Sapsali (Sapsari)

Two rows deserve attention of their own. The coat comes long and short — both forms are legitimate within the breed, and in the largest measured sample almost half were short-coated. And height is given separately for males and females: a single cross-sex range is wrong for this breed, because it quietly swallows the bitches.

The record in the state heritage register of the Republic of Korea is called “경산의 삽살개” — “the Sapsaree of Gyeongsan”; in the same agency’s English-language card it appears as Sapsari Dog of Gyeongsan. The category is 천연기념물, natural monument. The date in the register is 10 March 1992. Holder of record: the Korean Sapsaree Foundation; administrator: the city of Gyeongsan.

Three things that retellings fuse into one.

• This is not a “national treasure”. “국보” (national treasure) and “보물” (treasure) are conferred under the cultural heritage act and only on material heritage; natural monuments are designated under a separate act. A living population of dogs cannot in principle be entered in the register of national treasures — these are different registers, not synonyms.
• What is designated is not “the breed” but a population in a place. The tie to Gyeongsan in the name is not accidental: the agency’s standing guidance on naming natural heritage prescribes exactly the order “self-government unit + proper name”. The register also states the scope: “the proper number of head is 400, the registered number is 200”, with a note that this concerns Sapsarees kept within the city of Gyeongsan and registered as conforming to the standard.
• The record number is a historical reference today. On 19 November 2021 the heritage agency announced it was dropping numbers in designating national treasures, treasures, historic sites and natural monuments: the wording “designation number” was struck out of the subordinate legislation and the administrative forms. The figure 368 remains the key under which the state database still returns the record — its “revoked” field reads “no” — and the Foundation itself uses it. But primary here is the name, not the number.

Two other local dog breeds hold natural-monument status in Korea: the Jindo (record 53) and the Gyeongju Donggyeongi (record 540). All three records are in force. About its neighbour in the register we have written separately — the Korean Jindo Dog.

The date has a muddle of its own. The state database and the English-language card both give 10 March 1992. The government livestock institute gives 7 March. A regional encyclopaedia names 10 March — and in the same breath cites a newspaper report of the designation dated 8 March, that is, a day earlier than its own date. We take the date from the register and name the discrepancy rather than picking the convenient one.

Here popular texts offer two mutually exclusive statements — “export is strictly forbidden” and “Sapsarees are sold all over the world” — and each has a document behind it.

Since 17 May 2024, designated natural monuments in Korea have been protected by a separate Act on the Conservation and Use of Natural Heritage (passed on 21 March 2023; the designation of 1992 moved under it automatically by the transitional provisions). Its article 20 is headed “Prohibition on the export of natural monuments” and opens unambiguously: a natural monument may not be exported or taken out of the country. The exceptions are narrow: by permission of the head of the heritage agency, on condition of return within two years, and only for display abroad as part of a cultural exchange or for scientific research. Export of animals bred at a designated facility for research or exhibition is permitted separately. Article 34 adds a “keeping zone” regime — taking an animal outside it requires the provincial governor’s permission.

At the same time the Foundation openly describes the overseas side of its work: the breed’s first showing abroad in Washington in 2007, the role of mascot at a Korean festival in Los Angeles in 2011, a cooperation agreement with that festival’s organisers, and a network of dog placements — to several US states, and also to Poland, Australia and Singapore.

The two documents do not contradict one another once you notice what the ban is attached to. The act forbids the export of a “natural monument”, and that is defined as an object designated and published in the prescribed manner; the register, in turn, limits the designation to the registered head in Gyeongsan. So the checkable formulation runs like this: a Sapsaree registered in Gyeongsan as a designated natural monument may not be exported, other than by permission of the agency for display or research with return within two years. Where the line falls for the rest of the dogs of the breed does not follow from the available documents — worth bearing in mind when reading any categorical statement in either direction.

The commonest figure in popular texts is “more than three thousand”. It exists, it has an author and a date, and it is almost always quoted without either.

Three conclusions follow. First, “more than 3,000” is a 2010 figure, and later papers already give about four thousand. Second, 8,264 (or 8,256 in another paper) is not a head count but a cumulative pedigree over twelve to thirteen generations; anyone quoting that number as a count of living dogs is out by an order of magnitude. Third, the register’s figure is smaller than all the others because it measures something else: not every Sapsaree in the country, but the designated population in Gyeongsan.

Tellingly, one and the same register record gives two numbers within itself — 200/400 in a structured field and “about 500” in the descriptive text. We quote both.

The description in the state register puts it this way: the Sapsaree, an ordinary sight in every village, fell sharply in number after 1940, under Japanese rule — Japan made dogs a source of the hides needed for the war and killed them in large numbers — and came to the brink of extinction.

