Mark Torrance

🚨Just dropped: introduction to our Special Issue, by guest editors Chris Anson and Kirsti Cole.
✅importance of empirical evidence in gen AI and writing ✅methodological pluralism
✅wider, practical implications of the research
www.jowr.org/jowr/article...

Flow in mind, flow in fingers: Parallelism in written language productionJens Roeser∗, Associate ProfessorMark Torrance, Associate ProfessorDepartment of Psychology, Social Sciences, Nottingham Trent University∗Corresponding authorEmail:jens.roeser@ntu.ac.ukHomepage:https://www.ntu.ac.uk/staff-profiles/social-sciences/dr-jens-roeserSocial media:@sentwrite.bsky.socialAbstractLanguage production is known to operate on different levels of representation. “Flow” in writing results from parallelism in the coordination of subsequent planning units. In this article we discuss three points arising from Roeser, Conijn et al. (2025): (1) parallel processing results innon-additive effects, (2) study of the production of multisentence texts permits testing of questions around how language production is coordinated in real time, and, more generally, (3) statistical models must closely align with what we know about the cognitive process of what is being studied.

Flow in mind, flow in fingers: Parallelism in written language production - explore.bps.org.uk/content/bpsc... A short paper summarising some of the things we've been thinking about recently. With, we believe, application beyond written production. @sentwrite.bsky.social

Journal of Writing Research special issue: effects of AI use on writing process, product and practice - www.jowr.org/jowr/issue/v... A carefully curated collection of diverse robust empirical studies - controlled intervention evaluation, corpus analysis, computational linguistics, qualitative.

Psycholinguistic Databases, Stimuli, Utilities — Concepts & Cognition Laboratory

2026 big update to the psycholinguistic database page! If you know of corpora, lexical databases, or other resources that I've missed, please LMK. Trying to keep this thing relatively current and could use the help www.reilly-coglab.com/data

Martínez-Cano et al identified narrative characteristics in writing by people with schizophrenia to determine how these relate to positive and negative symptomatology. Results reveal that language difficulties occur in writing and allow for symptom differentiation.

www.jowr.org/jowr/article...

Evidence for the representation of non-hierarchical structures in language Nature Human Behaviour - Language is often thought to be represented through hierarchically structured units. Nielsen and Christiansen find that non-hierarchical structures are present across...

I'm very excited about this paper with @yngwienielsen.bsky.social just out in @nathumbehav.nature.com in which we provide evidence for the mental representation of non-hierarchical linguistic structure in language use.
🧵 1/4
Read the paper here: rdcu.be/eZ26u

Hello Bluesky! Journal of Writing Research is bringing the conversation to this platform. We'll start by highlighting some of the fascinating research in our most recent volume.

𝑫𝒊𝒔𝒕𝒊𝒏𝒈𝒖𝒊𝒔𝒉𝒊𝒏𝒈 𝒆𝒇𝒇𝒆𝒄𝒕𝒊𝒗𝒆 𝒘𝒓𝒊𝒕𝒊𝒏𝒈 𝒔𝒕𝒚𝒍𝒆𝒔 𝒊𝒏 𝒕𝒉𝒆 𝑷𝑬𝑹𝑺𝑼𝑨𝑫𝑬 𝒄𝒐𝒓𝒑𝒖𝒔- Morris, Crossley, Holmes, and Choi.
This article explores how high-quality persuasive essays differ in the linguistic features used, challenging the assumption that excellent writing is a single, uniform style. (vol. 17, no. 2)

My talk here: rpubs.com/mark-torranc... - Direct evidence (we think) that looking back into your text while writing serves to maintain output fluency. All the hard work done by Astha Singh (Iowa State University). Her ACL paper is linked from the final slide.

Meet multilangapp's younger sister Spellname App. We're still work in progress but try the "Density plots" tab and click on the legend keys. This is `plotly`.

There's a whole rant about the publication process for this paper, possibly also valuable comment about publishing cognitive papers on a topic that looks "educational", but mainly a rant, that may feature in a later post. Probably best avoided when it appears.

end/

GitHub - chukharev/cywrite: CyWrite CyWrite. Contribute to chukharev/cywrite development by creating an account on GitHub.

