I mean, we are living in two different realities now, and this really hasn't always been the case.

RIP Robert Redford, a Hollywood icon and GIF legend.

Happy Pythagorean triple day (9/16/25) y'all's

cultural evolution folks pretty much moved past "the Question on California & Paris," but it is still a valuable lens to think about sociology's core problems. here are some of my reflections on transmission & directionality. Thinking About California and Paris tkeskinturk.github.io/blog/attactors

editors: if you continue forcing us to add stupid stars to regression tables, we are *not* going to be able to resolve problems stemming from stupid stars in regression tables. thank you, thank you.

Eleanor Marx when someone says Byron is better than Shelley

Andrei Tarkovsky's Tropic Thunder.

casual causal language

a take: it should be perfectly acceptable for an author to use causal language when doing an observational research even if they simply condition on the observables. writing weasel words as lip-service is only making things worse and harming the entire conversation.

the idea of being on "the left" or "the right" has clear implications for many of us, particularly for predicting what people think about economic or cultural issues. but does this "alignment" show up across many countries? not really. a new blog post here: tkeskinturk.github.io/blog/leftright

Basically every regression to the mean since the 2010s, no matter how small, whether from young men or black people or latinos or whatever, is treated as a realignment

From IBM’s christening of the ordinateur to today’s platforms, our essay of the week highlights how datafied feedback loops sort, score, and steer people into an "ordinal society" that fundamentally alters our social life. By Marion Fourcade & @kjhealy.co in @aeon.co buff.ly/g9wd9Sx

Blueskyism, or, the cultural logic of late capitalism

a new working paper: osf.io/vsr5b I propose a three-stage model of cohortization where dynamics of cohort learning and political sorting serve as complementary engines of aggregate political change. I apply this to the case of the killing of George Floyd & the BLM. it's also my job market paper!

good research statement I think

Man

What a fragedy

not much of a sequence anymore; it's tragedy & farce at the same time:

there is in fact a pipeline issue in positivism: those deemed "positivists" by the intelligentsia are often poorly-read number crunchers while true positivists---by definition, deeply embedded in the humanities---are less likely to take such roles, creating a false image of political sorting.

one awkward piece of language in sociology papers is the frequent "I use such and such theory to understand such and such social outcome," which often reminds me of what Barry Barnes says here:

Wake up babe they’re gamifying a political campaign

hopping in to find the new Jean Jaurès---only to find the left's assembly line of David Brookses.

pretty much agreed with the spirit of this, which simply argues against short-term optimizing of politics based on polling. the truth is just-so stories on tabulated data is the modus operandi of popularist research, and a slightly false-guided humanistic correction of polling may work much better.

Whoops, I wrote an essay that introduces a new concept 🫣

Climate models since the 1970s nailed it—most predicted global warming almost exactly as it happened.

possibly one of the most exciting theory papers in recent times, and just perfect that it's out there in the best sociology journal (imho). from @acastroaraujo.bsky.social & @nicolasrestrepo.bsky.social:

In econ, caring about causal identification is centrist-neoliberal coded. In poli sci, it's conservative coded. On twitter, it's left coded.

a new working (i.e., for now, an idle) paper. I ask whether "dynamic constraint" may emerge when both credible information & cross-issue cues target political beliefs--lowering the cost of consistency-seeking memory search. I field an experiment and find that the answer is no: osf.io/yh2fm

research papers now too often (a) collect highly complex data, (b) engage in heavy-handed pre-processing to measure constructs, and (c) pile on multiple layers of statistical estimation. as a rule of thumb, I think a paper is much clearer when at most one of these is likely to raise eyebrows imho.

Another micro finding from my dive into the field of Philosophy: 20% of Princeton’s faculty are academic inbreds, way ahead of the rest

watching everyone going to the ASA from my room:

time to normalize a culture of "working papers" in soc. see why:

Pleased to share the latest version of my paper with Arthur Spirling and @lexipalmer.bsky.social on replication using LMs We show: 1. current applications of LMs in political science research *don't* meet basic standards of reproducibility...

I got inspired by a recent debate in the ASR on "religious habitus," so I'd like to call all fellow cultural sociologists to cancel the word habitus. Let’s Cancel Habitus Before It Explains Something Else tkeskinturk.github.io/blog/habitus

Too bad there won’t be a follow-up survey in a few years that asks about divorce.

apparently, the ANES conducted a short survey in 2024 of cohabiting spouses/partners of its sample, allowing us to observe the ideologies of both the focal individuals and their spouses/partners. naturally, that provides a wonderful excuse for procrastination:

contemporary semi-literary novels love to be like "here's 300 pages of how a protagonist heals from the trauma of what happened before the novel started, the language will be the prose equivalent of tap water, and all events will drop out of space onto the passive protagonist"

it's 2025, and the fact that sociology reviewers still want even more Bourdieu in an exponential citation trap implies that, in the not-so-distant future, sociology articles will be self-referential documents of the simple word "Bourdieu," with various years interspersed here and there.

ah, my nemesis, we meet again...

This ranks high among the favorite things I've managed to write.

to work at vogue in the 90s they gave you this list of concepts you had to identify on the spot to prove you're cultured... i recognized at least half, but i only ticked the ones i feel confident i could actually explain the significance of

Tolkien’s drawing of The Lonely Mountain.

most of one's intellectual socialization to sociology happens through textbooks, syllabi, or blogs that feature substantial divisions and debates between groups, and let me tell you it's *shocking* how few debates or disagreements there are in contemporary sociology.

certain branches of computational social sciences, the rise of LLM technologies, and new behavioral research clusters propel or intensify a new form of behaviorism among social scientists.