Posts by Author

From use of AI in language teaching to rhythm in second language speech production  0

Leiden University hosted two one-day conferences on second language learning within just three weeks: the Language Learning Resource Centre’s (LLRC) yearly conference, this time on AI in language education, was held on June 13th. On July 1st there was a one-day conference on rhythm and fluency in second language speaking, as satellite workshop alongside the 2024 Speech Prosody conference. As co-organizer of these one-day events, I got to experience both maximally. It made me realize once again the breadth and wealth of language learning research.

LLRC conference on AI in language education

For the 7th edition of the LLRC conference, presenters from Universities, Universities of Applied Sciences, a secondary school, and a global testing company from The Netherlands and Flanders showed their research and good practices to a crowd of researchers and language teachers from a similar wide range of different institutes, schools, and companies. With 110 participants, it was the busiest edition so far. Apparently, the theme of the day struck a note. Although AI has been around since the 50’s, it is only in the last two years that LLMs and intelligent tools are quickly and drastically changing language use, language learning, and language teaching. When the LLRC committee came up with the timely theme in the fall of 2023, we did not yet know it, but it turned out that June ‘24 was proclaimed as month of AI in education in the Netherlands, with many events throughout the country, and the one by LLRC was linked to it.

AI is not artificial and it is not intelligent

The keynote speaker, Esther van der Stappen (Avans University of Applied Sciences), introduced the theme by bringing together multiple perspectives and bridging the gap between computer science and education, shedding light on practical and ethical aspects of AI in education more in general. She certainly had our attention when she explained what AI is not: it is not artificial and it is not intelligent, to begin with.

Bastogne cookies and second language learning in both conferences

A little over a fortnight later, the one-day conference on prosodic features of learners’ fluency featured two keynote speakers: Lieke van Maastricht (Radboud University) and Malte Belz (Humboldt University, Berlin). Lieke showed how the use of hand-gestures during speaking in a second language should not be ignored, and Malte showed methodological issues that need to be addressed in research on measures of fluency in speaking such as pausing and filled pauses. This conference had some similarities to the LLRC conference, but the differences were more pronounced. To start with the similarities: on both days, technological advancements and tools played a big role; on both days, Bastogne cookies were served during coffee breaks and vegetarian bitterballen at the end of the day. Obviously, the biggest resemblance between both conferences concerned the broad topic of second language learning. But where the day on AI in the language classroom showcased research and practice on language teaching didactics (using AI) in the broadest sense, the conference on prosody and fluency showed research and research methods on very specific aspects that language learners need to master, namely on hesitations and rhythm in speaking.

Making worlds meet

To advance research and practice of language teaching, both types of exchanges among researchers and teachers are helpful, and in the end, we should strive to have both types of worlds meet. For instance, it is one thing to find out that gestures like arm movements should accompany the word or sentence stress, and that second language learners have trouble in timing gestures in this way; it is another thing to teach the timing of gestures to a classroom of 30 students. And indeed, in the closing session on July 1st, three “D”’s were recognized as Big Questions or New Challenges to tackle within the subfield of prosody in linguistics: 1) Differences between individuals in speaking and learning to speak, 2) the Dynamic nature of speaking processes, and, as third “D”: Didactics. So plans to have both types of worlds meet (more often) are already in the making! 

How are languages learned? A shared question for research at LUCL and ICLON  0

The title of this blog contains a Big Question: “How are languages learned?” I would say this is an intriguing question, considering how quickly children learn the vast amount of knowledge and skills needed for successful communication. They start with crying, giggling, and babbling some sounds, and by the age of three they are talking your heads off. This Big question is one of the fundamental questions of the broad research field of Linguistics. “How are languages learned” is also one of the big questions for practice-oriented research informing foreign language pedagogy, and the learnability question includes another relevant question for both fields: how language learning is the same or different for early versus late learners.

Fundamental questions in linguistics

If we consider the other three fundamental questions in the field of linguistics, we can even conclude that these questions are related to the first, learnability-question, and therefore also relevant for foreign language pedagogy research:

  • How do languages all over the world differ (are languages equally easy or hard to learn?);
  • how do languages change over time (how is language learning leading to language change?);
  • and how is language produced and understood in real time (how do cognitive processes lead to language learning and how are the cognitive processes the same or different for early and late language learners?).

