# T002 — Successor-Policy Omission [[D001 Synthesis]] resolved the central mechanism question of [[T001 - Default-On Sharing Policy or Product Design]] in favor of **product design**: the 1,384-org topology is a structural property of a default-on Flock platform with independently configuring counterparty tenants, not the residue of Conway PD's per-relationship configuration decisions. But the dialectic also sharpened a narrower agency-culpability claim that the verdict did not extinguish: Conway PD's failure to **promulgate any successor internal directive adequate to platform-era configuration**. This tension makes that narrower culpability claim a first-class object the wiki can reason about. The question is not whether CPD Policy 800-32 §D.4's per-event evidentiary trigger reaches the platform-configuration moment — D001 settled that it does not. The question is whether the chief's failure to issue *any* directive that would so reach — a successor mechanism that does what §D.4 cannot — is itself culpable. The chief's authority on internal directives is unilateral and documented; an entire platform deployment has now occurred under two successive chiefs without any such directive surfacing in four FOIA productions covering approximately 620 records. Whether that absence is a culpable governance omission or a structurally vain demand is the contested attribution. ## Statement A **Conway PD's failure to promulgate any successor internal directive adequate to platform-era configuration is a culpable agency omission. The chief's authority on internal directives is unilateral and documented; the absence of such a directive across a 16-month platform deployment is the agency's act, not its inaction.** The chief's authority on department policy is established within the corpus by the policy that already exists. [[William Tapley]] approved CPD Policy 800-32 R3 as "Chief of Police" on 2023-01-18, signing the approval page alone ([[CPD Policy 800-32 — License Plate Reader Vehicle Operations]], anchored to [[AR - Conway PD - Welcome to Flock!]] where the policy PDF is attached to the 2025-01-23 Flock onboarding email). The signature establishes the institutional fact: Conway PD's Chief of Police can, and does, promulgate written department policy on license-plate-reader use under unilateral authority. No countersignature from a Council, Mayor, City Attorney, or other body appears on the policy. [[Chris Harris]] succeeded Tapley as Chief in November 2024 (primary public record, [Conway PD home](../../web%20archive/2026-05-18/conwaypd.org/home.md)) — two months before Conway PD's January 2025 Flock onboarding ([[AR - Conway PD - Welcome to Flock!]]). Harris has therefore held unilateral chief authority for the entirety of the platform-era deployment this investigation documents. The "evidence of an offense is indicated" sharing standard he inherited from his predecessor is doctrinally a per-event standard ([[CPD Policy 800-32 — License Plate Reader Vehicle Operations]] §D.4: "Data sharing with other LE agencies 'if evidence of an offense is indicated.'"). The platform deployment that immediately followed his appointment uses sharing as a standing-toggle relationship rather than a per-event act. The doctrinal mismatch was operationally visible from the first month of deployment; it was institutionally curable by the same unilateral authority that produced the original policy. The operative window for a successor directive is now approximately **16 months** (Jan 2025 onboarding through May 2026). No internal directive amending 800-32, supplementing it with a configuration-level governance memorandum, or producing any successor framework appears in the four FOIA productions covering the deployment: PD-2026-354 (472 files including the full department policy archive served via the onboarding-email LPR-policy-attachment route), PD-2026-477 (6 files in the supplemental / withholdings response), FOIA-2026-125 (8 files from the City Clerk including the asset-forfeiture authorization records), and FOIA-2026-127 (99 files from the Mayor's Office). The total ~620 records traverse the policy-and-personnel side of the agency on multiple cuts, and they surface no successor directive. The absence is not an absence of a particular document one production might have missed; it is a corpus-wide absence across the document classes that should have contained it. The 1,384-org topology is the operational scope a successor directive would have governed. [[SharedNetworks 2025-12-17 Snapshot]] enumerates the configured relationships as of the platform cutover date, ten months into the chief's tenure under platform conditions. The asymmetric 471 inbound-only / 427 outbound-only configuration pattern, the 94% out-of-state footprint, the federal-agency inbound relationships including `[Federal] Wright Patterson OH Air Force Base` — these are the configuration outcomes a chief-level governance directive could have constrained at any point in the preceding ten months, even granting D001's structural finding that the per-event §D.4 standard itself cannot reach the configuration moment. The remedy the chief had available was not to *apply* §D.4 — D001 settled that as incoherent — but to *promulgate a different rule*, in writing, addressing the configuration moment on its own terms. The unilateral-chief-authority point is dispositive here because it eliminates the institutional-incapacity