LP-03 v0.4

2026 Okinawa Election — PMS 2.0 Re-Audit

Status: Post-Election Re-Audit
Date: 2026-09-18
Previous Version: LP-03 v0.3
Method: PMS 2.0 / Vocabulary Audit / Required Trace Re-Audit
Data Cut-off: 2026-09-18

Japanese ver. LP-03 v0.4｜2026年沖縄県知事選── PMS 2.0 Re-Audit

0｜Why Re-Audit?

LP-03 v0.3 was written before sufficient post-election cross-tabulations were available.

It therefore retained three Core Probes:

  1. lag-head

  2. the same configuration becomes simultaneously permissible

  3. directional synchronization

It also specified in advance the traces required for re-audit:

The purpose of v0.4 is not to explain the election retrospectively.

It is to ask a narrower question:

What happens to the PMS 2.0 vocabulary when the missing traces arrive?

The rule remains:

Do not fit the data to PMS.
Re-audit PMS against the data.

I｜What Was Observed?

1｜The Result

The 2026 Okinawa gubernatorial election produced a highly convergent electoral result.

Koja Genta received 423,124 votes, defeating incumbent Tamaki Denny and winning across nearly the entire prefectural map.

At the municipality level, Koja outpolled Tamaki in 40 of Okinawa’s 41 municipalities.

This establishes a large-scale electoral configurational convergence.

But:

40/41 municipalities ≠ 40/41 meanings.

An electoral map records votes.

It does not directly record the meanings, priorities, temporalities, or relations through which those votes became possible.

The post-election cross-tabulations therefore matter.

2｜Priority Issue × Candidate

The Kyodo News exit poll provides the clearest post-election cross-tabulation.

The survey was conducted at 60 polling stations, with responses from 2,909 voters.

Asked what they considered most important when choosing a candidate:

Among respondents prioritizing economic revitalization:

Among respondents prioritizing base issues:

A separate Yomiuri Shimbun exit poll, conducted at 96 polling stations with 2,839 respondents, found a similar differentiation using different categories.

Its leading priorities were:

About 70% of those prioritizing economic policy voted for Koja, while about 80% of those prioritizing base issues voted for Tamaki.

The two surveys should not be numerically merged because their questions and categories differ.

But both expose the same methodological point:

Different issue priorities retained different candidate propensities.

This is not evidence of priority convergence.

Yet the aggregate electoral configuration converged.

Therefore:

Priority convergence was not necessary for electoral convergence.

3｜Henoko Preference

The Kyodo exit poll also asked about relocation of the U.S. Marine Corps Air Station Futenma to Henoko.

The electorate therefore remained divided on Henoko.

A pre-election survey jointly conducted by QAB, Okinawa Times, and Asahi Shimbun on September 5–6 provides an additional trace.

The survey was commissioned to four internet research companies and obtained 1,260 valid responses from their monitors.

It asked both respondents’ position on Henoko relocation and their intended candidate choice.

The published analysis reported that:

The same qualitative relationship appeared for the strengthening of Self-Defense Forces deployment in the southwestern islands.

However, numerical candidate cross-tabulations were not publicly reported.

Henoko Preference × Candidate must therefore be classified as:

CROSS OBSERVED / NUMERICAL TRACE NOT PUBLIC

This distinction matters.

The missing percentages should not be reconstructed from the election result.

Nor should preference be substituted for priority.

Henoko Preference ≠ Henoko Priority ≠ Vote

The Anti-Overclaim Guard remains active.

4｜Party Support × Candidate

Partisan differentiation remained clearly visible.

According to the Kyodo exit poll:

The election therefore did not erase partisan configuration.

It occurred through a field in which partisan differentiation remained strongly observable.

Party structure persisted inside electoral convergence.

5｜Unaffiliated × Candidate

The unaffiliated vote requires particular caution.

In the Kyodo exit poll:

This alone prevents the aggregate result from being translated into a simple claim that unaffiliated voters uniformly moved toward Koja.

Other post-election survey reporting has produced different estimates of the unaffiliated vote.

Rather than selecting whichever survey best fits a retrospective explanation, v0.4 retains the disagreement as a methodological warning.

Survey estimate ≠ electorate itself.

And:

Aggregate victory ≠ uniform unaffiliated shift.

Survey disagreement is not something to erase in order to produce a cleaner story.

It is a Trace of the limits of observation.

6｜Age × Candidate

Age differentiation was also observable.

Kyodo reported Koja ahead of Tamaki among every age group through voters in their 60s.

Among voters in their teens through 30s, Koja received more than 70%.

The independent Yomiuri exit poll likewise found Koja ahead in every age group except voters aged 70 and above.

Among voters aged 18–39, Koja’s support was more than three times Tamaki’s.

The exact age boundary should not be overinterpreted across surveys.

The stronger conclusion is narrower:

Electoral convergence extended across multiple age groups while age differentiation remained observable.

No single generation can therefore be treated as the meaning of the election.

Age × Vote ≠ Common Meaning.

