The Selden Map of China

A large hand-painted Chinese merchant chart of about 1608 to 1624, showing shipping routes with compass bearings radiating from a Fujian port across the South China Sea to Japan, the Philippines, Java, Sumatra and the Indian Ocean beyond. It is not the territorial, China-centred map its cataloguers assumed for three centuries; it is the earliest surviving record of a private Chinese maritime trading network that appears on no other map of the period, official or otherwise.

Bequeathed to the Bodleian by the lawyer John Selden in 1659, it was studied once in 1687, dismissed as inaccurate by Edmond Halley some time after, and then effectively forgotten in the stacks for roughly three centuries. Historian Robert Batchelor recognised its significance in January 2008; a major conservation campaign followed in 2011; scientific analysis of its pigments in 2016 proposed a origin far from where most scholars had assumed. Where exactly it was made, for whom, and when its compass rose and route grid were set down relative to the coastline beneath them remain open and actively disputed questions.

This record treats every one of those disputes as live, names who holds which position, and states plainly what has not been settled.

A merchant's chart, not an empire's map

The Selden Map is a large sheet of Chinese paper, made up of smaller sheets joined together, painted in black carbon ink and watercolour and measuring approximately 158 by 96 centimetres. For roughly three hundred and fifty years it sat in the Bodleian Library catalogued, in effect, as an oddity: an unusually large, unusually decorative Chinese map of no settled purpose. What Robert Batchelor recognised in January 2008, examining it shortly before returning to the United States, was that the faint lines he could see radiating out from a point on the Fujian coast were shipping routes, plotted with compass bearings, and that nothing resembling them survived on any other Chinese map of the period.

The distinction matters because it changes what kind of object this is. Official Ming cartography of the period is centred on China itself, with the rest of the world reduced to a thin decorative margin of tributary and barbarian lands. The Selden Map inverts that convention. China occupies less than half of the sheet. The map's true centre is the South China Sea, and its subject is not territory but movement: a network of named ports from Japan to Sumatra and the Indian Ocean, connected by a mesh of routes that a merchant, not a magistrate, would need to read.

What the routes actually show

The shipping lines radiate outward from a point near the neighbouring Fujian ports of Quanzhou and Zhangzhou. From there the primary routes run northeast to Nagasaki and southwest through Hoi An, Champa and Pahang, with further branches reaching Manila, Java, Aceh at the northern tip of Sumatra, and on toward the Persian Gulf and Red Sea. Distances along the routes are given not in linear measure but in geng, or watches: a Chinese navigational unit of roughly 2.4 hours, ten to a day, derived from the time it took a length of incense to burn. That is a mariner's unit, not a cartographer's, and it is one of the strongest pieces of internal evidence that the routes were compiled from real rutters, sailing directions accumulated voyage by voyage, rather than drawn from an existing map.

The port network the routes describe is not the network of the official tribute system. It is a private, commercial network, evidently built up by people who actually sailed it. That is the single fact that makes the map historically important: it is direct physical evidence that an unofficial Chinese maritime trading world existed at this scale in the early seventeenth century, at a moment when European accounts and most surviving Chinese sources describe a state actively restricting private overseas trade.

How it was made: materials and a working method now partly visible

The support is Chinese paper, several sheets joined to make the full sheet, prepared and painted using landscape-painting conventions, ink and mineral colour. Scientific analysis carried out in 2016 by a team led by Sotiria Kogou at Nottingham Trent University, working with the Victoria and Albert Museum's science section, used non-invasive multispectral imaging alongside pigment and binder analysis and found some choices at odds with ordinary Chinese painting practice of the period. The binding medium identified was gum arabic rather than the animal glue conventional in Chinese painting; the green passages combine indigo with orpiment, a yellow arsenic mineral, together with a basic copper chloride consistent with atacamite, a pairing the team associated more closely with the Persian and South and West Asian painting tradition than with the Chinese one.

