Stefano Coretta

For me it’s a matter of internal philosophical coherence, but most researchers that identify as scientists don’t care about philosophy (or about building their own philosophical stance).

Honestly, my course has been basically a different course every year I taught it so far. I think I settled on most things now tho :D so 2026/27 should be pretty much like 2025/26 (but I am planning on extending the textbook past the course so still work to do ah!)

I feel you! I think it's unavoidable though (I am thinking about the stats course I teach in linguistics at UoE, which is also the only one we offer in the department). We do want to be cutting edge as much as possible when it comes to teaching especially when things move so fast.

We got a request as well like that, so it is probably the same team asking! Maybe it's a social experiment? :D

A lot of people don't quite know the difference in my experience (although at the same time, the terminology is not that stable yet: in link.springer.com/article/10.1... everything is "reproducibility", 5 types). I personally like the Turing Way framework.

homer simpson is holding a sign that says a on it ALT: homer simpson is holding a sign that says a on it

Psychologists, please abandon box plots. They are bad (why? where do I begin...).

Not really. It’s quite straightforward computationally. The problem is that we don’t have good baselines to choose parameter values from for simulation. So again, the problem conceptual theory, not methodological theory.

🎉 NEW PAPER with @gsakr.bsky.social

"Multivariate Analyses of Tongue Contours from Ultrasound Tongue Imaging"

👅📡🩻📉

doi.org/10.1177/0023...

Well, beta models that accept 0s and 1s are as complex. Ordinal beta is related to ordinal models so the complexity is similar. So I don’t think computational cost is really the issue here (I also think it should never be, but oh well…)

To decide how to measure and model acceptability, we need a neurocognitive and mathematical model of acceptability, although that’s a project of its own :D

Ratios cannot be used with 0,1 bounded variables, so one cannot say participant 2 has rated the sentence twice as acceptable. Ordinal models assume a latent continuous scale so it’s not as simple as ranking items in a categorical way.

Tools are instrumental; the emphasis is on relationality, enabling social and ecological justice concerns to be addressed more directly.

Our holistic approach transcends rigid divides such as structuralism, constructivism, and positivism, which we view as historical contingencies rather than essential frameworks.

🎉 - NEW MANUSCRIPT with @jess-hampton.bsky.social

Beyond the horizon: a more-than-human and holistic approach to language

lingbuzz.net/lingbuzz/009...

Maybe in the package vignette you can add code with list.files() so that people see an example. And you can stick to file or list of files?

I don't think there is a single way of doing that and as long as it is clearly documented, CRAN won't mind. tidyverse tend to accept a single file or a list of files, so for consistency I would do this, but you said you were thinking of dir for those who don't know how to get list of files.

a cartoon minion with a red light on his head and the words `` memory alert '' below it . ALT: a cartoon minion with a red light on his head and the words `` memory alert '' below it .

And the time comes when your tinsy work laptop has no more storage.

So Old Chinese *ɲ becomes Middle Chinese *ɻ (still /ɻ/ today in Mandarin). Does anybody know of a parallel change in other (non-Sino-Tibetan) languages?

My wish for 2026 is that I want to go back to my home planet. Take me back!!! Thank youuuuuu

Statement of commitment: Registered Reports in Linguistics will automatically offer Stage 1 in-principle acceptance (IPA) to any Stage 1 RR within the journal’s disciplinary scope that receives IPA at PCI RR, and will accept without further peer review any Stage 2 RR that has been recommended by PCI RR, subject to the manuscript meeting the journal requirements concerning bias control. Disciplinary scope: Research in linguistics, including all sub-fields.

Statement of commitment: Replication Research will automatically offer Stage 1 in-principle acceptance (IPA) to any Stage 1 RR within the journal’s disciplinary scope that receives IPA at PCI RR, and will accept without further peer review any Stage 2 RR that has been recommended by PCI RR, subject to the manuscript meeting the journal requirements concerning bias control and computational reproducibility. Disciplinary scope: Replication studies across a range of physical, life, and social sciences, including digital humanities, experimental philosophy, geoscience, linguistics, management sciences, marketing, medicine, mental health, metascience, neuroimaging, neuroscience, political science, psychology, and qualitative and quantitative research methods. Studies can be robustness reproductions using the same data and code, recreate reproductions using the same data and new code, close replications using new data and the same method, or conceptual replications using new data and a different method.

We're delighted to welcome TWO new diamond-OA PCI RR-friendly journals

Registered Reports in Linguistics, ed. by @scoretta.bsky.social ky.social & @jess-hampton.bsky.social
+
Replication Research @r2journal.bsky.social ed. by @aufdroeseler.bsky.social & team

rr.peercommunityin.org/about/pci_rr...

An image of two bonobos sitting in the grass next to a stream. Overlay text reads “1 million species of plants and animals at risk of extinction. - IPBES Global Assessment”

In 2019, IPBES #GlobalAssessment est. 1 million species of plants & animals are threatened with extinction.🥀

The figure is probably even higher, but what's important is the urgent need for #biodiversity preservation!

https://www.ipbes.net/global-assessment

I wish some editors and reviewers would understand this! Maybe for 2026, this is one of my wishes.

In this writing and reviewing season, I would like to kindly remind linguists to not hide your data or your observations behind what you think is "the main theoretical point of the paper." Cursory data and description are the linguistic impostor syndrome in writing. 1/2

All I want for Christmas…is a precisely defined research question - Trials Trials -

link.springer.com/article/10.1...

Ironically enough, the title of the paper doesn't include uncertainty and states the finding as a fact-based message :D

course schedule as a table. Available at the link in the post.

I'm teaching Statistical Rethinking again starting Jan 2026. This time with live lectures, divided into Beginner and Experienced sections. Will be a lot more work for me, but I hope much better for students.

I will record lectures & all will be found at this link: github.com/rmcelreath/s...

Vowel plot with F2 and F1 values. The formants are normalised Hz.

🎉 "Normalising formant values for plotting and modelling"

New blog post: stefanocoretta.github.io/posts/2025-1...

So we don't have a defined process model that can guide is in coming up with a bespoke statistical model.

An exmp of where theory is not strong enough. In the sentence “Pritesh emailed Themba. I saw him at the gym.”, the subjecthood hypothesis states that him refers to Pritesh, while under the syntactic parallelism hypothesis him refers to Themba. In reality, language users do either at varying degrees.

It includes an optimisation process in one of the three components, which makes regression unhelpful, for example.