Physical linguistics

This is re­ally my at­tempt at ap­proach­ing elimi­na­tive ma­te­ri­al­ism, and prob­a­bly read­ing Paul Church­land or Daniel Den­nett’s pa­pers would be bet­ter for you to get the point. I’m just writ­ing to or­ga­nize my thoughts.


There are three big prob­lems in sci­ence: uni­verse, life, and con­scious­ness. There is a good the­ory of the uni­verse on the macro and micro scale, and the prob­lem of its ori­gin. They are not the fi­nal word, but we have a good sense of any fu­ture up­dated the­o­ries would be like: mechanis­tic, math­e­mat­i­cal, prob­a­bly us­ing real, com­plex, and dis­crete num­bers.

A the­ory of life is still in the works, though there are en­courag­ing at­tempts. The phys­i­cal con­struc­tion of life and the de­scrip­tive the­ory of life is now com­plete ex­cept in the de­tails. We know that it would be some­thing made of evolu­tion, ther­mo­dy­nam­ics, chem­istry, and of course, math­e­mat­ics. The en­g­ineer­ing the­ory of life is still greatly miss­ing. We do not know how to cre­ate life, at most we can fork the ge­netic code and do lit­tle mod­ifi­ca­tions and mix­ings. We don’t even know if a robot is al­ive.

A the­ory of con­scious­ness is in an even ear­lier stage. There are some ba­sic stud­ies of the de­scrip­tion of con­scious­ness, and there are dozens of hazy philo­soph­i­cal the­o­ries that needs to be made quan­ti­ta­tive us­ing fu­ture data.


One prob­lem with con­scious­ness is its para­dox­i­cal qual­ities, cre­at­ing ques­tions that seem to both be com­pel­ling and de­formed:

  • “Why am I me in­stead of some­one else?”

  • “If Pinkie is copied, which one is the real Pinkie?”

  • “Is the feel­ing of blue same for ev­ery­one?”

  • “How does one freely choose?”

Now com­pare them with analo­gous ques­tions from uni­verse and life:

  • “Why is this rock this rock in­stead of that rock?”

  • “If this book is copied, which one is the real book?”

  • “Is this web­site the same web­site on ev­ery com­puter?”

  • “How does a slime mold de­cide which way to go?”

The analo­gous ques­tions lose their mys­tery and be­comes mun­dane, con­fused, or fas­ci­nat­ing but also sci­en­tifi­cally an­a­lyz­able.

Pos­si­bly the prob­lem is with the un­der­stand­ing of con­scious it­self, which is too con­fused. I pro­pose to re­move con­scious­ness from ex­pla­na­tions of life be­hav­ior (hu­man or not) as much as pos­si­ble. If it can be fully re­moved, then the prob­lem of con­scious­ness is solved. If it can’t be fully re­moved, then it con­cen­trates the effort for solu­tion.

A sketch

As a sketch of how such a re­moval might be done, con­sider a fully phys­i­cal ex­pla­na­tion of how hu­mans talk, which is cur­rently in­fested with con­scious­ness. The stan­dard ac­count is that there is a con­scious­ness that feels some­thing, then for­mu­lates that into words and sen­tences, then ex­presses them. Un­con­scious speak­ing is con­sid­ered non­sense, mean­ingless, noise. This doesn’t have to be.

The Heptapods

The Hep­tapods from Story of Your Life (Ted Chi­ang, 1998) are an ex­am­ple of a “free-will-free” form of life. Their lan­guage has de­ter­minism baked into it, just as hu­man lan­guage has free will baked into it.

What kind of uni­verse could pro­duce two kinds of life such that one is de­ter­minis­tic in lan­guage, but the other is free in lan­guage? And in such prox­im­ity too, such that they can ac­tu­ally meet each other and share the same phys­i­cal space and phys­i­cal laws?

To an­swer such ques­tions in a physics way, one would use a physics of lan­guage. What is a lan­guage ac­cord­ing to a physi­cist?

Phys­i­cal linguistics

What is a de­ter­minis­tic lan­guage, and what is a free-will lan­guage? How would a de­scrip­tion of free-will emerge in a de­ter­minis­tic sys­tem such as our uni­verse? And most im­por­tantly, how does a uni­ver­sal lan­guage, a sym­bolic sys­tem that can mod­els the phys­i­cal world that it is in, emerge in a de­ter­minis­tic world?

This is analo­gous to the prob­lem of zom­bie lan­guage: I once read that in a world with only philo­soph­i­cal zom­bie hu­mans, hu­man lan­guages would prob­a­bly have not evolved to talk about con­scious­ness and in­ner ex­pe­riences be­cause there is no such nonex­is­tent thing. This ar­gu­ment is dumb, since hu­man lan­guages already talk about many nonex­is­tent things, but it points at an in­ter­est­ing ques­tion: how would a de­ter­minis­tic sys­tem evolve a lan­guage that talks about things hap­pen­ing in it?

Phys­i­cal self-refer­en­tial science

In the same spirit, what kind of de­ter­minis­tic uni­verse would have lit­tle bun­dles of mat­ter in­side of it that be­haves roughly the same as some other patches of this uni­verse? We call these lit­tle bun­dles of mat­ter “com­put­ers run­ning phys­i­cal simu­la­tions”, or “a hu­man brain think­ing about sci­ence”, or maybe even “a lion brain think­ing about which way an an­telope is prob­a­bly go­ing to go next”.

If such ex­pla­na­tions can be done in de­tail, that would be a self-refer­ence in physics: a phys­i­cal sys­tem (our uni­verse) con­tain­ing a sub­stan­tial de­scrip­tion of it­self (the ex­pla­na­tion), as well as an ac­count for why it is likely for the de­scrip­tion to ex­ist in the first place (the ex­pla­na­tion about why a phys­i­cal world is likely to con­tain its own de­scrip­tion).