Add peer review of the VANGO report

PEER-REVIEW.md

@@ -0,0 +1,195 @@
+# Peer Review — VANGO: The Art Passport. A Research Report on a Digital Passport for Recording Visits to Works of Art
+
+**Reviewer role:** Peer reviewer, acting as a referee for a masters-level research report submitted under the Ethical Tech CoLab. I am assuming the bar of a design/systems-research venue (an HCI or digital-heritage workshop) that accepts prototype-and-argument papers but expects an explicit research question, honest positioning against prior work, and a conclusion that does not exceed the evidence.
+
+**Recommendation:** Major revisions
+
+**Date:** 22 July 2026
+
+---
+
+## Summary of the submission
+
+VANGO is a browser-based "digital passport" for art. A visitor standing before a
+participating work scans a QR code (or types a short code, or follows a link),
+and the application adds a dated, illustrated "stamp" to a personal passport book
+carried on the phone. The report (§1–§13) documents the prototype in deliberately
+non-technical prose: what it stores, why each stored value was chosen, and what
+the tool refuses to do.
+
+The submission's real contribution is not the software but an *argument about
+design restraint*, and it is worth stating generously. The report proposes that a
+record of attendance — "who went to see what, and when" (§1.2) — is a form of
+value in itself (§12.3), and that this value can be delivered at the cost of a
+printed page (§3.2) and with zero personal-data collection in guest mode (§3.3,
+§7.1). Several design decisions are argued with unusual care: the stamp's colour
+and fallback illustration are derived deterministically from the stamp's own
+identifier so the same work "wears the same colours in every passport in the
+world" (§5.5, §13.2); the tool declines to gamify (§6.1); the interface ships in
+four languages including Hausa (§8); and §9 pauses to consider the cultural-property
+implications of stamping a looted-class object. The report is also strikingly
+candid about its own defects (§10), including a placeholder passport number
+(§5.8) and the fact that the public demonstration cannot sign anyone in (§7.4).
+
+As a piece of design documentation this is well above average. My reservations are
+about whether it yet functions as a *research* paper, and about one ethical thread
+it raises but does not resolve.
+
+## Major issues
+
+**1. The conclusion asserts a finding the paper never tests.**
+*Location: §13.1 versus §1.7, §7.1, §10.5.* The Conclusion states that VANGO "asks
+whether [the national-park passport idea] works for individual works of art. The
+answer the prototype gives is a qualified yes" (§13.1). But the report also states,
+in three places, that VANGO "has not been tested with real museum visitors" (§1.7),
+that guest mode is the "whole application, working" only in a demonstration sense
+(§7.1), and that it has "never been placed in a real exhibition or tested with real
+visitors" (§10.5). A prototype that has never met a user cannot return an answer,
+qualified or otherwise, to whether the idea "works." The load-bearing claim of the
+paper — that an attendance record is "a form of value in itself" (§12.3) — is
+asserted, never evidenced. *Why it matters:* this is the gap between what was built
+and what is concluded, and a referee will read the mismatch as over-claiming. *Path
+forward:* choose one of two honest framings. Either (a) reframe the paper as a
+*design-rationale / systems-description* paper and change §13.1 to a claim about the
+design ("the prototype demonstrates that the mechanism is buildable at negligible
+cost and without data collection") — which the paper does support — deleting
+"works" and "qualified yes"; or (b) keep the evaluative claim and earn it with even
+a small study: 8–12 users given the tool in a gallery or a walkthrough, reporting
+whether the stamp felt worth keeping. Option (a) is achievable now and is the
+stronger paper.
+
+**2. There is essentially no positioning in the literature.**
+*Location: §2.2–§2.3.* The only prior work engaged is the US National Park Service
+"Passport To Your National Parks" programme (§2.2). The "gap" the paper claims —
+that individual artworks lack the physical stamp desk a park has (§2.3) — is
+asserted without reference to the substantial bodies of work that bear directly on
+it: museum visitor-engagement and repeat-attendance research; the gamification-of-culture
+literature (which the paper implicitly critiques in §6.1 but never cites); digital
+proof-of-attendance systems (location check-in apps such as Foursquare; the
+blockchain "proof of attendance protocol" / POAP, which is almost exactly VANGO's
+mechanism minus the ledger); and digital-badge / collectible-souvenir scholarship.
+*Why it matters:* without this, the reader cannot tell whether VANGO's framing is
+novel or is re-deriving known ideas, and the "gap" reads as manufactured by not
+looking. For an Ethical Tech CoLab research output this is the single biggest thing
+separating a report from a paper. *Path forward:* add a Related Work section (can be
+two pages) that contrasts VANGO against at least these four neighbours, and use it to
+sharpen the actual novelty claim — which I take to be *deterministic, privacy-preserving,
+zero-cost attendance records without a ledger or an account*. That is a defensible,
+narrow contribution; name it explicitly.
+
+**3. The Bura ceramics entry is an unresolved ethical problem, not a neutral illustration.**
+*Location: §5.11, §9.1, §9.4–§9.5.* The demonstration catalogue seeds a real
+looted-class object — Bura ceramics, which appear on the ICOM Red List (§9.2) — as
+sample data, attributed to a venue rendered only as the unexpanded initials "AABC"
+that "the repository does not expand" (§9.1). The report then says the tool "is
+silent on whether the object should be there" (§9.4) and adds "This is not a
+criticism of the design" (§9.5). For a paper produced under an *ethical* technology
+lab, this deserves harder self-scrutiny than a disclaimer. The prototype has chosen
+to put an at-risk cultural object, tied to an unidentifiable venue, into the one
+part of the app a first-time guest is guaranteed to see (§5.11 pre-seeds three
+stamps; the David and Bura are the two real works in the seven). *Why it matters:*
+the paper's own §9.5 correctly notes that "a document styled as a passport carries
+an implication of officialdom that the underlying record does not support" — and
+then illustrates exactly that failure with a Red List object. *Path forward:*
+(i) explain in the text *why* Bura was included and what AABC is, or remove the real
+contested object from demonstration data; (ii) promote the §9.5 suggestion (carry
+collection history / a link to the holding institution on the stamp page) from
+"a future version would do well" to an actual, if minimal, design element, since it
+is the paper's own remedy for its own diagnosed problem; (iii) reconsider whether
+"not a criticism of the design" is the right posture given (i).
+
+**4. The paper states no research question and no method, so the contribution is under-specified.**
+*Location: the report has no Introduction/Methods framing; §1 is an Executive
+Summary, §2 is Background.* The document moves from rationale (§2) directly to
+description (§4–§9) without ever stating the question it investigates or the way it
+would know an answer. The Objectives (§3) are *product* objectives ("give a visitor
+a durable record"), not *research* objectives. *Why it matters:* a reader cannot
+judge success against a target that is never set, and this is what makes issue 1
+possible — an untested conclusion slides in because no evaluative criterion was
+declared up front. *Path forward:* add a short statement of the research question
+("Can a low-cost, ledger-free, account-optional attendance record be designed such
+that its emblem is stable and shared across users, and what design tensions arise?")
+and a one-paragraph methods note describing this as a design-and-artefact study.
+This is largely reorganisation of material already present, and it resolves much of
+issue 1 for free.
+
+## Minor issues
+
+- **m1.** *§5.5 and §7.2.* "Hashing" is used for two very different things — a
+  non-cryptographic palette seed (§5.5) and password protection with bcrypt (§7.2).
+  A lay reader may conflate them. Add one clause distinguishing the fast, reversible-in-effect
+  seed hash from the deliberately slow cryptographic one.
+- **m2.** *§9.1.* The venue "AABC" is never expanded anywhere in the report. Expand
+  it or flag it as unknown in-text; leaving an unexpandable acronym on a real
+  object's record undercuts §9's whole argument about officialdom.
+- **m3.** *§7.4 versus §10.10.* §7.4 calls the account system "real, tested code,"
+  while §10.10 states "No automated tests exist in the repository." Reconcile:
+  presumably "tested" means manually exercised. Say so.
+- **m4.** *§1.4, §5.5.* "the same colours ... forever" (§1.4) overclaims: stability
+  holds only while the hash function and the five-palette count are unchanged.
+  Adding a sixth palette would reshuffle every stamp. Note this dependency.
+- **m5.** *§8.2.* The README documents only English/French/Italian and omits Hausa,
+  the report's headline inclusivity feature. This is a documentation gap the report
+  flags but does not fix; fixing the README is a five-minute change worth making
+  before submission.
+- **m6.** *§10.9.* Three versions of the app coexist in the repository with no
+  documented authoritative one. For reproducibility, state plainly which file a
+  future reader should build from (the report implies "current source" — say it in
+  §11 too).
+- **m7.** *§5.11.* Pre-seeded demonstration stamps dated "May and June 2026" are
+  fabricated visit records shown to every new guest. Low-stakes, but a passport-styled
+  app displaying invented dated entries is worth a one-line caveat in-app, not only
+  in the report.
+- **m8.** *§10.8.* The base64 profile-picture storage note is honest but belongs
+  near §7.3 (what the server stores) as well, so the privacy discussion is complete
+  in one place.
+
+## Things the report gets right
+
+- **Radical honesty about limitations (§10).** The paper enumerates its own broken
+  sign-in (§10.2), placeholder passport number (§10.3), fictional catalogue (§10.5),
+  and absent tests (§10.10). This candour is rare and genuinely valuable; it is the
+  paper's most trustworthy quality and should be protected, not sanded down.
+- **The deterministic-seed argument (§5.5, §13.2)** is the intellectual highlight.
+  The reasoning — that a random colour would make the stamp "feel arbitrary rather
+  than official," so the emblem must be a function of the work's identity — is
+  precise, correct, and well-told. It is the clearest instance of the paper
+  connecting a technical choice to a conceptual claim.
+- **Privacy by design (§3.3, §7.1, §7.3).** A guest mode in which "nothing is
+  transmitted anywhere" is a substantive ethical stance, and the report resists the
+  obvious gamified, data-hungry version (§6.1). This is the contribution most worth
+  foregrounding.
+- **Linguistic and cultural reach (§8).** Shipping Hausa alongside the European
+  languages, tied to the presence of a Nigerien object, is a thoughtful decision the
+  report is right to highlight.
+- **Prose and structure.** The plain-language, purpose-per-variable exposition (§5)
+  is a model of how to make a technical artefact legible to a non-technical reader.
+
+## Verdict
+
+**Major revisions.** The submission is an excellent piece of design documentation
+and a promising research paper, held back by three fixable things: an evaluative
+conclusion the work never tests, an almost-absent literature footing, and an
+ethical thread (Bura) that it raises but declines to resolve. None requires
+rebuilding the prototype.
+
+*Single highest-value revision:* Reframe the contribution honestly (issues 1 and 4
+together) — add an explicit research question and method, and change the Conclusion
+from "does it work? qualified yes" to a claim about what the *design* demonstrates.
+That one move converts an over-claiming report into an accurate research paper, and
+everything else — related work, the Bura remedy — attaches cleanly to it.
+
+## References
+
+[1] International Council of Museums. *Red List of West African Cultural Objects at
+Risk.* ICOM, 2016. (Bura-type terracotta statuettes and pottery are listed.)
+
+[2] UNESCO. *Convention on the Means of Prohibiting and Preventing the Illicit
+Import, Export and Transfer of Ownership of Cultural Property.* Paris, 1970. (Niger
+is a State Party.)
+
+[3] Washington Conference. *Washington Principles on Nazi-Confiscated Art.* 1998.
+
+*Note on format:* per the rendering tool used for this review, references are given
+here rather than as Word footnotes, and citations to the paper are located by its
+own clause numbering (e.g. §5.5).

README.md

@@ -43,7 +43,7 @@ Enter any of the codes below in the app (tap the **+** button → *Enter code*).
 | `HOLLOW21` | Hollow Choir | R. Venn | Studio Aperture |
 | `ECHO07` | Echo Garden | T. Lindqvist | Nomad Pavilion |
 | `VOID99` | Voidwalk | K. Amaro | Lattice Museum |
-| `BURA01` | Bura Ceramics | Niger | AABC |
+| `BURA01` | Bura Ceramics | Niger | Unspecified holder (illustrative) |
 | `DAVID01` | David | Michelangelo | Galleria dell'Accademia |
 
 ---
@@ -82,3 +82,31 @@ The app runs at `http://localhost:5173` by default.
 - [Vite](https://vitejs.dev) — build tool & dev server
 - [lucide-react](https://lucide.dev) — icons
 - Google Fonts — Fraunces, Space Mono, Inter, Nothing You Could Do, Caveat
+
+---
+
+## Peer Review
+
+The full independent academic peer review of this report is in [PEER-REVIEW.md](PEER-REVIEW.md) (also available as [Word](peer-review/vango-Peer-Review.docx) under [`peer-review/`](peer-review/)).
+
+**Recommendation:** Major revisions
+
+**What the review found:**
+
+- The conclusion that the idea "works" (S13.1) exceeds the evidence, since S1.7/S10.5 admit it was never tested with users. — **Fixed.**
+- The "gap" reads as manufactured: only the NPS passport is cited, with no engagement of gamification, museum-engagement, or proof-of-attendance literature (S2.2-2.3).
+- An unresolved ethical thread: a Red List looted-class object is seeded into demo data and then disclaimed as "not a criticism" (S5.11/S9.1/S9.5). — **Fixed.**
+
+**Noted strength:** An unusually candid limitations section (S10) and a precise, well-argued deterministic-seed design rationale (S5.5, S13.2).
+
+
+### Revisions applied (peer review, Tier 2)
+
+**The "it works" conclusion is withdrawn** (S13.1). VANGO has never been placed in a real exhibition or tested with a single museum visitor, so the prototype cannot answer whether the idea works — and the premise that an attendance record is "a form of value in itself" (S12.3) is asserted, not evidenced. S13.1 now claims only what is supported: feasibility. The mechanism is explainable in one sentence, a venue's cost is a sheet of paper, and guest mode runs without collecting personal information. Whether visitors would *want* it is stated as untested and named as the next piece of work.
+
+**The Bura Ceramics entry is kept, but made deliberate and honest** (S9.1, S9.5). It is retained because removing it would let readers believe provenance simply does not arise for an attendance record; a catalogue of only invented works could not make the boundary in S9.4 concrete. Two changes:
+
+- The venue **"AABC"** — an unexplained initialism that read as a real institution while attaching an at-risk object to an unidentifiable holder — is now **"Unspecified holder (illustrative)"** across the app, the printable stamp, and both catalogue tables. S9.1 states plainly that the entry is illustrative and that no real object, holder, or exhibition is referred to.
+- **"This is not a criticism of the design" is dropped** from S9.5. It deflected a criticism that lands: a document styled as a passport implies an officialdom the record does not support, and issuing one beside an object of contested origin lends it the appearance of a settled institutional context. The design does not cause that, but it does not avoid it, and the section now says so.
+
+Note: the three pre-seeded guest stamps (CHROMA14, FAULT02, HOLLOW21) are all fictional — Bura is catalogue-only and reachable by entering its code, not shown automatically.