Dangers of steelmanning /​ principle of charity

As far as I can tell, most peo­ple around these parts con­sider the prin­ci­ple of char­ity and its su­per saiyan form, steel­man­ning, to be Very Good Ra­tion­al­ist Virtues. I ba­si­cally agree and I in fact op­er­ate un­der these prin­ci­ples more or less au­to­mat­i­cally now. HOWEVER, no mat­ter how good the rule is, there are always ex­cep­tions, which I have found my­self in­creas­ingly con­cerned about.

This blog post that I found in the re­sponses to Yvain’s anti-re­ac­tionary FAQ ar­gues that even though the an­cient Ro­mans had welfare, this policy was mo­ti­vated not for con­cern for the poor or for a de­sire for equal­ity like our mod­ern welfare poli­cies, but in­stead “the Ro­man dole was wrapped up in dis­courses about a) the might and wealth of Rome and b) god­dess wor­ship… The dole was there be­cause it made the em­peror more pop­u­lar and demon­strated the wealth of Rome to the peo­ple. What’s more, the dole was per­son­ified as An­nona, a god­dess to be wor­shiped and thanked.”

So let’s as­sume this guy is right, and imag­ine that an an­cient Ro­man trav­els through time to the pre­sent day. He reads an ar­ti­cle by some pro­gres­sive ar­gu­ing (us­ing the ra­tio­nale one would typ­i­cally use) that Obama should in­crease un­em­ploy­ment benefits. “This makes no sense,” the Ro­man thinks to him­self. “Why would you give money to some­one who doesn’t work for it? Why would you re­ward lack of virtue? Also, what’s this about equal­ity? Isn’t it right that an up­per class ex­ists to rule over a lower class?” Etc.

But for­tu­nately, be­tween when he hopped out of the time ma­chine and when he found this ar­ti­cle, a ra­tio­nal­ist found him and ex­plained to him steel­man­ning and the prin­ci­ple of char­ity. “Ah, yes,” he thinks. “Now I re­mem­ber what the ra­tio­nal­ist said. I was not be­ing so char­i­ta­ble. I now re­al­ize that this po­si­tion kind of makes sense, if you read be­tween the lines. Giv­ing more un­em­ploy­ment benefits would, now that I think about it, demon­strate the power of Amer­ica to the peo­ple, and cer­tainly An­nona would ap­prove. I don’t know why who­ever wrote this ar­ti­cle didn’t just come out and say that, though. Maybe they were con­fused”.

Hope­fully you can see what I’m get­ting at. When you reg­u­larly use the prin­ci­ple of char­ity and steel­man­ning, you run the risk of:

1. Stick­ing rigidly to a cer­tain wor­ld­view/​paradigm/​es­tab­lished be­lief set, even as you find your­self will­ing to con­sider more and more con­crete propo­si­tions. The Ro­man would have done bet­ter to re­ally read what the mod­ern pro­gres­sive’s logic was, think about it, and try to see where he was com­ing from than to au­to­mat­i­cally filter it through his own wor­ld­view. If he con­sis­tently does this he will never find him­self con­sid­er­ing al­ter­na­tive ways of see­ing the world that might be bet­ter.

2. Falsely de­vel­op­ing the sense that your wor­ld­view/​paradigm/​es­tab­lished be­lief set is more pop­u­lar than it is. Pretty much no one to­day holds the same val­ues that an an­cient Ro­man does, but if the Ro­man goes around be­ing char­i­ta­ble all the time then he will prob­a­bly see his own be­liefs re­flected back at him a fair amount.

3. Tak­ing ar­gu­ments more se­ri­ously than you pos­si­bly should. I feel like I see all the time on ra­tio­nal­ist com­mu­ni­ties peo­ple say stuff like “this ar­gu­ment by A sort of makes sense, you just need to frame it in ob­jec­tive, con­se­quen­tial­ist terms like blah blah blah blah blah” and then fol­low with what looks to me like a com­pletely origi­nal thought that I’ve never seen be­fore. But why didn’t A just frame her ar­gu­ment in ob­jec­tive, con­se­quen­tial­ist terms? Do we as­sume that what she wrote was sort of a tele­phone-game ap­prox­i­ma­tion of what was origi­nally a highly log­i­cal con­se­quen­tial­ist ar­gu­ment? If so where can I find that ar­gu­ment? And if not, why are we as­sum­ing that A is a crypto-con­se­quen­tial­ist when she prob­a­bly isn’t? And if we’re sure that ob­jec­tive, con­se­quen­tial­ist logic is The Way To Go, then shouldn’t we be very skep­ti­cal of ar­gu­ments that seem like their ba­sis is in some other rea­son­ing sys­tem en­tirely?

4. Just hav­ing a poor model of peo­ple’s be­liefs in gen­eral, which could lead to prob­lems.

Hope­fully this made sense, and I’m sorry if this is some­thing that’s been pointed out be­fore.