Skip to content

Clarification Protocol

A response protocol — instantiates Premise–Action Decoupling

A response move that asks the sender to state the buried premise before answering, so the receiver can engage the real question without accepting it — and without escalating to a challenge.

Clarification Protocol is the front-line move for a receiver caught in a message whose premise they don't accept but whose surface task they still need to handle. Instead of answering (and appearing to concede) or refusing (and paying the cost), the receiver issues a clarifying request that pulls the buried premise into the open and asks the sender to make it explicit — and, in the same move, opens a non-ratifying response path: a way to keep the exchange going that neither affirms nor denies the premise. Its defining quality is that it is neutral inquiry, not challenge. It does not dispute the premise; it declines to accommodate it silently, forcing the sender to state it as a claim they now own. Where a challenge says "I contest that," the Clarification Protocol says "before I answer — what exactly are you assuming?" — which is often enough to make an unearned premise evaporate on its own.

Example

A support agent gets a ticket: "Why does your app keep deleting my files?" Answered head-on — "Sorry about the deletion, we'll investigate the bug" — the reply ratifies the premise that the app deletes files, logs a defect that may not exist, and routes the whole ticket down the wrong track. Refusing to engage — "the app doesn't delete files" — is a flat contradiction that helps no one and escalates.

The Clarification Protocol threads it. The agent replies: "Happy to get to the bottom of this fast — can you tell me which files, and how you noticed they were gone? Did you see a delete happen, or are they just missing from where you expected?" This engages the genuine at-issue problem (files the user can't find) at full speed, while carrying no position on whether the app deleted anything. The premise is now the user's to specify, not the agent's to inherit. When the answer comes back — "they're missing from my Desktop folder" — and it turns out a sync moved them, the false premise "the app deletes files" never entered the record as a fact, and the real issue is solved. The surface task got done; the premise was never conceded.

How it works

  • Spot the premise you'd be conceding — identify the specific proposition the straightforward answer would silently accept.
  • Ask, don't assert — pose a genuine question that requires the sender to make that premise explicit, framed as needed-to-help rather than as pushback.
  • Keep the surface task moving — pair the request with real engagement on the legitimate underlying need, so clarifying doesn't read as stonewalling.
  • Hold a neutral stance — neither affirm nor deny the premise; leave it as the sender's claim to state and, if they wish, support.

The distinguishing discipline is neutrality under engagement: it surfaces the premise by inquiry while remaining fully cooperative on the surface task, and it stops short of contesting. Disputing the premise, once surfaced, is a different and later move.

Tuning parameters

  • Directness of the ask — from a soft "just so I understand…" to a pointed "what are you assuming about X?" Softer preserves rapport and works when the premise is honest error; pointed is faster but edges toward challenge.
  • Engagement pairing — how much of the surface task you advance alongside the clarification. High pairing keeps cooperation visible and disarms the "you're dodging" read; low pairing is cleaner but can feel like a stall.
  • Persistence — how many rounds you'll re-ask if the sender keeps re-burying the premise. More persistence resists a determined presupposer; too much becomes its own kind of standoff.
  • Escalation threshold — the point at which a still-unstated or now-stated-and-false premise should be handed off to an actual challenge. Sets the boundary between clarifying and contesting.

When it helps, and when it misleads

Its strength is speed and low cost: a single well-formed question defuses a loaded framing without a fight, keeps the legitimate task moving, and shifts the burden of stating (and owning) the premise back to its sender — where an unearned premise often quietly dies. It is the least escalatory tool in the archetype and usually the first to reach for.

It misleads when it is stretched past its range. Against a sender who insists on the premise, endless clarifying becomes evasion of a decision that now needs to be made — at some point the premise must be accepted, tested, or challenged, not clarified again. And clarification can be weaponized into feigned obtuseness, asking "what do you mean?" in bad faith to stall a request that was perfectly clear. Its most important limit is that it is not a challenge: it surfaces a premise but does not refute one, so a premise that is stated plainly and false has outrun this tool. The discipline that keeps it honest is to clarify in good faith toward genuinely getting the surface task done, and to escalate to contestation once the premise is out in the open and still wrong.[1]

How it implements the components

Clarification Protocol realizes the inquiry-and-response components — the receiver's low-escalation moves:

  • clarification_request — its core act: a genuine question that compels the sender to make the buried premise explicit rather than let it ride.
  • non_ratifying_response_path — it keeps the exchange moving on the surface task along a route that neither affirms nor denies the premise, so engaging does not equal conceding.

It does not contest or refute a premise once it is stated (contestation_channel) — that stronger move is the Presupposition Challenge Marker; and it does not classify the message's force (response_coupling_rule, engagement_assent_coupling) — that is Force Classification.

Notes

Clarifying and challenging are distinct stages, and conflating them weakens both. A clarification asked as a disguised accusation loses the disarming neutrality that makes it work; a challenge softened into "just clarifying" fails to actually refute a premise that deserves refuting. The protocol's job is the first stage only — get the premise stated and owned — and to hand off cleanly to the Presupposition Challenge Marker when what's needed is not clarity but contestation.

References

[1] Presupposition accommodation (Lewis; Stalnaker) is the hearer's habit of quietly adjusting the conversational common ground to accept a speaker's presupposition so the utterance makes sense — the very reflex that lets buried premises slip through. A clarification request is a deliberate refusal to accommodate: it makes the speaker put the presupposition on the record instead. ↩