Two construction details recovered by later study bear directly on how the map was assembled rather than only what it depicts. Stephen Davies, analysing the surviving construction lines for the 2013 Imago Mundi special issue on the map, traced the drafting sequence of its framing elements: the bordered box was laid out first, then the scale bar at an angle to it, and only then the compass rose, built from a line drawn perpendicular to the scale bar. Separately, the 2016 Kogou team found that the shipping routes themselves were laid down before the coastline, cities and decorative landscape elements were painted in over them - the routes came first, the geography was fitted around them, not the reverse. Conservation work in 2011 confirmed the same point physically: when the nineteenth- and twentieth-century backing was removed, a full duplicate draft of the main Nagasaki-to-Pahang route, together with hash marks used to rule off distances, was found on the reverse of the sheet, matching the routes on the front.

Taken together, this is a map that was planned from its trade routes outward, revised as it was built, and finished with a European-derived compass rose and scale bar, drafted onto the sheet in a fixed internal order (frame box, then scale bar, then compass rose, per Davies) after the shipping routes had already been laid down over the coastline (per Kogou). Whether that sequence reflects one hand working in stages, or a second hand finishing what an earlier draftsman began, is not established by either study; see Open Questions.

The compass rose, the scale bar, and what they signal

Neither a compass rose nor a graduated scale bar of this kind is known from any earlier Chinese map. Their presence on the Selden Map is treated across the literature - by Batchelor, by Davies, and in the general secondary accounts that followed the 2011 conservation - as evidence that its maker had direct or indirect knowledge of European navigational chartmaking, most likely through contact with Portuguese, Spanish or Dutch shipping and the ports those powers used in Southeast Asia. That the compass rose is tilted rather than aligned to true north is itself informative: the tilt corresponds to magnetic, not true, declination, and its precise angle became one of the key pieces of evidence the 2016 pigment-and-declination study used to argue for a specific port of origin (see below).

What is not established, despite being an active line of inquiry, is whether the compass rose and the route grid belong to the same drafting campaign as the coastline and place names, or were added afterward, possibly by a different hand, once the base map existed. The construction-sequence evidence recovered by Davies shows a clear order of drafting marks for the box, scale bar and compass rose; it does not by itself establish whether that sequence took place within a single continuous campaign or after an interval, still less whether it was the same artist's hand throughout. No source consulted for this record states outright that a second hand added the compass rose or the route grid; the record's language matches the evidence, not the stronger claim.

Who it was made for: the disputed patron

No patron is named on the map, and the question of who commissioned it, and why, is unresolved. The most specific proposal in print is Robert Batchelor's: that the map was made for, or in the milieu of, the merchant-pirate Li Dan, patron of Zheng Zhilong (father of the Ming loyalist commander Koxinga) and a dominant figure in the Chinese trading diaspora of the 1610s and 1620s. Batchelor points to Li Dan's documented career - based first in Manila, later at Hirado in Japan after imprisonment by the Spanish - and to a cluster of bright red chrysanthemum motifs on the map near Hirado, a decorative detail unique on the sheet, as a possible marker of the patron's own base of operations. This is an argued case resting on circumstantial convergence, not a documented commission; Batchelor presents it as a hypothesis, and other scholars have not adopted it.

The more general and less contested reading, offered by multiple secondary accounts including the historian Hannah Theaker's, is that the map was made for an unnamed wealthy merchant, to be displayed - as the same accounts put it - in a reception hall or entrance, as a visible statement of the range of the owner's trading connections rather than as a working navigational instrument. Its scale, its decorative landscape painting, and the fact that its routes, while functionally derived from real sailing data, are laid out with an evident concern for visual balance across the sheet, all support a display object commissioned to be seen by visitors as much as read by a navigator.

Where it was made: three ports, no consensus

Three places have each been proposed in the scholarly literature as the map's place of origin, and none commands a consensus.

Timothy Brook, in his 2013 book on the map, argues for a Java origin, on the grounds that the southern half of the map, covering the archipelago, shows superior and more current geographic knowledge than the northern half. Batchelor, drawing on the port density around Luzon, Japan and Vietnam and on the Li Dan hypothesis above, argues instead for Manila or a source close to it. The 2016 scientific study led by Kogou proposed a third location entirely: Aceh, at the northwestern tip of Sumatra, based on a combination of pigment and binder chemistry unusual for the Chinese tradition but consistent with the Persian and South and West Asian palette that would have been available in a port with a substantial Islamic trading presence, together with the finding that Aceh is the only port marked on the map with a magnetic declination in the early seventeenth century matching the tilt of the compass rose. Aceh is also one of only six ports marked on the map with a red circle, a detail the same study reads as a possible sign of the map owner's principal base.

