this field is required

In response to my post on how not to think about cut­ting in line, Jacob Levy makes these impor­tant points:

I’m puz­zled. This seems like the eas­i­est of cases for a Kant­ian. I can­not will as a uni­ver­sal maxim that the cut­ter be allowed into line, because that would destroy the coher­ence of lin­ing up. It’s almost as clean a case as lying: the wrong­ness lies in the self-contradictory char­ac­ter. You can­not simul­ta­ne­ously will queues and cutting.

Or, to put it dif­fer­ently: the cut­ter him or her­self treats oth­ers as not-ends-in-themselves. Their time is not as valu­able or impor­tant as his/her time. The one per­son who uni­lat­er­ally, with­out the con­sent of every other per­son behind him or her, allows the cut­ter in thus *also* fails to treat those behind him/her as ends-in-themselves.

So basi­cally Jacob has sug­gested that I have mis­ap­plied Kant’s the­ory to our line cut­ting case, and that either form of the Cat­e­gor­i­cal Imper­a­tive can dis­pense with the prob­lem suc­cess­fully when prop­erly applied.

Let’s take CI1 first, the uni­ver­sal­iza­tion one. The first prob­lem with it has to do with maxim for­mu­la­tion. We have to fig­ure out a way of describ­ing the action we’re think­ing of tak­ing in order to give it the uni­ver­sal­iza­tion test. One good way of think­ing about max­ims is that they have an ACE form: I will do A(ction) in these C(ircumstances) for this E(nd). The trou­ble with max­ims is that any given action can be described using a vari­ety of dif­fer­ent max­ims. For instance, imag­ine that some­one is think­ing of rob­bing a store for some life­sav­ing med­i­cine. If you test the maxim “I will rob a store when I am broke to pro­mote my self-interest” for uni­ver­sal­iz­abil­ity, it fails, mak­ing the action morally imper­mis­si­ble. But if you test “I will rob a store when I am dying of can­cer in order to save my life,” that seems to pass the test, mak­ing the action morally per­mis­si­ble. But it’s really bad if Kan­tian­ism ren­ders con­flict­ing deon­tic ver­dicts on what is actu­ally just one action. And, in our queue­ing case, it’s not clear what maxim we ought to use, and whether using dif­fer­ent max­ims will gen­er­ate con­flict­ing ver­dicts as in the rob­bery case.

But Kant tends to give exam­ples with the broad­est maxim pos­si­ble, so to be char­i­ta­ble, let’s go with that. The broad­est pos­si­ble maxim for line cut­ting is prob­a­bly “I will let some­one into the line when they ask me, in order to help them out.” Jacob has sug­gested that “I can­not will as a uni­ver­sal maxim that the cut­ter be allowed into line, because that would destroy the coher­ence of lin­ing up,” mean­ing that let­ting cut­ters into the line is morally imper­mis­si­ble. Case closed?

Let’s back up a bit. Here’s what I wrote orig­i­nally: “it seems that mak­ing into uni­ver­sal law a rule allow­ing peo­ple to cut in line is not log­i­cally impos­si­ble nor cat­a­strophic in prac­tice.” There is way more packed into that claim than meets the eye. I was actu­ally say­ing that a maxim allow­ing cut­ters into line passes both parts of Kant’s uni­ver­sal­iza­tion test: first, that the world where peo­ple fol­low that maxim not be log­i­cally con­tra­dic­tory, and sec­ond that it not involve any con­tra­dic­tion in the will. Things get really sticky here, because it’s hard to argue about the qual­i­ties of coun­ter­fac­tual worlds, and Kant is pretty sketchy about describ­ing the weaker con­di­tion that max­ims not involve con­tra­dic­tions in the will. But, I’m happy to assume for the sake of argu­ment that I was wrong and Jacob is right and that, on Kan­tian­ism, the broad­est pos­si­ble maxim regard­ing let­ting peo­ple into a line is inco­her­ent, or log­i­cally impos­si­ble in practice.

Unfor­tu­nately, Kan­tian­ism still comes out look­ing bad. Why? Because an absolute pro­hi­bi­tion on let­ting some­one into a line is crazy. The line cut­ting case closely tracks a com­mon crit­i­cism of Kant (the ori­gins of which are unknown to me), the “mur­derer at the door” objec­tion. Here’s the thought exper­i­ment: you’re sit­ting at home, and a Jew­ish per­son comes to the door, says the police are after him, can he hide in your attic? You say sure, and no sooner than he hides than do the police come to the door and inquire as to his where­abouts. The com­mon­sense moral intu­ition is that you should lie, since the sit­u­a­tion is very high stakes and you can help save the man from cer­tain death at the hands of the police. But Kant thinks that max­ims per­mit­ting lying to achieve one’s objec­tives are log­i­cally impos­si­ble for every­one to fol­low, because in such a world mean­ing­ful com­mu­ni­ca­tion could not exist. So there is a per­fect duty to tell the truth and it is imper­mis­si­ble for you to lie, even to a mur­derer at the door, and (accord­ing to an objec­tor) that makes Kan­tian­ism implau­si­ble on its face. I think that the line cut­ting sit­u­a­tion is a close par­al­lel, and that any moral the­ory which for­bids let­ting any­one into a line even when the costs are very, very low or when the stakes are very, very high, is implau­si­ble. So, Kan­tian­ism has jumped out of the pot and into the fire, so to speak: get­ting over the maxim for­mu­la­tion prob­lem means that it is not inde­ter­mi­nate, but the murderer-at-the-door-type objec­tion shows that its deon­tic ver­dicts are sim­ply incorrect.