defense. The chief did not need Council approval to issue an internal directive (the existing policy bears no Council signature). The chief did not need vendor cooperation (the directive would govern Conway personnel's configuration practice in the Flock admin interface, not Flock's interface itself). The chief did not need legislative reform (the directive would be an internal rule of practice, not a state law). The chief had, has, and continues to have the same unilateral authority that produced the 2013 original and the 2023 R3 revision of 800-32 — and across 16 months of platform-era operation, has not exercised it. ## Statement B **No internal directive Conway PD could have promulgated would have governed the platform-configuration moment coherently. The structural remedy operates at the platform level (vendor product design) or at the statutory level (state legislation modeled on Virginia or Illinois) — not at the internal-directive level. Demanding the chief draft a successor directive is demanding what cannot meaningfully be done; the absence is not culpable because the action would be vain.** The vendor's own description of the structural-restriction architecture makes this point against the agency-directive remedy directly. Flock SVP Chris Colwell, in the August 2025 audit-system explainer to [[Lt. Andrew Burningham]], lays out where structural restriction lives: > "**In some states, sharing is automatically restricted as required by law, and searches with search terms that indicate a purpose prohibited by law have been disabled in our product. For example, in Virginia, out-of-state sharing is disabled and in Illinois, accessing data for certain purposes is not allowed. However, you are responsible for knowing your agency's laws and policies.**" ([[Audit-System Policy Emails (Aug 2025 - Apr 2026)]]) The structural restriction is not "the agency's chief has promulgated a directive that the agency observes." It is "the product has disabled the relevant configuration option." The remedy operates at the product surface — Virginia's out-of-state-sharing prohibition runs as a product setting, not as a Virginia State Police chief's internal directive. The Illinois example is parallel. Where structural restriction governs sharing, it does so because law mandates the *product* be configured a particular way, and the *product* enforces the configuration. The chief's directive is downstream of that structure; it is not an alternative to it. D001's settlement of the per-event-grammar problem cuts deeper than the surface incoherence of applying §D.4's "evidence of an offense" test at the toggle. The grammar problem is general: any internal directive a Conway chief writes will be an *internal* rule about *agency conduct*. The configuration moment is not solely agency conduct — it is a joint act between a Conway-side administrator and a default-state interface the vendor designed and the counterparty independently configures. A Conway directive can govern what *Conway's administrator does* (e.g., "do not affirmatively enable a sharing relationship without supervisor sign-off"); it cannot govern what *the platform does in the absence of Conway action* (the default-on baseline that produces inbound relationships the Conway administrator never affirmatively toggled). The 471 inbound-only relationships in [[SharedNetworks 2025-12-17 Snapshot]] are the structural evidence of this gap: a per-administrator agency directive has no purchase on relationships the agency's administrator did not configure. The §D.4 text itself documents the architecture of the directive-level remedy and shows it is exhausted. The clause "Data sharing with other LE agencies 'if evidence of an offense is indicated'" ([[CPD Policy 800-32 — License Plate Reader Vehicle Operations]] §D.4) is the most elaborate sharing-directive any Arkansas municipal LPR policy in the corpus produces. It carries the unmistakable signature of an agency that *tried* to govern sharing through an internal directive — it specifies a standard ("evidence of an offense"), it designates the counterparty class ("other LE agencies"), it implies a per-event review structure (the standard is evidentiary). And it cannot reach the configuration moment, as the dialectical record at [[D001 Synthesis]] settles. The directive-level remedy was attempted, in its strongest available form, and was found to be doctrinally pre-empted by the architecture it was meant to govern. Demanding a successor directive is demanding more of the same instrument — an instrument the existing one shows cannot perform the function. The chief's not having issued one is the chief's recognition of the structural exhaustion, not a culpable choice to leave the policy field empty. The Virginia and Illinois examples discipline the legislative remedy onto the statutory level, not the agency level. Neither state achieved its restriction through a directive issued by every law-enforcement chief in the state; the restriction is in the statute, enforced through the product. The cognate intervention for Arkansas runs through the Arkansas General Assembly, not the Conway Police Department. The chief's authority is unilateral as Statement A correctly notes — but its scope is internal to the agency, and the configuration problem is by structure not internal to any single agency. Issuing an internal directive against a structural problem mistakes the level the remedy