The event is real, and there is a primary trace of it. A peer-reviewed paper of 2018 on the history of veterinary medicine, devoted to rabies in colonial Korea, states it directly: the domestic dog population, on its reading, had been falling ever since dog hides began to be requisitioned as war material in the 1930s, and the Government-General considered it possible to obtain in Korea an average of 100,000–150,000 hides a year, at most up to 500,000. That estimate comes from a 1943 publication of the Government-General’s forest experiment station on Korean dogs and their fur.

🔴 Here is where the commonest error about this breed hides. In the document this is an estimate of procurement capacity — how many hides it would have been possible to obtain from all the dogs of Korea. In the retellings it turned into the number of Sapsarees killed per year. Two different statements, and not a single figure changes in the process.

Which gives a check. The same paper cites the Government-General’s veterinary statistics: the total number of dogs in Korea fell from 1,391,147 head in 1934 to 868,342 in 1940, roughly 38 % — and that is all breeds and all causes together. Against it, “a million and a half Korean native dogs killed” exceeds the country’s entire dog population as of 1940, while the English version “100,000–500,000 Sapsarees every year” would over seven years give between 700,000 and three and a half million Sapsarees alone — against a nationwide maximum of 1.39 million dogs of all breeds.

So we write it thus: the requisitioning of dog hides as war material is documented, the fall in the dog population is documented, and the number of Sapsarees killed is documented nowhere. Telling, too, that the historical paper just cited, whose author read the 1943 publication, does not mention the Sapsaree once in all its thirty-four pages.

One more detail breaks the convenient symmetry. It is often written that the Japanese authorities exterminated Korean native dogs in general. In fact the fates of the three breeds diverged: the Government-General designated the Jindo a natural monument in 1938, which the literature names as the reason it survived; the Donggyeongi declined for a quite different reason — tailless dogs were held “unlucky”; and the Sapsaree, in the breed foundation’s own words, was hit precisely because of its long coat and usable hide. This is not one story told about three breeds, but three different stories.

Here popular accounts make a typical mistake: they credit the start of the programme to one person — the best known. The state register puts it briefly and in the plural: in the late 1960s professors of Kyungpook National University collected and took into safekeeping about thirty Sapsarees. The regional encyclopaedia names names: the money for the collecting and the research was raised by professors Tak Yeon-bin and Kim Hwa-sik; 52 dogs were judged pure, and 30 of them were gathered. Ha Sung-jin — academic supervisor and patron to them both — later took their work over and kept the dogs on his farm near Daegu. Ha Ji-hong, the man the breed is associated with today, is his son.

⭐ Why every date in this story floats — there is an answer, and it is documented. The regional encyclopaedia says it plainly: Ha Sung-jin kept the dogs from 1970 to 1984, but had no clear intention of breeding from them, and no records survive. Fourteen years with no studbook — that is why the modern pedigree begins only in 1989, and why every retelling dates the collecting, the return and the number of founders differently.

The real number of founders of the modern population is eight, not thirty. Not our reconstruction: Ha Ji-hong writes it himself in his own peer-reviewed article — “in 1986 eight individuals resembling the Sapsaree in colour and build were collected around the country by local enthusiasts of the breed in Daegu”; the same number comes from the government livestock institute (in the spring of 1985 there were only eight of them on his father’s farm) and from the regional encyclopaedia. A genetic paper of 2019 repeats it thus: “in 1986 the Sapsaree population was restored from eight individuals selected for their resemblance to the original breed — by colour and body shape”.

The two “thirties” here are a trap: about thirty were gathered at the turn of the 1960s and 70s, and about thirty was the number again by 1989, this time starting from eight. The formula “from 30 to more than 3,000” is wrong at its base: the counting starts from eight. And the words “selected for their resemblance in colour and build” are worth attention: the founders were chosen by appearance, not by proven descent.

A great deal is written about the Sapsaree, and almost always approximately. Let us set the common claims against what stands in the papers.

The state register gives the meaning directly: 삽살개 is “a dog whose name means ‘the hound that drives away ghosts and misfortune’; the name itself is a native Korean word, and the dog appears often in songs, folk tales and paintings”.

That wording is worth taking literally, because it makes two different statements: that the name is natively Korean, that is, not borrowed, and that a dog under this name is present in the song, oral and painting traditions. Both belong to the register, and we relay them as its position.

The earliest dated attestation of the name leads not to spirits but to the coat. In the Hunmong jahoe — Choe Se-jin’s character dictionary, published in 1527 (訓蒙字會) — the vernacular form sapsalgahi stands beside the character for “dog”, with the Chinese equivalent 絡絲狗, “a dog of tangled threads”. So the oldest written explanation of this name is a description of the coat. That does not refute the meaning the register gives, but it shows well why we set out no morpheme-by-morpheme parsings of the name: they belong to a later interpretation, not to the earliest record.

