Qaeda threat before 9/11, according to Bar-Joseph and Levin (2009), is that there was a lack of communication between the branches of government, namely the Army and the Navy. In this case, there were multiple pieces of puzzle regarding the Japanese threat. The Army had some of these pieces. The Navy had other pieces. If the two branches of the military had worked together and put these pieces together, then they probably would have seen what was about to happen. However, this did not occur, so the pieces remained just that – discrete pieces of evidence about the Japanese threat that were never put together into a coherent whole. The reason why they did not share, according to Bar-Joseph and Levin (2009) is because of rivalries and competition over control of the intelligence. With the case of Iraq, there is evidence, according to Jervis (2006) of a deliberate attempt to distort information, which is above motivated bias. In other words, there was an attempt to distort the information to confirm what these nations already “knew” to true, and that is that Saddam Hussein had weapons of mass destruction. There was, in the word of Jervis, a conscious choice to politicize the information that was coming through. The Democrats in the United States who supported the war were the ones who stated that the evidence was politicized, according to Jervis (2006), because they were in the uncomfortable position of supporting a war that turned out to be based upon false pretenses, yet wanting to still make Bush look bad. So, they claimed that the evidence was politicized, and that they had been misled. However, for the Republicans in the United States, there was a resistance to the charge that the evidence was politicized. They were more likely to claim that the evidence simply had been misread, according to Jervis (2006). In other words, although the Republicans never would say this, their defense for going to Iraq, even though there were no weapons of mass destruction, would be that bias would have been the reason why the evidence was misread. George Bush and his officials were convinced that Saddam Hussein had weapons of mass destruction, so they “cherry picked” the evidence regarding this – giving credence to the evidence that he had WMD, and disregarding the evidence that he did not have WMD. Jervis (2006) goes on to to describe the reasons why the intelligence was incorrect for the WMD in Iraq. Each of these issues can be categorized as some type of bias. In fact, all the reasons that he states, specifically, would fall into the category of confirmation bias and mindset bias. The first category that Jervis (2006) states is the reason why the WMD intelligence was wrong was that there was too much certainty. The intelligence community was “certain” that Saddam Hussein had WMD. The preponderence of the evidence showed this. Therefore, the intelligence community became “certain” that Hussein had WMD. The reality was, according to Jervis (2006), that the bulk of the evidence did show that Hussein had WMD, but that it was far from certain, and there was other evidence that showed that he did not have WMD. As Jervis (2006) stated “the analysts overestimated the number of independent sources reporting to them, and they failed to carefully consider the significance of negative reports and the absence of evidence” (Jervis, 2006). In other words, they were guilty of confirmation bias – overestimating evidence that fit their conclusion, and underestimating evidence that does not. The next reason cited by Jervis (2006) was that there were no alternatives considered for the evidence. There were aluminum tubes that Iraq was importing in secret. The assumption was that these tubes were being used to transport uranium and other materials that would be used for WMD. There was not any alternatives discussed as to why Hussein would be importing these tubes. Hussein had also procured Unmanned Aerial Vehicles that had maps of the United States. The assumption was that the United States was threatened by Hussein because of this. No alternatives as to why Hussein would have these maps in these UAVs was discussed. There was a multitude of evidence pieces that confirmed the biases of the intelligence analysts, but all of these evidence pieces had other explanations for them, as well. There was never a discussion as to what the alternative explanations could be, however. Instead, all the pieces of information was seen by the intelligence analysts, and, perhaps more crucially, by the people using the intelligence analysis, as fitting their theory. None of the evidence was seen as fitting an alternative theory. Therefore, in this regard, this is yet more evidence of tunnel vision and confirmation bias. As noted above, confirmation bias may either be “cherry picking” evidence to suit their bias, or interpreting existing evidence to suit bias. In this case, it was interpreting existing evidence to suit the confirmation bias. The tubes in question were presumed to be used for enriching uranium, and there was not an alternative considered. The alternative would be, and this turned out to be correct, that Hussein was using the tubes to import material for conventional weapons. Since the UN sanctions were in place that dictated that Hussein could not import material for conventional weapons, this is what Hussein was forced to use to import this legitimate material. The fact that this was never even considered to be a possibility shows confirmation bias at work. And this is just one piece of the puzzle where alternatives were not considered. A lack of imagination is the next element that was cited by Jervis (2006) for why the intelligence failed in Iraq. This is related to the issue about not considering alternatives, but Jervis used the overall failure of imagination to impugn large mistakes, as opposed to more discrete ones that are the focus of the lack of alternatives argument. Specifically, the countries who invaded Iraq stated that one of the reasons why they went into Iraq was because