world/regions/bali/page/bali_S07.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.
REAL cultural architecture is reference-built, never raw-generated (§17.1); generation is
sanctioned for fantastical/vril structures only. Bali's built environment is this region's
strongest single claim on the rewards-intelligence pillar, and the claim is not decorative: the
island carries a documented, dated, still-working hydraulic system, a proportional building
system keyed to the human body, and a temple grammar whose every element has a stated reason.
The section's binding frame is the anti-fringe one the spine already fixes — the subak's order is
attested engineering and social science held as achievement, never lost-technology mysticism
[SRC: SRC_00157 — Eiseman, Bali: Sekala & Niskala, p. 2]. Everything below is buildable as real
mathematics, real statics, and real hydrology, and the player who studies it is meant to be paid
for studying it.
Two disciplines bind the whole section. The spine's Ch-4 lesson phrase for the subak is a negated
agent-noun construction, lexically the banned who-frame pattern even though nothing in this
material touches the antagonist; it stays design-layer and never reaches surfaced text, and the
affirmative form this page uses instead — an island that coordinates itself from the bottom up,
temple by temple and canal by canal — carries the same meaning and ships safely. And the source
that anchors most of this section is a 2011 World Heritage dossier that documents a living system
across a millennium, so its era strata are separated explicitly below: what the ~1200 CE present
frame carries is stated as present frame, and what the record dates later is carried as forward
horizon or authorial build material, never asserted as present-frame history.
The subak is a cooperative, multi-level water-temple irrigation system that governs the terraced
volcanic hills through shared allocation and ritual scheduling — collective-action engineering
owned in common by the cultivators who work it, and a living UNESCO-recognized achievement to be
rendered as attested engineering and governance [SRC: SRC_00157 — Eiseman, Bali: Sekala &
Niskala, p. 2]. Each subak is a specialized institution rather than a village, most of them
holding a written legal code, the awig-awig, that details the rights and responsibilities of
membership [SRC: SRC_00251, p. II-2 — Cultural Landscape of Bali Province, the UNESCO nomination
dossier (2011)]; by the eleventh century these irrigation jurisdictions already cross-cut village
boundaries [SRC: SRC_00251, p. II-55]. Irrigation water is held as the gift of the Goddess of the
Lake, Dewi Danu, with the subaks entrusted with the management of that gift and farmers
contributing a share of harvest to the rites of the subak temples [SRC: SRC_00251, p. II-2],
which is why the engineering and the ritual calendar are one system and not two.
The system's agronomic pay-off is geochemical and real: the Batur caldera eruption laid a
mineral-rich ignimbrite over southern Bali, which became the island's ricebowl [SRC: SRC_00251,
p. II-2], and monsoon rain leaches those nutrients out of the volcanic landscape and carries them
to the fields in the irrigation water itself [SRC: SRC_00251, p. II-4], so the canals are a
fertilizer-delivery network as much as a water supply. The volcanic-soil chain belongs to Section
1 and Section 9; what belongs here is the consequence — the hydraulic works are what convert a
volcano's output into a harvest, and a player who diverts water is moving fertility, not just
flow.
At every branching point the flow is split by stone proportional dividers, cut so that anyone
standing at the junction can tell at a glance how much water is running into each canal and so
verify that the division matches the agreed-upon schedule [SRC: SRC_00251, p. II-5]. This is the
system's most quietly brilliant piece of engineering and the most directly playable: the
allocation is not recorded in a ledger a headman controls, it is legible in the stone itself, so
verification is public, instantaneous, and available to every farmer on the branch. A schedule
enforced by geometry needs no enforcement officer.
Naming discipline: the dossier attests the dividers only by description, not by vernacular name.
The audit verified that neither of the two vernacular terms circulating in the harvest occurs
anywhere in the source, and supplying an unattested local word inside living-tradition substrate
is exactly the authenticity failure §17.1 guards against — it would read as authoritative
cultural detail while being uncited, and it would flow straight into the Section-17 localization
glossary. This page therefore carries the dividers descriptively until a source that genuinely
attests a term is acquired, and Section 16 takes the verify-list entry.
