A teaching dies the way a language dies. Not in one night — the last temple that practiced it burns, or the master leaves no heir, or the dynasty that funded it falls, and the chain of bodies that carried the method snaps. After that the method exists only on paper, if it exists at all, and paper is a poor teacher of a body-art. Dukes points at one place where, he believes, several Northern Chinese teachings that snapped on the mainland did not snap — they kept going, frozen and intact, just over the border. That place is Korea. The image worth holding is a refrigerator: things that spoiled everywhere else stayed fresh here because the door happened to be open at the right moment and shut again before the rot got in.1
This page records a single buried claim from Dukes's endnotes and the argument he hangs on it. The claim is geographic and archaeological. Recent archaeological discoveries — he dates them 1989 to 1990 — unearthed in Korea a previously unknown temple of an "Eastern Mountain sect," and Dukes proposes that this temple could be one connected to Chuan Fa practice, the Buddhist martial-meditative tradition his whole book is about.2 Around that one excavation he wraps a larger thesis: Korea's close physical proximity to Northern China meant that teachings which died out in Hopei and other northern areas — he names Sung Shan, the Sung Mountains where Shaolin sits — were preserved unchanged inside Korea.3 This preservation, on his account, is one of the reasons Chaan Buddhism, known in Korea as Son, grew so dominant there. And he closes with a prediction stated as near-certainty: as the years go by, "we shall find traces of many Chuan Fa practices and doctrines preserved in its temple records and monuments, which have been ignored by most researchers in the field."4
Strip it to the bone and it is a preservation thesis. The center of an art can decay while a margin of it survives, and the margin then holds the only living copy. Korea is cast as that margin for the Northern Chinese Buddhist material.
Think about how a thing survives by being left alone. A teaching at the center of a culture is the teaching most exposed to that culture's upheavals — its wars, its purges, its fashions in doctrine. Northern China was the center; it got the Hun invasions, the persecutions of Buddhism, the dynastic turnover. Each shock is a chance for a transmission chain to break. Korea sat close enough to receive the teaching early but far enough to dodge most of the shocks. So the same method that was being revised, suppressed, and forgotten at the center could sit unrevised at the margin. "Preserved unchanged" is the load-bearing phrase, and it cuts both ways: the margin keeps the old form precisely because the margin is not the place where the form is alive enough to keep evolving.
There is a second mechanism in Dukes's phrasing worth naming. He stresses temple records and monuments — paper and stone. A living lineage updates itself; a record fossilizes a moment. If Korea's value is archival, then what it preserves is not necessarily a still-breathing practice but a snapshot of one, the way a fly in amber preserves the fly's shape but not its flight. Dukes wants both — living Son dominance and fossil Chuan Fa traces — and the page should hold that he is reaching for two different kinds of survival at once.
This little endnote does real work for the larger Dukes corpus. It hands Bodhidharma: The Revisionist Biography a possible escape route from one of its own admitted gaps — that biography notes Korean Son is older than Japanese Zen and kept Chinese approaches Japan never received, and this page supplies the mechanism for why Korea would hold what Japan lost. It strengthens Chuan Fa Transmission Lineage by proposing a surviving deposit of the lineage outside the Sino-Japanese line everyone studies. And it gives the vault a clean instance of a general historiographic pattern — the periphery as archive — that the Wandering Acarya Transmission Mobility page can borrow: teachings move along the edges of empires, and the edges are where they get stranded and saved.
Watch how much weight Dukes puts on how little. The whole structure rests on a single dig dated 1989–1990, a temple of an "Eastern Mountain sect," and a verb he is careful to keep conditional: it "could be one connected to Chuan Fa practice."2 That is a real and honest hedge. But look at the next two sentences. The hedge evaporates. The proximity argument is stated flat — teachings "were preserved unchanged in Korea" — and the prediction is stated as confidence: "I am certain that, as years go by, we shall find traces."4 So inside one short endnote the temperature swings from "could be" to "I am certain." The excavation is the only piece of new physical evidence in the passage, and it does not actually establish the thesis; the thesis is older than the dig, and the dig is offered as a first confirming instance rather than as the ground of the claim. This is the page's whole lesson in miniature: a single ambiguous find is being recruited to corroborate a conviction the author already held.
