Causation as Bias (sort of)

David Hume called cau­sa­tion the “ce­ment of the uni­verse”, and he was con­vinced that psy­cholog­i­cally and in our prac­tices, we can’t do with­out it.

Yet he was fa­mously scep­ti­cal of any at­tempt to an­a­lyze cau­sa­tion in terms of nec­es­sary con­nec­tions. For him, cau­sa­tion can only be defined in terms of a con­stant con­junc­tion in space and time, and that is, I would add, no cau­sa­tion at all, but cor­re­la­tion. For ev­ery two events that seem causally con­nected can also, and with­out loss of the phe­nomenon, be de­scribed as just the first event, fol­lowed by the sec­ond. It’s re­ally “just one damn thing af­ter an­other”. It seems to me we still can­not, will not and need not make sense of the no­tion of cau­sa­tion (vir­tu­ally no progress has been made since Hume’s time).

There seems no need for an­other sort con­nec­tion be­sides the spa­tio-tem­po­ral one, nor do we per­ceive any. In philos­o­phy, a Hume world is a pos­si­ble world defined in this way. All the phe­nom­ena are the same, but no nec­es­sary con­nec­tions hold be­tween the sup­posed re­lata. Maybe one should best imag­ine such a world as a game of life-world, but with­out a fun­da­men­tal level gov­erned by laws and forces; or as a movie, made of frames that are not in­trin­si­cally con­nected to each other. So, how­ever strong the psy­cholog­i­cal forces that drive hu­mans to ac­cept fur­ther mys­te­ri­ous con­nec­tions: Shouldn’t we just stop wor­ry­ing and ac­cept liv­ing in a Hume world? Or are there ac­tual ar­gu­ments in favour of “real” cau­sa­tion?


Yes. There’s the prob­lem of or­der. What ac­counts for all the or­der in the world?It is re­mark­ably or­dered. If no spe­cial con­nec­tions hold be­tween events, why isn’t the world pure chaos? Or at least much more di­s­or­dered? When two billard balls col­lide, never does one turn into an pink elephant.To ex­plain this, men came up with laws of na­ture (self-sus­tained or en­forced by a higher be­ing).


So, there’s the para­dox: On the one hand, we have to pos­tu­late spe­cial con­nec­tions to ac­count for an or­derly world like ours; on the other, we can­not give a proper ac­count of these con­nec­tions.

In­fla­tion­ary cos­mol­ogy to the res­cue.

I won’t go into the de­tails (but see the non­tech­ni­cal ex­pla­na­tion and some fur­ther philo­soph­i­cal im­pli­ca­tions here).

Suffice it to say that

1) in­fla­tion­ary cos­mol­ogy is main­stream physics, and

2) it pos­tu­lates a spa­tially in­finite uni­verse in which ev­ery event with nonzero prob­a­bil­ity is re­al­ized in­finitely many times.

How does this help to solve our para­dox? The solu­tion seems straight­for­ward:

In an in­finite uni­verse of the right kind, or­der can lo­cally emerge out of ran­dom events. Our uni­verse is of the right kind.

So, we can ac­count for the or­der in our ob­served (lo­cal) part of the uni­verse.

Ran­dom events just hap­pen, one af­ter an­other, there is no need for mys­te­ri­ous causal con­nec­tions. We throw them out but keep the or­der.

Prob­lem solved.


Q: But if this is true, it’s the end of the world. Think­ing, ac­tion, sci­ence, bi­ases and many, many more con­cepts are causal ones. How can we do with­out them?

A: life is hard, get over it.

Q: But the the­ory is untestable?!

A: Falsifi­ca­tion­ism is dead; we have other ev­i­dences in favour (see be­low).

Q: But isn’t the the­ory self-defeat­ing?

A: It is cer­tainly odd to have a the­ory in­formed by ex­pe­riences and high-level physics that tells us that, strictly speak­ing, there are no ex­pe­riences or sci­ences. But it doesn’t seem in­co­her­ent to me climb the lad­der and then throw it away.

And, look­ing at the bright side:

In ad­di­tion to be­ing non-mys­te­ri­ous and con­cep­tu­ally sparse, this might al­low to solve some ad­di­tional (would-be?) hard prob­lems:

qualia, clus­ter­ing of tropes, time travel-para­doxes, in­de­ter­minis­tic pro­cesses: All easy or triv­ial when a thouroughly in­de­ter­minis­tic uni­verse is con­sid­ered.

So. What do you think – if you can?