Under the Wave off Kanagawa (Kanagawa oki nami ura), also known as The Great Wave, from the series Thirty-Six Views of Mount Fuji

There is no such object as The Great Wave. There is a design cut into blocks about 1830, perhaps eight thousand impressions pulled from them, and the surviving sheets, each with its own printing, condition and owners. This record describes one sheet: the Art Institute of Chicago's 1925.3245, bought from Yamanaka and Company by Clarence Buckingham in December 1905.

The British Museum scientist Capucine Korenberg located 111 impressions worldwide, reduced Roger Keyes's twenty-one states to eight, and placed this sheet in state 4. The Art Institute holds three impressions, in states 4, 6 and 7, and credits all three to the Clarence Buckingham Collection although Buckingham owned only one of them. Korenberg uses the museum's 1952 sheet as her published illustration of a late printing.

The record price for a Great Wave, USD 2,760,000 at Christie's New York on 21 March 2023, is a premium-inclusive total; Christie's does not publish hammer. A hammer of USD 2,250,000 can be derived, and is given here as DERIVED on a schedule this record could not read in Christie's own conditions of sale. That price belongs to a different sheet, and none of it is evidence of what this one is worth.

Which Great Wave this is

There is no such object as The Great Wave. There is a woodblock design, cut about 1830 and printed from those blocks in an unknown number of impressions over decades, and there are the surviving sheets. Each sheet is a separate object with its own printing, its own condition, its own colour and its own owners. Confusing the design with the sheet is the commonest error made about this image, and it is the reason this record exists in the form it does.

This record describes one sheet: the impression in the Art Institute of Chicago, accession 1925.3245, 25.4 by 37.6 centimetres, an oban colour woodblock print from the Clarence Buckingham Collection. It is signed Hokusai aratame Iitsu fude and carries the mark of the publisher Nishimura-ya Yohachi.

The Art Institute holds three impressions of this design. All three are catalogued under the Buckingham name. All three have different provenance. The museum's own catalogue states that all three are later impressions than the first state of the design.

What a later impression means, and why it is a price

A woodblock is a physical object and it wears. The keyblock carrying the outlines takes the most punishment: fine lines broaden, edges soften, and cracks open and print as white lines. Colour blocks are recut or replaced. An early impression and a late one from the same design can differ enough that specialists treat them as different objects, and the ranking is what the money follows.

The scholarship on this design is unusually good. Roger Keyes compared surviving impressions and deduced a sequence of numbered states, defining a new state as a change caused by woodblock damage, re-carving, or the use of a completely new block; he listed twenty-one. He was working from postcards, book illustrations and small black-and-white photographs. Capucine Korenberg re-examined the problem at the British Museum using digital images of 111 impressions, found that variations corresponding to certain of Keyes's states did not occur consistently, and reduced the sequence to eight states.

One of her incidental findings is worth more to a collector than the states themselves. When the 111 impressions were put in chronological order there was no sign of gradual woodblock damage preceding a break: the lines in the title cartouche did not thin before disappearing. The breaks happened suddenly, probably between printing events rather than during them, which supports the argument that printing is not the main cause of woodblock damage.

Where this sheet sits in the sequence

Korenberg's published table places the Art Institute's three impressions precisely. This sheet, 1925.3245, is a state 4 impression; she names it in her own text as an example of one. The museum's second sheet, 1928.1086, is state 6. The third, 1952.343, is state 7 of eight.

None of that contradicts the museum, which had already told the public in its own catalogue that all three of its impressions are later than the first state. What it adds is resolution: not later, but fourth, sixth and seventh of eight, in a sequence built from physical damage rather than opinion.

1952.343 has become the discipline's standard illustration of a late Great Wave. Korenberg uses it to show that the light blue areas of the sea in late impressions were not printed from the same block as in early ones, the shapes becoming more angular, and to show that a different yellow block was used for the boats. Sixteen of the 111 impressions she studied were printed with that replacement light-blue block. The very latest impressions have brownish pink boats and black rather than blue outlines, black ink having entered the series when ten further designs were added to it because of its popularity.

She also records, of that same Chicago sheet, that the colour of its clouds appears unfaded but may have been re-applied during a restoration. A conservation caveat published about a named accession is exactly the kind of fact that vanishes from a caption, and it is recorded here.

How many were made, and how many are left

None of the woodblocks used to produce The Great Wave in the 1830s survives. At least seven carved surfaces were used for the early printings, most likely four physical blocks with three of them carved on both sides.

Keyes's economics give the floor: a publisher had to sell at least two thousand impressions of a design to make a profit. For comparison, designs by Kunisada are estimated to have been printed three to four thousand times, and Hiroshige's Fifty-three Stations of the Tokaido between twelve and fifteen thousand. The working estimate for The Great Wave is up to eight thousand impressions.

