Learning Clinical Hebrew

A doctor walks into their clinic in Israel. He has been studying hard (the flash cards, the podcast, the online Ulpan… everything!) He introduces himself to his first patient of the afternoon. She smiles, leans forwards… within seconds, she is talking at full speed and throwing in words that sound entirely unfamiliar. He feels he catches about every fourth word. He nods. A lot. She finishes and waits patiently for a response.

It is not the first time we find ourselves wishing the hours of study had counted for more.

This post is about why the standard tools fall short, and what a different approach might look like. (Skip to the end if you are in a hurry, but the evidence is worth the read.)

The Real Problem

Learning a language to the point where we can practice medicine comfortably in it is genuinely hard. It takes time, the right tools, and a fair amount of patience with ourselves. And for doctors coming to work in Israel, whether from the UK, the US, France, Russia, or South Africa, this challenge is shared across languages, not just English.

It is not a new problem. A few references worth reading if you want to go deeper: Divi et al., 2007 and Karliner et al., 2007. We all know, intuitively, that getting the language right matters. The research simply confirms what we already feel on the wards.

We Are Not Learning a Language.

We Are Learning a Professional Language.

While I would love to be able to discuss in ivrit where my colleague went skiing, or their favourite local cuisine, when I made aliyah as a doctor, I did not need to. I needed to be able to work.

Thankfully, the language of work is more narrow, structured and learnable than the full mountain of Hebrew might suggest from a distance. Whether you are a radiologist, family medicine doctor or cardiologist, you need a specific set of vocabulary. The clinical Hebrew that matters to us is specific, transferable across presentations. It is built around a finite set of situations we encounter again and again.

This is not a small point. It can change how we think about the entire task.

What We Reach For

The go-to resources are familiar to most of us. Ulpan is the backbone, subsidised, structured, and a solid foundation in grammar and vocabulary. For medical professionals there is a dedicated medical ulpan stream with healthcare-specific content. Beyond that there are podcasts, textbooks, private tutors, and language exchange programmes.

These all have real value. But they share something in common: they teach us Hebrew. What they do not do, and cannot really do by their nature, is put us in front of a patient and ask us to practise medicine in it.

Ulpan in particular is a real commitment. A full first-stage course runs to 500 hours. For many doctors, taking months away from clinical work simply is not possible. So language learning gets threaded around the edges of a busy working life, and the gap between what we are studying and what we are actually doing at work stays wide. We learn vocabulary from lists, practise grammar from exercises, and then show up to work and feel we are back at square one the moment a patient starts talking.

As an evidence based community, we should be asking: are these tools truly the most effective route to clinical fluency? The evidence from language acquisition research suggests they are not sufficient on their own.

What the Research Actually Shows

We have to produce language, not just receive it.

Linguist Merrill Swain noticed something striking in the 1980s when studying French immersion students in Canada. Despite years of exposure to French in class, their ability to actually speak it remained well behind native speakers. Her Output Hypothesis (Swain, 1985) proposed that producing language, being pushed to speak and write and receiving feedback on the result, drives acquisition in ways that listening and reading alone simply cannot. We have to be placed in the position of needing to say something and having to work out how.

Feedback is what closes the loop.

Producing language without correction can actually entrench errors. When we say something incorrectly and nobody addresses it, we learn to say it that way. Corrective feedback, specific, timely, and targeted, is what allows learners to notice the gap between what they said and what they intended, and to adjust. (Lyster & Ranta, 1997)

Context is everything for vocabulary.

There is substantial evidence that words learned in a meaningful context are retained far better than words from a list (Nation, 2001; Nakata & Elgort, 2021). Encountering כְּאֵב in the middle of a conversation about chest pain, when we needed it and could not find it, is an entirely different experience from seeing it on a flashcard. The clinical weight of that moment creates a memory trace that a vocabulary list simply does not.

Realistic practice builds readiness for the real thing.

Medicine has understood this principle for decades. It is why simulation exists. Practising in conditions that resemble the real situation, with some of the pressure and uncertainty of an actual encounter, builds the mental pathways that allow us to perform when it counts. (Issenberg et al., 2005) Language learning works the same way.

The Confidence Question

There is something else worth saying, and it does not get discussed enough. Some doctors do not make aliyah at all, or delay it for years, partly because the language feels like an insurmountable obstacle. Hebrew looks impossible from a distance, especially when we are sitting in Manchester or Melbourne with a textbook and a vocabulary list and a growing suspicion that it is just too hard.

But those of us who have been through it know something that is genuinely difficult to convey from the outside: we can do this.

Confidence in clinical Hebrew does not come from being ready. It comes from doing it before we feel ready.

That is, in a way, what the app is for.

Introducing: Clinical Hebrew

Clinical Hebrew App grew out of a need I could not find anything to fill. I did not take months off for ulpan, and I did not find anything that helped me practise the actual conversations I was having at work. I was making the same mistakes and getting stuck, again and again.

The platform puts us inside realistic clinical scenarios: a 60-year-old with chest tightness in the emergency department, a parent with a feverish toddler, an elderly patient in primary care trying to understand a new diagnosis. The scenarios are built around the situations doctors actually encounter in Israeli hospitals and clinics. We respond by speaking or by typing, and the conversation develops from there. It is built for real life rather than time set aside for it. Each scenario takes five to ten minutes; short enough for a commute, a coffee break, or the few minutes before a ward round starts. Scenarios run from beginner, intermediate to advanced, so we can start where we actually are.

A clinical conversation in progress. 

After each scenario, we receive detailed feedback on our language, including specific notes on phrasing, grammar, and what a native speaker would naturally say in that situation. This is the corrective feedback loop the research points to.

Along the way, we can save vocabulary directly from our conversations, not from a generic list, but from the moment a word came up clinically and we needed it. That context is exactly what makes it stick.

The platform works across a range of language backgrounds. English, French, Russian, Spanish, and Portuguese speakers can all set their native language and receive explanations and feedback tailored accordingly.

It is one tool, not a complete solution. The more language practice we layer around it, the faster things move. But it fills a specific gap that most other resources do not: practising the clinical language we actually need, in a way that the evidence suggests actually works. Ideally started in the months before arriving in Israel, it is just as useful once we are already on the wards.

A Final Word

Unfortunately, as we know and come to realise, there are no shortcuts to clinical fluency. Most of the real learning happens on the wards, ready or not, and that part takes years. But the method matters. Practising real conversations, getting specific feedback, and building vocabulary in context will get us further than passive study, and will make the first consultation with a fast-talking patient feel a little less like we have been thrown in the deep end without warning. And with time, we will also be able to comfortably ask our colleagues about their skiing holiday and their favourite local cuisine.

Three free scenarios are available at clinicalhebrew.com. 

✷ As a member of the Anglo Doctors in Israel community, use code ADII at sign-up for 20% off, permanently. Valid for sign-ups in the next two weeks ✷

Dr David Gottlieb is a paediatric resident at Schneider Children’s Medical Center in Petach Tikva. He made Aliyah from London in September 2025 and founded Clinical Hebrew to help doctors navigate exactly this challenge.