Specialist Language Courses

Beyond the word list: what corpus research offers medical English teachers

Beyond Word Lists_ Clinical Language in Context

A new classroom model for teaching medical terminology builds on a large body of research. We look at what the evidence says, and what it means for EMP lecturers and OET teachers. 

A new model from Tashkent

In October, Diyora Isomiddinova of Tashkent State Medical University published a model for teaching medical terminology through data-driven learning (DDL). In DDL, learners study examples of real language taken from a corpus (a large, searchable collection of texts), looking at many examples together and analysing patterns to understand meaning and usage. 

As others who use corpora when teaching, Isomiddinova argues that medical students need more than definitions. They need to know how often a term is used, which words it combines with (its collocations), and how it behaves in real clinical and academic texts. Her model has five stages: Notice, Investigate, Interpret, Practice and Produce. At first, the teacher prepares concordance lines (lists of examples that show a word in its context). Later, students search a corpus themselves, with guidance. Finally, they use the language in medical speaking and writing tasks. 

The model is designed for classes with mixed levels of English and for learners who have not used corpus tools before. It supports the teacher’s explanations and develops students’ subject-specific language practice. 

The article is a review of earlier studies – it does not test the model with real students. However, the research it draws on is large and well established. 

Does data-driven learning work?

The strongest evidence comes from a meta-analysis by Alex Boulton and Tom Cobb, published in Language Learning in 2017. A meta-analysis combines the results of many separate studies. Boulton and Cobb analysed 64 studies with more than 3,000 learners. They found large positive effects. This was true when learners were tested before and after DDL, and when DDL groups were compared with control groups. 

In 2021, Boulton and Nina Vyatkina reviewed thirty years of DDL research. The idea is not new. It goes back to the highly influential work of Tim Johns and others at the University of Birmingham, who used the term in 1990. 

Comparing clinical English patterns together

Ready-made resources for medical English

Medical English teachers do not have to build a corpus themselves. Researchers have already produced several useful lists. 

  • Medical academic word lists. Wang, Liang and Ge’s Medical Academic Word List (2008) is an early example. In 2022, Tazik and colleagues built a new list from more than 18 million words of research articles in The Lancet, the BMJ and the New England Journal of Medicine. The list is open access and potentially very useful for students who need to read medical research – and the teachers who support them. 
  • Multi-word terms. In medicine, meaning is often carried by phrases, not single words. Florescu and Ohniwa at the University of Tsukuba analysed 28 million words of medical textbooks and identified 3,307 multi-word terms. Most of these terms were also frequent in a second medical corpus. Interestingly, almost none were frequent in general English, which shows how specialised medical phraseology is. 
  • Medical morphemes. As we know, many medical terms are built from Greek and Latin parts, such as cardio- or -itis. A study in English for Specific Purposes identified 136 common medical morphemes. Together, they make up 8.5% of a large web corpus of medical texts. This is a manageable learning target. 

From terminology to talk: why this matters for clinical communication and OET

Most corpus work in medical English looks at written texts. However, many students need help developing their clinical communication skills and as do healthcare professionals studying for OET. Here, the work of Shelley Staples at the University of Arizona is especially relevant. 

Staples recorded 102 nurse–patient interactions: 52 with internationally educated nurses (IENs) and 50 with nurses trained in the United States. She then compared the language that each group used (English for Specific Purposes, 2015). The main differences were not about medical vocabulary. They were connected to rapport (a good relationship with the patient) and patient-centred communication. Staples also showed that one language feature can do different jobs in different phases of a consultation. For example, nurses used first-person pronouns differently when they examined a patient and when they gave advice. 

Staples then used her corpus to design a Pronunciation for Nurses course (English for Specific Purposes, 2019). The corpus informed every step: the needs analysis, the materials and the assessment. Features such as pitch range became teaching targets. This is a clear example of a complete corpus-based curriculum in healthcare English. Staples presented this work in an EALTHY and SLC webinar in 2024. The recording, slides and references are all available in the EALTHY member’s area.  

For clinical communication and OET teachers, the message is simple. A corpus can show how clinicians speak: how they open a consultation, show empathy and give advice and thereby provide an effective model for learners to use.  

Getting started

  • Start on paper. Choose one term and print 10–15 concordance lines. Ask learners which words come before and after it, and what this tells them about its use. 
  • Use free software. AntConc, developed by Laurence Anthony, lets you search a small corpus of texts you are permitted to use, such as your own case notes or course materials. 
  • Use the published lists to decide which terms, phrases and word parts to teach first. 
  • Move from noticing to producing. End each activity with a role-play or a writing task, such as a referral letter, so that learners use the patterns they found. 

While Isomiddinova’s model has not yet been tested in the classroom, at least in terms of her research, it offers a clear route from a strong body of research into everyday teaching. 

References

 Boulton, A., & Cobb, T. (2017). Corpus use in language learning: A meta-analysis. Language Learning, 67(2), 348–393. Boulton & Cobb (2017) 
 Boulton, A., & Vyatkina, N. (2021). Thirty years of data-driven learning: Taking stock and charting new directions over time. Language Learning & Technology, 25(3), 66–89. Boulton & Vyatkina (2021) 
 Florescu, C. M., & Ohniwa, R. L. On the creation of a corpus-derived medical multi-word term list. Information, 16(2), 118. Florescu & Ohniwa 
 Isomiddinova, D. Q. (2026). Teaching medical terminology through data-driven learning: A corpus-informed model for English for Medical Purposes. Maktabgacha va maktab ta’limi, 4(10), 199–208. Isomiddinova (2026), Zenodo record 
 Staples, S. (2015). Examining the linguistic needs of internationally educated nurses: A corpus-based study of lexico-grammatical features in nurse–patient interactions. English for Specific Purposes, 37, 122–136. Staples (2015), ESP 37 – University of Arizona 
 Staples, S. (2019). Using corpus-based discourse analysis for curriculum development: Creating and evaluating a pronunciation course for internationally educated nurses. English for Specific Purposes, 53, 13–29. Staples (2019), ESP 53 – Pronunciation course for IENs 
 Tazik, K., Maniati, M., Afshar, M. R., & Biparva Haghighi, S. (2022). Developing a new academic wordlist for medical purposes – a viable tool for educators. GMS Journal for Medical Education, 39(1), Doc9. Tazik et al. (2022) 
 A corpus-based list of commonly used English medical morphemes for students learning English for specific purposes. English for Specific Purposes. Medical morphemes list (ESP) 
 EALTHY/SLC webinar with Shelley Staples 
More To Explore
Peter Rodway
News
SLC expands into Uzbekistan

Partnerships · International SLC director Peter Rodway recently spent a week in Uzbekistan, meeting university leaders, lecturers and students to explore opportunities for Medical English

Read More »