Friday, February 23, 2007

Evaluating Causal Directness

Jaworski and I are in the initial stages of an experimental philosophy project. We have started discussing this project on the EP blog, and I thought I might bring it up here as well.

Many of you may recall that when Joshua Knobe came to visit, he discussed his work on the effect of moral/evaluative considerations on determinations of intentionality. It turns out, according to these studies, that when people are confronted with analogous situations involving good vs. bad outcomes, they will attribute intentionality differently depending on their evaluation of the situation (the CEO story is the one referred to in both Knobe and Nadelhoffer's talks).

Jaworski and I are interested in discovering whether a similar effect occurs in determining the directness of an individual's causal relation with some event. Thus, we have the following cases:

CASE I

Bill and Ted are playing catch with a football in their backyard. In spite of being warned off many times by their mother, they are playing very close to the house. Bill throws a long pass, it sails over Ted's head and flies through the kitchen window, hitting a door, which swings open, before the ball rolls to a stop.

CASE II

Bill and Ted are playing catch with a football in their backyard. In spite of being warned off many times by their mother, they are playing very close to the house. Bill throws a long pass, it sails over Ted's head and flies through the kitchen window, hitting a plate on the wall, which falls and breaks, before the ball rolls to a stop.

What we wish to test is whether people will say, in the first case, that Bill caused the door to open (rather than that he opened the door) whereas in the second case, it will be said that Bill broke the plate (rather than that he simply caused the plate to break). If this is the case, it seems that evaluative concerns may have come into play in determining the directness of Bill's causal relation to the door opening vs. the plate breaking.

Anyway, Jaworski and I will be running this experiment at some point in the near future, and are looking for feedback, both general and in terms of the wording/content of these cases.

5 comments:

  1. True that breaking a plate seems very direct. But causing a door to open isn't necessarily indirect, it's just a little more ambiguous.

    rather than testing directness, it seems to me that a divergence in the cases would point to a difference in language usage when blame is involved. There may be a desire to use blame language when we see that someone has caused something bad to happen, but more relaxed language (though not necessarily indirect) when there is no need to blame.

    Is your test multiple choice? If so, can we see the wording of the answers?

    ReplyDelete
  2. Of course you can see the results and the wording of the questions (we'll probably post those up here before we run the experiment... just to make sure).

    You might be right about why we use certain words. That's a possibility anyways.

    ReplyDelete
  3. Something I'd recommend a little attention for is the difference between responsibility and causality. The former is something that we often assign on the basis of a looser sense of causality than what I think you're after in the latter term. Also, and this may be something Knobe already talked about (I wasn't there), I'd like to hear what the distinction between "Bill breaks x" vs. "Bill caused the x to break" amounts to, because I don't see it yet.

    ReplyDelete
  4. Ben: The more direct "broke" versus "caused to break" might be a way in which ordinary people insert normative responsibility into a causal chain. We predict that, when people believe someone is blameworthy for some state of affairs, they will use more direct language than when they believe that someone is not blameworthy, or are unsure.

    ReplyDelete
  5. I would have liked to see the Knobe talk. I have always thought if intention --> moral evaluation. Does he say that you look at the outcome and then assign some intent to the agent?

    That just seems weird. I would blame the kid for breaking the window because he had been warned not to play ball near the house, but I don't think I would blame him more for breaking the plate. I would also blame him more if I thought that he wanted to break the window and plate than I would if I thought he did it accidently.

    ReplyDelete