Ulatowski on Act Individuation

Imagine  a person who is pumping water and thereby poisoning the inhabitants of a nearby house.  Now consider the act of pumping water and the act of poisoning the inhabitants.  Are these two different actions, or are we just confronted here with two different descriptions of the very same action?

This highly abstract metaphysical issue was the topic of heated philosophical debate a number of decades ago.  Famously, Anscombe and Davidson argued that there was only one action under two different descriptions, while Goldman argued that there truly were two different actions. 

More recently, however, the issue has returned, this time in the form of experimental studies by Joe Ulatowski, and he has obtained some very surprising results. 

Apparently, people's intuitions about act individuation depend on the moral properties of the acts involved.  Thus, there seems to be an asymmetry such that people tend to adopt the Anscombe/Davidson view about morally bad actions while adopting the Goldman view about morally good actions. 

In Ulatowski's study, each subject was randomly assigned to one of two conditions.  Subjects in the 'poison condition' received the following vignette:

Smith’s job is to pump water into the cistern which supplies the water of a house.

 

One day Smith operates the pump and replenishes the house’s water-supply. The occupants of the house are healthy and have no health problems. Jones tells Smith that someone has found a way of systematically contaminating the water’s source with a deadly cumulative poison whose effects are unnoticeable until they can no longer be cured.

 

Smith says, “I don’t care about contaminating the water’s source; I just want to earn my pay.”

 

The occupants of the house drink the water. Sure enough, they are poisoned and die.

These subjects were then asked: 'Was Smith’s operating the pump the same thing as his poisoning the house’s inhabitants or were they distinct?'

Meanwhile, subjects in the 'savior condition' were given a slightly different vignette:

Smith’s job is to pump water into the cistern which supplies the water of a house.

 

One day Smith operates the pump and replenishes the house’s water-supply. The occupants of the house are sick and have severe infections. Jones tells Smith that someone has found a way of systematically purifying the water’s source with a cumulative antibiotic whose effects are unnoticeable until they cure someone who has a severe infection.

 

Smith says, “I don’t care about purifying the water’s source; I just want to earn my pay.”

 

The occupants of the house drink the water. Sure enough, they are saved and live.

These subjects were asked: 'Was Smith’s operating the pump the same thing as his saving the house’s inhabitants or were they distinct?'

The results showed a huge asymmetry.  A full 61% of subjects selected 'the same thing' in the poison condition, while only 13% selected this option in the savior condition. 

Ulatowski uses these results to argue that moral judgments are actually affecting people's intutions about the metaphysical problem of act individuation.  But if he is right here, it seems that we are still left with a further question.  That question is about why moral judgment would have that sort of impact. 

I must confess that I am completely baffled as to how to answer this question, and I was wondering if any of you might have any suggestions. 

[Ulatowski's paper is available for download at Ulatowski.doc]

Comments

[Cihan Baran]

Hey,

Isn't this result really a result of the Knobe effect?

Probably, the folk think that Smith intentionally pumps water. They also probably think that Smith does not intentionally purify the water, whereas Smith intentionally poisons the water.

Since purifying the water is not intentional, it is recognized as distinct from pumping the water, which is intentional.

I see the explanation in the Knobe effect.

Thanks.

[Josh Weisberg]

Is it the Knobe effect? Josh, it's your effect--what do you think? (Can we figure this out by experimentally polling all the Joshua Knobes? ;>) )

I think that event individuation is cheap--it follows after decisions about things people care about more: moral concerns of blame/praise, theoretical issues in philosophy/science, etc. In this case, subjects' intuitions are that Smith is to blame in the first case, and not in the second. So they choose the option that indicates this, without a thought to event individuation, qua individuation. By individuating as one event, they can blame Smith in the first case. But they do not feel he deserves praise in the second, so they individuate accordingly: two events.

Now why do they blame Smith in the first case? He could have prevented a horrible thing, and he did not. Plus, he was actively involved in the horrible thing--if he didn't act, the horrible thing would not have occurred. Once subjects' have determined that blame should be assigned, they individuate accordingly.

But perhaps I am missing what you're asking here, as this does not seem that surprising to me.

