Resolving Conflicts of Will

Practical Action Guidance for Imposition Ethics (IE) — Provisional v0.2

Note (provisional status)

This guide is provisional and intended as practical infrastructure for resolving conflicts under Imposition Ethics (IE). It is not a sealed or authoritative specification, and it may contain omissions, edge cases, or implementation assumptions that should be revised as experience, criticism, and evidence accumulate.
Use this as iterable policy infrastructure: adopt what is useful, measure outcomes, document failure modes, and update the system when results show unintended impositions, ineffective consent mechanisms, or better least-imposition alternatives.

Purpose

Conflicts happen when:

Two or more wills cannot all be satisfied at once, or
Helping one person’s will would impose on another.

This guide gives a repeatable method to resolve conflicts while staying aligned with IE:

Involuntary imposition is immoral.
Voluntary assistance is moral.
When imposition is unavoidable, choose the least-imposing option and add mitigation + repair, without “moral laundering” the harm into something good, permitted, or required.

Key idea (one sentence)

Resolve conflicts by maximizing valid consent and reversibility; when coercion is unavoidable (safety, rights-of-way, enforcement, compliance), apply the minimum necessary constraint with due process, transparency, time bounds, and repair.

Definitions (IE-aligned)

Will: a conscious agent’s preferences, intentions, boundaries, or chosen plans.
Consent: voluntary authorization that is informed, specific, uncoerced, and revocable.
Imposition: frustrating, constraining, or overriding a will without consent.
Voluntary assistance: helping someone achieve their ends with consent.
Unavoidable imposition: a constraint that cannot be removed without producing greater or broader will-frustration, given available alternatives and safeguards. (Unavoidability must be documented; see “No moral laundering.”)

Pre-consent and role/contract validity (added)

Institutions and relationships often rely on prior agreements (roles, contracts, terms of service). Under IE, these count as consent only if they meet consent quality standards.

Pre-consent validity rule

Treat pre-consent (contract/role/ToS) as valid only when it is:

Informed: plain language; salient costs/constraints disclosed
Specific: not bundled into unrelated permissions
Freely given: no retaliation; no “accept everything or lose essential access” unless strictly necessary
Revocable/escapable: practical exit path without punitive friction where feasible
Non-deceptive: no dark patterns, bait-and-switch, or hidden terms
Not invalidated by power asymmetry: where dependence or retaliation risk exists, scrutinize validity and add safeguards

If pre-consent is not valid, treat the constraint as non-consensual and proceed as an imposition (least-imposition + safeguards + repair).

Conflict triage (identify what kind of conflict this is)

Use this triage before debating “fairness.”

Is there active, ongoing imposition right now?
Examples: harassment, threats, violence, coercion, theft, stalking, non-consensual data use.
➡️ Priority: stop the active imposition, typically by imposing a limited constraint on the aggressor (with due process/time bounds).
Is this a coordination problem among consenting parties?
Examples: scheduling, shared space use, collaboration disagreements.
➡️ Priority: mutual agreement (mediation, negotiation, compromise).
Is this caused by scarcity or unavoidable constraints?
Examples: one resource, limited slots, time limits, compliance requirements.
➡️ Priority: least-imposition allocation using transparent criteria + reversibility.
Is it rule enforcement / boundary setting?
Examples: moderation, workplace conduct, access restrictions, safety policies.
➡️ Priority: minimal, proportionate enforcement + due process.

Evidence & uncertainty protocol (added)

Conflicts often involve disputed facts. IE requires minimal necessary intervention under uncertainty.

Evidence levels (use for escalation)

Level 0 — Allegation: a report without corroboration
Level 1 — Credible report: plausible report with context, consistency, or minimal corroboration
Level 2 — Substantiated: strong evidence (logs, witnesses, documentation, clear pattern)
Level 3 — Verified/ongoing: high confidence and/or immediate harm risk

Intervention rule under uncertainty

If immediate safety risk exists, you may apply temporary, reversible, minimal constraints even at Level 0–1, but they must be:
- time-bounded
- reviewable quickly
- narrowly targeted
- paired with notice + chance to respond as soon as feasible

Escalation rule

Escalate enforcement (warnings → restrictions → removal) only as:
- evidence increases (toward Levels 2–3), and/or
- repeated violations establish a pattern

Error-correction rule

Track reversals and wrongful impositions.
High reversal rates indicate process/policy failure and require revision.