Where a ridge stands between a river and the terraces that need it, the Balinese answer is to go
through the ridge. Each weir diverts the flow of a short stretch of river into a small irrigation
tunnel, usually no taller than a man and about a metre wide; the tunnels angle sidewise and
emerge a kilometre or more downslope to flood terraces carved into the flanks of the volcanoes,
and where a supply exceeds one subak's need the canals and tunnels are extended to carry the
excess to more distant subaks downstream. By these means small quantities of water are delivered
with astonishing accuracy to terraces as much as several kilometres from their source, and over
the centuries small teams of skilled workers drew the water out in long spidery threads across
the slopes of the sacred mountains [SRC: SRC_00251, p. II-3]. That is the present frame's
hydraulic vocabulary in the dossier's own words, and it is enough to build on without borrowing
anything later.
The historical detail that makes this section canon-grade is that the tunnel engineers are a
named, paid profession in the documentary record. The undagi aungan — irrigation-tunnel
specialists — appear in the Bebetin inscription of AD 896 in a list of professional artisans who
worked for pay, which the dossier reads as evidence that demand for irrigation construction was
already sufficient to support them as independent specialists [SRC: SRC_00251, p. II-54]. Tunnel
experts are named again in the Batuan inscription of AD 1022 [SRC: SRC_00251, p. II-55]. Bali in
the present frame is therefore not a place where irrigation happens to work; it is a place with a
three-centuries-deep tradition of professional hydraulic engineers who are hired, paid, and
recorded by name of trade. That is the fact to build the Trade-8 substrate seed on.
Craft reputation is regional in the record as well, though the attestation is late: the dossier
carries a 1921 Dutch colonial engineering report noting that the best tunnel builders are found
in Klungkung and enjoy a very good reputation all over the island, so their help is always
invoked for that kind of work [SRC: SRC_00251, p. II-73]. That is documentary provenance rather
than present-frame history, and the page uses it only as authorial licence for the pattern — a
tunnel craft with named centres of excellence whose specialists travel — never as a ~1200 CE
place-name claim. The long modern bores belong to the same later stratum: the dossier's 160 m
tunnels are a post-1960 figure for the Batukaru region, built once government aid and cement
arrived, and they stay out of the present frame entirely [SRC: SRC_00251, p. II-72].
Playable surface (Trade 8 Engineering, seeded here ahead of its Ch-27 home base): apprenticing
to a tunnel cut is a genuine statics-and-hydraulics problem — survey the fall of the land, choose
the tap-point on the river, hold the gradient across the bore, and the flow either arrives at the
far terraces or it fails [SRC: SRC_00251, p. II-3]. Solutions are graded by the real hydraulics
rather than by force. This is directed freedom with a correct answer the world will show you,
which is what the rewards-intelligence pillar asks for.
The island also learns to bank water at scale, and the record is explicit about when. The dossier
places the creation of larger dams at downstream locations in the second half of the second
millennium CE, under royal patronage and after the older high-elevation sites of Pakerisan and
Batukaru were long established, and it notes that the later Balinese kings supported the subaks'
practical and ritual work rather than taking direct control of irrigation [SRC: SRC_00251, p.
II-72]. The dams themselves are documented later still, by colonial engineers: the 1921 report of
O. W. Sörensen describes large earthen storage dams built by groups of subaks, with the labour
and the cost borne by those who would later profit from the dam, dams of great height up to 35 m
in the district of Gianyar, an earthen blocking dam built well above flood height with an
overflow canal dug on one bank, and a small water temple placed atop the larger ones; the old
indigenous dam of Badung is described as a roughly 40 m earthen dam with two overflow canals and
an outlet delivering water for about 478 bouw of sawah, the report's own land unit [SRC:
SRC_00251, p. II-73; p. II-74]. The Pejeng dam within the Pakerisan area is photographed under
construction by local subaks in 1925 and is a working modern waterwork under public management
[SRC: SRC_00251, p. II-74; p. II-35].
