Beyond control: Orientalist tensions and the history of the “upas tree” myth

Abstract

The “upas tree” is one of the most enduring European myths about Southeast Asia. Accounts of a tree so toxic that it renders the surrounding atmosphere deadly can first be identified in fourteenth-century journey narratives covering what is now Indonesia. But while most other such apocrypha vanished from later European accounts of the region, the upas myth remained prominent and in fact became progressively more elaborate and fantastical, culminating in a notorious hoax: the 1783 account of J. N. Foersch. This article examines the history of the development of the upas myth, and considers the divergent responses to Foersch’s hoax amongst scientists and colonial administrators on the one hand, and poets, playwrights, and artists on the other. In this it reveals a significant tension within the emerging “Orientalist” discourse about Southeast Asia in the early nineteenth century.

When Henry Yule and Arthur Coke Burnell were compiling Hobson-Jobson, their vast compendium of the colloquial language of the British Empire in Asia, in the 1880s, they introduced the entry for “Upas” as follows:

This word is now, like Juggernaut, chiefly used in English as a customary metaphor, and to indicate some institution that the speaker wishes to condemn in a compendious manner. (1996/1886: 952)

Upas — which, unlike juggernaut, has fallen out of general usage since the nineteenth century — was originally a generic Javanese term for a vegetable poison (Zoetmulder, 1982, qtd. in Dove, 2013: 29). In essence it refers to a particular species of tree with toxic properties, Antiaris toxicaria, found in Indonesia. However, the “upas” recorded in Hobson-Jobson was no mere botanical specimen; it was an artefact in the collective European imagination with a complex pedigree: the myth of an individual tree so terrifyingly toxic that it rendered its surrounds a poisoned wasteland and asphyxiated all who dared approach. In this article I examine the myth’s origins in medieval apocrypha, its elaborate culmination in the 1783 account of J. N. Foersch, and its subsequent emergence as a bone of contention between a colonial endeavour which saw empirically-grounded knowledge as analogous to political power, and a popular imaginative discourse which sought to preserve “the East” as an uncontrolled site of mystery.

The “upas tree” and its poison received considerable scholarly attention between the seventeenth and nineteenth centuries. Indeed, Yule and Burnell themselves dedicated seven pages of Hobson-Jobson to its discussion. A number of modern scholars have also considered aspects of the myth. Michael R. Dove and Carol Carpenter (2005) chart the history of upas as an object of scientific study, and argue that its fantastical elements reflect anxieties about the potential failure of the colonial project in Southeast Asia (188). They acknowledge that the publication of Foersch’s account marked the myth’s “separation from science and relegated it to the realms of literature” (2005: 193). However, they do not examine the reasons for that separation or the subsequent relationship between the divergent “literary” and “scientific” strands. Dove (2013), meanwhile, focuses on the seventeenth-century account of Rumphius, while also usefully highlighting the role of indigenous Indonesians in the construction of discourses around vegetable poisons. Taking a different approach, my study both charts the history of the upas tree as a myth within wider European discourses about Southeast Asia, and pays particular attention to the marked divergence of responses to the Foersch account. In this I am able to identify a previously overlooked pattern of increasing elaboration in the four century-long development of the myth, and a significant and specific tension within the emerging “Orientalism” of the early nineteenth century.

Medieval travellers and pseudo-travellers: Odoric, Mandeville, and Jordanus

One of the earliest accounts of a poisonous tree, growing in the forests of Southeast Asia, comes from the pen of one of the first European travellers known to have visited the region: Odoric of Pordenone, an Italian Franciscan who set out eastwards in the early fourteenth century. According to Odoric’s account of his journey — which was recorded in Latin around 1330 — he visited Persia and India, before travelling south by sea for 50 days to what is clearly the western part of Indonesia. Odoric describes visiting “a certain country called Lamori” (qtd. in Yule, 1913/1866: 146) and “another kingdom by name Sumoltra” (1913/1866: 149), both of which appear to be regions of Sumatra. He then continued to “a great island, Java by name” (1913/1866: 151). The account of Java contains political details which lend credibility to Odoric’s account, suggesting that he must have communicated at the very least with someone who had a good knowledge of its recent history. He describes a king of Java who had “many a time engaged in war” with the “Great Khan of Cathay” (qtd. in Yule, 1913/1866: 155) — seemingly a reference to the various clashes between the East Java polities of Singhasari and Kediri and expeditionary forces of the Chinese Yuan Dynasty in the late thirteenth century.

