Source

ISAC Chicago — Integrated Database (Egyptian)

Egypt, Nubia, Sudan, Ancient Near East (full ISAC scope is Mediterranean + ANE) · Predynastic through Islamic

500

Objects

147

Inscriptions

1,299

Cross-references

Browse all 500 objects → Visit source →

No imaged objects to show. Try the full browse view.

Source details

Type: museum_collection
Access: rest_api
Stack: ke-emu-solr
License: Public domain metadata; per-image rights per record
Modality: english_prose
ID system: ISAC IRN (internal record number) + accession number (varied)
ID tier: tier-2
Inscription richness: 10/10
Cleanliness: 9/10

Form breakdown

ostracon: 214
other: 71
papyrus: 60
funerary_equipment: 30
statue: 26
scarab: 25
jewelry: 18
stela: 17
seal: 14
architecture: 9
vessel: 6
cosmetic_object: 6
tool: 3
textile: 1

Notes

## Discovered schema (ISAC Chicago Integrated Database, surveyed 2026-05-23) ### HEADLINE: Brief hypothesized OCHRE; reality is KE EMu + Solr + Backbone - **NOT OCHRE.** Stack identified via static analysis of the SPA source: - Apache + PHP backend (Solr proxy at `/solr/index.php`) - Apache Solr index (numFound, response.docs, facet_counts in response; KE EMu field-prefix naming: Acc*, Adm*, Col*, Cat*, Loc*, Pro*, EAD*, Mul*, Bib*, Ins*/InscMark*) - Backbone.js front-end (`https://isac-idb.uchicago.edu/js/app/...`) - **dataUrl** is `https://isac-idb.uchicago.edu/solr/` (hidden in `/js/app/base/collections.js` line 11; lines 4 + 11 reveal a stale dev URL that helped identify the architecture) - This is a **third stack class** beyond MET/BM/BKM (proprietary) and Fitzwilliam (Axiell + CIIM). ### Hardened pre-flight result - Original pre-flight tagged ISAC `open` because `/api/` returned 200 JSON. - Hardened check: `/api/` returns `{"msg":"hello"}` (15 bytes). Every child path under /api/ (objects, items, v1, v2, docs, swagger) returns HTTP 404. FAIL the hardened test. - True data path discovered by tracing the SPA's Backbone Collection `getUrl()` method: `dataUrl + 'index.php?dataset=XX&{query}'`. Plain Solr query semantics. ### Six datasets exposed by Solr `dataset=` parameter - **MC** Museum Collection — 237,535 records total; **38,912 Egyptian**; **4,812 inscribed Egyptian** - **MA** Museum Archives - **ES** Epigraphic Survey — 21,502 records; *archival* (negatives, docs) not transcription content itself - **PA** Photographic Archives - **RA** Research Archive — 29,915 records; bibliographic with `BibUrl` to JSTOR / online PDFs - **CM** CAMEL — maps and remote-sensing (Iraq + Egypt aerial / Corona / Landsat coverage) ### Sample (n=166 MC + 19 RA) - stela n=14 - papyrus n=14 - statue n=14 - jewelry n=14 - vessel n=14 - architecture n=14 - bibliography n=19 (Egypt + IsOnline:Yes) ### Inscription schema (Test 2 from the brief) **Present and structured.** 8 fields per inscribed object: - `ColInscribed` (Yes/No flag) - `InsScript` — Egyptian Hieroglyphs / Demotic / Hieratic / Coptic / Greek / Arabic - `InsDialect` — Middle Egyptian / Late Egyptian / Sahidic / Koine / ... - `InsInscriptionSequenceNumber` — for multi-inscription objects - `InscInscriptionSubject` — subject classification - `InscMarkLanguage` — language tag - `InscMarkLocation` — physical location on object - `InscMarkType` — Letter / Literary / Administrative / Religious / Votive / Funerary - `InscMarkDescription0` (and `1`, `2`, ... for multi-inscription objects) — **English description / translation prose** This is the **third inscription-shape across the museum survey**: - **MET / BM / BKM**: no inscription field - **Fitzwilliam**: full MdC transliteration text per inscription - **ISAC**: structured metadata + English descriptions/translations, NOT transliteration In our 84-sample run, 49/166 flagged inscribed, 