33 Gilbert et al. 1993. 34
Which is not to say that beliefs aren‘t often ‗fragmented ‗(Egan 2008). I suppose that most beliefs are inferentially promiscuous to a degree, but don‘t actually interact with one‘s whole web of belief. Frankly, I bet most people‘s beliefs are highly fragmented and kept in context specific stores to facilitate not just further beliefs, but also to buttress one‘s psychological well-being (more on this in section 5.2.2).
39
One would think that if you knew that you were about to encounter false information and, as the Cartesian theory supposes, you had the ability to withhold assent, then you would not form beliefs based on the false information you subsequently encounter. However, the next two studies deal with situations in which people know that they are about to encounter falsehoods of certain sorts and yet still can‘t help but form beliefs based on the falsehoods. Which, to beat a dead horse, is just what the Spinozan, but not the Cartesian, would predict. 2.5.3.1 Belief Perseverance in the Face of Debriefing and Prebriefing
Another telling set of experiments comes from the literature on belief perseverance in the face of experimental debriefing. In a typical experiment, an experimenter asks
participants to read a bunch of suicide notes and to sort the real ones from the fakes. In Ross et al. (1975), participants encountered twenty-five pairs of notes and were told that one note from each pair was a real note, one note a fake. After seeing each pair participants would judge which note was real and which fake and were then given feedback on their
performance. After receiving the feedback the participants were (partially) debriefed. During the debriefing the participants were told that all the feedback they received was fictitious, it being arbitrarily determined beforehand regardless of the participants‘ responses. After the debriefing the participants were asked to estimate both how many times they actually
answered correctly and how many correct answers an average person would give. Sadly, the information in the debriefing session did not affect participants‘ opinions about their ability: if the participant originally received positive false feedback (e.g., twenty-four out of twenty- five correct) they believed that they were better than average at the task, and if they received negative false feedback (e.g., seven out of twenty-five correct) they believed they were worse than average at picking out real suicide notes from fake ones.
40
The aforementioned experiment is generally not taken to illuminate anything about belief acquisition per se. It seems that the participants formed their beliefs in a reasonable enough way, based on the experimental feedback. Once they are told that the feedback was non-veridical they may just have trouble updating their beliefs. Perhaps beliefs are ‗sticky,‘ in that once one has a belief, that belief is hard to get rid of. If so, then the debriefing effect wouldn‘t tell us about anything belief acquisition per se, but rather belief perseverance.
But what happens if the people are briefed before they take part in the study and receive false feedback (call such a technique ‗prebriefing‘)? What if before sorting the notes they are told that the feedback they are about to receive is bogus? The Cartesian view predicts that if we tell people beforehand that what they are about to read is false, and they have no reason to distrust what we tell them, then, ceteris paribus, they will approach the stimuli skeptically, withholding forming any beliefs about their ability if those beliefs are based on the bogus data. On the other hand, the Spinozan view predicts that since people believe everything they token, they‘ll be stuck believing propositions that they encounter even if they know beforehand that they are false.
As predicted by the Spinozan view, but not the Cartesian view, prebriefing the participants beforehand does not impact the participants‘ judgments about their ability.
Wegner et al. (1985) replicated the Ross study except the participants were told prior to the
task that the notes and the feedback are dubious. Yet even after the explicit prebriefing the participants continued to behave as if the feedback was veridical. They were unable to reject the feedback they received, even though they knew the feedback was bogus. These
perseverance effects are easily explicable on the Spinozan view: the knowledge of the feedback persists because the participants automatically believe the feedback when they hear
41
it, even though they know the feedback is false. The prebriefing effect helps to verify the first and third properties of the Spinozan theory. Equally importantly, these persistence effects are anomalous on the Cartesian theory, particularly casting doubt on the first property of the view. If the Cartesian theory were correct, then people would have the ability to withhold assent from propositions they encounter. Surely, if there was any time that this ability would rear its head, it would be in a situation where you know that you are about to encounter false information. Thus, the study just detailed gives us a perfect chance to allow the participants ability to shine, yet the people just could not reject the information they encountered. Such data should make one very wary of the Cartesian theory.
2.5.3.2 Never Trust a Fake Smile…At Least When You Know It’s Fake Before Seeing It Participants were presented with a series of smiling faces. They were told either before or after the presentation of the face that the face was either a fake smile (because the person had previously had a bad mood induced before being asked to smile) or a real smile (because the person has previously had a good mood induced before being asked to smile). Participants were also told that they would sometimes hear a tone and they‘d have to push a button indicating whether the tone they heard was high or low pitched. During this phase, subjects were asked to remember which faces contained real smiles and which ones
contained fake smiles. After the learning phase, subjects were re-presented with eight faces they had seen, four of which had been previously shown with the interrupting tone. In the test phase, subjects identified real smiles correctly 60% of the time when the smiles were
originally presented uninterrupted and 65% of the time when originally presented interrupted (an insignificant difference). However, the correct identification of fake smiles was at 65% when originally presented uninterrupted and only 35% when interrupted (a significant
42
difference). The results were the same whether or not the participants were told that the
smiles were fake before or after they saw the picture.35 Thus, even when subjects knew that
they were about to see a false smile they couldn‘t help but see it as a real smile. The moral to draw here is similar to the moral to be drawn from the studies mentioned above. If the Cartesian theory were right, then we‘d expect people to show no asymmetry in memory between remembering truths and remembering falsehoods. But once again we see that cognitive load sheds light on a deep-seated asymmetry between truths and falsehoods: being cognitively busy during the learning situation causes people to
misremember falsehoods as truths, but not truths as falsehoods. The study mentioned here is just more evidence toward showing that the time at which one learns that they are acquiring a falsehood matters not to the asymmetry in the way we process truths and falsehoods. This asymmetry is missed by the Cartesian theory, but not the Spinozan one.