world/regions/sumatra_java/page/sumatra_java_S16.md
CANON SUBORDINATION — this document is PROPOSAL-TIER: it serves canon and never outranks it.
The canon: the CVD · the T1 foundation docs · the T0 registries (registries/) · the spine
(docs/spine/CH_*.md) · the region pages (_source/02_Tier_2_Region_Pages/). Authority order:docs/DOC_MAP.md§ 0.
Canon served: NONE FOUND — this document cites no canon anchor anywhere in its body.
Do not build from it alone: open the spine node (docs/spine/CH_NN.md), the region page
(_source/02_Tier_2_Region_Pages/), and the T0 registry your work targets, and derive from those.
READ THAT CANON FIRST — open it and derive from it before you build anything from this document.
If this document disagrees with canon, CANON WINS and this document is the defect — fix the
document, never the canon. Nothing here is applied until it is ratified into canon.
Every citation carries its use_class (SUBSTRATE · ESOTERIC-PRIMARY · AUTHORIAL-REBUTTAL · REFERENCE ·
TONE · DUPLICATE · FRINGE-SKIP). Book-substrate citations arrive via the query layer with SRC_ ids, region-
filtered (query_substrate --region sumatra_java); the retrieval-provenance trail (corpus_query_log)
rides the page. The region filter is a correctness requirement, not a convenience: the corpus's
strong region hits are Batak and Sumatran, and an unfiltered semantic query for the Java monuments
leaked cross-region (a "Borobudur mandala relief" query ranked a Nanda Devi pilgrimage source), so
the Java-monument and Srivijaya-maritime substrate is honestly THIN in the processed corpus and is
carried as an acquisition gap below rather than back-filled with a false citation. Pass-2 did not
change that shape. It deepened the Batak side by half again and left the Javanese side exactly where
it was, which makes the imbalance sharper and the acquisition list below more load-bearing, not less.
One query-layer note rides at the top of this section because every builder hits it first. The
--region flag takes a CANONICAL region_id and expands it to the gazetteer's cultural slugs at query
time, so sumatra_java resolves to BOTH sumatra_batak and malay_world in a single call: one filtered
query returns the whole shelf, and a shelf name or a raw slug is not a legal --region value. All six
processed sources sit on the sumatra-batak shelf and carry the sumatra_batak substrate tag, so
Section 20's processing-status line stands as written. What does differ is RECORD metadata rather
than the query tag: five of the six EXR records read primary_region sumatra-batak and SRC_00158
alone reads primary_region sumatra_java. That anomaly carries a content rule and no query
consequence — SRC_00158 is shelved sumatra_java but a pustaha is a BATAK divination manuscript, and
everything that record supplies is attributed to the Batak, never to Java, regardless of the shelf
label.
Six sources are processed, deep-read and audited across the two record shelves — four carried from
pass-1 and two added by pass-2. All six are declared here, and a seventh shelf RECORD — a duplicate
conversion of the sixth source — is declared at the end of the list as the duplicate it is and never
as a seventh source. The Region Identity block's processed-source line names all six, and the
Section 20 session plan routes both additions.