Against that, Korenberg located 77 impressions in national institutions worldwide and gathered photographs of 34 more from auction sites, dealers, books and private collectors: 111 in all, of which 67 were available in high resolution. The United States holds 35 of the institutional impressions. The largest single holdings are the Museum of Fine Arts in Boston, then the Metropolitan Museum of Art with four, then the British Museum, the Art Institute of Chicago and the Tokyo National Museum with three each.

111 located is not the same quantity as how many survive, and the two numbers are routinely conflated in press coverage. A census of located impressions is a floor, not a total. Anyone quoting a survival figure for this design should say which of the two they mean.

How to tell an impression from a reproduction

Modern workshops in Japan reprint this design from newly cut blocks using traditional methods, and those reproductions circulate. Korenberg identifies two tests that separate them from Edo-period impressions.

The first is colour logic. In original impressions certain shapes are printed in the dark grey-blue or black used for the outlines, the cartouche and the signature; in the reproductions examined, those areas are a lighter blue, the colour used elsewhere for the medium and light blue shapes of the sea.

The second is a flaw. In high-resolution photographs of original impressions a small dark blue shape appears in a spot of the sea foam. The woodblock cutter David Bull attributes it to a sliver of wood the original cutter failed to remove when carving the block. It is present in the originals and absent from every reproduction Korenberg examined. A mistake made in the 1830s is now one of the more reliable ways to authenticate a sheet, which is a general truth about copies worth carrying away from this record.

What has faded, and what that does to a comparison

Edo-period prints were often made with light-sensitive colourants such as turmeric and orpiment, and most surviving Great Waves have been affected by light. The parts that go first are the yellow of the boats and the pink of the clouds. Of the 111 impressions located, no clouds at all were discernible in 26, most probably because the ink had faded completely, although it cannot be excluded that they were never printed.

That single figure should change how a reader looks at any two reproductions of this design side by side. Almost a quarter of surviving sheets have lost their sky entirely, and the difference between two photographs may be a century of daylight rather than a difference of state. Ultraviolet reflectance imaging can bring faded clouds back: a British Museum sheet with very little colour left in the sky shows them clearly under UV.

The pigment that made the series possible

The series began in 1830, when Hokusai was seventy. What made it look the way it does was a chemical import. The Art Institute points to Hokusai's copious use of the newly affordable Berlin blue in the skies and water throughout the series. Berlin blue, called Prussian blue in Europe and bero-ai in Japan, is a synthetic pigment that reached Japan through the Dutch trade and became cheap enough for commercial printing only shortly before this series was designed.

The blue is therefore not a stylistic choice made in a vacuum. A landscape series printed overwhelmingly in one new blue was, among other things, a demonstration of a newly available material, and the popularity of Japanese landscape prints that followed rests partly on a price change in a chemical works. The museum credits this series with establishing that popularity; before it the commercial print market was dominated by actors and courtesans.

What is actually in the picture

Three boats run under the wave. They are narrow, fast craft with crews at the oars, and the men are bent to their work; not one of them is looking at the wave. The crews are on a delivery run, which is why they are in open water in weather like this.

Mount Fuji sits small and still beneath the crest, at the point where the composition's violence resolves. Across the series the mountain is the protagonist of every sheet, seen near or far, in different weather and seasons and from every direction. Here it is the smallest element in the picture and the only motionless one, and the visual joke, if it is one, is that for a moment the wave looks bigger than the mountain.

Fuji was not scenery to the audience that bought these prints. It was a mountain people climbed as an act of devotion, and a series of thirty-six views of it sold into a market of buyers for whom it was a destination rather than a view. The subject of the series is a pilgrimage as much as a landscape.

Yamanaka and Company, and what happened to it

The first name in this sheet's chain is a firm: Yamanaka and Company, New York, which sold the print to Clarence Buckingham in December 1905. Yamanaka was the most important dealer of Asian art in the United States in the first half of the twentieth century and supplied a very large proportion of the Asian holdings of American museums and private collections in that period.

The firm's American operation did not end commercially. It ended by government action after the attack on Pearl Harbor, when its United States assets were seized under the Trading with the Enemy Act and vested in the Alien Property Custodian, and the stock was sold off. Every American museum object whose provenance begins with Yamanaka and Company sits upstream of that event, and it is one of the largest single interruptions of dealer records in the history of the American art trade.

It does not touch this sheet's title. Buckingham bought in 1905, thirty-six years before the seizure, and the print was in a museum by 1925. The reason to record it is that a reader tracing any Yamanaka provenance needs to know where the firm's American papers go, and the answer is that a large part of them go into a wartime custodianship rather than into a dealer's ledger.

Clarence Buckingham and his sisters

Clarence Buckingham, 1854 to 1913, was a Chicagoan who assembled one of the greatest collections of Japanese prints outside Japan. He bought this sheet in December 1905, at the beginning of the period when American collectors were competing seriously for ukiyo-e, and he bought at least one other sheet from the same series in the same transaction.