The description in the state register is short and remarkably concrete: the whole body is covered with long hair; the eyes are not visible behind the coat; the ears drop, the muzzle comparatively blunt rather than pointed as in the Jindo; the tail carried up, the head large, and in appearance the dog resembles a lion.

The Foundation’s description adds what the register lacks: the coat comes long and short, and the size is stated as “a medium and large dog”. A short-coated Sapsaree is neither a crossbred nor a fault but a full form of the breed. In a sample of 114 dogs there were 61 long-coated and 53 short-coated; in a sample of 378 the split also came out about 60 to 40, though that paper labels the two groups differently in its methods and its results.

⚠️ Two widespread details need correcting. First: the Sapsaree is often called wavy-coated. The one paper that counted coat structure — 882 dogs, three categories (straight, wavy, curly) — found 69 % straight-coated. Second: a tail curled in a ring over the back is presented as the breed’s silhouette, whereas the Foundation’s description makes something else the norm — “thick and long, freely extended in an arc” — and calls a tail curled along the spine a variant seen in some individuals.

Other details of the Foundation’s description that rarely reach translations: ears “mostly of drop form, but individuals with prick or semi-prick ears do occur”; eyes moderately large and round, mostly brown, with low frequency yellow and jade; canines large and well developed, the bite mostly normal; back straight, loin tucked up. In general impression the Sapsaree resembles an Old English Sheepdog, and its coat is closer in work to the dense fur of a Chow Chow than to a medium-sized herding breed’s.

Colour. The Foundation’s description lists four base colours:

• cheong (청, “blue”) — black-grey with a steel cast; together with red this is the most numerous colour;
• hwang (황, red-yellow) — from pale sandy to a deep red;
• baek (백, white) — occurs at low frequency;
• baduk (바둑, parti-colour) — spotted, also rare.

⚠️ The Foundation’s description sets no hierarchy between solid and parti-coloured dogs — it acknowledges a diversity of colours and merely notes which occur more often. The claim that solid dogs are “rated more highly” is absent from the standard. Worth knowing, too, that the Foundation’s English-language page lists chocolate among the rare colours, which appears in no Korean list, while the measured samples divide them more finely still: in the paper on 547 dogs the colours fall into five groups (blue 236, red 240, white 37, red-and-white 18 and chestnut 16), and in the paper on 882 dogs into four.

One more correction, important for anyone choosing a puppy: the nose leather is not always dark. The Foundation calls it black, glossy and moderately large, occasionally dark brown — and in white Sapsarees often pink.

⚠️ Of “blue” Sapsarees it is often written that their colour comes from a pigment dilution gene. A variant of such a gene (MLPH) has indeed been found in Korean native dogs, and in the Sapsaree its frequency came to 0.20 — but this rests on only five fully read genomes, and the authors themselves distanced it from colour word for word: “no cases of coat colour dilution or of the associated skin inflammation have been reported in these Korean breeds. This suggests that colour dilution depends on more than the MLPH mutation alone”. Korean popular sources describe the mechanism quite differently and far more simply: puppies are born very dark and with age white guard hairs mix into the coat — that mixing of black with white gives the “blue” cheongsapsaree, while the dark red hwangsapsaree gradually lightens. So it is greying rather than pigment dilution.

This is a case where it is more honest to show the disagreement than to give one “correct” number. We found seven sets — and no two coincide completely. Three were published by one and the same organisation.

The discrepancies are of different kinds, and worth telling apart. The register’s pair “51 and 49” is most probably a typical value rather than a boundary. But three tables from one foundation are already a classic source of the “fused” ranges in reference books: the lower bound from one page, the upper from another, and both “from the official site”. The two Korean tables were available at the same time, while the English-language one — the only one the Foundation calls the standard — differs most from both and even describes the breed as “medium”, whereas the Korean page says “a medium and large dog”.

The main point, however, lies elsewhere: the numbers that used to stand in our card (males 50–58 cm, females 48–55 cm, 18–28 and 16–24 kg) match none of these documents, and they are too small. Their origin can be traced: the introduction to a 2024 genetic paper carries a sentence about males “from 20 to 23 inches” and a weight “from 40 to 62 pounds”, and about females “up to 22 inches” and “from 35 to 55 pounds”. Converted, that gives exactly 50–58 cm and 18–28 kg. But that sentence cites a 2019 genetic paper — which says something quite different: “they are of medium body size (54–62 cm in height)”, with no inches, no pounds and no split by sex. So the citation contradicts the source it refers to, while the source agrees with the Foundation.

