What do our (qualitative interview) informants do? It’s a question that I have not asked myself much over the years. But when it hit me, I realised that I wasn’t alone. I think not many of us, qualitative researchers, stop to think about it.

It all started when I was preparing to write my book on suicide notes. Take this excerpt from a suicide note (in English translation from the Polish original):

If you can, buy me some flowers. OK?

What did the author of the note do? In other words, is the sentence a request, an instruction, a plea? Or perhaps something yet different written in a context I have no access to. Moreover, is it ‘anything’? Perhaps it’s just a sentence in the imperative mood, and sentences in the imperative are commonly, although certainly not exclusively, used to issue a ‘directive’ act of communication (speech act, if you prefer old money).

It’s worth pointing out that it’s not just a linguist’s fancy that took me to consider the matter. Directive acts of communication, that is to say those acts whose point is to get someone to do something, are located on a spectrum of the force of the instruction, which is related to the power of the instructor. And so, in an order the powerful speaker assumes that the addressee cannot not comply; in requests or advice, the speaker, who may but may not be of equal power, assumes the addressee may not comply; in pleas in turn, the speaker is at the mercy of the addressee who is under no obligation to comply at all, even though the speaker is at pains to get the addressee to do what the object of the plea is.

And here we see the issue. Interpreting the sentence in one way of or the other also interprets the relationship constructed by the speaker between her/him and the addressee. Moreover, as I label the acts of communication in question as instructions or requests, I also (presume) to decide what the writers had in mind when writing their notes or how they intended them.

Of course, the example above has a number of linguistic devices which make my life easier. The conditional form of the sentence together with the ‘OK’ at the end of the fragment, probably mitigate the imposition and suggest that the author of the note doesn’t want command the addressees. Except, if what was written were to be said, in a particular tone, that polite request, could very easily be translated into an irritated command.

So what’s the solution? Well, there is no solution. In the book, I step back from any interpretation and limit myself to saying that the notes contained ‘directives’. This is what the linguistic form tells and, in the context, it is likely to be used in its primary imperative meaning.

Now, as I flog myself with a methodological whip, I am not going down on my own. Time and again, I see researchers describe interview data in terms of verbs of speaking (such as ‘say’, ‘apologise’, ‘warn’, ‘deny’, ‘refuse’, ‘request’ and so on and so forth) without much reflection given to the issue. And the more I think about it, the more I have a problem with it. Moreover, it’s no only qualitative research where this is an issue, it’s also clinical practice.

So, let me take the verb ‘disclose’. Time and again, I read about questions which allow patients to

to disclose their interests, values, and preferences, and it gives the clinical team a chance to appreciate patients as full humans and not just as recipients of care.

as one article puts it (I’m not giving a link, I quote a type of argument); arguments about informants disclosing things are also aplenty.  And I really dislike the verb. It suggests what I say is hidden and, perhaps, secret, while my conversation with a medic is about sharing something which, normally, would not be shared. All of a sudden, the consultation about my cold is changing into a séance of mysteries.

There is more, though. I keep wondering whether if I simply said something, you would actually ‘appreciate me as a full human’, or do I not live up to your expectations, if I don’t go for the emotional full Monty. Incidentally, I do believe that it is really presumptuous on the part of any medic or researchers to suggest that somehow they get access to the ‘full me’ in an interview or a 10-minute consultation. Please, stop dreaming!

Just as in research practice, also in clinical practice, I would suggest using verbs of speaking sparingly and with much reflection. Well, also with some evidence to back it up! Telling you something doesn’t have to be disclosing or confessing, it also doesn’t have to be declaring or claiming, saying ‘no’ doesn’t have to be refusing or denying.

Now, I am not suggesting, of course, that it is impossible to use metalinguistic expressions. For example, if my interviewee said ‘I apologise’ or even ‘Sorry’, in a context, it is fairly reasonable to assume that the person apologised (though, I do stress the context). If I say ‘I promise’ I am likely to want to promise (in a given context). I am suggesting, rather, that verbs of speaking always interpret. And our interpretations do much more than offer a perspective on what people say. From setting up relationships (as with ‘disclose’ or ‘apologise’) to undermining veracity (as with ‘claim’), verbs of speaking can be hurtful, manipulative and very powerful. And they can also bend our arguments in a direction our fancy takes us.

There is no escape from language and its constructions. You can only replace one set with another. Go thinking and go replacing, safely. Stepping back.