Request for Steelman: Non-correspondence concepts of truth

A cou­ple of days ago, Buy­buy­dan­davis wrote the fol­low­ing on Less Wrong:

I’m in­creas­ingly of the opinion that truth as cor­re­spon­dence to re­al­ity is a minor­ity ori­en­ta­tion.

I’ve spent a lot of en­ergy over the last cou­ple of days try­ing to come to terms with the im­pli­ca­tions of this sen­tence. While it cer­tainly cor­re­sponds with my own ob­ser­va­tions about many peo­ple, the thought that most hu­mans sim­ply re­ject cor­re­spon­dence to re­al­ity as the crite­rion for truth seems al­most too out­ra­geous to take se­ri­ously. If upon fur­ther re­flec­tion I end up truly be­liev­ing this, it seems that it would be im­pos­si­ble for me to have a dis­cus­sion about the na­ture of re­al­ity with the great ma­jor­ity of the hu­man race. In other words, if I truly be­lieved this, I would la­bel most peo­ple as be­ing too stupid to have a real dis­cus­sion with.

How­ever, this re­ac­tion seems like an in­stance of a failure mode de­scribed by Me­gan McAr­dle:

I’m always fas­ci­nated by the num­ber of peo­ple who proudly build columns, tweets, blog posts or Face­book posts around the same core state­ment: “I don’t un­der­stand how any­one could (op­pose le­gal abor­tion/​sup­port a car­bon tax/​sym­pa­thize with the Pales­ti­ni­ans over the Is­raelis/​want to pri­va­tize So­cial Se­cu­rity/​in­sert your pet is­sue here).” It’s such an in­ter­est­ing state­ment, be­cause it has three lay­ers of mean­ing.

The first layer is the literal mean­ing of the words: I lack the knowl­edge and un­der­stand­ing to figure this out. But the sec­ond, in­tended mean­ing is the op­po­site: I am such a su­pe­rior moral be­ing that I can­not even imag­ine the cog­ni­tive er­rors or moral turpitude that could lead some­one to such ob­vi­ously wrong con­clu­sions. And yet, the third, true mean­ing is ac­tu­ally more like the first: I lack the em­pa­thy, moral imag­i­na­tion or an­a­lyt­i­cal skills to at­tempt even a ba­sic un­der­stand­ing of the peo­ple who dis­agree with me

In short, “I’m stupid.” Some­thing that few peo­ple would ever post so starkly on their Face­book feeds.

With this back­ground, it seems im­por­tant to im­prove my model of peo­ple who re­ject cor­re­spon­dence as the crite­rion for truth. The ob­vi­ous first place to look is in aca­demic philos­o­phy. The pri­mary challenger to cor­re­spon­dence the­ory is called “co­her­ence the­ory”. If I un­der­stand cor­rectly, co­her­ence the­ory says that a state­ment is true iff it is log­i­cally con­sis­tent with “some speci­fied set of sen­tences”

Co­her­ence is ob­vi­ously an im­por­tant con­cept, which has valuable uses for ex­am­ple in for­mal sys­tems. It does not cap­ture my idea of what the word “truth” means, but that is purely a se­man­tics is­sue. I would be will­ing to cede the word “truth” to the co­her­ence camp if we agreed on a sep­a­rate word we could use to mean “cor­re­spon­dence to re­al­ity”. How­ever, my in­tu­ition is that they wouldn’t let us to get away with this. I sense that there are peo­ple out there who gen­uinely ob­ject to the very idea of dis­cussing whether a sen­tences cor­re­spond to re­al­ity.

So it seems I have a cou­ple of op­tions:

1. I can look for em­piri­cal ev­i­dence that buy­buy­dan­davis is wrong, ie that most peo­ple ac­cept cor­re­spon­dence to re­al­ity as the crite­rion for truth

2. I can try to con­vince peo­ple to use some other word for cor­re­spon­dence to re­al­ity, so they have the nec­es­sary se­man­tic ma­chin­ery to have a real dis­cus­sion about what re­al­ity is like

3. I can ac­cept that most peo­ple are un­able to have a dis­cus­sion about the na­ture of reality

4. I can at­tempt to steel­man the po­si­tion that truth is some­thing other than correspondence

Op­tion 1 ap­pears un­likely to be true. Op­tion 2 seems un­likely to work. Op­tion 3 seems very unattrac­tive, be­cause it would be very un­com­fortable to have dis­cus­sions that on the sur­face ap­pear to be about the na­ture of re­al­ity, but which re­ally are about some­thing else, where the pre­cise value of “some­thing else” is un­known to me.

I would there­fore be very in­ter­ested in a steel­man of non-cor­re­spon­dence con­cepts of truth. I think it would be im­por­tant not only for me, but also for the ra­tio­nal­ist com­mu­nity as a group, to get a more ac­cu­rate model of how non-ra­tio­nal­ists think about “truth”