Intersectionality Finder begins with specific public issues, describes structural stakes, and treats every suggested relationship as a prompt for inquiry—not a verdict about people, communities, or political belief.
1. Start narrow. Choose a category, subcategory, and issue path rather than a demographic label.
2. Name the stake. Describe the institutional, material, civic, or safety condition at issue.
3. Test relationships. Use governed IntersectionEdge records for relationship analysis; tags and relatedIssueIds suggest candidates only.
4. Keep friction visible. Surface tensions, tradeoffs, no-strong-intersection findings, and evidence gaps alongside potential overlap.
5. Save only by choice. Active Finder work is session-first; device persistence requires an explicit Save on this device action.
The governed issue paths, trend lenses, and IntersectionEdge records have model-backed fields and relationship checks. Every current issue record remains unsourced and unassessed; no edge is reviewed, verified, or publishable. “Source-ready” names editorial structure, not evidence quality.
Recommendations are ranked deterministically from relatedIssueIds discovery aids, shared editorial bridge tags, shared trend signals, and cross-category widening. Public intersection conclusions require explicit IntersectionEdges, and every suggestion stays separate from identity, politics, affiliation, demographic category, intent, or support.
Finder 2.0 supports one issue for orientation or two to five for comparison. Optional scope is limited to broad institution/system, geography level, signal type, time horizon, and evidence threshold; it does not request identity, ideology, religion, wealth, citizenship, immigration status, affiliation, precise location, or support for an issue.
The product QA event taxonomy is documented for future review, but production measurement remains disabled. Build 4 adds no analytics provider, newsletter provider, contact integration, third-party script, cookie, pixel, fingerprinting, or remote behavioral profile.
These patterns are illustrative templates for future sourced analysis. They communicate through line, weight, spacing, shape, hatch, and typography—not identity colors—and each includes a text fallback.
Source status: illustrative and directional. The current taxonomy remains structured-but-unsourced unless a record explicitly advances through review.
The loom shows issue areas moving through systems, shared stakes, lived realities, and the point where overlap, tension, or divergence must be named.
| Sequence | Issue areas → structural systems → shared stakes → lived realities and differences → overlap, tension, or divergence. |
|---|---|
| Evidence boundary | Illustrative only; no sourced finding is asserted. |
Widths can encode issue-path count, never moral importance; selected paths become black while inactive paths remain accessible gray.
| Selected | Category → subcategory → issue → related intersection. |
|---|---|
| Inactive | Other paths stay visible in gray and are not ranked as less important. |
Cells encode strong overlap, meaningful overlap, partial overlap, tension, or insufficient evidence using monochrome marks.
| Area | Care | Safety | Voice | Evidence |
|---|---|---|---|---|
| Women’s rights | Strong | Tension | Meaningful | Insufficient |
| LGBTQ+ | Meaningful | Strong | Partial | Insufficient |
| Religious minority issues | Partial | Tension | Meaningful | Insufficient |
| Purpose | Compare categories by shared civic stakes while keeping unresolved evidence visible. |
|---|---|
| Status | Illustrative template, not a verified comparison. |
The axis keeps shared ground centered while making the unresolved tradeoff explicit.
| Structure | Concern A — shared ground — Concern B, with an unresolved question below. |
|---|---|
| Use | Name the disagreement without pretending it is solved. |
Each line represents a signal type and remains readable without animation or color.
| Signals | Legal, institutional, media, and civic trend lanes are separated as small multiples. |
|---|---|
| Status | Illustrative template; current issue cards do not claim verified trend measurements. |
The thread shows the intended future path from source to verified claim, classification, intersection, and implication; current unsourced cards stop before verification.
| Future sequence | Source → verified claim → issue classification → intersection → implication. |
|---|---|
| Current boundary | Current records remain unsourced/unassessed; no card is verified or publishable. |
Issue coverage is intentionally incomplete and changes through documented editorial review. Each record exposes source status, governance status, confidence, last review date, and this correction path. In Build 3, sources are empty, confidence is unassessed, and review dates are unset.
To propose a correction, identify the stable issue ID, the field in question, the proposed revision, and the strongest available primary source. A record can advance from drafted → source-ready → reviewed → publishable only after its evidence state supports that move.