How sure you need to be depends on what you’re doing (Or: As usual I think social media discussions are missing a more fundamental disagreement, this time about a SCOTUS nominee)

Brett Kavanaugh is being considered for a position on the Supreme Court of the United States. He’s also been accused of sexual assault. This information has been all over social media lately, and there seems to be, as there often is, a fundamental disagreement behind the arguments. On the surface, we see basic support versus opposition of the man. There are some straightforward statements of believe in Kavanaugh or else the women who have accused him, mostly prominently including Dr. Christine Ford. There is also quite the range of more general claims. For instance, some people are saying that you should always believe someone accusing someone of sexual assault. Others are taking the opportunity to speak up about what they take to be a worrying trend of false accusations. But these generalities are harder to grasp, so let’s look at the particular case at hand.

In this particular case, besides the basic disagreement about facts, there’s a prior disagreement about how sure either way you need to be to claim to believe in one side or the other. Or, more straightforwardly, to take one side or the other. What actions or consequences are at stake on a belief one way or the other change how easily we’ll take a side or make a belief claim.

A common way philosophers model sure-ness is by using what are called “credence levels,” numbers between zero and one that represent how confident one is in the truth of a statement.  I’m probably 99% sure it’s not going to snow in Phoenix tomorrow morning. Maybe even more sure than that. So I have a credence level of .99 for that. I think I have three decks of cards in my closet, but I’m not super sure. I wouldn’t even bet on even odds. My credence level is maybe .3.

These credence levels are also nifty for expressing how sure of something to be to act a certain way. For instance, in the US, a guilty verdict requires “proof of guilt beyond a reasonable doubt.” This is often expressed as requiring a credence level for the guilt of .99 or higher. But in a civil case, the standard is just believing the verdict is more likely than not. That is, .51 or higher. You can pick all kinds of cases. If you really hate rain, then maybe you only need a .1 credence level that it’ll rain tomorrow to bring an umbrella.

But before you can pick a requisite credence level in a given belief for a given action, some sort of goal is required. Or multiple goals. We have competing values that push us in either direction. I don’t want to get all wet walking to work, but I also don’t want to needlessly carry around an umbrella. I don’t care that much about staying dry, but carrying an extra umbrella annoys me a fair bit. So I need probably a .9 credence level that it’s going to rain. We want to have a functioning justice system, but we really don’t want to punish the innocent. Better a hundred guilty people go free than one innocent person get locked up. So we need that credence level of .99.

The difference in rhetoric of Kavanaugh’s supporters and detractors is revealing of entirely different focuses. The opposition is generally mostly focused on keeping a rapist out of the Supreme Court. For that end, you don’t seem to need to be all that sure. You might even, as I do, flip the proof of guilt beyond a reasonable doubt around to proof of innocence beyond a reasonable doubt. The cost of being wrong is much higher if he’s guilty. Meanwhile even if he’s innocent, a replacement can easily be found. Given the plethora of options and importance of the position, we should be really, really sure that we get someone really, really good. If there’s any reason at all to suspect a Supreme court nominee is a rapist, then we should just move on to the next option. So my opposition to Kavanaugh in this regard only requires a credence level higher than .01 for his guilt. (It is in fact higher than that, though I haven’t spent a lot of time fine-tuning my position. The credence level is far enough above .01 that not much farther thought is needed.)

On the other side, Kavanaugh’s supporters focus on a few things. Shouting “innocent until proven guilty” is one route, suggesting a demand for a .99 credence level for his guilt being needed to deny him the job. But digging a bit into it, there’s more of a focus on some notion of justice. The actual consequences are secondary to the importance of doing the right thing. This would move the bar probably at least to .51. Either Kavanaugh is deserving of the position or he is undeserving, but that fact has nothing to do with the actual results of him getting it. The question comes more down to “Is he a good guy?” as a quasi-factual question about his character.

Some do appeal to some notion of “ruining his life” being a bad thing, to which the standard response is that not being on the Supreme Court does not constitute one’s life being ruined. The standard response to that is that his reputation is being destroyed. I’d be really surprised if anyone hinged their judgement of him on how the Senate votes. The damage, deserved or not, is done. But this does bring up the presence of various actions to take or beliefs to have based on the credence level of a single statement. For example, while I only need a .01 credence level that someone is a rapist to say that they shouldn’t be a major government official, I do still think .99 is morally required to incarcerate someone. Given the former is the context usually at hand, .01 is the bar used to determine what to say I believe. In most of our lives, we have other contexts. If a friend shared a story about having been assaulted, my role would be to comfort, be confided in, or something along those lines. So the bar is pretty low for me to believe. Even if the evidence I have would seem a little suspicious, I don’t need to be very sure at all. On the other hand, if someone I don’t know accused a close friend, I’d probably need more convincing. (It’s a bit harder to pin this case down, though. I have plenty of evidence already built up leading me to believe that my friend wouldn’t do such a thing. So maybe I wouldn’t need that high a credence level in my friend’s guilt, but I would need a lot of evidence to get the credence level even to a medium level.)

These middle cases suggest to me that there’s some reasonable room for needing different amounts of convincing given different conditions. Whether you’re seeking to find someone good enough for a Supreme Court position, send someone to prison, support a friend, or achieve some notion of justice will determine how sure you need to be to take one position rather than the other. And how sure you need to be can vary from one extreme to the other. Given this fact, perhaps the prior questions need a bit more attention. We have the facts as they’ve been presented. Throwing them back and forth appears to be rather unconvincing. But maybe the sureness levels can be moved. If I think about it, I’m not super sure that .01 is the right bar to deny someone a spot on the Supreme Court. I could probably be convinced that I need to be more or less sure. Some people may also be able to be convinced on what the important values at hand are. This is often the route I take. I don’t think that I can convince people to believe Ford if they don’t already. I do think that I can convince a few people that the bar for being on the Supreme Court ought to be really high and so even if you think that Kavanaugh is probably innocent, you should still support moving on to someone who isn’t even accused.

There’s some back and forth from there. Usually the first defense is that a really corrupt political group might just get people to block everyone Trump appoints. I have responses to it, mostly appealing to the growing incredulity and subsequent increased scrutiny. But the point is that the badness of a rapist on the Court is much harder to deny, the badness is much harder to counterbalance, and even rather strong supporters of Kavanaugh are likely to have some non-zero level of suspicion. So this is the argument to have. It has another nice feature, too: The open acknowledgement of different contexts and aims requiring different levels of credence allows us to be more clearly supportive of people who have been assaulted and are in the current discussion being made afraid of coming forward for fear of widespread disbelief. It also satisfies some of the worries about the potency of false accusations. Sexual assault, among other crimes against others, is unfortunately often very difficult to be super sure either way what happened. But, we do at least have a fair bit of middle ground to work with. You need to be pretty damn sure someone did something wrong to lock them up. You don’t need to be so sure to just not give them a bunch of power. And you don’t need to be so sure to give someone some support when they’ve been hurt.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s