Juan Murillo Vargas

A paper with Vic Ferreira and Norvin Richards is now out

(1) Speakers syntactically encode zero complementizers as cognitively active mental object.

(2) No evidence LLMs capture cross constructional generalizations about null complementizers.

nam10.safelinks.protection.outlook.com?url=https%3A...

continuum of inductive potential from low (relatively minimal categories whose members are dissimilar) to high (coherent meaningful categories whose members are similar) above a cartoon child. an icon of a tiger appears under "high" inductive potential, with "closes eyes when happy" appearing as a feature of a tiger in a zoo, with an arrow pointing to the tiger icon, and a dashed arrow extending it to a tiger on a savanna. an icon of a pedestrian appears under "low" inductive potential, with "closes eyes when happy" appearing as a feature of a woman on a street, with Xs over arrows pointing to the pedestrian icon, and to a different pedestrian.

📣 new paper! people use some categories to generalize (e.g., we generalize something we learn about one tiger 🐯 to other tigers 🐅), but not others (e.g., we don't generalize from one pedestrian 🚶 to other pedestrians 🚶‍♂️). how do people learn what categories allow for generalization? 🧵

I have a lot of people to thank. But I want to publicly thank @thyacinth.bsky.social, @kennethblack.bsky.social, @matthiasmichel.bsky.social, and @justinkhoo.com for helping me at just about every stage of the paper. Any errors in the paper are in spite of all their help!

I end with some goodies about how liberalism opens up new avenues for thinking about concepts, the metaphysics of mental states, and non-propositional intentionality more generally.

This is already a long thread. But if you want to see 10k words about this stuff check out the pre-print! (7/7)

I argue liberalism best fits with what we know from natural language semantics, computational cognitive science, and developmental psych. I also argue it gives us a very clean solution to the similarity problem once we do away with the idea (myth?) that validity is about truth-preservation.
(6/7)

This gives us one solution to the similarity problem. "Um, uh," and their friends don't express propositions; so the conservative would predict they don't make into Lot.

This is not my view though. I defend "liberalism:" the view that LoT features expressions with all sorts of contents! (5/7)

The rest of the paper considers two answers to the similarity problem.

One is "conservatism." This is the cottagecore approach to LoT. On this view LoT is similar to natural language only insofar as it features expressions that (compose to yield expressions that) express propositions. (4/7)

This gives rise to what I call the "similarity problem." We need a story about what makes LoT similar *and* dissimlar to natural language.

These tasks pull in opposite directions. So it's not obvious what LoT's proponent should say.

(3/7)

I argue the answer isn't obvious. Natural language doesn't fully fit what people tend to mean when they say there's a language of thought.

For instance, natural languages feature words like "um, uh, ouch, oops," and so on. But these don't have the features LoTs are thought to have.
(2/7)

You may have heard talk of the "language of thought." Basic idea: some thoughts are represented in a language-like way. This is meant to explain putative similarities between thought and language.

But if there's a language of thought, it's fair game to ask: what makes it language-like?

(1/7)

Juan Murillo Vargas, How Language-Like is the Language of Thought? - PhilPapers The language of thought hypothesis (LoTH) claims that thoughts are underpinned by language-like vehicles. A central motivation is that there is a relevant similarity between language and thought expla...

Happy to announce that my paper "How Language-Like is the Language of Thought?" is forthcoming at Ergo!

Thread with a TL;DR below!
philpapers.org/rec/MURHLI-3

Can we perceive modal properties? Can we see only how things actually are, or are we also able to see how things could be? Much work in philosophy of perception assumes that our visual perceptual experience is restricted to the actua....

Such a readable and fascinating argument by @jessiemunton.bsky.social of @cambridgephilos.bsky.social To see a visual object is to see how it could be, not just how it is. #openaccess #philsky

where do they FIND these people?!

Evolving general cooperation with a Bayesian theory of mind | PNAS Theories of the evolution of cooperation through reciprocity explain how unrelated self-interested individuals can accomplish more together than th...

Our new paper is out in PNAS: "Evolving general cooperation with a Bayesian theory of mind"!

Humans are the ultimate cooperators. We coordinate on a scale and scope no other species (nor AI) can match. What makes this possible? 🧵

www.pnas.org/doi/10.1073/...

New from me and @esranur.bsky.social! In two exps with 3-4-year-olds, we find no differences in kids' reasoning about possible outcomes of an event in different temporal contexts; kids perform the same under physical and epistemic uncertainty psycnet.apa.org/record/2026-... #devpsy #psychscisky

We often have to judge who is knowledgeable—precisely when we are not. Can humans really do that? Our new paper in Psychological Science shows that, surprisingly, we can. drive.google.com/file/d/1b15E...

woah, crazy revelations at the coldplay concert

Thank you! Would love to get your thoughts.

(I also know the paper you attached--I really like it! I'm sympathetic to the view, but I have a paper where I argue that expressives don't have the right format to interface with logical operators. So was curious if you thought it generalized!)

Yep something like that! I was thinking something less fancy like "If Essman is involved I'll say hooray" or something like that, but your fancier gloss is probably better.

*meta-linguistically sorry!

I'm thinking that if the conditionals can't feature in inference the expressives are being used meta-linguistic (hence why the contents aren't inference-apt.)

Don't know of anyone whose worked on this, but quick Q: do you think these expressive conditionals license inferences? I.e., can you get discourse of the form "If Susie Essman is involved, hooray; Susie Essman is involved; so, hooray!"?

(1/2).

Kenneth Black, Common Ground, Conditionals, and _Ceteris Paribus_ Pragmatics - PhilArchive In conversation, we make use of shared information in order to communicate. But what shared information? In theorising about conversation, it’s popular to allot a privileged position to a single body ...

Pleased to announce that my paper, “Common Ground, Conditionals, and Ceteris Paribus Pragmatics,” is forthcoming in the Australasian Journal of Philosophy! 🧵

Here's a beautifully written piece on the saga: www.unclosetedmedia.com/p/a-philosop...

Beyond Argument: The Creative Craft of Philosophy Writing (guest post) - Daily Nous "I want to talk about the part about the part of philosophy writing that comes after the argument part: the bit where you work on expressing your idea clearly, delicately, even personally. I want to t...

“I want to talk about the part of philosophy writing that comes after the argument… the very specific work involved in infusing your writing with energy and life” — C. Thi Nguyen (@add-hawk.bsky.social) at Daily Nous on the creative craft of writing philosophy.

ROSE: A Universal Neural Grammar Processing natural language syntax requires a negotiation between symbolic and subsymbolic representations. Building on the recent representation, operation, structure, encoding (ROSE) neurocomputa...

New paper out today in Cognitive Neuroscience!
Proposing an explicit, causal-mechanistic, falsifiable and empirically grounded neural code for natural language syntax, and its innate basis.

ROSE: A Universal Neural Grammar

www.tandfonline.com/doi/full/10....

Happy to share this updated Stanford Encyclopedia of Philosophy entry on 'Associationist Theories of Thought' with
@ericman.bsky.social. Among other things, we included a new major section on reinforcement learning. Many thanks to Eric for bringing me on board!

plato.stanford.edu/entries/asso...

Excited about this in large part because Raphael is the absolute best

Transcranial Focused Ultrasound for Identifying the Neural Substrate of Conscious Perception Identifying what aspects of brain activity are responsible for conscious perception remains one of the most challenging problems in science. While progress has been made through psychophysical studies...

A roadmap for using transcranial focused ultrasound for identifying the neural substrate of conscious Perception: arxiv.org/abs/2507.08517. Comments welcome!