I don’t like psychological and psychiatric questionnaires. By reducing human experience into an item, a sentence or a dot on a sheet of paper, they take away what is crucial for experience: a story. As I say this at lectures or other speaking events, I am met with polite nodding and, oftentimes, reassurances that questionnaires are only of minor assistance, very readily discarded. I am met with some hostility, when I start unpicking the linguistic form of such instruments. And this is what this post is about. I want to show why the linguistic form of psychological instruments is important.
Before I start, I need to repeat what I said a number of times. Each language use carries its share of assumptions, beliefs and values, so that the reality represented is also constructed ideologically, from a particular point of view. There is no neutral language, there is no language without a perspective conveyed by the text. In this post I want to write about such a perspective in one of the most popular questionnaires designed to gauge the intensity of symptoms of depression, the Beck Depression Inventory (BDI).
The BDI (in its third version) has 21 items, all of which consist of 4 options, from 0 (sort of: no symptom at all) to 3 (plenty of symptom). In a way, you really don’t need to read the items in their entirety, you can simply remember – 0 means little, 3 means plenty and Bob (or maybe Dr Beck) is your uncle.
So, I want to look at what kind of communicative actions are imposed onto the respondent by the questionnaire, in what kind of positions those who complete it are placed in (I have written considerably more about the BDI in articles and the book Men’s Discourses of Depression, where you will find references). Consider the following zero options from a number of BDI items:
- 1-0. I do not feel sad
- 2-0. I am not discouraged about my future
- 5-0. I don’t feel particularly guilty.
- 6-0 I don’t feel I am being punished
- 8-0 I don’t criticize or blame myself more than usual.
- 9-0 I don’t have any thoughts of killing myself.
- 10-0. I don’t cry any more than I used to..
What is striking is that they are all in the negative. Thus, given the first person singular format, socially, they take the form of a denial. To explain: normally, when I make a positive statement, e.g. saying ‘I feel sad’, I assume that my addressee has not got the information and for one reason or another, would like to have it or I would like them to have it. This is the basis of Grice’s Cooperative Principle which stipulates, among others, that one should not say more than necessary. Things are different when we issue a negative statement. Normally, when I say that I don’t feel sad, I assume that my addressee might have grounds to believe otherwise. By volunteering a statement in the negative, I perform a denial.
Imagine a different scenario. Imagine coming back home that saying:
- I haven’t had an accident.
Such a statement is felicitous only if the person expects you to have had one. So, for example, after two days of having little bumps on the way home, as your third day is accident-free, you would plausibly say the above. Otherwise, you are considerably more likely to wait till you have a positive announcement to make, e.g. that you have had an accident.
And here is the one of the crucial issues with the BDI. The BDI doesn’t want the respondent to tell the clinician that they have ‘a symptom’. The BDI asks the respondent to deny having it. And in doing so, the BDI actually assumes and, in the process, co-constructs depression! How can you ever not be depressed, if the Beck Depression Inventory takes the point of departure of depression? It assumes that whoever completes it, actually has depression, or at least has certain symptoms which count as contributing to depression.
Yet, I hear you say, the BDI was designed to gauge the ‘severity of depression in psychiatrically diagnosed adults and adolescents’, then perhaps the assumption of depression seems no to so perturbing. However, the assumption constructs every diagnosis of depression as correct, a position untenable in view of a relatively large body of literature on the problems faced by nosological diagnosis, including Beck’s own classical studies! More disturbingly, the BDI is ‘one of the most widely accepted instruments (…) for detecting possible depression in normal [sic!] populations’. In such studies, quite obviously, the assumption of depression is simply untenable. It’s worth pointing out, incidentally, that the perspective of depression taken in the BDI is underscored by the fact that the nil score on the scale (i.e. the respondent denied having all of the symptoms tested by the instrument) still, according to the authors, indicates depression!
The questionnaire’s use of the first person singular adds to the problem. Reading the questionnaire, the respondent is put in a speaking position. Taking the perspective of ‘I’, the scale is putting words in my mouth. Thus, it is actually I who suspects that the world around me thinks that ‘I’ am depressed. ‘I’ might be tempted to deny it, but then, perhaps the world knows better, after all ‘I’ am completing the Beck Depression Inventory, given to me by my doctor… Perhaps I do have depression.
Whenever I make these arguments, apart from hostility, the most frequent argument I get is that, no matter what I say, it works! After all, people have different scores during and after therapy – indeed improving BDI scores are taken to be evidence of therapy’s effectiveness. So, the argument continues, people might not exactly know what the author meant, still, they are able to fill in the scale consistently and one is able to have insight into their depression.
I don’t like the argument because it relies on the patient’s ‘professionalism’. S/he knows what is involved and plays along so that the doctor can have the data they need. But patients understand very well what might be at stake – especially in a psychiatrist’s surgery – so they cooperate more or less willingly trying to give the ‘right’ answers.
The model of quantifying the experience of depression necessarily results in the dominance of medical discourse. The patient’s experience cannot surface as s/he plays the game, getting more ‘professional’ every time s/he is asked to complete a questionnaire. The diagnostic instrument becomes a goal in itself, and the situation is exacerbated by the fact that psychological and psychiatric testing does not account for the power relations involved in the activity. In my view, it is crucial to understand that completing the BDI, and indeed other such instruments, is a socially situated, context-dependent activity, with its particular configuration of power relations, particularly relevant in the context of a psychiatric examination.
And the fact that patients don’t challenge doesn’t mean that they cannot or don’t want to. In a piece of research, where people were encouraged to do what they wanted with the instrument, a number did actually correct it! Below is one such sheet where the respondent corrected the BDI’s ‘little assumptions’:
The BDI cannot be seen outside its anchorage in the dominant discourse of psychiatry and clinical psychology. It works within the parameters of such discourse, and so it successfully measures something, because it corresponds with the rules of what constitutes such measurement. And while it might identify the (Major) Depressive Episode (ICD F32-33 or DSM, 296.2-3), it is unlikely to pin down the experience of low mood, sadness, the experience of what we call ‘depression’.
And so, we come to my radical conclusion after which I tend to be shouted at. The obvious final question to be considered is the value of the BDI (or indeed any other standardised psychiatric instrument’s) score. Does it make sense to ask people to complete it? As a linguist, and I only speak as such, I cannot see the point of it. Completing the questionnaire is part of a set of practices that turn people into patients and professionalise their patienthood. The practice of administering such scales is not much more than a means of getting the patient to submerge themselves into the dominant psychiatric or psychological discourse. A clinical evaluation of this point is outside the realm occupied by linguistics.