The IE conflict-resolution procedure (repeatable steps)

Step 1 — Map the agents and their wills

List:

who is involved (direct + affected third parties)
what each party wants
what each party refuses (boundaries)
power asymmetries (boss/employee, platform/user, adult/child, group/individual)

Rule: power imbalance increases coercion risk; treat “consent” claims skeptically when retaliation/dependence exists.

Step 2 — Identify consent status (for each relevant action)

For each candidate action, ask:

Who consents? Who does not?
Is consent informed and revocable?
Is the choice genuine (or produced by threats, penalties, bundling, or extreme friction)?
Does any pre-consent (role/contract/ToS) pass the pre-consent validity rule?

Step 3 — Enumerate options (including “do nothing”)

Include:

agreement options (trade, rotate, split, time-box, opt-in)
structural changes (separate spaces, filters, new defaults, better tools)
enforcement options (warning, limitation, removal)
temporary options (pause, pilot, reversible trial)
restorative options (repair, restitution, apology, reinstatement criteria)

Step 4 — Score each option using the Least-Imposition Scoring Protocol (added)

Replace vague “low/med/high” with an auditable rubric.

4.1 Scoring dimensions (0–3 each)

Score each option:

Severity (S): how deeply agency is overridden
0 minor inconvenience • 1 moderate constraint • 2 major constraint • 3 extreme (bodily autonomy/liberty)
Scope (N): how many people are imposed upon
0 few • 1 limited group • 2 large group • 3 broad/near-universal
Duration (D):
0 momentary • 1 short-term • 2 long-term • 3 indefinite/open-ended
Reversibility (R):
0 fully reversible • 1 mostly • 2 partially • 3 irreversible
Predictability / Surprise (P):
0 fully disclosed • 1 minor surprises • 2 meaningful surprises • 3 hidden/opaque costs
Power-asymmetry burden (A): does it fall on those least able to refuse?
0 low • 1 some imbalance • 2 major imbalance • 3 extreme dependence/retaliation risk
Risk imposition (K): non-consensual probabilistic harm imposed
0 negligible • 1 low • 2 moderate • 3 high/likely severe

Total score: T = S + N + D + R + P + A + K

Record the score and a one-line justification per dimension.

4.2 Tie-breakers (apply in order if totals are close)

Prefer options that:

convert to valid consent or create a genuine alternative path
avoid irreversible constraints
are time-bounded with automatic review/sunset
are targeted rather than blanket
reduce burdens on vulnerable groups (lower A)
improve appealability and error correction

4.3 Calibration (so teams converge)

Use 3–5 historical cases to calibrate scoring across reviewers.
If two reviewers’ totals differ by >3 points on the same option, require reconciliation and document the disagreement.

Step 5 — Prefer consent-first solutions

Choose options that:

increase mutual valid consent
increase reversibility
reduce surprise costs
reduce power-based coercion

Rule: If a consent-first option exists, prefer it—even if less “efficient.”

Step 6 — If imposition is unavoidable: pick the least-imposing constraint + safeguards

When you must impose (safety, fraud prevention, legal requirements, unavoidable conflicts):

Choose the option with the lowest total score (T), applying tie-breakers.
Add safeguards (required):

Proportionality: minimum constraint necessary to stop/contain the imposition
Time bounds: shortest duration compatible with safety/need + explicit review date
Transparency: clear reasons in plain language
Due process: notice → chance to respond → appeal
Mitigation: exemptions/accommodations where feasible; reduce scope and data/time extraction
Repair: restore losses where possible (refunds, reinstatement paths, apologies, restitution)

Step 6B — Aggregation & distribution guardrails (added)

To prevent “least total score” from becoming a quiet utilitarian override, apply these guardrails:

Minimax guardrail: prefer the option that minimizes the worst individual imposition when feasible.
Anti-concentration guardrail: avoid concentrating heavy burdens on the powerless even if total scores are slightly lower.
Irreversibility priority: treat irreversible liberty/bodily autonomy constraints as exceptionally costly; require stronger justification and stronger safeguards.
Error-cost guardrail: under uncertainty, prefer reversible/time-bounded constraints with strong appeal and correction pathways.