23/166 had populated `InscMarkDescription*` entries, 0/166 have multiple inscriptions per object, and 0/166 contain MdC-diacritic transliteration text. (Across the full dataset: 4,812 inscribed Egyptian objects; 2,469 with Egyptian Hieroglyphs as script.) ### MdC transliteration (Test 2 follow-up) **NOT present in InscMarkDescription0 content** (English prose). The ISAC inscription field is a translation/description, not a transcription. The transliteration-fuzzy-match cross-ref path remains confirmed only for Fitzwilliam (N=1). However, ISAC inscriptions can still join to TLA/TM via: - **InscMarkLanguage** = "Egyptian" + **InsScript** = "Egyptian Hieroglyphs" — language/script metadata for filter joins - **Site + InsScript + DatPeriod** triplet — likely uniquely identifies many inscribed artifacts when cross-matched against TM-text records ### Excavation linkage (Test 3 from the brief) **Yes — strongly modeled.** Five provenance fields per MC record: - `ProCountry` — country (Egypt, Iraq, Iran, ...) - `ProRegion` — geographic region (Upper Egypt: Thebes, Faiyum, ...) - `ProSiteLevel1` — site name (Medinet Habu, Abydos, Dendara, el-Lahun, Umm el-Qaab, Ramesseum, ...) - `ProLocus` — specific locus within the site (Tomb of Anpy, Tomb 24 Room 19 the Great Temple, Foundation Deposit, Tomb of Djer) - `ProCulturalAffiliation` — culture (Egyptian, Coptic, Meroitic, Napatan, A-Group, ...) In our sample: 166/166 have a populated site, 78/166 have a specific locus (tomb / room / grave). The locus granularity is **archaeologically usable** — this is the level that lets us link an artifact to a published excavation report. **Bibliographic linkage to PDFs (RA dataset)**: - `BibUrl` — direct URL to published reference (often JSTOR) - `IsOnline` — boolean flag - `BibCallNumber`, `BibAuthorEditorLocal`, `BibType` (Journal Volume Article / Book / Festschrift / Conference / ...) - Bibliographic hierarchy: `BibParentTitleLocal` → `BibGrandparentTitleLocal` (article → journal volume → journal title) Sample BibUrl values from our 10-record RA sample: - Aeragram → http://www.aeraweb.org/wp-content/uploads/2015/01/aeragram8_1.pdf - A History of Egypt. Vol. IV → https://books.google.com/books/pdf/A_History_of_Egypt_Under_the_Ptolemaic_D.pdf?id=VnI01LsSRtoC&output=pdf&sig=Ro7YOWlhRhIHLFuGk3dd9aM4PXw - A History of Egypt. Vol. V → https://books.google.com/books/pdf/A_History_of_Egypt_Under_Roman_Rule.pdf?id=-x8NAAAAIAAJ&output=pdf&sig=Ayv886GEnoxLXRegBPluG9KSiFE This is the **bridge to print-OCR work** the brief asked about — once we have an RA record for a published volume + an MC record with a matching ProSiteLevel1 / ProLocus, we can cross-reference the artifact to the volume that publishes it, and download the PDF (where IsOnline=Yes) for OCR-based transcription mining. ### Demotic Dictionary check **NOT a sibling resource of the IDB.** Lives at: `isac.uchicago.edu/research/publications/demotic-dictionary-oriental-institute-university-chicago` This is a static publication landing page — a published-print dictionary with downloadable PDFs of completed letter volumes. Paths `oi-idb.uchicago.edu/dictionary/` and `isac-idb.uchicago.edu/dictionary/` both return HTTP 404. The Demotic Dictionary needs its own future survey targeting its PDF chapters as text-source, not as a queryable database. ### Six-way museum comparison (museum pattern N=6) | metric | MET | BM | BKM | Fitz | Leiden | **ISAC** | |---|---|---|---|---|---|---| | total Egyptian (sampled / available) | 77 of 27,972 | 13 of ~100k | 4 of ~80k | 80 of ~16,801 | 0 of ?? | **166 of 38,912** | | has inscription field | 0/77 | 0/57 | 0/4 | 6/109 | gated | **49/166** | | with content/text | 0 | 0 | 0 | 6 (MdC) | gated | **23** (English) | | multi-inscription/object | n/a | n/a | n/a | yes (Fitz) | gated | **yes (0 observed)** | | MdC transliteration | 0 | 0 | 0 | yes | gated | 0 (English instead) | | structured find spot | prose | no | no | yes + coords | gated | **yes (166/166 site, 78/166 locus)** | | with structured period | partial | prose | prose | yes (71) | gated | **yes (157/166)** | | with structured dynasty | yes | n/a | n/a | n/a | gated | **yes (43/166)** | | direct API access | yes | no (CF) | no (Vercel) | yes (key) | no (CF) | **yes (Solr endpoint)** | | stack | proprietary | proprietary | proprietary | Axiell + CIIM | WordPress + CF | **KE EMu + Solr + Backbone** | | coverage on direct path | 100% | <0.1% | <0.01% | 100% | 0% | **100% (38,912 Egyptian!)** | | PDF/publication linkage | no | no | no | publications field | gated | **yes (RA `BibUrl`)** | ### Flag interpretations - Flag 1 (object vs inscription modeling): CONFIRMED structured-inscription modeling at ISAC. The decision to use a separate `inscription` table (1:N) is well-validated now (Fitz + ISAC both). - Flag 2 (transcription presence): VARIES by source. Fitzwilliam has MdC transliteration; ISAC has English description. Both are useful — they populate different columns of the `inscription` row. - Flag 3 (cross-ref gap): NOT a gap for ISAC's RA → publication PDFs. RA's BibUrl directly links to JSTOR / online PDFs of OIP volumes, giving us the print-OCR-transcription bridge the brief asked about. - Flag 4 (image licensing): "Public domain metadata; per-image rights per record" per ISAC's how-use-idb page. ### Implication for the museum-pattern verdict (N=6) 4 of 6 museums surveyed give us direct programmatic access (MET, Fitzwilliam, ISAC, + 2/3 gated via WAF — BM/BKM/Leiden). Of the 4 with direct access: - 1 has no inscription content (MET) - 1 has MdC transliteration (Fitzwilliam) - 1 has English description (ISAC) - 1 has nothing modeled Inscription modeling is **institutional choice, not vendor-locked.** Fitzwilliam is Axiell+CIIM with MdC. ISAC is KE EMu+Solr with English. Both expose structured inscription fields; the field is populated to different research-community conventions. ### Operational gotchas / future use - **Solr endpoint pattern**: `https://isac-idb.uchicago.edu/solr/index.php?dataset={MC|MA|ES|PA|RA|CM}&q=...&rows=N&fq=...&wt=json` - Solr query syntax: standard. `fq=ProCountry:Egypt`, `fq=ColInscribed:Yes`, facets via `facet=true&facet.field=...` - Multiple `fq` parameters AND together; chain in URL or pass as list. - No rate limit observed during this run; politeness sleep 0.4s used. - Session cookie `isac-idb` is set on first hit; not required for /solr/ calls (we confirmed cookies-vs-no-cookies returns the same data). - The Backbone `/id/{N}` URL pattern works for human browsing but returns the SPA HTML, not data — use `/solr/index.php?...&q=irn:{N}` for the JSON record. ### Open questions / follow-ups - The ES (Epigraphic Survey) dataset has 21,502 archival records about the project but the actual epigraphic content (transcriptions of Medinet Habu inscriptions, Karnak, etc.) is likely in the published OIP volumes, not in ES. The ES dataset lists the photo-negatives + docs used to make those volumes. Means a side survey of OIP volumes might surface the transcription content. - The Chicago Demotic Dictionary's PDF chapters (A, B, F, H, I, S, T, W, Y, Z complete) live at isac.uchicago.edu. Worth ingesting as a separate text source later. - ISAC's IRN format (numeric, no prefix) is the cleanest cross-source join key we've seen — easy to validate and dedupe. - The BibUrl-IsOnline pairing on RA can be filtered to find ALL ISAC Egyptian-related publications with online PDFs; that's a discrete next task that complements this survey.