Linguists take different angles when tackling these issues. The question about how languages differ is mainly a structural question, and linguists have described languages at different levels: from the level of sounds to the level of text or discourse organization. How and why languages change is mainly a social question, as languages change through generations, through contact between languages due to migration, and because of people’s need to belong to social groups. The question how languages are processed in real time is cognitive, where linguists investigate how and which cognitive processes are put into action, for instance for the enormous task of listening and speaking with about 5 words per second. The same structural, social, and cognitive viewpoints towards language are also relevant for research into foreign language pedagogy, to answer the more direct practice-oriented question “How to teach language?”.

A unique position for foreign language classroom research

This means that for such practice-oriented research, an enormous amount of relevant research as well as a vast tradition in research methods is available. If we would compare this to research into pedagogy of other school subjects, it becomes clear that this situation for foreign languages as school subjects is unique. Take Biology, for example. It is not the case that the research field of Biology busies itself mostly with the question “How is Biology learned?” Likewise, the main research question of History is not “How do people learn History?”, etcetera.

Practice-oriented research and fundamental research both inform foreign language pedagogy

My conclusion for this blog is that both fundamental and practice-oriented research are needed to better understand and promote language learning and teaching in the classroom. For instance, to investigate successful didactics for teaching verb forms in Spanish, researchers first need to know which verb forms are problematic and why. Or for (research into) teaching and assessing speaking skills, it is first necessary to understand the construct of speaking in a fundamental way. Because foreign language learning research is partly built upon the vast research tradition of Linguistics, I would argue that we can and must make use of this unique situation and interact with researchers who work in linguistics on the more fundamental parts of the big issues to inform practice-oriented research into foreign language teaching. At Leiden University, interaction between the Leiden University Centre for Linguistics (LUCL) at the faculty of humanities and the practice-oriented research at ICLON is already present. I hope we can make the collaboration in the future even stronger, for instance through the Language Learning Resource Centre as well as through working together as teachers on the two-year educational MA for (foreign) languages.


Buiksprekers in onderzoek naar tweedetaalverwerving: Emily Felker’s dissertatie “Learning second language speech perception in natural settings”  0

Op 10 juni 2021 verdedigde Emily Felker haar proefschrift aan de Radboud Universiteit Nijmegen. Ze kreeg moeilijke vragen, en beantwoordde die stuk voor stuk met verve. De meeste vragen gingen over hoofdstuk 6 van haar proefschrift, het hoofdstuk waarin onderzocht werd of expliciete instructie over klanken nu nut heeft voor bewustwording en herkenning van die klanken door tweedetaalleerders. Kort gezegd is de uitkomst “ja” en kunnen we hieruit (weer eens) concluderen dat bewustwording door middel van expliciete instructie zinnig is voor het leren van een tweede taal, ditmaal op het gebied van klankperceptie.


Dit soort onderzoek is winst voor de taalonderwijspraktijk. Ik hoef leraren(opleiders) niet uit te leggen waarom we dit zouden willen weten. Maar dat geldt niet direct voor de andere hoofdstukken in het proefschrift van Felker. In de corona-tijd sprak ik mijn collega’s vaker buiten, op een terras, wel zo ontspannen. En zo kwam het voor dat mijn collega MVT-lerarenopleider mij vroeg naar het boekje dat ik op dat moment aan het lezen was. Enthousiast begon ik te vertellen over de ingewikkelde, nieuw-uitgevonden experimentopzet uit hoofdstuk 3 en 4 van het proefschrift…


Het buiksprekerparadigma

In dit experiment zit een echte proefpersoon tegenover een niet-echte, namelijk een bondgenoot binnen het onderzoek, een soort medeplichtige van de onderzoeker. Samen moeten ze puzzels oplossen en hebben ze daarvoor ieder een deel van de informatie op hun eigen computerschermen, en dit doen ze in het Engels. Daarbij moeten ze woorden aan figuren verbinden. Nu is de crux dat die bondgenoot een vreemd accent heeft; ze zegt telkens /i/ in plaats van /e/, dus bijvoorbeeld ‘bitter’ in plaats van ‘better’, en bovendien zegt ze /ie/ waar anderen /i/ zeggen. Als de proefpersoon dan de geschreven woorden “bitter” én “better” heeft om uit te kiezen voor het oplossen van de puzzel, dan moet de proefpersoon dus gaandeweg het experiment begrijpen dat dit een accent is (met e -> i en i -> ie), want anders lukt het niet om de puzzel correct op te lossen. Om ervoor te zorgen dat de medeplichtige een constante uitspaak heeft van dit accent, gebruikt de medeplichtige een heel arsenaal aan vooraf opgenomen zinnetjes en woorden. Om mijn collega beter uit te leggen hoe dit werkte, deed ik voor hoe de medeplichtige zich dan telkens kort achter haar computerscherm waarop de puzzel te zien was verborg, om te doen alsof ze in de microfoon praatte, maar in feite drukte ze dan op een knopje. Als een soort buikspreker.