North Sumatra (manuscript cultures 14, CSMC Hamburg, 2020) — THE region keystone and the
best-anchored source on the shelf: the Batak history and sub-group map, the datu and the bius
village-cluster order, the pustaha bark-book craft, the full divination taxonomy from the porhalanan
calendar table through the four-instrument oracle ladder, the tondi and sahala material (held in
pass-2; only the anchor correction is recorded below), the pangulubalang and porsili ritual
constructs, the Surat Batak script, and the Srivijaya-to-Majapahit port history
[SRC: SRC_00226] · SUBSTRATE
— the pustaha's interior read as a working instrument: the porhalaan and pormesa zodiac table, the
spirit diagrams, the datu's ritual arsenal, the multi-hand book, and the manuscripts' own collection
history [SRC: SRC_00158] · ESOTERIC-PRIMARY
Museum of Art, 1994) — the material-culture anchor for Sections 3, 6, 7 and 10: the saddle-roof
house and its singa guardian, the si gale-gale effigy, the piso sanalenggam sword, the gunpowder
and pupuk recipes, the kulcapi lute, and the Island-SEA maritime and Srivijaya-Majapahit frame
[SRC: SRC_00446] · SUBSTRATE
443, 2021) — the Section-7 sculptural anchor and the shelf's only coast-to-highland economic
witness: the Pasaguan ancestor images in Toba tuff on the Aek Tarum bank in Asahan regency, the
anisan grave post, and the Batubara port network the trail fed [SRC: SRC_00152] · SUBSTRATE
Manuscripts in Leiden University Library", Sari 25: 9-21 (2007) — NEW IN PASS-2. The collection
history told about itself, plus the three-medium writing ecology (bark, bamboo, European paper), the
literary-genre and divination-method inventory, the hadatuon chain-of-teachers naming teacher,
author and commissioner in the colophon, the Dairi high-creator material, and the torhan-torhan
riddle form [SRC: SRC_00228] · SUBSTRATE
surat basic signs and the anak ni surat diacritics with their four functions and regional
restrictions, the letterform genealogy and the Na-letter dating chain, the regional-variant analysis
that lets a hand be placed, and the printing-type history [SRC: SRC_01758] · SUBSTRATE
the same work already carried as SRC_01758, twenty-seven chunks, pending de-duplication upstream.
It is declared so that a builder who meets the id on the shelf knows what it is, and it is never
cited as corroboration: a second hit on the same Kozok text is the same evidence counted twice and
not a second witness. Every script claim on this page cites SRC_01758, and nothing on this page
rests on the duplicate [SRC: SRC_01914] · DUPLICATE
use_class is declared per source above, but two of these sources carry interior registers that a
builder must handle differently from the source's body, so those registers are declared separately
here rather than being smuggled in under a SUBSTRATE token.
the interior Batak as "Kataran ... cannibals of a peculiarly ferocious disposition" is a named East
India Company officer's slur in a named year. The audit RULED against the timid handling: it is not
authorial-context-only, it is attributed-and-contradicted. It stays in the substrate, attributed on
its face to Anderson and to the Company's commercial interest in the people it was trading with,
reachable by a player who reads the colonial record, and visibly refuted by everything else on that
same page — Batak cotton grown in the country, woven at home, dyed with home-grown indigo, coloured
with forest woods, and moved through a port working six hundred perahus
[SRC: SRC_00152, p. 12]. It is never repeated as fact about the community. Deleting it protects
nobody and hides how the defamation was manufactured; attributing it does the teaching.
itself frames critically as Western writing about a mysterious interior. The honest object is the
CLAIM and what the claim did: a European account, wrong in its framing, that circulated for
centuries and helped justify intervention. The build makes no counter-assertion the sources do not
support either: SRC_00158 says only that the missionary and traveller reports were greatly
exaggerated, and "greatly exaggerated" is not "never happened", so a flat in-world negative is as
much a departure from the record as the slur is. The corrective instrument is the literacy, the
book-craft and the calendrical science, which this shelf now supplies in quantity.
[SRC: SRC_00446, PDF p. 10 = print p. 9]. Object data from that catalogue ships; the diffusionist
frame does not, and the catalogue repudiates its own older discourse at PDF p. 20 = print p. 19.
Nothing on this page derives Sumatran or Javanese achievement from an external superior source.
South-Sumatran proto-script origin in its own text and the harvest row that stated it as fact was
CUT for exactly that reason. The surviving weave carries the intermediary as a leading hypothesis
and the page states it no more strongly than that.
No source on this shelf earns TONE or REFERENCE, and DUPLICATE falls only to the duplicate record
declared above.
