A three-hour conversation on the 6th of March 2026 in a studio outside Tegallalang with one of the Bali architects who has spent 17 years building, refusing, and walking away from villa projects on rice paddy land. He has logged 24 completed villas across the Ubud highlands and the surrounding sub-village band. He has walked away from a further 19 projects that asked him to convert active subak terraces into lawn. The subak is the 1,000-year-old water-management system that UNESCO inscribed in 2012 across 19,500 hectares of Bali. Roughly 4,200 hectares of that listed area sit within an hour of an Ubud villa. The architect's manifesto is what he says he will and will not draw on that land. The manifesto reads like an editor's style sheet, not a sermon.
By The Villas For Kings desk
The rice paddy is not just a landscape. It is a water grid, a labour system, and an inherited contract between a farmer, a temple, and a downstream neighbour. The subak organises the irrigation of every active terrace in central Bali through a network of weirs, canals, and ceremonial water priests. The system was already 800 years old when the Dutch arrived. The UNESCO inscription in 2012 was the international recognition of a working agricultural infrastructure, not a museum piece. The terraces visible from a villa terrace at sunrise are working farms.
The architect we interviewed has asked us to mark him and his practice until our autumn 2026 follow-up. He works in the same Ubud register as Guy Morgan, whose Villa Sawah completed in 2003 set an early benchmark for villa design that referenced the agricultural building rather than the resort. He has trained in the practice of one of the Australian-Balinese studios working out of Sayan, and he has refused to take on Bamm Architects' rice-paddy-edge brief on the grounds that the brief itself was incompatible with the subak.
What follows is his manifesto, in the four refusals he has logged on his standard project brief, the three design moves he has worked out instead, and the three pressures that will, in his view, decide what Ubud villas look like by 2030.
"The first refusal is the lawn. The client arrives with a brief that includes a 600 to 1,200 square metre lawn for events, dogs, or children. The lawn requires the conversion of an active rice terrace into a graded grass surface. The conversion breaks the subak grid. The downstream farmer loses the water flow he is entitled to. The temple loses the offering cycle that ties the field to the village. The owner gets a lawn that 18 of his neighbours can no longer flood when their turn comes."
"I do not draw lawns over active terraces. I draw lawns on the rare plots that were never subak land, or on the lower terrace of a villa where the rice cycle has already been abandoned by the previous owner. In 17 years I have drawn perhaps four large lawns. Each one sits on legitimately retired ground. The other 19 clients who came to me for the lawn went to another architect. I have no problem with this. The other architect is the architect making the decision."
"The renter who arrives at a villa with a 1,000 square metre lawn surrounded by working terraces is the renter who has been sold a fiction. The fiction is that the lawn is the local landscape. The local landscape is the terrace. The lawn is the conversion."
"The second refusal is the cantilevered infinity pool that projects over the irrigation canal. The canal is the artery. The cantilever pool fixes a 35 to 60 tonne concrete mass above the canal, casts a shadow on the water, and concentrates seepage where the canal cannot accept it. The downstream pressure drops. The terrace below the villa, which the renter photographs at sunrise, dries by week three. The villa staff fill it cosmetically with a hose. The terrace is no longer an active terrace. It is a stage set."
"I have walked through eight villas where the downstream pressure has been damaged by the pool above. The owners have, in every case, been unaware. The fault was in the engineering, not the intention. The fault could have been avoided at the design stage. The architect who drew the cantilever did not run the hydrology study. The hydrology study is not optional on subak land. The hydrology study costs 4,400 to 7,200 dollars. The villa it protects is a 1.4 to 3.6 million dollar build. The economics are obvious."
"The third refusal is the multi-villa compound on a single subak parcel. The client buys two adjoining lots that the village had farmed as one continuous terrace. The brief asks me to design a four-pavilion compound: principal villa, two guest villas, a wellness pavilion. The brief is a hotel programme in private skin. The subak parcel cannot absorb that much hardscape. The terraced footprint required is roughly 2,200 square metres of cut and fill, and the irrigation grid will not survive the cut."
"I have drawn two-pavilion projects on parcels that historically supported two compounds. I have not drawn a four-pavilion compound on a single parcel. The client goes elsewhere. The compound gets built. The neighbouring village writes a letter to the regency administration. The administration sits on the letter. Two years later the irrigation downstream fails, and the cause is traceable to the compound. The neighbouring village then writes a second letter. The administration responds the second time. By the third year a partial demolition order has, in two cases I know of, been issued."
"The renter who books one of these compounds will sometimes notice the village resistance. The villa staff will be quieter than expected. The neighbours will be less friendly than expected. The villa is sitting on contested ground. The renter cannot fix that, and the platform marketing the villa will not mention it. Our work on Passed On in Bali 2026 names four properties in this category."
"The fourth refusal is the imported materials palette. A client arrives from Mykonos or Mustique and wants whitewashed render, polished concrete, large-format porcelain, smoked oak. None of these materials belong to the central Bali highland. The local register is volcanic stone, terracotta-fired bata bricks, hand-finished merbau and bengkirai, woven bamboo, alang-alang thatch on secondary structures. The imported palette will look right in photographs taken in the dry season. The same palette in the rainy season will mildew, stain, and read as misplaced. The renter who arrives in October sees the photograph. The renter who arrives in March sees the reality."
"The argument I make to the client is climate, not aesthetics. The Ubud average rainfall is 2,500 to 3,200 millimetres a year. The humidity does not drop below 75 percent for nine months of the year. The materials that survive that load are the local materials. The materials that fail are the imported ones. The imported palette is not wrong because it is foreign. It is wrong because it does not last."