Each theory rests on a different category of evidence - documentary and geographic argument from Brook, patron biography and port density from Batchelor, hard material science from Kogou's team - and none of the three has been retracted or has displaced the others. This record treats the question of where the map was made as genuinely open, and lists it as such under Open Questions rather than adopting any one position.

Dating the map: 1608 or 1619

Internal evidence narrows the map's creation to a window between 1607 and 1624, but two scholars have argued for materially different dates within it. Timothy Brook contends that the English East India Company captain John Saris obtained the map in 1608, while trading in Southeast Asia, and brought it to England in October 1609, which would place the map's creation at or before 1608. Robert Batchelor argues for a later date, around 1619, on the basis of the detail with which Taiwan is depicted, which he reads as reflecting geographic knowledge that only accumulated through the 1610s. The two datings differ by roughly a decade and rest on different kinds of evidence: Brook's on a specific proposed transmission event, Batchelor's on internal cartographic content. The Bodleian's own conservation project page, for its part, dates the map to the 1620s without adopting either scholar's argument directly. No single date is treated as settled in the literature consulted for this record.

How it reached England: an East India Company connection, but not a settled chain

John Selden's own words are the earliest documented reference to the map's presence in England. In a codicil to his will dated 11 June 1653, Selden describes bequeathing 'a Mapp of China made there fairly and done in colloure together with a Sea Compasse of their making and Devisione taken both by an englishe commander.' That sentence is the primary source for two facts and one open question: the map was accompanied, at least at one point, by a Chinese-made compass; and it was 'taken' - a word implying capture or seizure rather than purchase - by an English commander, whose identity Selden does not give. Brook's Saris hypothesis is the most specific candidate offered for that unnamed commander, tied to Saris's 1608 to 1609 voyage as chief factor of the East India Company's mission and his return passage to England, but Selden's codicil itself does not name Saris, and the identification is Brook's inference, not documented fact.

What is solidly established is the East India Company context generally: the map passed through hands connected to the Company's trading and shipping network in maritime Southeast Asia at some point between its creation and its appearance in Selden's possession by 1653. The precise mechanism of capture or acquisition, the ship involved, and the date it left Asian waters are not documented in any source consulted for this record.

Selden, the Bodleian, and three centuries of obscurity

John Selden, the jurist and maritime-law scholar best known for Mare Clausum, his rebuttal to Grotius's Mare Liberum, died in 1654. His library and his collection, including the map, passed to the Bodleian under the terms of his will, reaching the library in 1659. For nearly three decades afterward the map's Chinese text was simply illegible to anyone at Oxford. That changed in 1687, when Michael Shen Fuzong, a Chinese convert to Christianity travelling in Europe, visited Oxford and worked with the Bodleian's librarian and orientalist Thomas Hyde to add Latin marginal translations of the map's place names and legends - the only layer of annotation added to the map after it left its original maker's hands, and itself now a distinct historical layer worth reading as evidence of late-seventeenth-century European curiosity about China rather than as part of the original object.

After 1687 the map's documented trail thins. Later secondary accounts describe it on display as a curiosity in the Bodleian's Anatomy School by the early eighteenth century, and describe the astronomer Edmond Halley examining it and dismissing its cartographic accuracy at some point thereafter, a judgement credited with helping consign it to obscurity. Neither of these two claims was independently verified against a primary source for this record; both derive from the Wikipedia article on the map, which is flagged in Sources for replacement with a citation to Batchelor's or Brook's own account of the map's institutional history. What is certain is the outcome: by the twentieth century the map was mounted flat as a wall hanging, rather than stored and handled as a manuscript, a decision that would go on to cause serious physical damage, and it remained outside serious scholarly attention until 2008.

Rediscovery, January 2008

Robert Batchelor, an American historian, examined the map at the Bodleian in January 2008, shortly before leaving Oxford to return to the United States, working with David Helliwell, the library's curator of Chinese collections. What Batchelor identified - the shipping-route network, the compass bearings, and the map's status as a document of private maritime trade rather than official territorial cartography - had gone unrecognised for the length of the map's time at Oxford. The rediscovery set off the research programme that produced Batchelor's own 2013 Imago Mundi article and 2014 book, Brook's 2013 book, Davies's construction study, and eventually the 2011 conservation project and the 2016 scientific analysis, none of which could have proceeded without the map first being correctly read for what it was.

