Discourse analysis for medics

This post is a response to a challenge. A few months ago, two GPs threw a gauntlet for me to write a crash course of discourse analysis. And so, here it is. Discourse analysis for the medic. This is the riskiest blogpost I have written. Inevitably, there will be gaps, omissions, contexts of which I will not think. Perhaps what I say below will be obvious and not worth saying, perhaps, and hopefully, it will be worthy of note. And so, with some trepidation here is the first and very tentative stab at it.

Before I continue, there is some clearing of the field to be done. DA is very rarely done on the hoof, as you listen to the person speaking. In fact, as you listen to someone speaking, you rarely notice anything. We, linguists, work with transcripts, so we can spend some time unpicking what was said. And so, from many years’ experience as an interviewer, I know how difficult it is to do analysis just on the basis of listening. I had many an interview after which I thought the informant ‘said nothing’ only to discover later that the interview was full of interesting things. But this is because I had the benefit of the transcript and the time to read it, code it, analyse it. And I could find all the nominalisations, passive voices, agencies and whatever else.

A doctor will have no such luxury. Any insight into how the patient says things must be observable here and now, as they say it. Moreover, in the short space of time, the doctor’s priority will always be ‘the symptoms’, regardless of whether seen as experiences or objective entities. Because I understand it, below I write about three things I believe you can focus on during a conversation, at least to an extent. What I am about to write can be nuanced, extended, detailed. But I intend this post to be the so-called ‘food for thought’, I want to suggest options and certainly do not wish (as if I could) provide easy and clear solutions.

And so, here come the three things a discourse analyst recommends that you, medics, pay attention to.

1. The first thing I would suggest considering you questions. There is no doubt that doctors are aware of the kind of questions they ask, but focus mostly on what they ask. I’d like to draw attention to how they are asked. As an example, I would like to quote a study of psychiatrists’ questions by Justyna Ziółkowska.

She considered the positions (in Davies and Harré’s understanding – plenty of references on the Internet) doctors’ questions projected on their patients and it turned out that doctors ask questions in three ways. First, they positon their patient as an ‘observing assessor’ who will make a detached assessment of themselves or their problems. Ziółkowska shows questions such as:

  • Did you notice…?
  • Do you feel X or Y?
  • Did X increase or decrease?

Second, it’s the positon of the ‘informing witness’, asking the patient to verify information:

  • Have you got thoughts of resignation?
  • Do you remember…?
  • When did it start?

Incidentally, I have written about the ‘when questions’ – I continue to be puzzled by them.

The third positon is that of the ‘experiencing narrator’. This is the only position which allows the patient to speak for themselves, so to say. These are simply questions when the doctor asks me to tell them something, tell them a story. Interestingly, the request to tell a story was by far the rarest of the questions asked in Ziółkowska’s study.

As it can be seen, I think, the three kinds of questions open up different spaces for the patient in which to tell ‘their story’. You can immediately ask which is the ‘good/best way’. Of course, there is no ‘best way’. It all depends on what you want to achieve with your question. What I am only suggesting is that, as medics, you should understand what question you ask and what kind of answer you can get on the basis of it. And while you might want to ask, say, an ‘observer question’, if this is useful at a particular juncture of your consultation, ask it, because this is the question you do want to ask.

There is always more to be said about questions and let me mention a couple more things. We, linguists, say that questions have a focus, they show your interest in the thing you ask. This is why asking questions is not easy. If your first question is, for the sake of argument:

What’s troubling you in particular?

you focus on different things than if you ask

What are your symptoms?

Incidentally, I have been asked the latter a zillion times and I am yet to be asked the former question (or something similar). And here comes a non-medical story.

Quite a few years ago I was doing a research project on the Polish-German border, which is a river. We arrived in the Polish town of Zgorzelec, which in 1945 was created out of the eastern part of the German town of Goerlitz. One of the things I wanted to ask the inhabitants of the Polish town was whether they went to….exactly, where? Germany? Goerlitz? Across the border, across the river, the other side, I had no idea. But I was conscious that if I ask ‘my way’, I will lose ‘their way’ and I was interested in ‘their way’.

My question is: What do you lose when you ask about ‘my symptoms’? You see, when I ask questions, my priority is to hear something without actually asking about it. In such a way, I realise that it is (narratively) important. The moment I have to ask the question, I lose the opportunity that the person sitting in front me might actually tell me themselves. But this means I always have a strategy of asking questions. I start with the most general, then go to the more detailed questions and then end with general ones. There are other strategies, of course, but whatever I do, I have a plan. I accept that I might be slow(ish), but I am yet to see a plan in the consultations. You just ask questions, as if they were just pushing buttons. They don’t, so how about pausing to think, I’ll wait.

