Mathew Gillings

After our very succesful Corpus Linguistics 2025 conference hosted by @robbielove.org at Aston, CL is going on tour! Our next conference will be Easter 2027 in Hong Kong at HK Poly U. In summer 2029 we return to Europe when @michamahlberg.bsky.social will host CL at Erlangen!

Florent Moncomble’s corpus tools A collection of tools for collecting and analysing corpora of online media discourse

#Linguistics
2 new additions in my #corpus toolbox, geared towards the study of the meta-discourse of press coverage: the Guardian and Figaro Comments Corpus Builders can be used to collect corpora of article comments from those two online newspapers.
1/3

Diaries at the ready... the next Corpora & Discourse International Conference will take place from 23-25 June 2026, at Lancaster University. We are also very excited to launch our website, with plenary speaker info, key dates, and much more besides: wp.lancs.ac.uk/cad-2026/. Please share widely!

Congratulations Nathan!! 🥳🥳

An excellent article (use link).
Its time to abandon nonverbal behaviour as a deception cue. 50 years of research show no support for it. Polygraph, voice stress analysis, microexpressions, eye tracking - its all pseudoscience.
#Deception

crestresearch.ac.uk/download/490...

Corpus-Assisted Discourse Studies Cambridge Core - Applied Linguistics - Corpus-Assisted Discourse Studies

It seems that the Cambridge Elements in Corpus Linguistics series is free to download for the duration of #CL2025 again — including my own. I’m never one to pass up a freebie, so do take a look: www.cambridge.org/core/element...

What great news, congratulations Stef!!! 🥳

If you’re going to #CL2025 and interested in either health communication, or methodology papers more broadly, our work here should be of interest. And, more importantly, it’s @janadeclercq.bsky.social’s first corpus conference, so do say hello if you see her! 😉😀

I am the other Mathew-with-one-t, pleased to meet you! 🤝

Now time for a summer holiday along Australia’s east coast. All completely new for us so recommendations welcome!

That’s a wrap on #IPrA2025! We organised a panel on regional (im)politeness variation, presented our own work on variation in the USA, and then gave a methods paper looking at discourses of chronic pain from both an interactional sociolinguistics and a CADS perspective. What a great few days. 🇦🇺

The word is out! Next IPrA is in our (Finnish) neck of the woods, in Helsinki 🇫🇮 Pragmaticians, converge!

We visited a couple of days ago!! Aren’t they cute? We were really surprised to see how many there were!

This week I’m in Brisbane for the #IPrA2025 conference. It was straight into it last Saturday, giving a pre-conference workshop on Corpus Pragmatics with participants from 11 countries. A real delight! 🇦🇺

International Pragmatics Conference just kicking off at the University of Queensland in Brisbane. Always a great opportunity to pick up conversations with so many friends and colleagues as if no time had passed since we last met.

Thanks to those posting from #ICAME46 this week so we can follow along vicariously. 😀 This workshop yesterday on speech act development sounded really great.

A few photos from Bergen last week during the Association for Business Communication conference where I presented some work on the discourse of corporate wrongdoing. Stunning city with some particularly friendly four-legged residents who don’t seem to mind the chilly summers! 🇳🇴

🆕 CADSbib annotated bibliography

👩‍💻 Do you work with corpora & discourse?

📚 500+ references annotated with corpora, language, tools, methods & more.

🔗 nfdykes.github.io/blog/2025/ca...

#CADS #DiscourseAnalysis #CorpusLinguistics #DH

This CfP for a special issue of Corpus Pragmatics (ed. by P. Madella) looks really great. It promises to make a significant contribution to a more linguistically-informed approach to deception (detection) - something much needed!

New paper out in the International Journal of Business Communication on corporate wrongdoing -- specifically how Boeing navigated the 737 Max crisis a few years ago. Check it out open access here: doi.org/10.1177/2329...

Travelling to the UK this morning to attend and give a plenary at the Manchester Forum in Linguistics. I’ll be talking about using corpus methods to study variational (im)politeness - really looking forward to being back north for a few days. 😄 mfilconf.wordpress.com

After a soft launch in the last couple of weeks, #wmatrix version 7 is now completely freely available worldwide for academic users plus UK government and NHS email addresses. This is the largest revision of the software since it was created 25 years ago: ucrel.lancs.ac.uk/wmatrix/

WU (Vienna University of Economics and Business)

I was recently interviewed as part of @wuvienna.bsky.social's "Meet our Researchers" series. They put together a nice video about using corpus linguistics to study climate change discourse, based on my paper with Carmen Dayrell. Really happy with how this turned out! www.wu.ac.at/en/research/...

YouTube video by Wirtschaftsuniversität Wien Climate Change and the Media | Meet Our Researchers: Mathew Gillings

What do you think of when you hear the term "climate change"? Whatever it may be, it's probably partly shaped by the media.

WU researcher @mathewgillings.bsky.social has looked into the discourse around climate change in UK newspapers. In our newest video, he shares some interesting results:

One of the architects of LIWC @ryanboyd.io got in touch about our paper – he’s written this blog post about

www.pancakes.wtf/20250110/lan...

First we’d like to say that Ryan makes interesting points in his rejoinder and we absolutely welcome this…

Key findings:
- Approaches with both high and low contextualisation produce high interanalyst agreement
- Topic frequency does not equal saliency for humans. They identify topics that machines do not, e.g. "Diversity & inclusion"
- Humans seek to identify distinct discourses; machines blur the lines

Specifically, we:
- Retrieve topics via topic modelling, then label them via LLM prompting, eyeballing and concordance analysis
- Retrieve topics via closing reading (by asking two analysts to read an entire corpus)

We then:
- Compare results between analysts
- Compare results between methods

My paper with Sylvia Jaworska on methodological triangulation is now out in Applied Corpus Linguistics! We look at how the chosen analytical method affects the identification and labelling of topics in a corpus. How does the amount of available context change interpretation? doi.org/10.1016/j.ac...

Can’t believe the algorithm hadn’t steered me towards this. Truly A+ content!