Social and Critical Psychology

Discourse and Subject Position Laboratory

Six months in a fictional person's life, written up five times. The events are identical in every version. What differs is the kind of person the account produces, what she may legitimately do next, who is entitled to say what is happening, and where responsibility lands.

Fictional case - the person, the service and all five accounts were written for this activity

Learning objective

By the end you should be able to identify the subject position an account makes available, and say which actions it renders reasonable and who it authorises to speak.

About 30 minutes. Nothing you do here is saved or sent anywhere.

Two things this is not

It is not a word-frequency exercise. Nothing here depends on which words appear most often; what matters is the position an account makes available to the person it is about, and what that position permits.

It is not a claim that the events are made of language. The ledger of events is fixed and stays on screen throughout, and one of the entries in it is a budget decision that only the advocacy bulletin states outright; the commissioning report acknowledges a reduced-hours model but recodes it as implementation performance, and the remaining three genres have no field for it at all. Analysing how something is written up is not the same as saying that only the writing is real.

How to use this

Read the ledger of events. It does not change.
Commit to an answer to the question below.
Choose an account. It stays pinned beside you while you answer four questions about it.
Read the feedback, then choose another account. The comparison builds as you go.
Analyse at least three, then read the synthesis.

R, the service, the assessment and all five accounts are invented for teaching. No real person, service, record or document is reproduced, and nothing here is a clinical description of anybody.

This part needs JavaScript

The laboratory presents accounts and evaluates your analysis in your browser, so it cannot run with JavaScript switched off. The ledger, the debrief and the limitations are readable without it.

Five accounts of the same six months are about to appear. What is the most important thing that will differ between them? Tone - some will be sympathetic and some will not Which facts are included and which are left out What kind of person R is made into, and therefore what she may do next How accurate they are - one of them will be closest to the truth

Key terms

Discourse: A patterned way of talking about something that makes certain descriptions readily available and others hard to produce.
Subject position: The kind of person an account makes its subject into, and what that person can reasonably be expected to do.
Legitimacy: Whose account is treated as authoritative. A description can distribute it without ever saying that it is doing so.

No account selected

Accounts you have analysed

For each account: the subject position it produces, what becomes legitimate, who is authorised to speak and where responsibility lands
Account	Kind of person	Legitimate next action	Who may say what is happening	Responsibility

An account produces a kind of person

"Non-compliant", "service user exercising choice", "self-advocate", "person taking ownership of her recovery" and "customer" are not five words for R. They are five positions, and a position comes with a specification of what someone in it can reasonably do. Somebody positioned as non-compliant who writes to a manager is escalating. Somebody positioned as a self-advocate who writes to a manager is doing the thing self-advocates do. The letter is the same letter.

Legitimacy is distributed by the description

Notice how the same act - starting a peer group in the library - appears in the five versions. It is disengagement, it is unmonitored risk, it is community organising, it is recovery-oriented self-management, it is a customer taking her business elsewhere. Each of those makes a different response by the service obviously appropriate, and only one of them makes asking R what she wanted obviously appropriate.

Every account authorises somebody

Accounts do not only describe; they nominate whose description counts. A case note authorises the clinician, a risk register authorises the safeguarding lead, an advocacy bulletin authorises R and people like her, a recovery narrative authorises R in a carefully bounded way - she may narrate her journey, but the service still decides what a journey is. Asking who is entitled to say what is happening is often more revealing than asking what is said.

Responsibility moves without anyone moving it

Between the case note and the recovery account, responsibility for the six months travels from R's illness to R's choices. In neither version does it arrive anywhere near the decision, in ledger entry 2, to halve the programme's hours. That decision is in the ledger and is not disputed by anybody, yet only the advocacy bulletin states it outright; the commissioning report acknowledges a reduced-hours model but keeps neither its cause nor its consequences; the case note and the recovery account, like the risk register, have no place for it at all. The most consequential thing a discourse does is often not what it says about a person but what it has no place to say about anything else.

Language, institutions and what can be done

None of this is about wording in the sense of politeness. The positions are held in place by forms, templates, review meetings, funding conditions and the professional training of the people who write them. That is why changing the vocabulary alone rarely changes very much, and why the recovery account - which changed the vocabulary comprehensively - can leave the distribution of authority almost exactly where it was.

What this does not demonstrate

Real discourse analysis is bigger than this. It works with much longer stretches of talk and text, and attends to what accounts are doing in the interaction they occur in. Critical realists and discursive psychologists disagree about a great deal while agreeing about that.
No account is simply the villain. A case note exists so that a clinician who has never met R can act safely at two in the morning. A risk register exists because services have harmed people by not keeping one. The costs are real and so are the reasons.
Recovery language is not a trick. It was fought for by service users against exactly the case-note position, and it made real differences. What the comparison shows is that a vocabulary can be won and then absorbed, which is a different and sadder point.
Five accounts is a simplification. A real file contains dozens, they contradict each other, and the same author shifts position within a paragraph.
Nobody is assessed. R is fictional. The tool asks nothing about the person using it and is not a method for analysing a real person's records.

The ledger: what happened

First, a judgement

The laboratory

Five write-ups, one set of events