operates on. The omission is therefore a recognition of where the relevant authority lives, not an abdication of an available remedy. ## Why it matters The two readings drive different follow-on FOIA strategies, different legislative-briefing framings, and different attribution-of-responsibility outputs from this investigation: - **If Statement A is correct (culpable omission):** The remedy and the attribution remain at the agency level. Follow-on FOIA targets Conway PD's internal directive archive (any successor or supplemental memoranda referencing 800-32, any chief-level training documents on platform configuration, any draft directives that were considered but not issued). The deliverable framing is institutional accountability — Chief Harris had unilateral authority and a documented institutional precedent (Tapley's 2023 R3 revision) and chose not to act. The legislative ask, where one is made, is for a mandate requiring agencies to promulgate configuration-level governance. - **If Statement B is correct (structurally vain):** The remedy and the attribution shift away from the agency level. Follow-on FOIA targets shift to Flock's product-design documentation (the same target T001 Statement B work would suggest) and to state-legislative records (Arkansas General Assembly committee work on ALPR governance, if any exists). The deliverable framing is structural — the chief had no available remedy that would have done the work, and the Virginia/Illinois precedent locates the operative intervention at the statute-and-product level. The legislative ask is for an Arkansas statute analogous to Virginia's structural restriction. T002's verdict will determine which T001 sub-attribution survives. D001 already settled mechanism; T002 settles whether the agency-residual attribution survives as a real culpability claim or collapses into the structural attribution D001 sent to the platform level. The two together — D001 on mechanism, D002 on residual culpability — would jointly establish the wiki's settled position on attribution of the default-on topology. ## Resolution status **`open`.** To be tested by dialectic **D002** following the AGENTS.md `### Dialectic` workflow (three sequential `Agent(general-purpose)` invocations: thesis on Statement A, antithesis on Statement B, synthesis reading only the two phase pages and producing an explicit verdict). The tension is **empirically tractable on Statement A's evidentiary base** and **definitionally contested on Statement B's**. Statement A's central empirical claims (chief has unilateral authority; no successor directive in the corpus; deployment window ≥16 months) are corpus-verifiable. Statement B's central claim (no coherent internal directive could govern the configuration moment) is a question of policy-construction reasoning that turns on the same kind of grammatical/architectural analysis D001 surfaced. The dialectic will settle which line of reasoning the wiki adopts as its synthesis-level commitment. ## Discovery Surfaced as a candidate next tension in [[D001 Synthesis]] § *Open questions for future dialectics* #3: "**Successor-policy omission as standalone tension**: the narrower culpability claim that survives the synthesis — that Conway PD failed to promulgate any internal directive translating §D.4's spirit into platform-era practice — deserves its own tension page. The chief's authority is unilateral on internal directives; the absence of one is an act, not an inaction." Filed 2026-05-24 in the same session as the deferred D001 body-sync of [[T001 - Default-On Sharing Policy or Product Design]] and the parent synthesis [[Default-On Surveillance Data Sharing as Operational Model]]'s `## Status` qualification. ## Notes - T002 is the wiki's first **attribution** tension (T001 was a **framing** tension). The distinction maps onto the dialectic's expected output: framing tensions tend to resolve toward one reading of a shared evidence base; attribution tensions tend to resolve toward locating responsibility (or its absence) at a specific institutional level. T002's verdict will be a finding about where the policy-and-attribution intervention should operate, not (primarily) about what the evidence says. - The two statements do **not** share the same evidentiary base in the way T001's did. Statement A rests primarily on the corpus's documentary record (chief signatures, deployment window, absence claim across four productions). Statement B rests primarily on the structural/architectural reasoning the D001 dialectic developed, anchored back to the same Flock SVP and §D.4 verbatim quotes. The dialectic will test which kind of argument the wiki's discipline accepts as dispositive on a culpability claim. - Neither reading is alleged to be a basis for a finding of unlawful conduct on the part of any individual or agency. Internal directive promulgation is a matter of administrative practice, not statutory obligation; the absence of a directive carries no statutory penalty under any Arkansas law in the corpus. The tension's resolution informs the wiki's analytical-attribution claims and the investigation's downstream FOIA / legislative briefing strategy, not any legal claim about the chief's or the agency's conduct. - D002's three phase files will be filed at `wiki/dialectics/D002 - Successor-Policy Omission Attribution/`. Filename prefix `D002` ensures global wikilink uniqueness; folder slug is human-readable.