On his death in 1913 the collection passed by descent to his sisters, Lucy Maud Buckingham, 1870 to 1920, and Kate Sturges Buckingham, 1858 to 1937. Lucy Maud died seven years later. Kate Sturges Buckingham outlived them both by a quarter of a century and gave the prints to the Art Institute of Chicago in 1925.

What kind of collector she was shows in a different sheet's provenance. In 1928, three years after giving the collection away, she bought another impression of the Great Wave from the German collector Alexander G. Mosle and gave that to the museum too. A donor who goes on buying for a collection she has already handed over is behaving as a curator rather than a benefactor, and the Art Institute's Japanese print department is what came of it.

Three impressions, three chains, one museum

Set the three Art Institute sheets side by side and the central problem of this archive is visible inside one collection.

1925.3245, this sheet: Yamanaka and Company, New York; sold to Clarence Buckingham, Chicago, December 1905; by descent to his sisters Lucy Maud and Kate Sturges Buckingham, 1913; given to the Art Institute in 1925.

1928.1086: Alexander G. Mosle, 1863 to 1945, Germany; sold to Kate Sturges Buckingham, Chicago, 31 May 1928; given to the Art Institute in 1928.

1952.343: Frederick Standish Colburn, 1878 to 1957, Evanston, Illinois; sold to the Art Institute in 1952.

Three sheets of the same design, identical to a tenth of a centimetre in the museum's own measurements, carrying the same signature and the same publisher's mark, and every one of them credited to the Clarence Buckingham Collection although only one of them was ever owned by Clarence Buckingham. A credit line records who paid, not who owned. Anyone reading the Buckingham credit as an ownership history will get two of these three wrong, and a caption reading simply the Art Institute's Great Wave identifies nothing at all.

The German collector in the middle chain

The one name in the group a diligence reader should stop on is Alexander G. Mosle, 1863 to 1945, described in the museum's provenance only as of Germany. Mosle assembled a substantial collection of Japanese art in Germany and sold this impression to Kate Sturges Buckingham on 31 May 1928.

The date is the answer to the question the name raises. A 1928 sale by a German collector to an American buyer is a pre-1933 transaction and falls outside the period covered by the Washington Principles and the national frameworks built on them. Nothing in the published record suggests it was anything but an ordinary sale and this record does not imply otherwise. It is set down because German provenance in the interwar decades is the sort of line a checklist flags, and the useful thing an archive can do is give the date rather than leave the flag hanging.

The most reproduced image in the world

Fame has a manufacture and the manufacture is documentable. This design reached Europe in the second half of the nineteenth century with the wave of Japanese imports that produced Japonisme, and European artists and designers absorbed it faster than any other Japanese image. It has since become a book cover, a wallpaper, a tattoo and a character in the Unicode standard.

The consequence for a provenance archive is unwelcome. Universal reproduction means that most people who believe they have seen a Great Wave have seen a photograph of an unspecified sheet, and that captions in books, in the press and online almost never say which impression is illustrated. The image is everywhere and the objects are anonymous, which is the exact inversion of the situation this publication exists to correct.

The record price, and why it is not this sheet's price

The auction record for any impression of this design was set at Christie's New York on 21 March 2023, in a sale of Japanese and Korean art that made USD 11,413,992 in total, 253 per cent hammer above low estimate, with registered bidders from twenty-two countries. The Great Wave brought USD 2,760,000. A Korean white porcelain moon jar in the same sale made USD 4,560,000, a record for that category.

Both of those figures are premium-inclusive totals. Christie's publishes what a lot fetched and does not publish what the hammer fell at, so the hammer must be derived, and a derived hammer is only as good as the premium schedule assumed. This record could not read Christie's own conditions of sale for March 2023; the house's website refuses automated requests, and the only schedule this record verified directly is a 2016 one that was superseded.

Assuming a New York schedule of 26 per cent of hammer to USD 1,000,000 and 20 per cent above it, the total resolves to a hammer of USD 2,250,000 and a premium of USD 510,000. Two things support that assumption and nothing else does. Under it the moon jar in the same sale resolves to USD 3,750,000, and both figures are ordinary bidding increments. Under the 2016 schedule the same totals give USD 2,293,750 and USD 3,793,750; under the schedule Christie's adopted later in 2023 they give USD 2,239,669 and USD 3,727,273. Neither pair is a bid anyone made.

So: hammer USD 2,250,000 and premium USD 510,000, both DERIVED, on a schedule this record has not verified in the house's own terms. The estimate was USD 500,000 to 700,000.

None of this is a value for the sheet in Chicago. That one has never been offered at auction: it went from a dealer to a private collector in 1905, passed by descent, and entered a museum in 1925. The two prices that ever existed for it are private invoices that have not been published, and the record impression of 2023 is a different object in a different state.