Kurt Gray
kurtjgray.bsky.social
Kurt Gray
@kurtjgray.bsky.social
Prof at UNC, studying morality, religion, AI.
Director: Deepest Beliefs Lab; Center for the Science of Moral Understanding.
Author: The Mind Club; Substack; forthcoming book
Take home: "If you want to understand why your father-in-law votes differently from you, or why a foreign-exchange student emphasizes different values, your best bet is to understand their perceptions of harm."

with @sampratt99!
March 18, 2025 at 6:57 PM
Taxonomies: there are lots of different ways to count moral concerns. We cover taxonomies of 3, 4, 5, 6, 7. But distinctness between these themes exists more in the minds of researchers than in the minds of the people we study.
March 18, 2025 at 6:52 PM
Cognition: Moral pluralism is obviously true, but our mind uses a harm-based template to make more judgments, explaining why perceived harm almost perfectly predicts condemnation across different acts.
March 18, 2025 at 6:47 PM
Politics: Political disagreement is connected to assumptions of vulnerability, a new (under review) idea from our lab. Libs vs. con see different people as especially vulnerable to victimization.
March 18, 2025 at 6:41 PM
Culture: We draw from Shweder's classic work on "ontologies of suffering" how different cultures understand what causes harm. If you think defiling sacredness causes suffering, you think it's immoral.
March 18, 2025 at 6:36 PM
In this new review, we explore the idea of moral pluralism, and how it can arise from psychological (and cultural) concerns about suffering. Moral differences arise when different people make different assumptions about harm.

fulltext: https://tinyurl.com/moralityinourmind
March 18, 2025 at 6:31 PM
While creating HLLMs (historical LLMs) can be challenging, we believe they provide a novel research opportunity to overcome the present-focus of current work and give voice to societies of the past.

Authors: Michael Varnum, Nicolas Baumard, & @MohammadAtari90
March 18, 2025 at 6:36 PM
LLMs have successfully replicated human responses (like moral judgments & cognitive biases) in present participants.

Training them using historical texts would allow researchers to examine non-WEIRD populations, cultural change, & generalizability of psych phenomena.

2/3
March 18, 2025 at 6:31 PM
Big thanks to amazing lab managers/RAs @blakey_will & @sampratt99 for all their help with the manuscript and throughout the process.
Orange/harms figure credit: Kevin House
Amazon pre-order: https://a.co/d/bF9bQJf
Amazon.com
a.co
March 18, 2025 at 6:52 PM
It won't come out until January--after the election, but just in time for the inauguration. Hopefully, by then, people will be more interested in understanding each other than winning. But I'll be covering its contents in the substack.
March 18, 2025 at 6:47 PM
March 18, 2025 at 6:41 PM
I'm really pleased that it's (mostly) done. The proposal "the victim within" was rejected by everyone, until Edward Kastenmeier at @PantheonBooks saw its promise. He had me (repeatedly) reframe and rewrite it into something much better.
March 18, 2025 at 6:36 PM
Liberals and conservatives seem to have different morals, but I argue that we all share a harm-based moral mind. Our evolutionary past makes us worry about harm, but people today disagree about which harms are most important/real, creating moral outrage and political disagreement
March 18, 2025 at 6:31 PM
March 18, 2025 at 6:31 PM