Condition before conservation, and the damage a well-meaning remounting caused

By the time of its rediscovery the map was in poor physical condition. Bodleian conservation records describe it as having been remounted, apparently in 1919, with a thick paper border and a cloth lining intended to let it be displayed and handled as a wall hanging. That intervention, whatever its intent, caused severe cracking of the original Chinese paper support: fragments of the painted surface were falling away with every unrolling of the map by the time conservation work began in 2011. The 1919 remounting is a clear example of a well-intentioned repair actively harming the object it was meant to protect, and it is the single largest physical threat the map faced in its documented history.

The 2011 conservation project

Conservation ran across 2011, ahead of the map's public unveiling that September, led by the Bodleian's Robert Minte and Marinita Stiglitz, working with Keisuke Sugiyama of the British Museum and Mark Barnard of the British Library. The project deliberately combined Western paper conservation with Far Eastern painting conservation techniques, on the reasoning that a Chinese paper support required treatment methods that a Western-trained conservator working alone would not have applied. Practically, this meant gradual humidification of the paper from both sides using traditional dahlia-style sprayers, which drew out historic discolouration and allowed the distorted sheet to be flattened; removal of the twentieth-century textile lining and thick paper border; and the application of new infills and a new lining suited to the object's structure rather than to wall-hanging display.

Removing the old backing exposed what conservation had not been designed to find: a complete draft of the main Nagasaki-to-Pahang shipping route on the reverse of the sheet, along with hash marks used to rule off distances, confirming that the routes were laid out methodically and checked, not painted freehand. This is the physical evidence behind the construction-sequence conclusions reached separately by Davies and by the 2016 Kogou team. Only the earliest paper border, carrying the 1687 Latin annotations, was judged original enough to retain; every later intervention was removed.

Unveiling, scholarship, and what conservation made possible

The conserved map went on public display for the first time in the Bodleian's autumn 2011 Treasures exhibition, running from 30 September to 23 December that year, alongside a one-day colloquium that brought together around fifty cartographic and East Asian specialists to examine it directly. Contemporary reporting placed its insurance valuation at roughly four million pounds at that point, a figure reflecting institutional insurance practice rather than any market transaction; the map has never been sold and no export question has ever arisen, since it has remained in the Bodleian's custody without interruption since 1659.

The conservation project is also what made the 2016 scientific study possible: without the backing removed and the paper stabilised, the multispectral imaging and pigment sampling that underpin the Aceh hypothesis could not have been carried out. The map subsequently travelled outside Oxford for the first time in its documented history, appearing in a joint loan exhibition with the National Library and the Asian Civilisations Museum in Singapore that ran into early 2020, alongside a parallel display at the Bodleian's Weston Library.

What is happening to it now

The map remains in the Bodleian's permanent collection under its historic shelfmark, MS. Selden supra 105, viewable in full through the Digital Bodleian platform and through a dedicated project site, seldenmap.bodleian.ox.ac.uk, both offering a zoomable IIIF-compliant image. It is not, so far as this research established, on continuous physical public display; access for study is through the digital surrogate and, for scholars, through the Bodleian's manuscript reading rooms by appointment.

Research on the map continues. In 2021 Wesa Perttola, of the University of Helsinki, published a computational study modelling the sailing durations implied by the map's routes against reconstructed historical monsoon wind data, using sequential least-cost path analysis - a modern maritime-archaeology method applied directly to the map's own encoded voyage data, testing whether its routes are navigationally plausible under real seasonal wind conditions rather than only visually coherent. The map has also acquired a documented afterlife in contemporary politics: it is cited, including by commentators critical of the practice, in discussions of China's historical claims in the South China Sea, precisely because it is genuine seventeenth-century evidence of Chinese maritime activity in those waters - evidence that nonetheless documents private merchant trade, not territorial sovereignty, and that an international tribunal at The Hague explicitly declined to treat as establishing modern historical rights when it rejected China's nine-dash-line claims in 2016. Citing the Selden Map in that debate uses it for a claim - sovereignty - that the map itself, a merchant's chart, was never made to support.