You are reading the endnotes of a martial-arts history at midnight because the main text raised more questions than it answered, and you hit Note 70b. The phrase "I am certain" stops your eye. You stop and do the historian's reflex: separate the find from the belief. On the left margin you write what was actually dug up — an unidentified temple, an "Eastern Mountain sect," a date. On the right you write what Dukes wants it to mean — a surviving deposit of Chuan Fa, ignored by everyone else. You draw a line between the two columns and label the line "could be." Then you ask the cold question the excitement was skipping: has the temple been identified by anyone other than Dukes? Is "Eastern Mountain sect" a sect that scholars recognize, or a translation only he uses? You don't have those answers tonight. So you do the only honest thing — you file the whole claim with a flag on its forehead: single source, speculative, one dig doing the work of a proof. The thesis stays in the vault, because a margin-as-archive idea is worth keeping. It just doesn't get to wear the word "certain" without quotation marks.
Evidence base: Dukes's Note 70b alone — a single endnote, no external citation for the excavation, no published archaeological report named, no identification of the temple beyond "Eastern Mountain sect." [single-source; speculative] governs this entire page. Nothing here has independent corroboration in the vault or in the passage itself.
The honest part of the claim is the part that survives scrutiny: Korea's proximity to Northern China and the antiquity of Son relative to Japanese Zen are real, and the periphery-as-archive pattern is a recognized one in the history of religions. The unsupported part is the specific bridge — that this temple, from this dig, connects to Chuan Fa. The page preserves the gap between the two and does not close it.
Open: Dukes predicts future finds in Korean temple records and monuments. As of his writing the prediction is unfulfilled by anything but the one ambiguous dig. The open question is whether any of his "certain" future traces ever materialized — and the vault cannot answer that from this source.
Reading this beside Dukes's own Bodhidharma biography is uncomfortable in a productive way. There, Dukes is the careful demythologizer — he watches legends assemble, dates sources against each other, refuses to fill gaps with "must have." Here, in the endnotes, a different Dukes shows up: one who fills a gap with "I am certain" on the strength of a single unconfirmed excavation. The two are the same man at different distances from his own enthusiasm. The Korean thesis is exactly the kind of claim his Bodhidharma method would interrogate if someone else had made it. That is not a gotcha; it is the recurring shape of the whole book. Dukes's warrant throughout is his own temple's privately-held material and his own network of informants — the Kongoryuji line — and the same warrant that lets him see what outsiders miss also lets him believe what outsiders can't check. Geoffrey Samuel, the vault's critical-historical anchor, would let the proximity argument stand and quarantine the dig instantly: an unidentified temple is data about the future of the question, not an answer to it. Put Samuel and this endnote together and you get the rule the page enforces — a find is not a confirmation until someone other than the person who needs it confirms it.
Protestant Template Distortion of Asian Religions — that page describes how Western scholarship privileges text over living practice, treating "the sources are late" as if it settled the question of whether a transmission survived. Dukes's Korean thesis is the mirror-image error from the opposite side of the same coin. Where the Protestant template under-trusts embodied transmission, Dukes over-trusts it: he is so committed to the idea that practice survives off the page, stranded at the margins, that he treats a single unread excavation as near-proof of survival. Put the two together and you get an insight neither produces alone. The reliability of a transmission claim does not depend on whether the critic favors text or practice — it depends on whether the claim is checkable by someone outside the lineage. The text-skeptic who dismisses Korean Son and the practice-enthusiast who romances it are both substituting their prior for the evidence. The vault needs a stance that is neither: trust the periphery-as-archive pattern, demand identification of the instance.
Bodhidharma: The Revisionist Biography — the biography page argues that legend is compressed information and the historian's job is decompression, not deletion. Apply that engine here and it cuts against Dukes himself. The Korean thesis is, structurally, a legend in the making — a single ambiguous object accreting a confident story about lost northern teachings preserved intact. Decompressing it, you find what the story is load-bearing for: it lets the Chuan Fa lineage survive every documented break in the Chinese record by relocating the survival somewhere conveniently hard to inspect. The insight the pairing produces is that the same author can be a rigorous decompressor of other people's legends and an active composer of his own, and that a vault built from one author's work has to run his decompression method back over his own confident asides. The refrigerator thesis is Dukes's Daruma doll — a real object growing a story larger than itself.
Sharpest implication: if a tradition's last living copy can survive only at the geographic margin of the culture that birthed it, then the history of any "lost" art should be searched at the edges first — and the search has to be run by someone with no stake in finding it, because the person who most wants the survival is the person least able to confirm it.
Generative questions: