Source

Brooklyn Museum — Egyptian, Classical, Ancient Near Eastern Art

Egypt and Sudan (and Classical, Ancient Near East) · Predynastic through Coptic

2,786

Objects

13,739

Images

5,572

Cross-references

Browse all 2,786 objects → Visit source →

Sample objects

Northern Group of Pyramids at Meroe (Kabushiyah) as seen, in 1921, from the NNE Prior to Their Excavation

74.48

Recarved Plant Scroll with Snakes and Bird Heads

68.3 · El Behnasa (Oxyrhynchus), Egypt · Late Egyptian Antique Period and Modern

Three Busts on a Capital

72.10 · Egypt · Modern

Paralytic Raising His Bed

62.44 · Antinoe (El Sheikh Ibada), Egypt · Modern

Recarved Plant Scroll with Human Figure and Lion Head

67.176.2 · El Behnasa (Oxyrhynchus), Egypt · Late Egyptian Antique Period and Modern

The Holy Family(?)

77.129 · Egypt · Modern

Roundel with Human Head

60.212 · Antinoe (El Sheikh Ibada), Egypt · Modern

Source details

Type: museum_collection
Access: wikidata_sparql_fallback
Stack: proprietary-cms-vercel-gated
License: Mixed (BKM site terms; Wikidata gateway data is CC0)
Modality: none
ID system: BKM accession number (often year-prefixed) + Wikidata Q-ID
ID tier: tier-2
Inscription richness: 0/10
Cleanliness: 3/10

Form breakdown

statue: 1,144
jewelry: 563
other: 373
vessel: 294
papyrus: 167
architecture: 99
stela: 61
funerary_equipment: 49
cosmetic_object: 18
tool: 8
textile: 7
document: 2
seal: 1

Notes

## Discovered schema (Brooklyn Museum via Wikidata fallback, surveyed 2026-05-23) ### HEADLINE FINDING: Brooklyn API is gated; documented endpoint no longer accessible The brief asked us to verify the current API endpoint at `api.brooklynmuseum.org/api/v2`. Result: that subdomain is **NXDOMAIN** as of 2026-05-23. The historical OpenCollection API lived at `www.brooklynmuseum.org/opencollection/api/v2/` (documented as recently as February 2025 per Wayback) and is now uniformly behind **Vercel Security Checkpoint** — every path including /api/v2/object, /opencollection/api/v2/object, /sitemap.xml, and /robots.txt returns the 33KB Vercel JS challenge page (HTTP 429). No programmatic access. Probe outcomes are stored as `_probe:*` samples. ### Brooklyn joins BM in the "gated" column With Brooklyn confirmed gated, **2 of the 3 surveyed museums are programmatically inaccessible**. The museum-pattern verdict has flipped: open APIs are now the exception, not the rule. MET is the outlier. Implication for the remaining ~25 sources: budget realistically for paid aggregators or direct museum partnerships. ### Fallback: Wikidata SPARQL Survey via `https://query.wikidata.org/sparql`, filtering BKM items by their specific Wikidata property — **P4740** "Brooklyn Museum artwork ID" (NOT P195 which museums commonly use; BKM is exceptional here). ### Coverage reality - Wikidata items with P4740 (any BKM object): 1,001 - Of those, Egyptian-likely by ANY filter: 4 - BKM holds ~80,000 Egyptian objects. **Effective coverage <0.01%.** - Cross-refs: TM 0/4, TLA 0/4. ### Sample by form category - statue n=3 - other n=1 ### Schema observed (Wikidata entity JSON; per-Q-ID — same as BM) - P4740 BKM artwork ID - P217 inventory number (accession, when populated) - P31 instance-of (multi-valued) - P186 material; P189 location of discovery; P575 time of discovery - P495 country of origin; P571 inception - P2048/P2049/P2610 dimensions in mm - P18 image (Commons filename → Special:FilePath/{filename}) - P282 writing system, P407 language (when present) - P8532 TM text ID — none in our sample - P12185 TLA object ID — none in our sample - P973 described at URL — sometimes the BKM permalink ### Four-flag comparison: MET vs BM-via-Wikidata vs BKM-via-Wikidata MET (#3) BM (#4 via WD) BKM (#5 via WD) total surveyed 77 57 (13 Egyptian) 4 (4 Egyptian) inscription text in any field 0/77 0/57 0/4 TM cross-ref 0/77 2/57 0/4 TLA cross-ref 0/77 1/57 0/4 writing-system tag n/a 2/57 0/4 image 77/77 48/57 4/4 collection coverage on path 100% (API) <0.1% (WD) <0.01% (WD) direct API access yes no (Cloudflare) no (Vercel) TM/TLA seeds for fuzzy match none ~3 high-value pairs none ### Flag 1 — Object vs inscription modeling (CONFIRMED across all 3 museums) The MET conclusion (object and inscription as separate one-to-many tables) holds universally. No museum surveyed so far models inscriptions as content. ### Flag 2 — Transcription presence (CONFIRMED ABSENT across all 3 museums) Museum-curated data simply doesn't carry transcription. Locked in: the inscription corpus comes from TLA/TM/EpiDoc projects, joined to museum objects by IDs. ### Flag 3 — Cross-ref gap (BKM is the worst case) BKM has zero TM/TLA cross-refs in its (tiny) Wikidata coverage. BM had a couple of high-value seeds (Rosetta Stone, Papyrus of Ani). MET had none. So the fuzzy-match pipeline can't be seeded from BKM at all — we'll need TLA/TM-side records that mention BKM accessions to bootstrap that direction. ### Flag 4 — Image licensing Wikidata-mediated images come via Commons (uniformly free). BKM-side rights are unknowable while their API is gated. BKM historically had a permissive non-commercial metadata license but image rights varied per object — same per-object model as MET. ### Operational gotchas (in addition to BM survey) - BKM Wikidata property is **P4740**, not the more common P195 (collection) or P217 (inventory). Wikidata property choice is inconsistent across museums — verify per-museum (see meta.wikimedia.org for property catalogue). - Vercel Security Checkpoint vs Cloudflare Turnstile: same effect (programmatic access blocked), different signatures. Vercel = HTTP 429 + 33KB; Cloudflare = HTTP 403 + ~5KB. The "Apply for your API key" page is itself behind the gate, so even getting a key requires solving a JS challenge. - Wikidata SPARQL endpoint is flaky on multi-filter queries — saw an HTTP 502 mid- survey on the label-regex filter. Treat 502 like the BM 504: split into smaller queries, retry transient failures. ### Open questions / follow-ups - Contact BKM directly about API access (Tina Hertel / their tech team). BKM had a strong open-data culture historically; the Vercel gating may be incidental to a recent platform migration rather than policy. - Check Wayback CDX for older cached BKM API responses (raw JSON). Even a small cached corpus would give us schema confidence for when we DO get API access. - DPLA has BKM content; get an API key. - **Bigger picture**: 2/3 museums gated. Time to do an access pre-flight for the remaining ~25 sources before committing to per-source scripts. Logged separately.

Brooklyn Museum — Egyptian, Classical, Ancient Near Eastern Art

Sample objects

Head of Akhenaten Made in Two Pieces

Upper Half of Statue of Akhenaten

Cast of a Relief of Akhenaten, Nefertiti, and Princesses

Inlay Head of Akhenaten

Scarab