{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp05_complete_proposal_portfolio20_20260803","cell_id":"negative_space_design__rhetoric","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_id":"negative_space_design__rhetoric__protected_warrant_silence","proposal_index":1,"version":0,"title":"Protected Warrant Silence for Live Argument Testing","problem":"During live testing of a proposed policy argument, moderators often fill the first silence after asking participants why a conclusion follows. They restate the claim, offer an example, or disclose the intended rationale before participants have articulated their own inference. The added speech occupies the interval in which the audience's actual warrant—or inability to supply one—could become observable.","actors":["Advocacy or public-information communication team testing an argument","Moderator conducting the message-testing session","Participants recruited from the intended audience","Research observer or analyst coding the session"],"observable_state":"In session recordings, the moderator supplies a restatement, example, candidate reason, or approving interpretation before a participant states why the claim follows. Participant responses then reuse the moderator's language, and the record does not reveal whether the warrant originated with the audience or was introduced by the moderator.","consequence":"The communication team cannot reliably distinguish an audience-supplied inference from moderator-assisted agreement. It may therefore advance an argument whose apparent coherence depends on explanatory material that will be absent in actual delivery.","affected_objective":"Elicit interpretable evidence about whether intended audiences independently connect a stated policy claim to its conclusion and which assumptions they use to do so.","intervention":"Use a Protected Warrant Silence protocol. After presenting a claim, the moderator asks one neutral question—“What, if anything, makes that conclusion follow for you?”—and displays a minimal cue stating “Thinking time; you may write, speak, or pass.” For a predeclared interval of up to 12 seconds, the moderator offers no examples, paraphrases, approval signals, or candidate warrants. Participants may record an initial response privately. The interval ends when participants indicate readiness or the limit is reached; the moderator then uses nonleading clarification questions. The intended warrant and illustrative explanations remain in recoverable moderator notes and may be introduced only after initial responses have been captured and labeled as post-capture material.","structural_mapping":[{"archetype_element":"Attention Competition Map","domain_realization":"The moderator's restatements, examples, facial approval, and candidate reasons are identified as elements competing with participants' own inference formation and expression."},{"archetype_element":"Omission Candidate","domain_realization":"Moderator-supplied warrants, examples, and reparative rephrasing are removed from the pre-response turn, while the original claim and neutral question remain."},{"archetype_element":"Protected Empty Space","domain_realization":"A bounded quiet interval follows the warrant question and cannot be filled by the moderator merely because no participant answers immediately."},{"archetype_element":"Positive Form Relationship","domain_realization":"The silence is tied specifically to the claim-warrant relation: its purpose is to make the participant's own connecting premise, uncertainty, or nonresponse observable."},{"archetype_element":"Absence Boundary","domain_realization":"A minimal thinking-time cue, a maximum duration, and explicit prohibited moderator behaviors distinguish the interval from accidental dead air."},{"archetype_element":"Meaning-of-Absence Check","domain_realization":"Silence is not coded as agreement. The moderator distinguishes thinking, confusion, refusal, accessibility need, and discomfort through participant-controlled readiness signals and later neutral clarification."},{"archetype_element":"Reintroduction Trigger","domain_realization":"Moderator speech resumes when participants indicate readiness, an initial response is captured, clarification is requested, distress appears, or the 12-second limit is reached."},{"archetype_element":"Accessibility and Recoverability Guardrail","domain_realization":"Participants may write, speak, request more time, ask for the claim to be repeated, or pass; omitted explanatory material remains available for a clearly labeled post-capture debrief."},{"archetype_element":"Clarity or Effect Test","domain_realization":"Recordings and first responses are coded to determine whether a warrant appeared before moderator explanation and whether participants understood the silence as deliberate thinking time rather than judgment or malfunction."}],"mechanism_mapping":[{"mechanism_slug":"editorial_cut","role":"Cuts candidate warrants, examples, and redundant restatements from the moderator's pre-response turn while preserving them in recoverable notes for post-capture use.","counterfactual_removal":"Without this cut, the moderator can occupy the nominally quiet interval with content that seeds the inference, so participant-originated and moderator-originated warrants remain confounded."},{"mechanism_slug":"facilitation_silence","role":"Requires the person with the greatest conversational authority to withhold their own contribution long enough for participants to supply, reject, or fail to supply the missing connection.","counterfactual_removal":"Without protected facilitator silence, conversational pressure is resolved by the moderator, and the audience's independent inference is not exposed."},{"mechanism_slug":"blank_or_rest_frame","role":"Uses a minimal thinking-time cue and a defined exit to make the quiet interval read as an intentional segment of the protocol rather than abandonment, disapproval, or technical failure.","counterfactual_removal":"Without the framing cue and exit rule, participants may interpret the absence inconsistently, making reactions to the silence inseparable from reactions to the argument."