[Jussi Suikkanen]

This is interesting. I wonder if there is a significant difference in the questions. Poisoning seems to be ambiguous between the non-moral description *putting poison* in their water and the morally loaded description *killing* them. Saving, on the other hand, is not ambiguous in the same way. It is definitely the morally loaded opposite of killing and not the opposite of 'putting poison'.

So, I wonder if you would get more symmetrical results if you used either:
a) something like 'Was Smith’s operating the pump the same thing as his killing the house’s inhabitants or were they distinct?' as the question of the first vignette or
b) something like 'Was Smith’s operating the pump the same thing as his giving medicine to the house’s inhabitants or were they distinct?'
as the question of the second vignette.

[Justin Sytsma]

It strikes me that the two questions are not symmetric in a way distinct from moral concerns: Poisoning the water is temporally coextensive with operating the pump... but saving the house's inhabitants is not. This reading of "poisoning" (as opposed to reading it as "killing the inhabitants") is, I think, reasonable. If people follow the principle that the same events have the same temporal extents, then the responses seen make sense to me.

I think the way to remove this confound is to follow something like Jussi's (a) and (b). I would like to see two conditions, one where both "events" are temporally coextensive:

"Was Smith's operating the pump the same thing as his poisoning the house's water?"

"Was Smith's operating the pump the same thing as his decontaminating the house's water?"

And another where they are not:

"Was Smith's operating the pump the same thing as his killing the house's inhabitants?"

"Was Smith's operating the pump the same thing as his saving the house's inhabitants?"

[stephen]

OK, Josh W, but if he deserves blame in the first case why doesn't he deserve praise in the second? The asymmetry in act individuation may well reflect an asymmetry in ascription of praise and blame in these cases: but that raises the question, why the latter asymmetry?

I'm not sure, but I think it's got something to do with the counterfactual in each case. In case 1 the guy doesn't have the *excuse* that 'pumping the water is different from poisoning the family' because he would have pumped the water whether or not it was poisoned (or purified, for that matter). In case 2 he can't *take the credit* for curing the family because he would have pumped the water whether or not it was purified (or poisoned). Having an excuse for a bad action is symmetrical with taking the credit for a good one. Perhaps respondents are anticipating a kind of excuse commonly given in such cases ("I wasn't doing X. I was only doing Y"), by point out that X = Y. And they are anticipating a kind of falsely claimed credit in the other case ("I'm good, I just did Y"), by pointing out that X /= Y.

But there I've just restated the asymmetry, not explained it...

I don't think it's an easy problem.

We get the blame for reckless harm, but not the credit for unthinking or uncaring good. ('Reckless good' is not even a familiar concept.) We have a moral duty to not to inflict undeserved harm, but no moral duty to promote general good. So we have a duty to take care not to harm, but no duty to take care that we promote good. So we get blamed for failing in our duty to take care not to harm, but if we fail to take any notice whether we promote the general good then we deserve neither praise nor blame. In case 1 the guy failed to take care not to harm the family. In case 2 he merely didn't care whether he did them any good. Hence the asymmetry in praise and blame. Then by Josh W's argument, the asymmetry in act individuation follows.

Maybe that works. Maybe not.

[N. Patterson]

What if we injected some new facts into the scenario where Smith poisons the occupants. Here, noone has told Smith or Jones that he occupants who are at risk of posioning are the very same people who contaminated the water.

Do you think we would see the same results in this test if we included a scenario where the victims are not blameless? Better yet, would it matter?

Maybe people are not really "giving their intutions about the metaphysical problem of act individuation", but are subconsciously identifying with the victims and responding on an emotional level? IOW, they are not really giving the type of response they believe and/or claim they are giving. Did anyone ask the respondants to explain why their assignment of moral culpability were affected as so? How would the average person explain the double-standard when challenged? Whatever the reasons, these results are interesing.

[N. Patterson]

I just realized I completely overlooked the question as to why the results are asymetrical. Sorry for the double-post. I thought I should at least try to address actual question put to us. Perhaps people are identifying with the victim, and would hope that others would actively want to aid them and or refrain from them harm. As so, they are equally peeved at Mr Smith, as in either case, he doesn't seen to care at all about their welfare.