world/regions/sumatra_java/page/sumatra_java_S15.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 (scanned from this document's own citations — widen it by hand where it is thin):
_source/02_Tier_2_Region_Pages/angkor_khmer.md·_source/02_Tier_2_Region_Pages/bali.md
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.
Connections, each citing its Cross_Cultural_Link_Graph row where one exists AND carrying that row's
own link_type — the characterization is per-link, never a blanket. Only the parallel-independent
links are a shared category, never a diffusion claim; the registry types others as documented
transmission, cognate-descent, or catalog-convergence, and the page says so per row. The substrate
weaves the deep read added are written only where both ends are named and cited in the record that
carries them; those whose target region has no page here are written as originating at Sumatra-Java
with the reciprocal end still owed, so a claim is never made without naming the machine tie it does
not yet have. This region's harvest was large and its cross-regional yield was NOT: the honest
count, the shapes that failed, and the places where the weave is genuinely thin are reported at the
end of this section rather than padded out of sight.
from below the Uluwatu sea-cliff with Captain I Wayan Dwija, carrying Shell (W1) and Kira (F1),
and five links arrive with them. The conjured-killing-spirit predator parallel is received first —
leyak → begu ganjang, LINK_0147, parallel-independent, a shared cosmological category of a
vital-force predator and never a diffusion claim, the tall begu ganjang a different sub-motif from
the flying-viscera leyak. Four weaving-progression edges land at the Borobudur crossing with it:
the Solar-Hero seed LINK_0082, the architecture-proportion LINK_0120, the guardian-at-the-gate
LINK_0128, and the unity-weaving LINK_0138
Hero toward South India, Ch 7), LINK_0121 (architecture-proportion toward the Egyptian peak,
Ch 15), LINK_0129 (guardian-at-the-gate toward South India, Ch 7), and LINK_0139 (unity toward
Tewahedo syncretism, Ch 9) — all typed weaving-progression, all opened by beats this page stages
Port Master Nakhoda Iskandar (CHAR_0037) across the Bay of Bengal, and three edges resolve at the
next node — the Batak tendi → Vedda tendi vital-force parallel LINK_0001 and the begu ganjang →
Sinhala disease-demon predator LINK_0148, both parallel-independent, plus the excursion's
hydraulic-achievement LINK_0187. Ch 6 is the read node for all three
Naga Padoha's fetters-and-earthquakes cosmology set beside Azi Dahaka bound under Damavand, an
Austronesian pole distinct from the Iranian one and never a diffusion claim
world-axis LINK_0185 is knowledge-transmission and the page writes it at that strength — the Indic
Meru cosmogram reached the Khmer court along the documented trade web, and each realization is
locally authored — Angkor the largest Khmer realization of the shared axis mundi, an author of the
form and not a passive recipient (Angkor → South India, Ch 7). The flying-viscera-witch
LINK_0186 (Bali leyak → Khmer krasue/ap) and the hydraulic-achievement LINK_0187 (Angkor baray →
the Sri Lanka tank-cascade, Sri Lanka predating) are both parallel-independent. The
knowledge-transmission LINK_0185 and the parallel-independent LINK_0187 originate on the excursion
page and LINK_0186 is received there from Bali; they are listed here as the excursion tie only
Six of the fourteen surviving substrate weaves are about script and the book, and they are the
strongest cross-regional material the page has. Treat literacy as a playable system: every row below
is something a player can read off a letter-shape, a diacritic name, or a fold of bark.
