Unlike McKnight, Day could never quite get in to cognitive-behavioral therapy. The contradiction at the center of it was just too gargantuan and obvious to take seriously. McKnight saw the contradiction too, but was so desperate by the time he started CBT that in a very non-McKnightian way, he found himself not caring. At that point, relief was the only thing he cared about, logic and contradiction be damned.
Basically, the contradiction was the following: CBT was structured around making depressed people like Day re-assess their thoughts and beliefs to show just how errant those thoughts and beliefs in fact were. So, for example, Day frequently thought things like “I’ll never get Iressa, there’s no chance.” Thoughts like these were a manifestation of a deep-seated belief of his own total fucking incompetence. So Dr. Royer would probe this belief with questions like, “OK, Day, are you really incompetent? Let’s look at the evidence on both sides.” And then they’d list ten reasons why Day wasn’t incompetent, which was almost enough to make Day a believer in his own newfound competence.
Except here’s where the contradiction came in. One of the things Dr. Royer was implying here is that Day obviously was not incompetent. He was obviously, in fact, highly competent. Truth be told, when they decomposed the problem and looked at the evidence, it was extraordinarily obvious. But, Day still thought he was incompetent. So wouldn’t the fact that Day still thought he was incompetent in spite of enormous amounts of evidence that he was competent actually prove his point, that he was incompetent? As in, his thinking is so incompetent that he can’t even distinguish between deep competence and total incompetence? And wouldn’t that mean the only logical possibilities are that 1.) Day was indeed incompetent, and recognized that. So he was fundamentally correct about the assertion, or 2.) Day was highly competent, and his belief was so off-the-reservation that his even thinking it must make him, at the end, incompetent.
He raised this to Dr. Royer, who didn’t seem thrilled with this part of the tete a tete. Day felt pretty smart about this whole thing, until Dr. Royer raised him one: in the first scenario (Day = incompetent, and Day accurately assesses it as such), wouldn’t, by Day’s logic, that mean that since Day accurately identifies his own incompetence, that he is in fact competent, thereby making him ultimately competent at the end of it all?
After a somewhat belabored discussion of the hermeneutics of competence, they agreed that the only way to truly resolve this, if one took “ability to distinguish reality from non-reality” as a provisional definition of competence, (which Dr. Royer had serious objections to, but nevermind) was for 100 or so unbiased observers (this being the number required to power the study sufficiently) to rate Day’s competence on a one-dimensional scale, 0–10, and compare the average to Day’s own rating. If the median answer from the observers was close to Day’s score, that meant Day was competent, and if it was far, incompetent. But they then also agreed that to truly test this, a single instance wasn’t sufficient — they’d have to do this over the course of several months, taking a measurement each month, to control for temporal variations (i.e. what if Day had eaten something slightly depressive for breakfast that day that had nothing to do with his competence and artificially depleted his own self-score, or what if a stranger had by chance told him he was gorgeous a few moments before and artificially inflated his self-belief.) But if they did all this, they’d be able to come to a pretty accurate assessment of Day’s competence. But then Day further and correctly pointed out that they’d have to cycle out observers between sessions to avoid either the anchoring effect or emotional exhaustion that swayed the scores in one direction or another. Dr. Royer agreed here, and so the n of people would ultimately have to be 1200, if they were doing it for a whole year. But then Dr. Royer asked what about his own anchoring effect and emotional exhaustion, wouldn’t that affect Day’s self-rating, and might that, in fact, be maybe a hint of what this depression is? Yes, Day agreed it would affect his self-rating, but he was the operand here, not a thing to control, and so they’d just have to accept his behavioro-economic failings as an unavoidable part of the trial, and no, it didn’t help illuminate his depression because that would just be admitting he gets emotionally exhausted faster than other people and is thus incompetent. And then finally they also agreed that since each observer likely wouldn’t know Day extremely well, and thus couldn’t be guaranteed to have strong opinions about his competence, that a one-dimensional scale was insufficient, and what they’d actually need was a two-dimensional scale, magnitude 0–10, and confidence 0–10, which would distinguish between people who had strong vs. weak beliefs about Day’s competence vs. incompetence.
So, after mapping this path out, they decided that such a study was probably not only infeasible but likely to get bogged down in so many pointy-headed considerations of statistical rigor that it would probably exacerbate everyone’s feelings of incompetence, not to mention probably not provide sufficient return on time, money, emotional investment, etc. So they decided to go back to talking about what Day’s recent interest in MILF/son big dick porn said about his latent Oedipus complex.