Every one of the six carries content a player of the frame can walk into; none is a laboratory result
that only governs from outside the fiction.
philologist of Batak manuscripts, written as public outreach so that students can read original
Batak manuscripts, and where the other five touch letterforms it wins. Four consequences bind: the
diacritics have FOUR functions, not three — replace the inherent /a/ with schwa, /e/, /i/, /o/ or
/u/, add the nasal /ng/, add the fricative /h/, delete the /a/ — with the schwa restricted to Karo
and Pakpak and the /h/ to Karo, Pakpak and Simalungun, and a Karo or Pakpak literacy puzzle is
unsolvable without those restrictions; the basic-sign inventory is NINETEEN ina ni surat, stated
flatly and twice in this source, so the nineteen-to-twenty range that varies by dialect is cited to
SRC_00158 (fn 21 at print 168) where it actually lives and never to Kozok, while what does vary
inside SRC_01758 is the anak ni surat count — five in the opening statement, six in Toba, eight in
the cross-regional set the conversion's dropped table would have carried — and that spread is not
standardized away either; the script spread SOUTH to NORTH, from Mandailing to Toba and Simalungun
and only afterwards to Pakpak and Karo; and Kawi influence at the early stage is documented on real
epigraphic reasoning, the Batak Ja being Da plus a horizontal stroke, which holds in Kawi but not in
Pallava. That last point ships alongside the shared-ancestry framing rather than being suppressed
for a cleaner Sumatra-independent narrative — dropping a documented influence to protect a tidier
story is its own kind of scrub, and it costs the player the best piece of real epigraphy on the
shelf [SRC: SRC_01758 — Sejarah aksara Batak, Ina ni Surat; Anak ni Surat; Analisis Ina ni Surat].
most reliably anchored source here. Where it and a thinner source disagree on Batak ritual practice,
it governs — with one standing exception recorded so a later reader does not "correct" a true line
into a false one: the calendar table is spelled porhalanan as the catalogue's principal form and
porhalaan as an attested variant IN THE SAME BOOK, and both spellings are cited rather than one
being normalized away. Its independent count of the basic signs, the nineteen main Toba characters
at print 76, agrees with Kozok and is the second attestation of that number on the shelf.
where the build wants certainty and those hedges survive: the closing string on a pustaha was needed
to carry and store the manuscript, PROBABLY open, hung on the house wall; the calendar's horizontal
ruling is twelve OR THIRTEEN lines, which is the alternation the whole rabbit hole is about; the
wizard figure in a spirit diagram is PROBABLY the datu himself; and the basic-sign inventory is a
RANGE of nineteen to twenty varying by dialect (fn 21 at print 168), which is this source's own
formulation and is not transferred onto the paleography. Its source_year is recoverable as
2022 from the DOI printed on its own first page, and that typed field is null pending a metadata
pass; its title and author fields, alone on this shelf, are populated.
reports a collection's history with its own actors named. Its typed title, author and year fields
are null while its provenance note carries all three; the gap closes at the merge step.
own 1975 site visit, published as a 2021 field note, with no measurements taken, no GPS, the exact
location now lost, and the author hedging his own key claims throughout — "appears to exhibit", "may
be", "possibly", "would be", "perhaps". It is cited as OBSERVATION, never as settled scholarship.
Two of its own references cannot be resolved from it: a "Bartlett (1927)" citation carrying the
Pardembanan attribution, against a bibliography listing only Bartlett 1921 and 1930/31, and an
unfilled placeholder for a Sibeth plate number. Any build beat turning on the Pardembanan
attribution consults Bartlett directly first. Its pigment claim is likewise an inference from
looking at weathered paint, not an assay: the colouring WOULD BE haematite and POSSIBLY charcoal or
lamp black, and no player is ever told the stone was analysed.
Smithsonian introduction — and nearly every BATAK ethnographic claim it carries traces through the
catalogue to a small pre-WWII colonial corpus: Warneck 1909, Loeb 1935, Bartlett 1934, Giglioli
1893, and Brenner 1894. Using the technical observations while declining the framing is the correct
handling and the catalogue itself repudiates that discourse. The operating rule for this page:
anything from SRC_00446 that would harden into a civilizational mark or a science-grounding row
carries the provenance visibly and takes a second source first.
Three governing facts fall out of the shelf as a whole and bind every section that cites it.