the Indian family of writing and its most influential ancestor is the Pallava script, which the
source states flatly and without hedge, adding that every indigenous Indonesian script descends
from it [SRC: SRC_01758, line 69 — Kozok, Sejarah aksara Batak]. Typed knowledge-transmission,
frame documented-transmission, Thread 17, lens A4, and the anti-fringe guard is the point of the
row rather than a rider on it: the script is credited to its documented Brahmi ancestry and to its
own Batak builders, never to a lost race or a mystical proto-alphabet. Two disciplines bind. The
proto-Sumatran intermediary of the Srivijaya era is a LEADING HYPOTHESIS in the source's own words
and is written as one, never as settled fact. And the far end is Pallava — a script — not Tamil
and not Grantha; attaching this descent to the Tamil people would credit a living people with a
script they did not write, and this page does not do it. That correctness matters at the seam,
because the Ch-7 South India node is where Chelvan Kanakkan's Tamil-Chola thread and this script
row will meet and must not be confused for each other
owed: Surat Batak's nearest living kin are the neighbouring highland scripts — surat ulu and the
Kerinci surat incung, with Rejang-Lebong, Lembak, Lintang, Pasemah, Serawai and Lampung in the
same Ka-Ga-Nga group — sharing the Ka, Ga and Ha letterforms almost exactly and much of Da besides
[SRC: SRC_01758, lines 107-111]. Typed linguistic-cognate, frame documented-transmission on shared
documented descent, Thread 17, lens A4, and classified CONTESTED rather than monograph on purpose:
the single-proto-script origin behind the shared forms is hedged by the source at every mention and
closes on the concession that the details are not known and may never be. Ship the shared
letterforms as fact and the common parent as inference. One instruction for the hand that cuts the
South Sumatra page: this group is written here as ONE target because the source groups it, but it
must be split into per-people rows there, since Kerinci's script carries its own name,
Rejang-Lebong's carries a second, and Lampung's differs from surat ulu in the source's own account
Batak and the Hanacaraka scripts of Java, Sunda and Bali share a common Kawi and Pallava ancestor
rather than one descending from the other — the Batak Na-kuno still closely resembles the Kawi Na
[SRC: SRC_01758, line 299]. Typed documented-transmission with divergence, Thread 17, lens A4,
and the framing is deliberately sibling-branches and never Java-teaches-Sumatra. But the row
carries the source's own further finding rather than suppressing it for a cleaner independence
story: the same passage concludes there was Kawi influence on the early development of the Batak
script, and the evidence is a single letter — Batak Ja is Da plus a horizontal stroke, which holds
in Kawi and does NOT hold in Pallava [SRC: SRC_01758, line 301]. That is the single best piece of
real epigraphic reasoning in the whole harvest, and it is the substrate under the Borobudur
two-inscriptions scene (SC_CH05_two_inscriptions, CH05_B01): the child who already carries one
people's writing holds it against a second people's and finds cousins, and a player who looks
harder finds a letter that proves which cousin
asserted: the Batak word for their own script, surat, is the same word used in Kerinci, Lampung
and Rejang, and reaches the Philippine scripts as surat or sulat in Tagbanwa, Tagalog and Mangyan
[SRC: SRC_01758, line 3]. Typed linguistic-cognate, Thread 17, lens A4, frame catalog-convergence
and NOT documented-transmission: the source catalogues the same term across a named set without
claiming a mechanism, and no Austronesian etymology for surat is asserted here, because the source
gives a Sanskrit derivation only for aksara and none at all for surat. The catalogue is what the
evidence carries, and the catalogue alone is a strong rabbit-hole
the vowel marks are cognate outward — the Mandailing talinga against Javanese taling, Lampung
keteliling and Rejang katiling [SRC: SRC_01758, lines 321-323], and the Mandailing amisara against
the Indian anusvara, which the source calls a very close sound-match [SRC: SRC_01758, line 383].