Learning in natural settings?

Na mijn lange maar enthousiaste uitleg over dit nieuwe experimentele paradigma, reageerde mijn collega droogjes met “maar waarom zou je dat allemaal doen?”. En omdat de rest van de terrastafel al andere meer prangende onderwerpen besprak, bleef mijn antwoord op deze vraag uit. Daarom bij deze: waarom zou je een buikspreker-experiment uitvoeren? De titel van het proefschrift is “Learning second language speech perception in natural settings”. Hoezo is een buikspreker-paradigma een natuurlijke setting?


Dat heeft te maken met de eeuwige tegenstelling in onderzoek: zo natuurlijk mogelijk enerzijds, en zo gecontroleerd mogelijk anderzijds. Vóór het paradigma van Felker was er nog nooit gecontroleerd onderzoek naar het leren van accenten in taal in live-interactie. Nu kon Felker onderzoeken hoe tweedetaalleerders wennen aan een voor hen tot dan toe onbekend accent, en hoe ze dat accent gaandeweg leren met expliciete feedback (Buikspreker: “No not that one, you need bitter not beeter”).


Nieuwe mogelijkheden voor onderzoek

Dit paradigma opent tal van nieuwe mogelijkheden voor onderzoek naar tweedetaalverwerving: als je gecontroleerde gesprekken kunt inzetten in onderzoek, dan kun je bijvoorbeeld preciezer uitzoeken welk type feedback, op welk aspect, op welk moment, en door wie het beste werkt in interactie, zowel voor directe uptake als voor langetermijn-leereffecten. Je kan dit bijvoorbeeld inzetten om meer onderzoek naar feedback te doen bij perceptie, zoals in Felker’s onderzoek, maar het is uiteraard ook te gebruiken om uit te zoeken hoe je feedback in interactie op uitspraak, of op andere aspecten van productie het beste kunt inzetten. Vervolgens is er ook een heel scala aan experimenten te bedenken om uit te zoeken of het uitmaakt wie die feedback geeft: kan dat een peer zijn, of werkt het beter als het een docent is; in hoeverre maakt het uit of het een moedertaalspreker is of niet, et cetera.


Dus hier kwam mijn enthousiasme vandaan, het is niet direct toepasbaar in de praktijk, maar het onderzoek dat mogelijk is gemaakt kan ons wel weer meer leren over de mogelijke positieve effecten van feedback in interactie, toch één van de meest gebruikte vormen van instructie in een les moderne vreemde talen. Los van de theoretische en praktische implicaties van het nieuwe paradigma gun ik het de toekomstige onderzoekers die als medeplichtige in zo’n experiment meedoen: hoe grappig moet het wel niet zijn, om door de juiste combinatie van knoppen met opgenomen zinnen alle proefpersonen te doen geloven dat er een echt gesprek plaatsvindt.

Intercultural Communicative Competence in the Language Classroom: Is this where language education is headed?  0

Authors: Nivja de Jong and Tessa Mearns (members of Meesterschapsteam MVT)


One time is not enough, two times is a trend, and three times is a tradition

In June 2017, the LLRC (Language Learning Resource Centre) was ‘launched’ with a full-day programme by and for language teachers, language teacher educators, and language learning researchers. Most of those present were from Leiden University, unsurprising as the LLRC has as its central aim to bring together language teachers and researchers from the different Leiden University institutes involved in teaching languages: LUCL, LIAS, ATC, ICLON and LUCAS. On June 7th, 2019, the third full-day conference was hosted by the LLRC, this time on the topic of intercultural communicative competence with a focus on higher education. Professor Michael Byram (emeritus professor, Durham University) was the invited speaker of the day. Anyone who has ever heard of the term “intercultural communicative competence” knows the stature of Prof. Byram. Although again, most people came from Leiden University, both presenters and attendants were now more diverse.


“Why should the biologists get all the fun?”

This was a question posed by Prof. Byram during his keynote. He referred to the fact that in CLIL (Content and Language Integrated Learning), as the acronym already clarifies, it is usual to teach content whilst teaching language. So if biologists teach about biology whilst teaching language, what (fun) topics can and should language teachers use?

