Wednesday, May 9, 2012

Confirmation Bias and Enthymeme Mistakes

Confirmation bias is the human tendency to notice evidence that confirms what we already believe and to not pay attention to data that doesn’t fit into our preconceived notions. I’ve seen mention of this and accounts of research showing it often. But I haven’t noticed any research that address distinguishing between distaste for a contrary view per se, and distaste for arguments that had a bad enthymeme or a bad stasis.  
Image by Derrick Tyson
The assumption seems to be that any tuning out of contrary arguments is purely the result of natural resistance to the idea that you are wrong, ignoring the fact that some contrary arguments might be might be dismissed as irrelevant not because of their content but because of their framing. But it is important to consider there are two types of framing problems that could contribute to argument being tuned out: arguing about the wrong thing,wrong stasis, or starting the argument at the wrong entry point, wrong enthymeme.
    First let’s take stasis, which exact facet of the issue in question you address your argument to, usually divided into 4 possibilities. Stasis can be in fact “I didn’t steal your car, I never touched your car”, in definition “I didn’t steal your car, I borrowed it”, in quality, “I had to use your car to take grandma to the emergency room, she might have died.” or in procedure “car theft is a matter for the courts, you shouldn’t be pursuing this argument yourself but should leave it to the proper authorities..” A 5 minute harang focused on the physical evidence of my having taken to car is going to turn me off if my difference of belief is actually a statis in definition or quality, Likewise it would be offensive to hear a well reasoned argument about how what I did was wrong if I know you have the fact of what I did wrong.
Second let’s look at where you start your argument, the enthymeme. Enthymemes are the unstated parts of an argument, in a specific sense they are unstated constituents of syllogisms. In the broadest sense they are propositions commonly agreed on by everyone involved before the argument starts. They seem so obvious that they don’t need to be stated. In fact not stating them produces a sense of community and encourages people to engage in the conversation. But in any case you have to pick some point to start your argument at and whatever logically proceeds that point is your enthymeme.
Here’s a very simplified example of enthymemes and how they can go wrong:
Low interest rates always have a potential to produce high inflation. Therefore, the Federal Reserve must always reserve the option of raising interest rates if high inflation seem imminent.
If you look at this as a syllogism you will see it is missing a premise. What premise is need to make this a valid argument? “High inflation must be avoided.” So that is the enthymeme. If that statement seems obvious and certain to you then the argument will seem cogent to you, even if you disagree with the conclusion. But if the enthymeme seems wrong or even just highly doubtful, the argument will seem to be missing the point, to be at least be not considering the situation fully.
I’m not saying that confirmation bias doesn’t exist or is not a major factor. But if you want to convince people to change deeply held beliefs all you can do about confirmation bias is rail impotently about it. But you can do something about choosing a good enthymeme and the correct stasis. And especially if already facing confirmation bias, neglecting to do so can doom your argument to failure. Blaming polarization and lack of consensus on confirmation bias put all the burden on people who are wrong. Which is a group people seldom identify themselves with. If there is something that people who are right should be doing we need to recognize that too.
So we need to ask ourselves: How much of our tuning out arguments we disagree with is simple confirmation bias and how much is arguments being irrelevant due to a badly chosen enthymeme or the wrong stasis being addressed?

No comments:

Post a Comment