Typed linguistic-cognate, documented-transmission, Thread 17, lens A4. This is the sharpest of the
script rows for play because it is the source's OWN instrument for tracing which way the script
travelled: the diacritic names are the evidence, so a player who collects them is doing the
philology rather than being told its result
is not the Batak's alone — the same manuscript form is used by the South Sumatran Lampung, and the
oldest known bark folding-book is a South Sumatran manuscript on bintangor bark [SRC: SRC_00226,
print p. 66 — The Power of Writing: Batak Manuscripts]. Typed parallel-independent in the harvest
and framed as catalog-convergence at audit, Thread 17, lens B2. It widens the region's writing web
from a script into a material culture without asserting one-way borrowing, and it is the craft
half of the literacy system: the pustaha the player learns to read is a form with neighbours
target end owed: the Toba Batak twelve-constellation oracle pormesa na sampulu dua is named from
its first constellation, and mesa is a Sanskrit loanword — Aries — with the other eleven following
in the same borrowed order, morsoba from vrsabha, metuna from mithuna. The panggorda oracle takes
its name from garuda and the pormamis deity-names from Mahesvara, Visnu, Sri, Kala and Brahma
[SRC: SRC_00226, print p. 70-71]. Typed reception-transmission, frame documented-transmission,
Thread 12 The Oracle / Divination, lens C4, both ends named and cited in one passage. This is the
one place in the harvest where a transmission question is both genuinely open and genuinely
attested: Indic astronomical vocabulary was received and re-authored into a Batak divination
system that is nobody else's. The local authorship is the anti-diffusion guard and it is
documented rather than asserted — the porhalanan, the catalogue's principal spelling with
porhalaan its attested variant, never had a computational function and its only use was divination
[SRC: SRC_00226, print p. 77-79], which is not what the borrowed vocabulary was for at its far
end. No Cross_Cultural_Link_Graph row exists yet; minting one is a registry action and the Indic
end has no page here to receive it
pormamis na lima, the five-part-of-day oracle read off a lettered grid, is very similar to the
most common divinatory technique in Malay manuscripts, where it is called ketika lima [SRC:
SRC_00226, print p. 71]. And the ari rojang, the oracle of thirty named days most of them named
for animals, was read by Voorhoeve as closest to the Malay rejang chiefly on the similarity of the
animal names, with a further Southeast-Asian variant blending Hindu naksatra, the Chinese animal
zodiac and Persian elements [SRC: SRC_00226, print p. 74-75]. Both typed parallel-independent in
the harvest and both framed at audit as catalog-convergence, Thread 12, lens C4 — and the framing
is the honest one in BOTH directions: the source asserts similarity, which forbids calling this a
transmission and equally forbids calling it independent invention. Voorhoeve only suspected a more
direct relation. Palembang is where these two manuscript worlds physically meet on this page, and
a player who reads a Batak grid and a Malay grid side by side there is looking at a real open
question rather than a settled one
singa, tenoned to the house beams and composed of water buffalo, horse and naga, and the Bornean
aso, the serpentine dog motif used to ward off evil, are presented by the source as two regional
realizations of ONE shared naga category [SRC: SRC_00446, print p. 24; print p. 39 — Art of Island
Southeast Asia: The Fred and Rita Richman Collection]. Typed parallel-independent in the harvest
and CORRECTED at audit to catalog-convergence, Thread 2, lens C2 — asserting independent invention
where the source asserts a shared category would be a Frame-A error in the permissive direction,
so the page does not make it. The aso is claimed here only as a Bornean motif; nothing sites it on
longhouses or on hampatong, because no cited page does
Sanskrit for lion and naga from the Sanskrit for snake, and what the Batak built with the lion-word
is a creature of water buffalo, horse and serpent [SRC: SRC_00446, print p. 24; print p. 39].