of the way that Hussein was acting. Hussein was appearing to not take the steps necessary to avoid war. The explanation for why Hussein was not taking the threat from the United States seriously, according to Jervis (2006) was that Russia and France had apparently promised Hussein that they, as countries, would restrain the United States from invasion. Jarvis (2006) states that the United States erred in not even considering that this was a possibility that this would be the reason why Hussein was acting recalcitrant. This was a lack of imagination, and it was also an example of confirmation bias – the fact that Hussein was not acting like he wanted to prevent war meant he had something to hide, and the United States decided that this was further evidence of WMD. Of course, this does not make much sense – if Hussein was acting like he didnt care if there was a war should mean that he had less to hide, because a war would uncover this dirty secrets. That the United States used this as another reason to invade shows the extent of their confirmation bias and the extent that they were willing to distort the evidence in their favor. Excessive consensus is the next culprit that Jervis (2006) notes was the reason for the failure of intelligence with regards to Iraq. This is related to groupthink, although Jervis (2006) states that groupthink was not really at work, because groupthink is what occurs in small groups – there is a need to conform to group consensus, so the individuals in small groups tend to think like the group thinks so that they would be in line with the group consensus. In the case of Iraq, there were not really small groups, but a large group, and the small groups that existed were always changing in composition. However, there was still the same motivation at work, and that was that the people who were looking at the evidence had a desire to conform to the stated premise, because everybody else was thinking the same way. Jervis said that this was understandable, as the members of the intelligence community who believed that Iraq was reconstituting WMD were people who were highly respected and intelligent, not to mention conscientious. In other words, if the evidence was good enough for them, it was good enough for everybody in the intelligence community, according to Jervis (2006). What these individuals think became the conventional wisdom, so this is what everybody thought. This kind of conformity is related to the confirmation bias and the mindset bias, because what people think is reinforced by what other people think, and if there was somebody in the group who was a naysayer, this person would be most likely ignored. So, the mechanism is the same as for confirmation bias. The last reason why Iraq was thought to be accumulating weapons of mass destruction was what Jervis (2006) termed the “failure to challenge assumptions” (p. 22). This was another classic example of confirmation bias and tunnel vision, not to mention mind set. The intelligence community made assumptions based upon previous judgments. However, there were also uncertainties that were present at the same time that these judgments were being made. These uncertainties, however, were not brought forward with the judgments. The general assumption, according to Jervis (2006), was that Husseins program was “consistent, coherent and unchanging” (p. 23). Since he had tried to get WMD in 1991, according to the prevailing wisdom, it was assumed that he was still trying to get WMD even over ten years later. However, as Jervis (2006) notes, Hussein never did have a coherent plan, and he didnt even have complete control of the country. Moreover, his regime was not competent. However, the intelligence community had assumed that Hussein had a coherent plan, that he had complete control over his country, and that he had a competent regime. Everything that was not supportive of these assumptions was disregarded. Conclusion Confirmation bias, tunnel vision and mindset are powerful psychological biases that make things seem apparent, then they are actually anything but. Like the 2012 election, where the assumption on the right was that Romney was going to win, and everything that supported this was used to forward this assumption, and everything not supported was discarded or ignored, the same thing happened in Pearl Harbor and Iraq. The US assumed that the Japanese knew better than to start a war with the United States, but they were not correct. Therefore, the signs that Japan was going to attack were dismissed as chatter. In Iraq, the US assumed that Saddam Hussein had WMD. Therefore, all intelligence was cherry picked to bolster this assumption. While these are powerful biases, humans should learn to overcome them, in particular when they are dealing with something as sensitive as intelligence. Because the consequences of these biases in cases like Iraq and Pearl Harbor are disastrous and tragic. Bibliography Ask, K. and Granhag, P. (2005) “Motivational Sources of Confirmation Bias in Criminal Investigations: The Need for Cognitive Closure.” Journal of Investigative Psychology and Offender Profiling, vol. 2: pp. 43-63. Bar-Joseph, U. & Levy, J. (2009) “Conscious Action and Intelligence Failure,” Political Science Quarterly, vol. 124, pp. 461-488. Findley, K. and Scott, M. (2006) “The Multiple Dimensions of Tunnel Vision in Criminal Cases.” Wisconsin Law Review: pp. 291-397. Heuer, Richards J. Psychology of Intelligence Analysis. History Staff, Center for the Study of Intelligence, Central Intelligence Agency, 1999, http://murd0c.net/reprimander/PsychologyofIntelligenceAnalysis.pdf (accessed December 10, 2012). Heuer, R. (2005) “Limits of Intelligence Analysis.” Orbis (Winter): pp. 75-94. Jervis, R. (2006) “Reports, Politics and Intelligence Failures: The Case of Iraq,” The Journal of Strategic Studies, vol. 29, no. 1, pp. 3-52. Nickerson, R.S. (1998) “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises.” Review of General Psychology, vol. 2, no. 2: pp.175-220. Read More