The architect's refusal pattern is not aesthetic. It is operational. The Ubud register, like the Vietri tile and pietra di Furore register on the Amalfi Coast, is a list of materials that have already passed a 30-year endurance test on the specific climate the villa stands in. Our work on the architect redesigning the Amalfi Coast villa walks the same pattern in a different protected landscape.
"The Balinese vernacular is a courtyard of separate small pavilions, not a single large house. The principal sleeping pavilion is the bale daja. The principal living pavilion is the bale dauh. The kitchen is a third pavilion. The shrine is a fourth. The walls between the pavilions are the open courtyard. The vernacular is the design solution for the climate. I draw the modern villa as a contemporary version of the courtyard. The renter gets a sequence of pavilions. The cross-ventilation does the work of the air conditioning. The footprint stays small. The terrace stays around the pavilions, not under them."
"The imported joglo from Central Java is the second move. The joglo is a 19th-century Javanese pavilion structure. The teak frame is mortise-and-tenon, dismantleable, and survives the transit to Bali without metal fasteners. I have reassembled six joglos over the past 11 years as the principal living pavilion of a villa. The joglo is not Balinese, and I tell the client this. The joglo is a Javanese answer to the same climate question. The renter accepts the answer because the spatial experience is correct. The architect who pretends the joglo is Balinese is a different argument."
"The infinity edge I refuse to draw above the canal is the infinity edge that aligns with the subak gravity flow. The pool I draw instead is set on a 30 to 60 degree off-axis from the canal, so the pool overflow goes back into a recovery sump rather than into the irrigation grid. The pool reads correctly from the principal terrace. The pool does not read as flush with the rice line. Both are conscious decisions."
"The off-axis pool is a smaller pool. 11 to 14 metres long rather than 18 to 22. The off-axis pool is also a more expensive pool, because the recovery sump and the filtration loop both have to be over-engineered to handle the load that a canal-aligned pool sheds into the grid. The owner pays roughly 18 to 26 percent more for the off-axis specification. The owner gains a property that the village will defend over the long horizon. The arithmetic, again, is obvious to anybody who has watched a contested compound fail."
"The third move is the dedicated staff line that runs behind the principal pavilions and is invisible from the courtyard. The Balinese household has always had separate service and ceremonial flow. The villa with a single corridor for both is the villa where the staff and the renter cross paths every hour. The villa with a dedicated staff line allows the staff to move between the kitchen, the principal pavilions, and the storage spine without ever entering the renter's sightline. The renter notices that the villa runs without intrusion. The renter does not notice why."
"The cost of the staff line is roughly 6 to 9 percent of the build cost. The owner who has budgeted at the bottom of the range will skip it. The villa they get back is a villa that lives 30 percent worse than the photographs suggest. The villas in our work on the villa staffing shortage receipts are, almost without exception, villas that skipped the staff line at the design stage."
"Three. The tourist tax is the first. The 10 dollar per visitor tax introduced in February 2024 is a small signal of a larger regulatory direction. The provincial administration is preparing zoning amendments that will, in draft form, require a hydrology certificate for any new build on subak land. If the amendment passes in 2027, half the villa projects currently in pre-construction will fail the certificate. The 19 projects I have refused will become the 19 projects that, retrospectively, made the right call."
"Water is the second. The dry season aquifer pressure in the Ubud band has dropped each of the past four seasons. The cause is partly climate, partly extraction by the villa and resort sector. The 8-bedroom villa with a 100,000 litre underground tank, two staff showers, and an irrigation system runs through 4,800 to 6,400 litres of water a day in peak occupancy. The terrace below it runs on a third of that. The maths is not sustainable. The regulatory cap is coming. The owner who has not invested in rainwater capture is the owner who will be forced to retrofit."
"The third is the local resistance. The village banjar is the unit of local government in central Bali. The banjar can refuse a project at the social-licence stage even after the regency permit is issued. The banjar has done so on four villas I know of in the past three years. The banjar's decisions are not appealable in any court the foreign owner can reach. The smart owner spends 12 to 18 months in conversation with the banjar before the architect is hired. The owner who skips this step is the owner whose villa never opens."
"One specification. Ask whether the villa has a hydrology certificate, and ask whether the terraces visible from the master suite are still being farmed. If the answer to either question is no, the villa is part of the conversion problem rather than part of the working landscape. The renter who books anyway is making a choice with full information. The renter who books without asking is being sold a marketing line."
The manifesto is, in the architect's own framing, a list of refusals. The list is short. The list also explains why his book of completed projects is 24 rather than 60. He has refused more brief than he has accepted. The brief he has accepted, he says, is the brief the renter actually wants to sleep inside. Our work on the best villas in Bali ranks against the manifesto. Our work on the Bali destination guide walks the wider buyer question.
What is the subak? A 1,000-year-old Balinese irrigation and farmer-priest coordination system, inscribed by UNESCO in 2012 across 19,500 hectares of central Bali rice landscape.
Why is a lawn a problem? A lawn over an active terrace breaks the subak water grid and removes a downstream farmer's irrigation entitlement.
What materials does he specify? Volcanic stone, fired bata brick, merbau and bengkirai hardwoods, woven bamboo, alang-alang thatch on secondary structures.
What is the joglo move? Importing a 19th-century Javanese teak pavilion and reassembling it as the principal living pavilion. A climate-appropriate spatial answer, distinct from the Balinese bale.
What should a renter ask? Whether the villa has a hydrology certificate and whether the terraces visible from the master suite are still under active rice cultivation.
Our sister sites cover the hotels, restaurants, and bars that frame the same Ubud landscape the architect has worked across for 17 years.
One email a week. Architect interviews, planning intelligence, and the villas we pass on. Subscribe to the buyer’s brief.
Last updated 2026-05. We have not adjusted our editorial for the commission rate. See how-we-make-money for the full disclosure.