...research and methods around tracking writers' eye movements within their ever-changing emerging text. Particularly github.com/chukharev/cy... and dating back to link.springer.com/article/10.1...
3/

Understanding written composition as fluent, semi-parallel process with occasional hesitations - work led by @sentwrite.bsky.social. We might possibly have already mentioned doi.apa.org/doi/10.1037/...

And...
2/

When writers produce multi-sentence text they periodically pause and look back into what they've written. We describe this behaviour in adult L1 and inexpert L2 composition.

journals.plos.org/plosone/arti...

This brings together methods / ideas from two research streams...

1/

Paper in Language and Cognition with @marktorrance.bsky.social and @seriousstats.bsky.social

We show that semantic contrast shapes timing of pre-planning in speech and writing.

doi.org/10.1017/lang...

If you're into how context shapes how we plan language, check it out!

More written production poster excitement at AMLaP2025:

Thursday PM #200 English / Norwegian cognates and false friends sites.google.com/view/helenes...

Friday AM #167: Spelling difficulty now disrupts future lexical retrieval rpubs.com/jensroes/aml...

Spelling can be tricky but imagine switching between two spelling systems where some words are spelled differently ("hånd" in Norwegian) and others look similar ("gift" in Norwegian isn't nice).

Check out Helene's research (poster 200 on Thursday Afternoon, 4 Sept. 2025 at #AMLaP in Prague).

When do we actually think about what we want to say and what words to use? Interestingly our mind does a lot of this work while we're writing text. In fact, we demonstrated (psycnet.apa.org/record/2026-...) that even young writers often don't stop before starting a new sentence. Way to multitask!

This paper, of which I'm inordinately proud, is now out in JEP-General doi.apa.org/doi/10.1037/.... Huge credit to Jens Roeser @sentwrite.bsky.social for conception and all the heavy lifting.

Image showing research design from poster linked in text.

Writers composing multi-sentence texts often pause briefly, glance back at isolated words or short phrases, then continue writing. We think we have evidence that this cues what to say next. rpubs.com/mark-torranc... or, better, talk to me on Thursday morning at #AMLAP25.

a cat is looking at a laptop computer screen with a lot of text on it . ALT: a cat is looking at a laptop computer screen with a lot of text on it .

Ever wondered what happens in our mind when we write simple messages, posts, or full essays? Also how can psychologists tests theories about writing? English is known for it's tricky spelling rules which allows psychologists to study what's going on, when things are going wrong.

New paper out 🎺 introducing process-based measures of automatisation in L2

The elicited imitation task is widely used as a measure of automatized L2 knowledge. However, the scoring of the task relies exclusively on product-based measures (i.e., accuracy of L2 production).

And to be clear, the multiple spelling errors in this thread are there for illustrative purposes, and to subliminally advertise our mail order ham side hustle.

This looks massively useful. Thanks so much to bsky.app/profile/mari... and team.

We don't have any new data to add just at the moment, though we do have an English written image naming dataset for difficult-to-spell words to write up. In the meantime, though, this is fun and potentially useful jens-roeser.shinyapps.io/multilangapp/. Short thread here bsky.app/profile/did:...

So please pass jens-roeser.shinyapps.io/multilangapp/ on to anyone you think might find psycholinguistic joy in exploring these data.

I was reminded to post about this bsky.app/profile/jami... from @jamiereillycog.bsky.social. Our database features there alongside much, much more.

Or compare typing speed in different countries / languages. These are English and Italian writers. Lots of scope for post-hock hypothesizing around effects of orthographic transparency. /6

jens-roeser.shinyapps.io/multilangapp/ will also help you find stimuli for you written picture naming experiment, or just feed you prejudice about which Europeans are the worst spellers. Below are the top 9 hardest-to-spell Snodgrass image names. Norway taking a surprise first place. /5

Collected face-to-face. Oh how naive we were. The plot in the first post shows correlations among naming diversity, spelling diversity, mean time onset latency (RT) and mean interkey interval once output commenced. /4

Data from samples of at least 60 adults naming all of the images from the colourised Snodgrass and Vaderwart picture set in their first language (Bulgarian, Dutch, English, Finnish, French, German, Greek, Icelandic, Italian, Norwegian, Portuguese, Russian, Spanish, and Swedish). /3