Typed reception-transmission, frame documented-transmission, Thread 2, lens C2. The context is
local elites importing foreign religion into their own courts, which is how the source puts it,
noting that documented evidence of civilizing migration or conquest is lacking [SRC: SRC_00446,
print p. 17] — a reception, not a civilizing wave arriving. This is the page's clearest
anti-hyperdiffusion exhibit and it does double duty as a guard on a named at-risk claim: the
Sanskrit-derived singa and the Indic naga met at the Borobudur gates are a DIFFERENT relation from
Naga Padoha, who is the Batak's own bound Austronesian underworld serpent on an independent pole
(LINK_0156). The arc's serpents are not one undifferentiated parallel and this section does not
collapse them
gale-gale funerary puppet danced as a likeness of the dead [SRC: SRC_00446, print p. 60], the Nias
adu zatua [SRC: SRC_00446, print p. 78] and the Sa'dan Toraja tau-tau of South Sulawesi [SRC:
SRC_00446, print p. 87] belong to one Island-Southeast-Asian family of ancestor images. Typed
parallel-independent in the harvest, framed at audit as catalog-convergence, Thread 20, lens C1 —
and the row carries the source's own CONTRAST rather than smoothing it, because the contrast is
the interesting part: unlike most other ancestor figures in the region the tau-tau functioned more
as a portrait of the deceased than as a receptacle for the soul. So this is a family of related
solutions to one problem, not one uniform practice. A live scholarly tension rides with it and is
not resolved here: the source attributes the portrait reading to contemporary Toraja Christians
and notes it is probably shaped by Christian doctrine, while the Toraja term for the soul that is
seen points the other way
Toba Batak call basapi and the Karo Batak kulcapi is named in the same family as the kutyapi in
Java and the kudyapi in the southern Philippines, which the source offers explicitly as linguistic
and formal affinities indicating close cultural relations between groups [SRC: SRC_00446, print p.
64]. Typed linguistic-cognate, catalog-convergence, Thread 21, lens E1. Two limits ship with it:
the source names PLACES, so the far end is Java and the southern Philippines and not any named
people, and the widely-used Batak term hasapi is not attested in this source and is not cited to it
Malays of Perak place the origin of Hantu Pemburu, the Spectre Huntsman, at Katapang in Sumatra,
and the article argues the dreaded Malay demon derives from the Batak divine huntsman Haji Guru,
helper of the high creator Debata i datas langit [SRC: SRC_00228, print p. 18; print pp. 16-17 —
Brakel-Papenhuyzen, Van der Tuuk's collection of Batak manuscripts]. Typed reception-transmission,
frame documented-transmission, lens C2, peer-reviewed basis — the only peer-reviewed footing on
this page. Frame A is safe here in the strongest available way — the claim of Sumatran origin is
the Malays' own, not a scholar's imposition — and the design value is the inversion: a good divine
servant on the Batak side is a feared demon on the Malay side downstream, one figure crossing a
cultural border and flipping moral valence. Four disciplines bind and none of them softens the
encounter. Hantu Pemburu is named as a PERAK MALAY being wherever it appears and never presented
as Batak folklore. It is held strictly distinct from the honored Haji Guru. The crow and the seven
dogs are the BATAK creation myth's, sent to accompany Haji Guru, and they are not transferred onto
the Malay figure — the derivation is the author's twice-hedged argument, not an established
identity, and merging the two figures' attributes would convert one into the other. And the being
is sited on its own tradition's ground, the Sumatran coast and the strait, never on the Javanese
volcano slopes. Within those lines it is a dreaded demon in its own tradition and is staged as its
own tradition fears it
This is a cross-regional connection of a different kind and it is not a Cross_Cultural_Link_Graph
row: it is not a weave between two cultures' motifs but a physical route by which one culture's
sacral books left the region. The page owes it because it is true, because it is the honest answer
to why the player can read these manuscripts at all, and because a pustaha-recovery beat whose
ground truth is that most pustaha are elsewhere is live material rather than an awkwardness.
national and private collections, and roughly two thousand now sit in collections worldwide [SRC:
SRC_00158, print p. 164-165 — Between Mind and Soul: The Pustaha Manuscript]. That record is
shelved to this page's region id, but a pustaha is a BATAK divination manuscript and every claim
in this block is attributed to the Batak, never to Java
the datu's role in society by taking objects out of circulation, with collectors and missionaries
destroying those objects or acquiring them for ethnographic museums and for the tourist market
[SRC: SRC_00226, print p. 17]. One complete chain is documented end to end — Johannes Winkler, a
mission doctor, studied the language and collected the artefacts [SRC: SRC_00226, print p. 11],
and acquired one artefact around 1910 and sold it to the ethnological museum in Hamburg in 1912
[SRC: SRC_00226, print p. 81]. Suppression, extraction, sale, display, with dates
der Tuuk to translate the Bible and to describe the Batak language for the instruction of
Protestant missionaries, a project that moved one hundred and fifty-two pustaha and twenty-five
bound volumes to Leiden — and that same collection saved an important part of the Dairi cultural
heritage from oblivion [SRC: SRC_00228, print p. 10; print p. 19]. Both are true at once and the
page states both; van der Tuuk is also named in the pustaha literature as the collector-linguist
whose collecting and analysis produced the sub-Toba classification that literature still cites
[SRC: SRC_00158, print p. 169]
and whom he then employed to write and copy them; the Kelasen storytellers whose repertoire was
written down for the first time; and Ompu ni Manungkol Langit, head of Aek na Uli, who would not
part with his pustaha until he was paid one European dog as large as a calf, a dozen flasks of
jenever, ten Spanish mats and three padang rusak [SRC: SRC_00228, print p. 11; print p. 19]. A
Batak owner driving a hard bargain with a Dutch collector is the beat, and it is the opposite of
a community to whom things merely happened. The internal restriction is documented too and is not
a colonial invention: pustaha knowledge was held to the sons of wealthy people, clan leaders and
close relatives of a datu
was formed before Christianity, Islam, modernization and globalisation together changed the
world-view and the way of life of the Batak of North Sumatra, and the manuscripts sit unread today
partly because the number of scholars of Batak languages is dwindling [SRC: SRC_00228, print p.
19; print p. 10]. The antagonist, where there is one, is the colonial and mission apparatus —
never the Batak community, and never a single corrupt individual invented to carry an institution's
weight
museums, no Bible Society and no accession numbers, and a player at the chapter's present cannot
investigate any of it. It is forward time-fluctuation and modern-revisit stratum only, exactly as
the page treats Candi Sukuh and Pagaruyung, and it is never dressed as L2 depth available to the
era-present player
Cross_Cultural_Link_Graph row: low Ice-Age sea levels repeatedly joined Bali, Java and Sumatra
across exposed land until the Last Glacial Maximum, which is why the three islands' tigers are one
recent lineage rather than three separate arrivals. This page RECEIVES it as the same non-link
fact, mints nothing, and types nothing — common descent in an animal is not a parallel between
cultures. The anchor is the bali page's own, at its own Section 15, and no page anchor is claimed
here that this page's records do not carry. Two consequences bind: Khan the Sumatran Tiger (G1,
bonded at samosir_highlands) rests on a genuinely shared animal rather than a resemblance, and
nothing on this page is Wallacean — Sumatra and Java sit on the Sunda Shelf, west of the Wallace
Line, as Bali does
Reciprocation discipline binds: each claimed link is reciprocated at its target as the target pages
are cut, and the LINK_ rows are the machine tie, so a page-side claim without a reciprocal target
row is a defect. That is exactly why every substrate weave above is written as owed rather than as
done — minting a Cross_Cultural_Link_Graph row is a registry action, and Borneo, Nias, Sulawesi, the
South Sumatran highlands, the Philippines, the Perak Malay coast and the South India script end have
no region page here to receive one. The spine's own eleven-edge origination-and-resolution set stands
unchanged above it: LINK_0001, LINK_0083, LINK_0121, LINK_0129, LINK_0139, LINK_0147 received,
LINK_0148, LINK_0156, and the excursion's LINK_0185, LINK_0186 and LINK_0187.
The count is the point of this closing block and it is not flattering. The deep read proposed thirty
cross-regional weave candidates across the six audited records. Fourteen advanced and all fourteen
are written above. Nine were cut at audit and seven are held pending fixes, and not one of those
sixteen rows is consumed here in any form, including paraphrase. Two of the six records — the Aek
Tarum image field-note and the pustaha-manuscripts article — contributed NO surviving cross-regional
weave at all. The pustaha article still reaches this section, but through the collection-and-removal
history above rather than through any weave row; the Aek Tarum field-note contributes nothing here
at all, and its terrain, dye-craft and trade substrate is queued to the sections that own those
surfaces, with no claim made here about whether they took it.