Build handling: this is a real and un-softened indigenous achievement and it stays on the page at
full specificity, but it is a LATER stratum. A forty-metre earthen dam raised by village labour
is a silhouette worth building, and its home is the forward time-fluctuation window and the
world-state strata, not the ~1200 CE present frame, where the attested works are weirs, tunnels,
canals, aqueducts and terraces. Section 18 owns the window; Section 19 owns the strata.
The engineering claim is not inferred from surviving stone; it is dated by royal inscription. The
record the dossier assembles is the backbone of every date this section asserts.
| date | inscription | what it attests | anchor |
|---|---|---|---|
| AD 882 | Sukawana A I, the earliest dated Balinese royal inscription | irrigated rice fields (huma) already in the royal record | p. II-54 |
| AD 896 | Bebetin A I | the undagi aungan, irrigation-tunnel engineers, in a list of artisans who worked for pay | p. II-54 |
| AD 962 | the Manukaya stone, at Manukaya village | a Warmadewa king's construction or improvement of a bathing place at Tirta Empul, with a dam and two pools | p. II-62 |
| AD 1022 | Batuan | tunnel experts again, the first of many mentions of sawah, and a water-allocation official, the makaser | p. II-55 |
| AD 1071 | Pandak Bandung | the first appearance of the word subak, as the root of kasuwakan | p. II-55 |
| AD 1072 | the charter of Er Rara I | a royal order to remeasure one kasubakan's fields, listing irrigated areas in at least eighteen communities | p. II-55 |
| AD 1181 | the charter of Udanapatya | nineteen kasuwakans whose names differ from the villages in the same charter, so their jurisdictions cross-cut the villages | p. II-55 |
Alongside the inscription sequence the dossier records a time-reckoning system keyed to the
rice-growth cycle [SRC: SRC_00251, p. II-4] — the calendar and the hydraulics are the same
instrument read two ways, and the calendrical half is Section 5's. Against a ~1200 CE present
frame this table is the region's chronological floor: the works the player walks are already
between a generation and three centuries old, the vocabulary for them is already in writing, and
the Bebetin engineers of AD 896 and the 914 CE Blanjong pillar named in the page header stand
within a single generation of each other. The 1181 charter is the closest anchor of all, nineteen
years short of the present frame.
Design consequence worth binding: population-genetic work reported in the dossier reconstructs a
budding-deme spread in which irrigation began around the springs and rivers highest upstream, the
easiest sites to develop with traditional technology, and budded downstream as population grew,
placing the Pakerisan and Sebatu subaks probably among the oldest in Bali and originating before
the twelfth century [SRC: SRC_00251, p. II-68; p. II-70]. That finding is modern laboratory
science and is authorial-only: a ~1200 CE player cannot investigate a chromosomal diversity
gradient, and the finding never appears as era-present investigable lore. What it does govern is
the build — the upstream Pakerisan-side terraces are dressed as the ancient ones and the
downstream fans as the newer work, so the landscape itself carries a true history the player can
feel without being told. The same passage sets a second build constraint: most of the western
expansion into what became Bali's ricebowl, the Batukaru and Tabanan side, occurred after the
close of the Classical era in the fourteenth century [SRC: SRC_00251, p. II-68; p. II-70], so the
western terraces are dressed young and thin at ~1200 CE against the deep, dense Pakerisan
landscape. The lineage framing stays strictly settlement history; any drift toward
bloodline-mystical reading is out (the vril firewall applies to every lineage claim).