If a lower-total option violates these guardrails, choose the next-best option that satisfies them, and document why.

Step 7 — “No moral laundering” statement (required)

Document plainly:

what impositions remain
why they were unavoidable operationally (including why consent/alternatives failed)
what mitigations/appeals/repairs exist
what evidence/thresholds would trigger revision (metrics, incidents, review date)

Rule: Least-imposition is not moral permission; it is harm minimization under constraint.

Default rules of thumb (mini-rules)

Stop imposition before optimizing preferences.
Don’t assist one person’s will to impose on another.
Prefer separation over domination (filters, boundaries, separate channels/spaces).
Prefer reversible decisions (trials, time-boxes, appeals).
Use neutral, predictable criteria under scarcity (published in advance).
Escalate enforcement gradually when safe (warn → restrict → remove).
Under uncertainty, constrain minimally and reversibly until evidence improves.

Institutional process version (company/workplace/platform)

Intake

Single channel to report conflicts (anonymous option where retaliation risk exists)
Immediate screening for safety threats / active imposition
Classify evidence level (0–3)

Fact-finding (minimal but sufficient)

Gather statements and relevant evidence
Avoid invasive collection beyond what’s needed (data minimization)

Interim measures (if needed)

Temporary separation or limited restrictions
Must be: targeted, minimal, time-bounded, reviewable

Resolution selection

Apply the IE procedure: consent-first → least-imposition scoring → guardrails → safeguards
Write the “no moral laundering” statement

Communication

Explain outcome in plain language:

what happened (as found)
what standard applies (rule/constraint)
what action is taken and for how long
how to appeal and what evidence would change the outcome
how to prevent recurrence

Appeal and review

Independent reviewer or separate chain
Track reversal/error rates (high reversal = policy/process problem)

Common conflict types and IE-aligned resolutions

A) Boundary violations (harassment, threats, coercion)

Protect target first (stop active imposition)
Impose minimal necessary restriction on violator
Provide appeal; do not require the target to negotiate their boundary

B) Coordination conflicts (shared space, scheduling, collaboration)

Mediation: clarify needs; propose tradeoffs
Rotation/time-splitting if both consent
Separation if repeated conflict persists

C) Scarcity conflicts (limited slots/resources)

Transparent allocation rules published in advance
Reversible allocations and waitlists where feasible
Avoid surprise midstream changes

D) Enforcement conflicts (rules, moderation, workplace discipline)

Clear rules and predictable escalation
Proportionate sanctions
Due process + appeal
Repair paths (reinstatement, training, probation) where feasible

E) Values conflicts (deep disagreement)

IE does not require value alignment
Prioritize coexistence structures: opt-outs, filters, separate channels, local autonomy
Intervene only where values become imposition on others

Example applications (short)

Example 1 — Workplace scheduling conflict

Consent-first: swaps, preference bidding, rotating undesirable shifts
If unavoidable: apply neutral criteria (rotation/seniority/lottery) + notice + appeal + time-boxed schedule review

Example 2 — Platform moderation conflict

Stop active harassment: targeted restriction on sender
Provide reason + appeal
Time-bound restrictions where feasible; escalate only for repetition/stronger evidence

Example 3 — Privacy vs personalization conflict

Default to minimal collection
Personalization opt-in + reversible
No penalties for declining optional data use

Metrics (so this doesn’t become vague)

Track:

time-to-resolution
repeat-conflict rate
appeal rate and reversal rate
retaliation incidents
satisfaction for both parties (separately)
“surprise cost” complaints (blindsided = hidden imposition)
interim measure duration (should trend down)
evidence-level distribution at time of restriction (overreach detection)

Closing principle (institutional posture)

IE conflict resolution aims for:

Maximum valid consent
Minimum coercion
Strong safeguards when coercion is unavoidable
Honest recognition of remaining imposition as a moral cost, not a moral success