When teaching language, students and teacher(s) use language, and this language that is used is always about something. Teachers can to some extent choose the content for which the language will be used. If we agree that being a communicative competent user of a language, also means to understand (something about) the (sub-)culture of the language, it becomes apparent that a form of content in the language classroom can and should be “culture”. So biology teachers do NOT get all the fun. Language teachers can claim an enormous amount of interesting topics to teach, and this in turn will lead to learners that are more interculturally competent and language-aware.


A shared vision

The same note was struck in the last lecture of the day, by the Meesterschapsteam MVT, who showed their model on language teaching for the Dutch curriculum: intercultural communicative competence is the core competence, and this core evolves from three overlapping learning outcomes: cultural awareness, language awareness, and language skills, as depicted in the figure below.


Figure: Learning outcomes for a content-rich Modern Languages curriculum (adapted from  Meesterschapsteam MVT’s vision on the future of Modern Languages education)


A question of ‘perspective’

Apt content to foster these competences can be specified by approaching ‘language’ from different academic perspectives: language as a structural phenomenon, language as a cognitive phenomenon, and language as a cultural phenomenon. Insights from such theoretical models have direct and fun implications for language teachers in their everyday practice. Instead of borrowing topics from other subjects (reading and talking about “panda’s eating bamboo”, “(made-up?) hobbies”, “the environment”), teachers can choose topics that are relevant for language: therefore choosing topics that are about structures of languages (from phonetics to discourse), about cognitive phenomena (from top-down processing while listening to Zipf-distributions or child language learning), about social phenomena (such as status of dialects, the effect of language and its power), and about cultural phenomena (ranging from stereotypes about cultures to literature and other forms of art). Biologist teachers get all the fun? No way! Language teachers get all the fun!


Support for a content-rich vision

The rest of the day’s presentations illustrated through practical examples how language teaching can include teaching about culture and developing intercultural competence (ICC). From the multilingual school to partnerships across international borders, language-and-culture-integrated literature teaching, assessment of ICC, to ICC training for university staff and students, the day presented us inspiration from a whole host of research findings and good practices.

Reform can only succeed if there is support at grassroots level. If the attendance and enthusiasm with which this conference was received are anything to go by then perhaps the vision of content-rich language education is not so far off the mark.


Next year’s LLRC Day will be held in June 2020 and centre around the theme ‘The interplay between practice, research, and theories in L2 learning’. Keep an eye on the LLRC website for more information.


Do you use Cronbach’s alpha to check internal consistency? Don’t! Use Summability.  0

Do you sum questions of your exams to get final scores of students? Do you use a questionnaire with likert-scales? Do you analyze these questionnaires by taking the means of these questionnaire-items? Do you use the mean of questions in evaluation-forms? Do you average response times to items in experimental settings?


If you have answered yes to any of these questions, you may (or may not but should) have wondered whether the items in your questionnaire, exam, test, or experiment are (sort of) measuring the same thing, the construct you had intended to be measured. If so, it is more than likely that you have calculated Cronbach’s alpha and (if the value was over .7) happily reported that indeed, the items were internally consistent. If so, you have calculated and reported the wrong measure and you are not alone. Despite the fact that methodologists have shown numerous time that Cronbach’s alpha is not suitable for measuring internal consistency (see Sijtsma, 2009, for instance), in handbooks Cronbach’s alpha can still be found as the prime choice measure to be calculated. Because the intention of questionnaire and test constructers is to summarize the test by its overall sum score, Jelle Goeman (and myself) advocate summability, which is defined as the proportion of total “test” (questionnaire-subset, exam, evaluation) variation that is explained by the sum score.


Our paper recently came out in the journal Educational Measurement: Issues and Practice, in which we show summability to be a stable measure across a number of variables (including test or questionnaire length). From the few examples that have been calculated until now, and from insight in the mathematic formula, we can assume that a summability of .5 can be considered “high”. As yet, however, more experience has to be gained on summabilities of tests in various fields before definite recommendations can be given.


Therefore, I end this blog with a “Call for Calculations”: please go to ( and calculate summability yourself, for an existing test, exam, questionnaire, or experiment. You can download the R-code from the website, or use the link to the shiny-app. All you need is a table with items as columns and participants as rows, filled with participants’ scores on the items, supposedly measuring your (one) construct. The table can be in plain text-format or it can be an SPSS-file. Report your scores through the form available on the website. In this way, we will be able to gain a fast accumulation of knowledge of what constitutes “high,” “moderate,” and “low” summabilities. Thank you!