Lombok. Flores is a Wallacea island; these are not. None of the six sources on this shelf is a
biogeography source, so any Sunda-Shelf, land-bridge or Wallace-Line statement on this page is
carried from the peer-reviewed anchor declared on the bali page's own Section 16 and is never
sourced to a Batak manuscript record or to a museum catalogue.
and the page's present frame is ~1200 CE. Some of them look much further back inside that record —
SRC_01758's paleography runs to the ~400 CE Kutai and Ci Aruten inscriptions and the seventh-century
Srivijayan Old Malay, and SRC_00446 and SRC_00226 carry the Srivijaya-to-Majapahit sequence — but the
reaching-back is modern scholarship's, not the era-present player's. The catalogued objects are
mostly nineteenth- and twentieth-century; the manuscripts were collected from the 1850s onward; the
field visit was 1975; the scholarship is 2007 to 2022. What carries back into the 1200 CE present is
the deep layer these things encode — the cosmology, the craft, the calendrical science, the script
tradition. What does NOT composite into the present frame is any dated object, any collection event,
any lapsed practice, or any modern scholarly result. Those belong to Section 18's forward strata or
to authorial notes. This is the single failure mode that most reliably slips past a gate: a
genuinely interesting modern finding admitted as depth the era-present player investigates.
records carry anchors that are NOT in print space, and one carries a MIXED convention inside a
single record.
Each source paginates differently and each was verified against its stored text during audit. Citing a
raw PDF index as a printed page is the single most common provenance defect in this lane and it has
already produced real drift on two shelves. Builders use the conventions below, and an anchor outside
a verified set is checked before it ships.
across sixteen sampled pages. This record carries a MIXED convention internally and it is the
dangerous one: the token "p. 19" resolves as print 19 in the oracle-day and great-serpent rows and
as PDF 19, that is print 17, in the threatening-letters row — the same string meaning two pages
inside one record. The corrected print anchors are: the writing-bamboo initiation and the letter to
the missionary at print 76; the threatening-letters sub-genre at print 17; the tondi and sahala
material at print 80-81; the protective formula at print 79, with the paired invocation names at
print 68; the Toba-for-pustaha and Karo-for-bone distribution at print 60; the nineteen main Toba
characters at print 76; the Sanskrit loan pair at print 70; the itinerant datu at print 60. Two
further corrections are not offset errors but range errors: the calendar table runs print 77-79, not
74-79, and the bullet-warding bone amulet is print 82 for the figure plus print 83 for the
description. The oracle ladder at print 70-71 and the pustaha craft at print 65-67 resolve correctly
as print and need no adjustment.
every anchor in the record resolves in print space. This is the only source on the shelf where the
naive reading is the correct one.
ride this record rather than its content. The chunk metadata stores page_start and page_end in
PDF-index space 1-16, so anything citing from chunk metadata rather than from the running heads
produces an anchor off by 163; and the chunking splices footnote text into body text at page
boundaries, most consequentially at the print 166/167 boundary, where footnote 6's continuation runs
into the body sentence and yields a tail belonging to Marsden's tribe list rather than to the datu
profession — which would produce a confidently wrong claim about a real people's institutions if a
downstream extraction ran over it. One anchor in the record is also genuinely misplaced and is
corrected rather than inherited: the genealogy, Hinduism and lexical material sits on print 165, not
print 166.
found exactly four PDF-index leaks in the record and no fifth: three anchors reading "p. 4", which
is print 12, and one reading "p. 2", which is print 10. Any anchor outside the 9-21 band is a PDF
index by construction, because the article has no page outside it.
PRINT = PDF minus one, proved programmatically across twenty-six anchors by the printed folio
surviving as the last line of each extracted page. Every anchor resolves correctly IN PDF SPACE, so
no content claim is wrong on this account — but a human opening the physical catalogue lands one
page off. Until the record is relabelled, every SRC_00446 anchor on this page carries an explicit
space label, as the Heine-Geldern citation above does.