And, while I granted it for the sake of argu­ment, I don’t think that the line cut­ting maxim is inco­her­ent, I actu­ally think it vio­lates merely the weaker con­di­tion of being con­tra­dic­tory to will. That would mean that let­ting peo­ple into the line is an imper­fect duty (a sometimes-duty) instead of a per­fect duty (an always-duty). And, like I said in the orig­i­nal post, Kant doesn’t say any­thing too use­ful about when & where we should ful­fill our imper­fect duties. So that’s unsat­is­fy­ing, too.

Now, let’s look at Jacob’s pro­posal for how CI2 can han­dle the line case: that we view both the cut­ter and he who per­mits cut­ting to be fail­ing to treat oth­ers as ends-in-themselves. Each seems to be sub­ject­ing the peo­ple in the back of the line to a plan of action to which those peo­ple would not or could not con­sent (that might not be true, but it’s unim­por­tant). That means that cut­ting in line is imper­mis­si­ble, because it nec­es­sar­ily involves fail­ing to respect peo­ple in this way.

Here’s what I said pre­vi­ously about CI2:

I agree with Fred Feld­man that, while the moral intu­ition that using peo­ple is morally bad is surely cor­rect, it is not a par­tic­u­larly infor­ma­tive guide to action (”On Treat­ing Peo­ple as Ends in Them­selves: A Cri­tique of Kant”). The most plau­si­ble inter­pre­ta­tions of CI #2 read it as requir­ing that either one help oth­ers achieve their goals, or specif­i­cally their ratio­nal goals. But, pre­sum­ably, both a per­son who wants a place in line and the peo­ple behind you in line have goals, even ratio­nal ones, which con­flict. You can’t help both. So Kan­tian­ism doesn’t seem to gen­er­ate any kind of spe­cific answer to our question.

I’ll elab­o­rate a bit. “Respect” is great and all, but if respect­ing per­sons is to be our guide to action, we need a good way of fig­ur­ing out what respect amounts to, in prac­tice. The best under­stand­ing of “respect” on offer is that it con­sists in not merely refrain­ing from using peo­ple, but actu­ally tak­ing their own goals seri­ously and help­ing to pro­mote them at least some of the time. Also, note that Kant thinks that every sin­gle human per­son has infi­nite moral value. As such, Kant for­bids any attempts to make moral trade­offs, say, by harm­ing one per­son to the ben­e­fit of a larger group.

These con­sid­er­a­tions leave a Kant­ian very lit­tle to work with in the queue­ing case. On the one hand, she is sup­posed to respect every­one all of the time, and to man­i­fest this some­times by ful­fill­ing an imper­fect duty to take up oth­ers’ (ratio­nal) goals as her own. On the other hand, she may not decide whose goals to pro­mote by assign­ing greater moral value to the group of line-waiters who will be incon­ve­nienced by let­ting some­one in, or by assign­ing a lesser moral value to the sin­gle line-cutter, because every­one con­cerned has infi­nite moral value, accord­ing to Kant. What the heck is she to do? I have no idea. So, while Kan­tian­ism isn’t nec­es­sar­ily false,  it is not ter­ri­bly use­ful (in the sense of being action-guiding) in the line case.

So, in con­clu­sion, I think Kan­tian­ism is great inso­far as it suc­cess­fully cap­tures two very impor­tant moral intu­itions: that moral rules need to be uni­ver­sal­iz­able, and that dis­re­spect­ing peo­ple by using them for one’s own ends is imper­mis­si­ble. But there are some seri­ous and, in my opin­ion and to my knowl­edge, unre­solved prob­lems with Kan­tian­ism as a stand­alone the­ory of nor­ma­tive ethics.  Jacob’s com­ments and the avail­able replies go to show that these two moral intu­itions by them­selves are per­haps nec­es­sary but not suf­fi­cient for think­ing about how to act, in the line case and in general.

• Book­mark on Delicious
• Digg this post
• Like on Facebook
• Buzz
• share via Reddit
• Share with Stumblers
• Tweet
• Sub­scribe to the com­ments on this post
• Email a friend