The final thing that comes with the symptoms question is whether your patient actually understands you. I am referring to the work of Mishler who noted that you speak ‘medicalese’ and I speak ‘lifeworld’ (which makes your job, to some extent, translating). When you talk symptoms, how do you know that I understand your question the way you do? Once, when I was (again) asked about the symptoms, I said I didn’t know the ICD-10 diagnostic criteria. The doctor’s jaw dropped and said – prepare for your jaw to fall – ‘How do you know about the ICD?’. I collapsed. They really didn’t understand…

And finally, do you know what I would like to be asked? I would like to be asked something like:

What are the three things that trouble you the most?

How about helping me (and you) order things, prioritise. You will know what to write in the notes…


2. Now, let’s focus on what your patient says. The first thing I would like to suggest is that you look at  who is talking. There are two aspects of this. The first aspect is the formal-linguistic construction of the speaking subject. That is to say, do I speak as ‘I’?

Let me exemplify it by my research on men’s depression. One of the most striking aspects of the research was that the men talking to me never spoke of themselves as ill. Whenever they spoke of their illness, they chose to use pronouns such as ‘we’, ‘you’, ‘one’ etc. Never “I”. I interpreted it as their unease with their illness, trying to avoid speaking directly about depression. I was also wondering about insight, which in depression is said to be almost universal, but then is it? What also struck me was that I was the first to actually observe it!

Needless to say, there isn’t a clear and obvious way of interpreting such ways of speaking.  I just want to say that in my view it is significant that I might come along with ‘my problem’ and choose not to speak as ‘I’, regardless of whether you can make sense of it or not.

The second aspect of who speaks is who they speak as. Regardless of common assumptions, a patient can take a number of social positions, from a man, through a father, all the way to an ‘old professor’. Incidentally, the men in depression mostly spoke as ‘breadwinners’, for them getting better meant going back to work. And yet,  for some inexplicable reason they were never helped to go back to work, but doctors insisted they should be helped with improving their mood. Go figure.

There is one more thing here. If you go back to the two hypothetical questions above, they also project what I should speak as. Should I speak of myself, or should I speak of these objective symptoms? Which one do you prefer?  I am more than likely to tell you different things.

I must admit that it also always irritates me when you refer to yourself with ‘we’, doctor.


3. The third aspect of what your patient says is how certain they are. For linguists, certainty is to do with modality, yet, I suspect this is not exactly a place for describing it in any detail. Let me just say that when you are at your most certain, you simply say things in the third person. So thing like this:

  • God exists.
  • Water boils at 100 degrees.
  • This is the treatment for your condition.

No qualifications, no hesitation, no personal perspective – the third sentence, incidentally,  I heard a couple of months ago from a doctor and I was quite taken aback by it. These sentences simply describe the world as it is. Outside these, language can render a spectrum of certainties. From adverbs such as ‘probably’, or ‘certainly’  through ‘I think’, all the way to modal verbs such as ‘may’, ‘might’ etc.

And here, let’s just go back for a while to the ‘observing assessor’ questions I briefly discussed above.  Apart from putting me in a position of an assessor (as if I had this self-monitoring device which provides me with data on myself), they also positon me to speak with relatively high certainty. In other words, I’m supposed to say what ‘I noticed’ and not what ‘I think I noticed’ or ‘I might have noticed’, let alone what ‘I felt.’, all of which change significantly the certainty with which I speak and, indeed, wish to speak.

From my own experience, I know that I modulate how certain I want to appear, sometimes, I simply don’t know. However, I am yet to see clinical notes which make reference to what patients say recording at the same time the certainty with which they speak. I’m not a medic, yet, I would imagine that if patients change levels of certainty, they must think it important and so should you. No?

There is, of course, a question for you. With what certainty do you speak, doctor? Do you stop to think about it?

Now, I want to stress that there aren’t any easy interpretations of the phenomena you notice. They most certainly (can you see how I lowered the certainty?) mean something, but what they mean will always be context-dependent and may remain unclear. After all, you don’t know how your patient talks outside the surgery. So whatever you do, interpret with caution. Do remember, however, that you can actually say things like: ‘But you don’t sound certain?’.


As I finish this post I keep wondering whether there is more to be said about ‘medical DA’.  At the moment I still think the three things I described are likely to be the three things that can be looked at by a medic. But there is a counterpart of this post. A linguist’s view of medical consultation. How we (whoever that may be) see it. So, I shall also tell you what I, a linguist-patient, pay attention to.




  1. Martin Stockler

    Excited and honoured: I think it’s my first question to be answered in its own blog. And sorry about the typo for q3. It Should of course be:

    ‘Are there some other things that you think we should talk about?’