so a page number in an SRC_01758 citation is a fabrication by construction. The headings in use,
reproduced exactly as the record cites them, are Ina ni Surat, Anak ni Surat (also written Anak ni
Surat / Tanda Diakritik), Aksara batak dan sejarahnya, Silsilah Aksara, Analisis Ina ni Surat,
Persamaan Surat Batak, Surat Ulu, dan Surat Incung, and Huruf Cetakan Zending & Landsdrukkerij. One
standing rule rides with it: the ingested file carries the same material twice, an earlier draft
block and a later REVISED and EXPANDED block that diverges from it across forty-two line pairs. Cite
the revised copy; treat the first as superseded. Before any re-index, check the chunker — a dedupe
that keeps the first copy silently ships the less accurate text. The same source also carries
conversion typos, including a Kalimantan Barat that should read Kalimantan Timur for the Kutai
inscription; where the page corrects the source it says so in place, so a later verifier who checks
the anchor and finds the typo does not read the correction as fabrication.
on the page. All six processed sources are Batak and North-Sumatran; not one carries Borobudur or
Prambanan at build grain. Sections 1, 5 and 7 and the monument beats stay gated on this.
relieved at frame grain only: SRC_00446's introduction carries the Srivijaya-to-Majapahit sequence
and SRC_00226 the port history, neither at the grain Sections 8, 9 and 18 need for a declining port
economy.
Records — UNRELIEVED, and pass-2 made this asymmetry worse in a way worth naming: the inscription
thread's BATAK half is now deeply sourced and its OLD-JAVANESE half has no processed anchor at all,
so the two-inscriptions comparison currently rests on one well-sourced end and one unsourced end.
Flagged to the sections that own that scene rather than resolved here.
this entry changes rather than disappears. What remains wanted is a securely paginated edition with
the letterform plates intact: the corpus copy has no print pages, its conversion dropped at least
one letter-inventory table, and it carries typos. Anything that needs a citable plate or a stable
page reference takes the published edition.
and no further; see the verify list for the mechanism claim that must not ride it.
SRC_00228 carry the high-god and creation material at catalogue and article grain, which is enough
for ambience and not enough for the structural account the cosmology beats assume.
the folklore lane already needed AND one of the pre-WWII sources SRC_00446's Batak claims trace
through, so reading it directly serves the provenance check at the same time.
raised by pass-2 and specific: SRC_00152's Pardembanan attribution rests on a reference that cannot
be resolved from its own bibliography, and the grave-post material is exactly what Section 7 builds
on.
cannot be cleared on SRC_00228 alone; a Perak-Malay demonology anchor is the acquisition that would
clear them. Nothing from those rows appears anywhere on this page until it does.
plate caption reading "Datu with pustaha, medicine horn and magic staff"; the vernacular names
commonly attached to them appear nowhere in that article. The names are real and well attested in
the wider Batak literature, so this is a citation gap and not a content gap, but SRC_00158 must not
stand as their citation in any weapon or object registry row.
deep-research obligation.
"manufactured by the welding together of thin strips of steel" and NOT pattern-welding, which is a
specific decorative-structural technique the source never claims. The source also calls the
technique a later development documented for late-nineteenth and early-twentieth-century examples,
so a 1200 CE build does not inherit it as ancient [SRC: SRC_00446, PDF p. 67 = print p. 66].
rose in smaller eruptions continuing for thousands of years, and "resurgent doming" appears nowhere
in it. Either the book's own wording or a real geological citation, never a substituted mechanism
riding an anchor that does not support it.
nothing on that question ships on this page in any form until it clears. Recorded so the gap is
visible rather than quietly filled from general knowledge.
not an attested fact: the source says only that the blue dye is obtained from a species of indigo,
tarum, abundant throughout the island, and never comments on the toponym [SRC: SRC_00152, p. 12]. A
curious player can be told it is a reading; the page does not state it as attested.
only in their provenance notes — SRC_00152, SRC_00226, SRC_00228, SRC_00446 and SRC_01758 — and
SRC_00158 alone has typed title and author and needs only its year (2022, from its own printed DOI).