7｜Previous Vote × Current Vote

Kyodo reported one of the strongest direct traces of electoral reconfiguration:

30.8% of respondents who said they had voted for Tamaki in the 2022 gubernatorial election voted for Koja in 2026.

This directly observes movement across a previous electoral boundary.

It does not, however, establish why each voter moved.

The same transition in candidate choice may have been produced through different priorities, relations, temporalities, or judgments.

Therefore:

Previous Vote Shift ≠ Common Meaning Shift.

The trace establishes electoral movement.

It does not establish a common direction behind that movement.

II｜Municipality-Level Re-Audit

8｜The Geography of Convergence

Official municipality-level results provide a different kind of evidence.

Unlike exit polls, these are not sample estimates.

They are complete administrative vote counts.

In 2022, Tamaki received the largest number of votes in 28 of Okinawa’s 41 municipalities.

In 2026, Koja outpolled Tamaki in 40 of 41 municipalities.

More importantly, when the 28 municipalities carried by Tamaki in 2022 are compared directly with the 2026 results:

27 of the 28 switched to Koja.

The only municipality in which Tamaki retained a plurality over Koja was Ogimi Village.

Thus:

2022 Tamaki plurality municipalities: 28/41

↓

27/28 switched to Koja in 2026

↓

2026 Koja plurality municipalities: 40/41

The transition was therefore not confined to one city, one base-hosting area, one age cohort, or one partisan enclave.

It crossed most of the municipal configurations that had produced a Tamaki plurality only four years earlier.

This establishes the geographical extent of electoral reconfiguration.

It does not establish a common reason for that reconfiguration.

Municipal convergence ≠ Semantic convergence.

Nor does municipality-level aggregation identify the direction through which individual voters arrived at the same electoral configuration.

Geographical convergence ≠ Directional synchronization.

III｜Required Trace Re-Audit

9｜Trace Matrix

The Required Trace specified before the post-election data became available can now be re-audited.

Required Trace Status Re-Audit Result
Age × Candidate OBSERVED broad Koja advantage across younger and middle age groups; age differentiation persists
Party Support × Candidate OBSERVED strong partisan differentiation persists
Henoko Preference × Candidate PARTIALLY OBSERVED qualitative cross published; numerical cross not public
Priority Issue × Candidate OBSERVED sharply different candidate propensities by priority
Previous Vote × Current Vote PARTIALLY OBSERVED 30.8% of previous Tamaki voters shifted to Koja
Unaffiliated × Candidate OBSERVED Kyodo shows Tamaki advantage; cross-survey estimates require caution
Municipality-level Swing OBSERVED / COMPLETE ADMINISTRATIVE DATA geographically extensive reconfiguration

The central empirical pattern is therefore not disappearance of heterogeneity.

It is:

heterogeneity remaining observable alongside large-scale electoral convergence.

IV｜Core Probe Re-Audit

10｜Probe A — lag-head

Verdict

RETAINED / NOT IDENTIFIED

The new data expose several strong differences:

But none of these can simply be identified as lag-head.

A large percentage is not a lag-head.

A decisive issue is not necessarily a lag-head.

A preference is not a lag-head.

A vote is not a lag-head.

Therefore the Vocabulary Audit remains intact:

Issue ≠ lag-head.
Preference ≠ lag-head.
Vote ≠ lag-head.

The available election data do not provide sufficient Trace to identify a single lag-head.

11｜Probe B — simultaneously permissible

Verdict

SUPPORTED BY NEW TRACE

This is the principal movement from v0.3.

The post-election traces show persistent heterogeneity across:

Nevertheless, one electoral configuration became available across sufficiently many heterogeneous electorates to generate large-scale convergence.

This does not empirically measure permissibility.

Permissible remains a PMS analytical term rather than a survey variable.

The data therefore do not prove the proposition directly.

But they strengthen the probe that:

the same electoral configuration can become simultaneously permissible across heterogeneous relations without requiring identical meanings.

The question mark retained in v0.3 can therefore be narrowed.

Not:

CONFIRMED

but:

SUPPORTED BY NEW TRACE

12｜Probe C — directional synchronization

Verdict

NOT ESTABLISHED

This is the principal restraint introduced by v0.4.

The aggregate election result looks strongly synchronized.

The cross-tabulations do not establish synchronization of direction.

Economic-priority voters and base-priority voters showed sharply different candidate propensities.

Partisan divisions remained strongly visible.

Henoko preferences remained divided.

Previous Tamaki voters did not move uniformly.

The available unaffiliated-voter evidence does not justify treating that category as a single directional bloc.

Therefore:

Vote convergence ≠ Directional synchronization.

The same electoral configuration may be reached through different relations, priorities, temporalities, and meanings.

The ballot records the configuration selected.

It does not record a common direction behind it.

V｜Revised Transition

LP-03 v0.3 provisionally retained the following transition:

heterogeneous lag

↓

multiple permissible configurations

↓

multiple directions

↓

the same configuration becomes simultaneously permissible ?

↓

large-scale configurational convergence

↓

directional synchronization ?

Post-election data require a revision.