The coordination layer is built infrastructure, not an abstraction. It nests upward from the
field shrine (bedugul) at the corner where irrigation water first enters a farmer's fields, to
the subak temple, to the Ulun Swi at the head of the terraces, to the regional Masceti, to the
crater-lake temple at the top of the watershed [SRC: SRC_00251, p. II-6]. Every weir or spring
carries its own shrine to the Deity of the Weir, and each canal's block of terraces — usually a
kilometre or more downstream from the weir — its Ulun Swi, the Head of the Rice Terraces temple:
the tutelary presences are attached to the hydraulic components themselves, one shrine per piece
of working plant [SRC: SRC_00251, p. II-45]. Membership follows the water and nothing else: a
temple's congregation is exactly the farmers who share its source, so the larger the source the
larger the congregation, and because the crater lake is regarded as the ultimate origin of every
spring and river, the congregation of the supreme crater-lake temple, Pura Ulun Danu Batur, is
all of them [SRC: SRC_00251, p. II-5; p. II-19]. The dossier's documented modern scale for that
rule is a congregation of more than 250 subaks and a redistributive centre whose priests are
consulted on water rights and on the founding of new irrigation systems [SRC: SRC_00251, p.
II-5]; the present frame carries the rule and a proportionally smaller island, never the modern
count. Physically this means the player can read the scale of a watershed off the size of the
temple standing at its head, which is a navigational and inferential tool the world hands over
for free.
A symbolic axis rides the same plumbing and is load-bearing for the build: upstream water is
life-giving and is gathered in sacred vessels; downstream water is cleansing and is left running
to carry impurities to the dissolving sea [SRC: SRC_00251, p. II-7]. Water direction is therefore
meaning as well as gradient, and a canal's place in the sequence tells you what it is for.
Two routing statements the page owes plainly. The authority half of this hierarchy — who
arbitrates, whose lawbook governs a dispute, and what standing is won or lost by forcing a share
— is Section 8's material and is queued there, not settled here. And Pura Ulun Danu Batur sits on
the Batur crater rim, which the Zone Catalog does not currently carry; a crater-rim water-temple
zone is a pending Zone Catalog ruling, so the apex temple is held at region grain and is
deliberately NOT routed into besakih_mother_temple. Staging one community's temple, goddess and
documented sacred history at a neighbouring community's temple would be a §17.1 misattribution,
and an honest gap is better than a wrong zone.
Balinese building runs on an anthropometric proportional system, asta kosala kosali, a complex
series of rules that dictate the layout, size and proportions of every building in the house
compound and of traditional structures generally, on the principle that the home is a living
thing and must be harmonious with the body of the family patriarch [SRC: SRC_00157 — Eiseman,
Bali: Sekala & Niskala, p. 9]. The system is administered by the undagi, the traditional
architect, who before anything else takes measurements from the body of the head of the
household: the length of the index finger, the width of the little finger, the width of the three
middle fingers, and the width of the second joint of the index finger, from which three notched
bamboo sticks, the gagulak, are cut as the measuring rods for the pillars — the pillars governing
the size of the buildings. The compound's own layout is set from the length and width of the
owner's foot, and the outside wall dimensions from the span of his outstretched arms and the
distance from elbow to outstretched thumb [SRC: SRC_00157, p. 9]. This is the same professional
class whose tunnel-cutting branch the inscriptions pay by name, and it is the single most
distinctive architectural fact the region carries.
What it means in play is that no two compounds are the same size, and the difference is not
random — it is a measured record of the person who commissioned it. A player who learns the
system can pace out a doorway and read the builder's household off the stone, which turns
architecture into a legible archive and gives the Thread-13 sacred-geometry anchor a mechanic
rather than a motif. It also gives the procedural layer a correct generator: compound geometry is
parameterized on a household anthropometry seed and derived, so variation is authentic by
construction rather than by noise.
The Ch-5 join is already typed and is copied here, never invented: the architecture-proportion
progression LINK_0120 carries out of this chapter into the Borobudur crossing, with LINK_0119
arriving from the Manggarai builds behind it (Section 15 owns the link characterizations). Bali's
seat in that chain is the body-derived module — the proportion system that measures a building
against the person who lives in it.