The failures did not reduce to one shape, and recording them at the right level of abstraction is
what stops a later pass re-proposing them. Six distinct defects account for the sixteen: an end A
that was a bare stylistic category rather than a named culture, which cannot be checked at all
because there is nothing at that end to check, compounded by a documented-transmission frame laid
over a thrice-hedged 1975 visual judgement; a far end naming a real culture that the cited pages
never actually describe; a script attributed at its far end to the wrong people, which would credit
a living people with a writing system they did not write — held for a one-field correction rather
than cut, and the correct attribution is carried above from a different record; an uncited culture
riding inside a multi-culture anchor; a claim of independent invention where the source asserts a
shared category, which is a Frame-A error in the permissive direction; and a hedged scholarly
hypothesis written as settled fact inside a row bound for the science-grounding surface. The remedy
for most of them is research rather than authoring: give the far end a name and its own page anchor
from its own region's shelf, then re-audit.
Where the weave is thin, plainly. It is thin over JAVA. All fourteen surviving weaves carry a Batak
source end; the harvest is Batak and Sumatran through and through. Java stands at the far end of
exactly three of them — once as a script relative, once inside the diacritic-name cognate set, and
once as a place-name in a lute-name catalogue — and in none of them as its monuments. Borobudur and
Prambanan carry no page-side substrate weave in this harvest at all, and their four
weaving-progression edges are spine-copied and stand on the spine's authority, not on any anchor
this section can show. That is the same gap Section 16 records at the corpus level, and it is stated
here rather than covered by inventing a monument link the records do not carry.
It is thinner in DEPTH of relation than the transmission count alone suggests. Seven of the fourteen
are catalog-convergence rows — honest but weak claims, asserting a shared category rather than a
route: the two Malay day-count oracles, the leporello bark book, the surat word-catalogue, the
singa-and-aso naga category, the ancestor-image family, and the lute-name family. The other seven
carry frame documented-transmission with both ends named and cited: the Pallava script ancestry, the
highland Ka-Ga-Nga cognate group, the Kawi influence on the early Batak letterforms, the
diacritic-name cognates, the Sanskrit zodiac vocabulary inside the pormesa oracle, the Sanskrit
singa and naga received into Batak house-guardian craft, and the Spectre Huntsman's Sumatran origin.
Two qualifications ride that second set rather than being smoothed away: the highland-script row is
classified CONTESTED at audit rather than monograph, because its common parent is inference and only
its shared letterforms are fact; and the Spectre Huntsman row is the only peer-reviewed row on the
page, which is why it is the one whose far-end discipline is written out at greatest length above.
Those seven are the rows worth building a beat on.
Anchor convention, recorded so a later verifier does not read a correct citation as drift. Anchors
in this section are PRINT pages throughout. The Power of Writing record mixes conventions across its
own anchor set — the audit re-derived print as PDF minus two and corrected a cluster of anchors that
were stored as PDF indices — so no record-wide rule is asserted here; what is asserted is that every
anchor this section carries from it was checked individually and resolves as PRINT as given. The
Richman-collection record stores PDF indices and every anchor above has been converted to print at
PDF minus one, verified across twenty-six anchors by the printed folio surviving in the extracted
text. The Van der Tuuk article paginates 9 to 21 and its anchors are print as given. The Kozok
script history is a line-numbered text conversion rather than a paginated book, so its anchors are
LINE numbers, and they are taken from the revised second copy of the doubled block, never the
superseded first copy.
Serpent-thread continuity: Naga Padoha the bound Austronesian chaos-serpent carries Thread 2 forward
as an independent pole, distinct both from the Sanskrit-derived singa and naga received into Batak
house-guardian craft above and from the protective Indic naga met on the excursion. The arc's larger
corruption pattern is held dark at page grain — Section 11 carries that metadata, nothing in this
section names it, and the first enemy who reads back is legible to the player only as an
unexplained, unnamed unease.