The two that are already recoverable in full are SRC_00228 (Brakel-Papenhuyzen, 2007, Sari 25: 9-21)
and SRC_01758 (Kozok, Sejarah aksara Batak). None is an audit-owned key; all close at the merge step.
the same shelf is almost certainly a second conversion of the same Kozok work and needs
de-duplicating before re-index; and SRC_00158 carries two chunk-level traps of its own — its chunk
metadata is in PDF-index space, which is the off-by-163 trap named above, and its chunking splices
footnote text into body text at the print 166/167 boundary, where footnote 6's continuation runs
straight into the body sentence and yields "The profession was hereditary / in Angola and Toba, and
three in Mandiling", a tail belonging to Marsden's tribe list rather than to the datu profession.
A downstream extraction running over that boundary would produce a confidently wrong claim about a
real people's institutions.
a rabbit hole because the harvest had left it on the table. The bones of it are cited and it is
ordinary substrate told honestly: the manuscripts are framed as objects functioning in parallel in
Western and Eastern societies, and from colonial times were often brought to Europe as souvenirs and
kept in national or private collections, with roughly two thousand now in collections worldwide
[SRC: SRC_00158, p. 164-165]; van der Tuuk is named there [p. 169] and the article that carries his
Leiden collection records that the Dutch Bible Society employed him to translate the Bible and
describe the language for the instruction of Protestant missionaries [SRC: SRC_00228, p. 10]; the
mission civilisatrice took as its goal the annulment of the datu's role in society by taking objects
out of circulation [SRC: SRC_00226, print p. 17]; the mission doctor who studied the language also
collected the artefacts [print p. 11], acquired one around 1910 and sold it to the ethnological
museum in Hamburg in 1912 [print p. 81]. Suppression, extraction, sale and display is a complete
documented chain and it is the honest answer to why a player can read these manuscripts at all. Two
disciplines bind it: the antagonist is the colonial and mission apparatus, never a community; and
Batak agency is SHOWN rather than routed around — the datu who taught a European collector to
interpret pustaha and was employed to write and copy them is the community acting, and a collection
can be extraction and preservation at once without either half being deleted.
omitted. The paragraph that gives the caravan route its entire economic substance — six hundred
perahus at Batubara, rattans, salt-fish, horses and silk cloths as the principal exports — also
records that enslaved people and horses were brought down from the interior in considerable numbers
[SRC: SRC_00152, p. 12]. The true-cause clause governs how it is staged: build the INSTITUTION and
its real causes, debt-bondage and captive-taking and elite brokerage at both ends of the route under
real economic pressure, with highland communities as agents inside the network rather than only as
its cargo. Concentrating a documented regional institution into one corrupt individual would spare
the historical polities their own agency and teach the player something false, which is precisely
the scrub the clause forbids. Note also that salt is the field-note author's own hedged speculation
and is absent from the 1823 export list, and that the port sits slightly north of the site — the
route is a legitimate synthesis of two adjacent real facts, and it is ours rather than the source's.
reading, then the eight animal-deities of the eight compass directions, then the seven deities of
the seven weekdays, then the five deities of the five parts of the day read on a five-by-five
lettered grid, each escalation costing time and offering-materials and sharpening the answer
[SRC: SRC_00226, print p. 70-71]. The constellation names are Sanskrit loans — Indic astronomical
knowledge received and localized, credited to both ends.
computational function, its only function being divination [SRC: SRC_00226, print p. 77-79;
SRC: SRC_00158, p. 170]. The twelve-or-thirteen line alternation is the interesting part and is not
flattened to twelve.
extended, worked with a balobas ruler and a panggorit knife in black, red and yellow ink, between
carved wooden covers with a rattan closing band [SRC: SRC_00226, print p. 65-67]. The bark is the
same tree that yields agarwood, which is a trade-and-chemistry thread of its own.
tree-bark in a pan-Batak spelling meant to be read across dialects, stories and laments and
love-charms on bamboo in one dialect's phonetic spelling, and personal or narrative texts on
European paper produced at a collector's request [SRC: SRC_00228, p. 11, p. 13, p. 15]. The
load-bearing point is that letters inscribed on bamboo prove reading and writing were not entirely
the privilege of the religious experts [p. 13].
chain and the regional-variant analysis [SRC: SRC_01758 — Silsilah Aksara; Analisis Ina ni Surat].