The tape measure supplies the check. In the sample of 378 dogs the measured maximum is 64.5 cm, so the card’s previous ceiling of 58 cm cut off the upper part of the real population; while the minimum across both sexes together is exactly 50.0 cm, so “48 cm for females” is below anything ever measured. At the same time the means sit near the lower edge of the described range: the mean male in the sample of 882 dogs is 53.4 cm against a stated minimum of 54. The practical conclusion for the reader: a typical Sapsaree is closer to the bottom of the standard than to the top — but dogs of nearly 64 cm exist, and they did not fit into the old card.

So the card on this page now carries the set from the Korean breed-description page, and necessarily separately for males and females: it is the only one that covers the measured data from both ends. On weight it is worth bearing in mind that two small feeding papers of 2026 (18 dogs each) gave mean adult weights of 28.63 and 23.49 kg — so the upper edge of the stated range is rather a typical large male than a record.

This part rarely reaches foreign descriptions, and it is what explains why the breed could be held on to at all after 1986.

• Dokdo island. Since 1998 the Foundation has kept “sentry” dogs on the island; by its own account this is already the ninth generation of that line.
• Heritage sentries. In 2014–2016 twenty trained Sapsarees served at sixteen temples, among them Jeondeungsa on Ganghwa.
• Assistance dogs in therapy. Since 1999 the Foundation has selected its most sociable dogs for work in care institutions, homes for the elderly and rehabilitation centres; since 2021 it has worked with one specific psychiatric clinic — Daedong Hospital in Daegu. This is not “an ever wider practice in Korea’s hospitals” but twenty years of work in social institutions plus one medical partner.
• Outside Korea. The first showing abroad was in Washington in 2007; in 2011 the breed was the mascot of a Korean festival in Los Angeles; since 2019 the Foundation has taken part in shows under FCI auspices in pursuit of international recognition.

On the therapy side it is worth knowing the strength of the evidence too — and here the Foundation behaves more honestly than many, because it itself lists the studies it relies on, together with their samples. And the samples are very small: five patients in a 2021 programme, six children (three in the test group and three in the control) in a study of 2000, fourteen patients in another. That is grounds for saying that the direction exists and has cautious positive results — not that its effectiveness has been proven.

The Sapsaree is one of the few breeds cloned with conservation stated outright as the aim, and the result was published with the numbers in full. The nucleus donor was a nine-year-old male; oocytes from 38 donors were used. 309 cloned embryos were transferred to fifteen recipient bitches; two proved pregnant, each giving birth to a single puppy weighing 530 g — a yield of about 0.65 %. One of the two puppies died: it was injured by its surrogate mother. Both matched the nucleus donor genetically, while their mitochondrial DNA came from the oocyte donors.

Two details make this work more interesting than simply “the breed was cloned”.

The first. The authors note that the surviving clone “apparently showed signs of hip dysplasia… and this had also been found in the somatic cell donor dog”, adding: “although we cannot explain the reasons for this, it indicates that a genetic problem rather than the cloning procedure itself or an epigenetic modification may be responsible”. Three hedges in two sentences — and at the same time one more independent hint that the hips are a subject in this breed.

The second. The work incidentally confirmed a breed feature often quoted as an “interesting fact”: the Sapsaree’s colour changes naturally from puppyhood to adulthood. The clones’ coats at first differed from the donor’s, and with maturity gradually became like his. Here the clones served as an unintended control.

One clarification, because it is always asked: this work was done by a group at Seoul National University, and it has not been retracted. The publications of that same laboratory that were retracted in their day concerned human embryonic stem cells, not dogs.

• A dog named by a place. What stands in the state register is not “the Sapsaree breed” but “the Sapsaree of Gyeongsan”: the protected thing is the population of one particular city, as the agency’s naming guidance requires.
• It resembles someone other than you would think. By whole genomes its closest relatives are not the Korean neighbours but the North Chinese native dog and the Tibetan Mastiff; its West Eurasian share is 40.4 %.
• A short coat is a Sapsaree too. The Foundation’s description lists both forms, and in the sample of 114 dogs examined there were 61 of one and 53 of the other.
• The oldest record of the name speaks of the coat. A dictionary of 1527 gives the vernacular form of the name with the Chinese equivalent “a dog of tangled threads”.
• The colour changes with age — recorded not in a retelling but in the cloning paper, where the clones’ coats “arrived” at the donor’s colour as they matured.
• A breed with a post. Since 1998 Sapsarees have served as “sentries” on the island of Dokdo; in 2014–2016 twenty dogs served at sixteen temples.
• It was the mascot of a world championship. Sarbi, the mascot of the 2011 World Championships in Athletics in Daegu, was drawn from a Sapsaree.
• Rarity does not make it an exception. Its survival story brings the Sapsaree close to the Rampur Greyhound — also a national asset that nearly vanished; and in how hard it is to obtain outside its homeland it recalls the Sage Koochee.