The Balinese architectural vocabulary is small, consistent, and reason-bearing, which makes it
excellent build material: each element is a rule the player can learn once and then read
everywhere. The temple itself is the frame — a three-part enclosure running downslope from the
jeroan, the walled inner sanctum holding the shrines, through the transitional jaba tengah, to
the outermost and least sacred jaba where secular life is permitted [SRC: SRC_00157, p. 8].
between the middle and outer courtyards and at the threshold of the compound [SRC: SRC_00157,
p. 8]. It is the chapter's Gate-Naga threshold site and the Puzzle variant of the agility course
(copied from the spine's typed layer).
leering Bhoma face above its opening, whose fangs and bulging eyes keep evil forces out of the
holiest part of the temple [SRC: SRC_00157, p. 8; p. 254].
tradition itself: bhuta-kala are said to be unable to go around corners, so the screen breaks
the line into the courtyard, in house compounds and again immediately inside Uluwatu's kori
agung, where celebrants must turn left or right to pass it [SRC: SRC_00157, p. 213; p. 254].
This is the clearest case on the page of architecture that is a working defensive system in its
own cosmology — the geometry is the ward. Its tactical use in the leyak defense and the
shadow-duel is queued for Sections 5 and 13; what Section 7 owns is the form and the reason for
it.
[SRC: SRC_00157, p. 254].
set-dressing). Its cosmology is the received Meru mountain: the supreme god who resides on
Mount Meru broke apart the summit and sent the pieces to Bali, his son becoming the first male
god of Bali and taking up residence atop the larger fragment, which became Agung, while on the
floor of the crater lake in the smaller fragment, Batur, his daughter built a palace and took
the name Dewi Danu, Goddess of the Lake [SRC: SRC_00251, p. II-19]. The tower is a mountain
built to scale, so the build reads tier count as rank, and the terraced hillside outside the
temple wall is the same figure in earth.
of the above (Section 9 owns the material sourcing).
At the Uluwatu sea-cliff temple the grammar reaches its most elaborate form: rare winged
(bersayap) split-gates, their tops curving out and back to a point like a bird's wingtip, step up
the approach toward the kori agung, a massive white-coral structure carrying the Bhoma face and
distinguished by a gapura arch, an architectural device almost never used in Bali; larger Ganesha
dwarapala flank it, and stone boats at the Dalem Jurit shrine mark both the boat-shaped peninsula
and its defence against what arrives from the sea [SRC: SRC_00157, p. 252; p. 254; p. 255].
Uluwatu is one of the sad kahyangan, among the six most important temples on the island and a
living sacred site of the highest rank [SRC: SRC_00157, p. 250]. The build terms are binding: the
agility course runs the cliff traverse and the approach thresholds, never over consecrated
shrines, and the winged gates are read rather than clambered.
Downstream from Tirta Empul the Pakerisan valley deepens into a gorge whose sides are shaped into
narrow rice terraces, and at its bottom royal candi and priests' meditation niches are carved
directly into the living basalt of the riverbank. Because they were cut into living rock instead
of built from stacked blocks, parts of the complex have survived the centuries in exceptional
condition relative to the best-known free-standing monuments of Java, and ancient rock-cut water
channels still run to spouts positioned in front of each candi [SRC: SRC_00251, p. II-32]. This
is subtractive monumental architecture — the monument is the cliff with stone removed — and it
sits in the same watershed as the irrigation works, sharing their hydraulic idiom: even a royal
funerary complex is plumbed. The dossier also makes a design-derivation claim in its own words,
that the candis' design shows a strong link to the Classical states of Java [SRC: SRC_00251, p.
II-32]; that is substrate here and reciprocates toward the Ch-5 Java crossing, with the link
class and any graph row belonging to Section 15 and the link-graph lane, not minted on this page.
The gorge carries a readable inscription in Kadiri Quadrate characters naming the king who passed
away in the Jalu and his two children [SRC: SRC_00251, p. II-33; p. II-65]. The dossier
identifies that king as Udayana, who reigned in Bali from 989 to 1021, and the two sons as
Marakata and Anak Wungsu, and dates the complex to the eleventh century [SRC: SRC_00251, p.