}],"causal_chain":["Moderator additions are identified as competitors with participants' independent warrant formation.","Candidate warrants, examples, and approving interpretations are withheld but preserved for later recovery.","A neutral question opens a visibly framed, bounded quiet interval.","Because the moderator cannot immediately resolve the silence, participants have room to formulate their own connection, express uncertainty, or decline to infer one.","Initial responses are captured before explanatory material is reintroduced.","Analysts can compare the stated claim with participant-originated warrants and distinguish apparent comprehension from moderator-assisted ratification.","The communication team can revise, retain, or reject the argument based on the bounded test without treating silence itself as assent."],"baseline":"A conventional semi-structured message-testing session in which the moderator asks why a claim is persuasive but may quickly restate it, offer examples, probe with candidate reasons, or verbally reassure participants when the room is quiet.","nearest_rivals":["A moderator checklist prohibiting leading questions, which constrains wording but does not reserve or frame a protected response interval.","A private written survey administered before discussion, which preserves response independence but removes the live rhetorical exchange and its observable hesitation, clarification, and recovery dynamics.","Cognitive interviewing with think-aloud prompts, which elicits reasoning but can add continuous verbal demand rather than using bounded absence to expose whether the warrant arises without prompting.","General moderator training in neutral affect and wait time, which may encourage restraint but does not preserve omitted explanations, define an absence boundary, or specify reintroduction triggers."],"remaining_contrastive_claim":"The proposal's testable distinction is a live, claim-specific diagnostic interval in which moderator explanation is deliberately absent, that absence is explicitly framed and bounded, and the omitted rationale remains recoverable only after participant-originated responses are captured.","authority_safety":{"decision_authority":"The research lead may set the protocol for a voluntary message-testing session; the moderator controls moment-to-moment delivery, and each participant retains authority to speak, write, request clarification, extend the interval, or pass.","authorized_first_step":"Run the protocol only in a recorded mock session using non-sensitive sample claims and consenting volunteers before applying it to consequential public communication.","excluded_actions":["Treating silence or nonresponse as agreement","Withholding safety, consent, compensation, or participation information","Using prolonged silence to pressure disclosure or compliance","Silencing participant objections, dissent, or requests for clarification","Deleting contradictory responses or the recoverable moderator notes","Deploying the protocol in crisis, trauma, disciplinary, or compulsory settings without separate review"],"halt_rollback":"End the silence immediately if a participant requests clarification, signals distress, encounters an accessibility barrier, or the maximum interval expires. State that the pause was procedural, offer the claim again in neutral language, allow a pass or written response, and revert to the standard nonleading interview protocol while recording the deviation."},"negative_tests":{"strongest_counterevidence":"Participants consistently interpret the framed silence as evaluation or pressure, producing socially desirable rationales rather than independent warrants; a private pre-discussion response method yields cleaner origin evidence without that reaction.","problem_falsifier":"Session recordings show that participants ordinarily articulate their own warrants before moderators add content, or moderator additions occur only after initial responses have already been captured, so premature filling is not causing the inferential ambiguity.","intervention_falsifier":"Blinded coding finds that the protocol does not improve attribution of warrant origin or claim-boundary understanding relative to baseline, while participant reports show greater confusion, discomfort, or perceived coercion.","risks":["The pause may amplify status differences or be read as moderator disapproval.","A fixed duration may be too short for some participants and uncomfortably long for others.","Writing before discussion may privilege participants comfortable with rapid verbal or written formulation.","Removing examples may expose a genuinely underspecified claim but could be misread by the team as participant deficiency.","The thinking-time cue may itself alter how carefully participants process the claim, limiting transfer to ordinary reception."]},"next_evidence_step":"Conduct two small mock sessions of four to six consenting volunteers each using the same non-sensitive claims and a counterbalanced baseline/protocol order. Preserve timestamps and verbatim first responses; have a coder who did not moderate label whether each warrant originated before any moderator explanation, whether the claim was understood, and what each silence appeared to mean. Collect an anonymous post-session item on pressure, clarity, and preferred response mode. Treat the result only as a feasibility and measurement check, not as an effect estimate.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"No cross-proposal comparison was performed; the candidate is realized entirely as a live rhetorical-testing procedure within the supplied cell.","revision_record":{"parent_version":null,"progress_targets_addressed":[],"conceptual_changes":[],"operational_changes":[],"evidence_changes":[],"claim_changes":[]}}