Its ceiling is part of the interest: six writers from six districts within the Toba sphere already
differ measurably, no firm eastern-western Toba boundary exists, and too few manuscripts have secure
provenance, so the honest result is a RANGE and a pinpoint is unattainable by construction.
World War, nobody still fully commanded the writing tradition, dozens of invented letterform
versions circulated, and students taught the new forms could not read original Toba manuscripts — and
the corrective was built from the manuscripts themselves, by mapping letterforms across a hundred
Batak naskah and choosing the most frequent form [SRC: SRC_01758 — Huruf Cetakan Zending &
Landsdrukkerij]. The fact carries at full weight and the Batak-led corrective carries as agency; the
scholar's aesthetic contempt for modern revivalist forms is not the game's voice to repeat, and
declining it drops no fact.
market for pustaha has run for at least a century and some pieces reached museum collections through
middlemen, while bone amulets were often faked outright, certain regional specimens are to be
regarded with suspicion, and faked inscriptions are meaningless rather than merely wrong
[SRC: SRC_00226, print p. 65, print p. 83-84]. Real hands can be told from imitations of hands, which
is the same skill the letterform analysis teaches.
with access restricted to sons of wealthy people, clan leaders, or close relatives of a datu
[SRC: SRC_00228, p. 11, p. 12]. Much of that knowledge is genuinely lost today, the manuscripts
surviving while few can read them and the Dairi language endangered [p. 19, p. 20], and the causes
the record names are plural — Christianity, Islam, modernization and globalization [p. 19], together
with a dwindling number of scholars of Batak languages [p. 10]. The multi-cause account is the honest
one; single-cause routing erases both the record and Batak agency. This material is handled as
heritage weight and is never recruited to do reveal-ladder work.
black in haematite and probably charcoal or lamp black, while the woven cloth described in 1823 is
chiefly dark blue with red and white lines intermixed [SRC: SRC_00152, p. 5, p. 7, p. 12]. The images
depict cloth in the pigments available to a stone carver, not in the dyes available to a weaver, and
that distinction is a better craft observation than the smoothed-together version.
The gate rule runs both ways: no SRC id appears here that is not cited in a body section, and no
source cited in a body section is absent here. SRC_00152, SRC_00158, SRC_00226 and SRC_00446 are
already carried across the body from pass-1. SRC_00228 and SRC_01758 are declared here AND cited here
with resolvable anchors — SRC_00228 at p. 10, p. 11, p. 12, p. 13, p. 15, p. 19 and p. 20, SRC_01758
at four named section headings — so the both-directions gate is already satisfied for both ids on the
page as it stands. What remains open is body-section CONSUMPTION of their seeds, not a citation gap.
They are queued to the sections that own their material — the literacy, script and divination
surfaces, the ritual and cosmology sections, and the questline and trade sections that consume the
pustaha craft and the oracle ladder. This section cannot see whether those sections took them, so the
assembly pass verifies that rather than assuming it. Books named as MUST-ACQUIRE carry no SRC id by
design; they are not in the corpus, and inventing an id for an unowned book would make the query layer
lie.
The manifests governing this page mark ninety-seven rows CONSUMABLE, twenty-six HELD and eleven CUT
across the six records. A held or cut row does not reach this page in any form, including paraphrase,
and where a hold removes something the page wants — the Malay huntsman material, the script-displacement
weave, the tondi and sahala arc, a second-source-pending entity — the gap is declared in the lists
above rather than filled from general knowledge. That is the firewall working, not a shortfall.
Twenty-two mint_flags ride these six records. Each is a RULING PENDING and this section writes around
every one of them: a reference section is the last place to resolve a vocabulary question by using it.
One clarification belongs here because the flags themselves are easy to misread. Several are phrased as
though no quest-type vocabulary existed; that is not the case. The quest_type enum is ratified ACTIVE
T1 canon, and what those flags actually record is that a particular seed_kind or minigame class has no
MAPPING into it yet. Per-seed typing is the questline section's stub to fill against the ratified enum,
never this section's, and no quest_type value is invented anywhere on this page.