Léo Varnet @leovarnet@qoto.org

Thinking about kinds of analysis errors. There are errors of design, doing the wrong analysis intentionally, & also errors of execution, doing the wrong analysis by accident. Like this 2016 example in which country codes entered as continuous by accident. https://www.cell.com/current-biology/fulltext/S0960-9822(16)30670-4

Working in ways to prevent such errors is important. I simulate synthetic data to validate my analysis. Even in simple contexts, accidents can happen. Because computers suck and are out to get us.

I swear, I've been submitting articles to journals for almost 10 years now and I still have no idea what to write in the cover letter. Like, what can I say that the editors won't get from just skimming the abstract and looking at the contact info associated with my profile in the submisison system?

WE SHOULD BE PAST THE NEED FOR COVER LETTERS, DAMMIT! THIS IS THE FUTURE!

#academia #linguistics #psycholinguistics

This quote, estimating that at least 50% of the literature might be a Type 1 error due to inflated false positive and publication bias, was published 38 years before Ioannidis (2005) making the same point in 'Why most published research findings are false'

A question that is reminiscent of both Douglas Hofstadter and Daniel Dennett's work (and also of some of @gregeganSF books): what sort of things can be truly generated by simulation? As John Searle said: "No one would suppose that we could produce milk and sugar by running a computer simulation of the formal sequence in lactation and photosynthesis". Similarly, Dennett stresses that a simulated hurricane causes no floods and devastation... Or does it? Maybe if you manage to integrate an AI in the simulation, then from its point of view this would look like a real hurricane. But at least it is not a hurricane at the "level" of the programmer. Now there are also things that are "truly" generated by simulation at the level of the programmer. Think of mathematical proof. If you design a computer program to produce mathematical proofs, then it won't generate simulated mathematical proofs but true mathematical proofs. The same goes for language or music. Now the big question is : on which side of this divide does the mind fall? Is it more like milk or more like language?

#AcademiaAntiracistFightClub ✊24 juillet 2023

Série de l'été #DelaraceESR🌞

"Rejeter la thèse bien étayée de l’existence d’un racisme systémique (...) ne devrait pas être le fait de personnes qui  exercent un métier scientifique".
par @anthrofuentes
https://academia.hypotheses.org/50733

Job alert! I'm hiring a postdoc to work on the dynamics of speech rhythms in #French and #German. This is part of the #ANR-#DFG project #DRhyaDS (https://qoto.org/@leovarnet/109550769028496021).
Please share widely!
https://mycore.core-cloud.net/index.php/s/nVjBSw9H9OChpMF
@linguistics @psycholinguistics

Grâce à l'import des médailles d'argent du CNRS sur #Wikidata, on peut à présent répondre à cette question : à partir de quand la parité dans la remise de ces récompenses est-elle atteinte ?

#SPARQL #Wikipédia
https://wikif.hypotheses.org/140

What is the cost of science being conducted in English? @tatsuya_amano &co reveal that non-native English speakers spend more effort than native speakers in conducting scientific activities (reading, writing, preparing presentations...) #PLOSBiology https://plos.io/46Rfptc

Have you read the JASA Express letter, Speech beyond the binary: Some acoustic-phonetic and auditory-perceptual characteristics of non-binary speakers, by Brandon Merritt? Check it out at https://doi.org/10.1121/10.0017642
#PrideMonth

Welcome 🔥

Holy _shit_ this paper, and the insight behind it.

You know how every receiver is also a transmitter, _well_: every text predictor is also text compressor, and vice-versa.

You can outperform massive neural networks running millions of parameters, with a novel applications of _gzip_.

https://aclanthology.org/2023.findings-acl.426/

I feel like this comic has been my whole career as a linguistics researcher. Eg "progressing" from "should we be talking about Noun Phrases or Determiner Phrases?" to "what is a Noun really"?

So... good?

https://xkcd.com/2797

#linguistics @linguistics

Thank you #elonmusk. From the deep of my hearth.

Finalement, il reste une question : est-ce que toutes les formes d'#EcritureInclusive sont aussi efficaces pour lutter contre le biais masculin ? Il y a encore peu d'études sur le sujet, mais il semblerait globalement que non : de façon surprenante, les formes neutres (p.ex. "l'architecte" qui peut désigner un homme ou une femme) engendrent elles aussi un biais masculin (Lindqvist et al. 2019) et les langues sans genre grammatical comme le #turc et le #finnois ne sont pas non plus immunisées contre ce biais (Renström et al 2022, merci @elmerot de m'avoir pointé cet article). Si l'on souhaite susciter des représentations homme/femme égalitaires, le meilleur outil est la forme double (par exemple "les chercheurs et les chercheuses" ou les chercheur·euse·s). Nous menons actuellement une étude pour étudier cet effet en français. (6/X) @psycholinguistics @linguistics #MasculinGénérique #stéréotypes #LangageInclusif #linguistique