The easier ask.

Same wage, not the same shape

The founding insight of functional analysis is that behaviour works for the person: it earns something. A replacement therefore has one job — to earn the same thing more cheaply, more safely, and with less cost to the person’s life. It does not have to look like the behaviour it replaces. It has to pay like it.

This is also the guideline position. NICE NG11 recommendation 1.6.1 asks services to develop an alternative behaviour that achieves the same function — not a quieter version of the same behaviour, and not a rule against it.

The failure mode the law exists to prevent is familiar: a replacement chosen because it looks better to the team, which pays a different wage or no wage at all. It is taught, it fails, and the failure is attributed to the person.

The ask is a plan step, with its own dials

• The ask is taught as its own step — node kind behaviour_response, strategy teach_fct — so it appears at the moment it is needed rather than in a document read once a year.
• It is paired with reinforce_success: honour the ask the moment it lands, on first uses, every time. A new ask is competing with a behaviour that has always worked; if the ask is ignored even occasionally, the person learns the old form pays better.
• The acknowledgement is thinned as the ask becomes habitual — the same fading discipline the assistance and cue rails use, applied to the payoff rather than to the help.
• The height of the arc is the ask the person starts themselves, unprompted. In the live record that shows as the cue rung falling to none on dated rows for the taught step.

How equivalence is proven rather than declared

The system never asserts that two things are equivalent. It holds three conditions and checks them against the person’s own episodes:

• Same occasion. The new form appears in the situations that used to produce the old one — not in easy moments chosen for the demonstration.
• Different form. It is genuinely a different act, recorded as such, rather than the same behaviour relabelled.
• Same settle. Afterwards the person comes back to their own baseline in the way the old behaviour used to achieve. Their history is the control; there is no population norm anywhere in this.

All three come from rows the carer already recorded, which is what makes the claim examinable. A practitioner can look at the same evidence and disagree — that is the point of keeping it in the record instead of in a conclusion.

Consistency is a palette problem, not a person problem

Teaching already happens. Staff swap behaviours for easier asks constantly, informally and often brilliantly — and the record loses it, so the next carer invents a different swap and the person is taught two competing answers to the same problem.

Behind the palette sits an authored strategy library: forty evidence-attached entries covering ten functions across four kinds of response, each carrying an example, a non-example, and its citations. The entries are an authored distillation of published practice guidance — principles cited, sources dated, wording ours — not copied text. The non-example discipline is the part that does the work: each entry states plainly what would not count, which is how a palette stops drifting into a list of nice ideas.

The evidence behind the easier ask

Functional communication training (Carr & Durand 1985) is among the best-evidenced behavioural interventions in the field: teach a communicative form that earns what the behaviour earned, honour it, and the behaviour loses its job. It descends directly from Carr’s 1977 reframing of self-injury as operant rather than pathological — behaviour that works for the person.

NICE NG11 carries it into UK guidance at recommendation 1.6.1. At the pre-verbal end, Intensive Interaction (Nind & Hewett) supplies the lineage for asks that are not words at all: responsiveness, imitation and shared rhythm are the medium in which a first ask can appear.

What this puts in front of a practitioner or a panel

“Prove the intervention works” is the first question a funding panel asks, and it is the question a teaching arc is unusually well placed to answer — if the arc was recorded as it happened.

• The taught ask as a plan step with a date, so the intervention has a start rather than a recollection.
• The occasions it was used on, beside the occasions the old form used to occupy — the comparison the equivalence claim actually rests on.
• What happened afterwards each time, in the person’s own terms, so “it works” is a pattern in the record rather than a sentence in a review.
• The behaviour-support plan itself — recognise, read, respond, by phase — printing from the same substrate, so the plan a panel reads is the plan the shift is running.

Questions, answered the way we answer them

The rest of the grammar

• An ABC whose A can be counted — the context pieces that make antecedents computable.
• Fifty years to find the function — where the equivalence law came from, and the chapter still open.
• The fade — the same fading discipline applied to support rather than to payoff.