II-65] — which makes the gorge a memorial of the page's own present-frame dynasty, roughly two
centuries old when the player descends it. Jalu means keris, which is why the river below is the
Pakerisan, the River of the Dagger — a place-name that is a lineage-death memorial, and a
literacy stop the chapter's growing reading can be spent on. The naming discipline is explicit:
this is an etymology of a river name and a royal memorial, and no named legendary blade is
derived from it here. Section 6 already reserves the region's pusaka keris line for the build
pass, and a name for it is not this page's to mint.
Zone standing: the Pakerisan gorge is not in the Zone Catalog, and a gorge exploration zone is a
pending Zone Catalog ruling. The material is therefore carried at region grain and takes no zone
assignment in this pass. Care terms: Gunung Kawi is a protected living sacred site, built as a
type-accurate referenced composition and never as a high-poly replica of the actual monuments,
and the rites performed today in the temple standing before the ancient niches are live practice,
not ruins [SRC: SRC_00251, p. II-33].
Sudana CHAR_0025): the whole hydraulic argument made legible in one descent — dividers, weir
shrines, terrace fan, and the temple at the head of the source.
glance, route a boundary-crossing tunnel to a block of terraces that cannot otherwise be
watered, and hold the gradient [SRC: SRC_00251, p. II-5; p. II-3]. Taking more than the agreed
share starves a subak downstream and is felt as a community read (FORK_CH04_A), never as a
progression gate.
of terraces starves rats, insects and rice disease of habitat at watershed scale, because a
pest that cannot migrate to a standing crop has nowhere to go [SRC: SRC_00251, p. II-4; p.
II-7]. It is regional pest control achieved by timing alone, and it is a scheduling puzzle whose
correct solution is the historical one.
themselves, coupled to ecological feedback, produces near-optimal island-wide water sharing and
pest control — the coordination is achieved by the subaks temple by temple and canal by canal,
and it is a documented case in the complexity literature the dossier cites [SRC: SRC_00251, p.
II-4; p. II-5; p. II-5 footnote (Lansing and Miller 2005); p. II-19 footnote (Lansing 2006)].
This is the intellectual pay-off of the chapter and it grounds to SCIENCE_GROUNDING.md, never
to a lost-technology frame.
divider geometry, all seeded here ahead of the Ch-27 home base.
forms, and the terrace geometry, read as one proportional discipline rather than three motifs.
No quest_type is asserted for any of the above. The controlled vocabulary is a pending ruling and
this page writes the surfaces without force-fitting one.
sacred sites: referenced, type-accurate compositions, never high-poly reproductions of the
actual temples (§17.1).
and the subak fan is load-bearing rather than decorative — the water has to read as running
downhill through the dividers the player is asked to set.
seed derives the compound's proportions, so procedural variation is culturally correct by
construction. This is the one place on this page where generation is sanctioned on real cultural
architecture, and only because the tradition's own rule is the generator.
portal and bore, canal and spout, field bedugul, subak temple, Ulun Swi, and Masceti. The large
earthen storage dam with its overflow canal and summit shrine is authored to the same kit but
flagged later-stratum, for the forward window and the world-state passes rather than the ~1200
CE dress.
canals, aqueducts and terraces in clay, stone, timber and bamboo. Large storage dams, cement,
and long modern bores belong to the later strata and stay out of the ~1200 CE build.
architecture-vertical ascent), with a candi-bentar split-gate Puzzle variant, per the spine's
typed course row.
are left running. A player who reverses that is doing something legible and wrong, which is the
cheapest possible teaching moment.
entry).
water-temple zone and a Pakerisan gorge zone. Both bodies of material are carried at region
grain until those rulings land.
are carried with their true provenance (a later stratum documented in 1921 and after) and are
queued to Sections 18 and 19 for the forward-window and world-state passes; they are not
present-frame history.
the AD 882–1181 inscription sequence as the section's chronological spine, so the Section-16
reference list and the Section-20 population plan take the same-pass entry.