LP-03 v0.4

heterogeneous traces

↓

different preferences / priorities / partisan positions persist

↓

the same electoral configuration becomes simultaneously permissible

↓

large-scale configurational convergence

↓

without demonstrated semantic convergence or directional synchronization

This is a narrower claim.

It is also more tightly constrained by the available Trace.

VI｜What the Election Does Not Show

The 2026 Okinawa gubernatorial election does not demonstrate that:

Those claims exceed the available Trace.

The official result establishes convergence in electoral configuration.

The cross-tabulations establish persistent observable heterogeneity beneath that convergence.

But another guard is necessary:

absence of demonstrated semantic convergence ≠ proof of semantic divergence

v0.4 does not claim that common meanings were absent.

It claims that common meaning is not established by the available evidence and is not required to describe the observed configurational convergence.

VII｜Working Proposition after Re-Audit

LP-03 v0.3 proposed:

Political transition does not require semantic agreement.

The 2026 post-election data do not prove this proposition universally.

They do, however, strengthen its plausibility in this case.

A more precise formulation is:

Political transition can occur through configurational convergence across heterogeneous electorates without demonstrated semantic agreement.

Or, in its shortest form:

The configuration converged.
The meanings did not have to.

The second sentence is deliberately modal.

It does not say:

the meanings did not converge.

It says:

their convergence is not required by the available Trace to account for the observed electoral configuration.

VIII｜Remaining Trace

Two important traces remain incomplete.

Henoko Preference × Candidate

The relationship was observed qualitatively in the pre-election QAB / Okinawa Times / Asahi survey.

But a numerical candidate cross has not been publicly confirmed.

The election-day Kyodo poll published the overall Henoko distribution but not a numerical Henoko Preference × Candidate cross.

Status:

CROSS OBSERVED / NUMERICAL TRACE NOT PUBLIC

2022 Vote × 2026 Vote

Kyodo provides an important partial trace:

30.8% of previous Tamaki voters → Koja

But a complete transition matrix covering all previous and current choices has not been publicly established.

Status:

PARTIAL CROSS

These gaps should not be filled by inference.

They remain open Trace requirements for any future revision.

IX｜Evidence Status

Administrative Data

Okinawa Prefectural Election Administration Commission

2022 and 2026 municipality-level official vote counts.

Status:

COMPLETE ADMINISTRATIVE TRACE

Kyodo News Exit Poll

Used for:

Status:

PRIMARY POST-ELECTION SURVEY BACKBONE

Yomiuri Shimbun Exit Poll

Used as an independent comparison for:

Its question categories differ from Kyodo and should not be numerically merged with it.

Status:

INDEPENDENT CHECK

QAB / Okinawa Times / Asahi Shimbun Pre-Election Survey

Used for:

Because this is a pre-election internet-panel survey rather than an election-day exit poll, it is kept analytically separate.

Status:

PRE-ELECTION QUALITATIVE CROSS

X｜Final Re-Audit Verdict

Post-election cross-tabulations do not dissolve the heterogeneity observed before the election.

Different issue priorities retained sharply different candidate propensities.

Partisan divisions remained visible.

Henoko preference remained divided.

Previous Tamaki voters did not move uniformly.

Age differentiation persisted.

Nevertheless, the official result shows a geographically extensive convergence in electoral configuration.

Therefore:

Heterogeneity remained observable.

Electoral configuration converged.

But:

Configurational convergence ≠ Semantic convergence.

Configurational convergence ≠ Directional synchronization.

The post-election data strengthen the PMS 2.0 probe that the same electoral configuration may become simultaneously permissible across heterogeneous relations.

They do not establish that those relations acquired a common meaning or direction.

Re-Audit Status

lag-head

→ RETAINED / NOT IDENTIFIED

simultaneously permissible

→ SUPPORTED BY NEW TRACE

configurational convergence

→ OBSERVED

semantic convergence

→ NOT DEMONSTRATED

directional synchronization

→ NOT ESTABLISHED

The resulting proposition is narrower than a theory of electoral consensus:

Political transition can occur through configurational convergence across heterogeneous electorates without demonstrated semantic agreement.

Or:

The configuration converged.
The meanings did not have to.

This is not a claim that meanings were absent.

It is a claim that common meaning is not required by the available evidence to account for the observed electoral configuration.

Do not infer meaning from convergence.

Do not infer direction from vote.

Keep the heterogeneity visible.

Re-Audit closed: 2026-09-18.

LP-03｜Development Notes v0.1

LP-03｜政治転換はいかにして起きるのか ──2026沖縄県知事選にみる lag-window と configurational convergence（v0.3）

LP-PMS-00｜政治的意味空間2.0の語彙監査 ── lag-head / lag-window / configuration / synchronization の再定式化

EgQE — Echo-Genesis Qualia Engine
camp-us.net

© 2025 K.E. Itekki
K.E. Itekki is the co-composed presence of a Homo sapiens and an AI, and a Hokkaido dog,
wandering the labyrinth of syntax,
drawing constellations through shared echoes.

📬 Reach us at: contact.k.e.itekki@gmail.com

| Drafted Sep 18, 2026 · Web Sep 18, 2026 |