  2. PS I have just notice the other comment from Martin
    In teaching with GP trainees we often talk about openings
    I walk out to the waiting room to call the patient in If I don’t know then I usually check I have used/pronounced name correctly and introduce myself as we walk down the corridor into the consultation room I then usually gesture to the patient I sit and await their opening gambit without a question from me unless they don’t start talking – feedback on this approached welcome (but awaited with a little trepidation)

  3. Thank you so much for this Dariusz As one of the 2 GPs who threw down the gauntlet I feel honour bound to respond in a timely fashion However I am struggling as the approach you are describing and explaining is so very different from the medical literature on communication/consultation skills that I am in a degree of culture shock I think I may need to follow some of the background reading you suggest So here are some initial thoughts – which I may revise and reject!
    Firstly I am troubled you have never been asked what is troubling you in particular I reflect back to a Twitter conversation about open questions and I think I now understand why I didn’t understand what you meant then I try and do (and try and teach) to start with an “open” question like your “what’s troubling you” in order to start the consultation with the intial narrative as presented by the patient I then try to go with the flow of what the patient brings so if they respond with symptoms then I will go down the symptom route first if they respond with psychosocial concerns I would go down that route first
    I appreciate that I may fail to so this often and also that this strategy is opaque to the patient Hence why I try and recommend to trainees to signpost the next direction of the consultation and to summarise regularly
    I totally agree with the need for the patient to say themselves my strategy at the start of a consultation is to try and get the patient talking to tell me in their own words what is going on
    Your description of starting with general questions then becoming more detailed is what I would call the open to closed cone (as described in Calgary Cambridge model)
    I think there is one important area which may explain some of the aspects you are puzzled by for example the “when question” that is the clinical decision making process that is going on inside the doctors head which is driving the questions that puzzle you The when question is so important to us because it is vital to the diagnostic process as especially in general practice it has such power to predict for example a cough that has gone on for one week has a completely different differential diagnosis to one that has lasted 3 months To be honest we often just can’t stop ourselves from asking that question early on because it helps so much but IMO it often distrupts the patients narrative when used too early on
    The other issue is that cognitive load of trying to decide what is going on medically at speed and emotional labour of dealing with misery
    I think often drives out what could be both more humane and more effective communication strategies
    For example I have never seen it written in any book but every working GP knows the freely volunteered information is usually more valuable than the yes/no to a closed question
    I’m going to stop my rambles there and go away and think some more
    Thanks again

    1. Dariusz Galasinski

      Alan, thank you very much for your response. I do appreciate it. Here is my response, though I would like to leave some things out as I will write about them in a couple of posts.

      1. Yes, I agree, there is much difference between what ‘we’ say and what ‘you’ say. I suspect the difference is because ‘you’ want to teach clear communication rules, ‘we’ think they are very hard to come by. And so, we are hardly of use, when we always say ‘it depends’.
      2. I am not troubled at all. In my academic life I meet doctors who are way more reflective on their practice and understand it much more than I would have assumed. In real life things are not that wonderful and rosy. But I am still unclear whether I would prefer a poor communicator whom I perceive as caring or a wonderful communicator who is cold and heartless.
      3. The example of the above point is the ‘Calgary Cambridge model’. When I started being interested in medicine and medical communication, I started looking at what doctors do. and I am yet to see the strategy. And I think we, patients, more or less consciously, cope with it trying to help you out. That, needless to say, creates a very interesting space for the patient and her/his role.
      4. For me the crucial bit you mention and one which I am painfully aware of is the cognitive load. Seeing me for a few minutes, you have a huge task, which must be underpinned by your own anxiety to help me in my misery. And so, when I tell you that you should also pay attention that this, that and the other, you may well tell me to go away. There is only so much you can deal with. And I understand that fully and this is precisely why I only raised three things.

      And I have no solution. As non-medic, I think that paying attention to what and how your patient says things can be useful. But I am not in your seat dealing with a grumpy and clever-smug-irritating patient like myself.

  4. Thank you for the wonderful post. I’m a medical oncologist interested in what and how things are discussed with people affected by cancer.

    I was heartened to see that your preferred opening is.similar to mine: ‘what are the main things troubling you?’

    I’d value your opinion on my two other favourites:’how can I help you today’, and ‘are the some other things that you think we should talk about’.

    I’m very much looking forward to your next installments.

    Apologies for clumsiness: will not do this on my iPhone next time!

    1. Dariusz Galasinski

      Thank you very much for your kind words. I hope you don’t mind waiting for a couple of days for an answer. I think the issue of the ‘first question’ is worth a longer response, so I thought I would write a blog about it.

Comments are closed.

Loading ...