As well as Java, Odoric describes a nearby country called “Panten”, or “Thalamasyn”. The most obvious identification for this territory is Banten, which occupied the ethnically distinct western quarter of the island of Java. It is here that the poison tree makes its modest first appearance:

Here be found trees that produce flour, and some that produce honey, others that produce wine, and others a poison the most deadly that existeth in the world. For there is no antidote to it known except one; and that is that if any one hath imbibed that poison he shall take of stercus humanum and dilute it with water, and of this potion shall he drink, and so shall he be absolutely quit of the poison. (qtd. in Yule, 1913/1866: 155–57)

It is hardly a fantastical account, other than in the singularly disgusting remedy — repetitions of which continued for centuries, despite its having no obvious indigenous source (Dove, 2013: 30). ¹

Odoric’s account was subsequently plagiarized in one of the most influential of all medieval travel narratives: the wildly inauthentic Travels of Sir John Mandeville, which seems to have been composed sometime in the mid-fourteenth century. Mandeville was purportedly a knight from St Albans, though no such knight ever actually existed, and the identity of the real author is unknown. Recent scholarship suggests that the Travels’ composer may have been French (Higgins, 1997: 29). The sources from which “Mandeville” composed his narrative — which takes in the Holy Land, Ethiopia, and much of Asia — are many and varied, but when it came to Southeast Asia he appears to have directly plundered Odoric’s account of Sumatra and Java. He included the mention of a poisonous tree but, perhaps uncharacteristically, he toned down the noxious antidote. Mandeville’s cure was simply “to take of the leaves of the same tree, crush them, soak them in water and drink it, or else a man will die very quickly, for neither treacle nor any other medicine will help” (qtd. in Moseley, 2005/1983: 132). Mandeville also diverges from Odoric by adding one of the sporadic allusions to personal experience which bolster the apparent authority of what is not, for the most part, a first-person travelogue: “With this preparation [from the tree] the Jews once thought to have poisoned all Christendom, as one of them confessed to me; but, blessed be Almighty God, they failed in their purpose” (qtd. in Moseley, 2005/1983: 132; emphasis added). Besides being a notable piece of medieval antisemitism, this insertion adds a particular frisson, absent from the original Odoric account. Odoric’s poisonous Javanese tree was securely located “elsewhere” as far as his European readers were concerned. By raising the possibility of “all Christendom” being dosed with the toxin, however, Mandeville artfully extended its reach beyond an imaginative “East”, and beyond the pages of the book. The poison, he implied here, was portable and conceivably a threat to readers in France and England. In light of this, Mandeville’s modification of the only antidote gains an alarming aspect: Odoric’s version of the cure could clearly be prepared anywhere, whereas Mandeville’s version required access to the tree itself — access entirely unavailable to any poisoned inhabitants of “Christendom”. It is a rare example in the wider “upas” discourse of the tree’s poison being presented as potentially a direct threat to readers in Europe.

Modern scholarly accounts relating to the upas myth, such as those by Dove and Carpenter (2005), and Daniel Carey (2003), typically mention Odoric and Mandeville as its earliest textual iterations. But in this they fail to acknowledge that two distinct elements fed into its development. The first, as referenced by Odoric and Mandeville, is a simple fact: that there really are several species of highly poisonous plants in Indonesia — as indeed there are in most other parts of the world — and that some of their poisons have been used by humans in hunting, warfare, and perhaps assassination. The second component of the myth is a very specific fantasy: of a particular species of tree so toxic that it poisons the surrounding atmosphere. The earliest European reference to that — apparently overlooked by most previous scholars — appears to come from Odoric’s contemporary, the Dominican traveller Jordanus of Severac, who spent several decades in Asia and whose Mirabilia Descrpita was also written around 1330.

Jordanus seems rather less likely than Odoric to have visited Indonesia in person. It is unsurprising, then, that his account contains a number of more fantastical elements:

There is also another exceeding great island, which is called Jaua [sic], which is in circuit more than seven [thousand?] miles as I have heard, and where are many world’s wonders. Among which, besides the finest aromatic spices, this is one, to wit, that there be found pygmy men, of the size of a boy of three or four years old, all shaggy like a he goat. They dwell in the woods, and few are found.

In this island also are white mice, exceeding beautiful. There also are trees producing cloves, which, when they are in flower, emit an odour so pungent that they kill every man who cometh among them, unless he shut his mouth and nostrils.

There too are produced cubebs, and nutmegs, and mace, and all the other finest spices except pepper.

In a certain part of that island they delight to eat white and fat men when they can get them. (qtd. in Yule, 1863: 30–31)

Here the poison tree is nestled amongst other tropes which would repeat in many subsequent European accounts of Indonesia: hairy, forest-dwelling humanoids, a bounty of tropical produce, and cannibalism. But most significantly for this study, Jordanus’s tree is not simply a source of a deadly toxin; it specifically poisons the surrounding air. This is the true earliest textual trace of a tall travellers’ tale which would reach its apogee in the final years of the eighteenth century.

Botany, conquest, and anxiety: Seventeenth-century accounts

After Odoric, Mandeville, and Jordanus, European accounts of Indonesia remained few and far between for several centuries. It was only after the arrival of the Portuguese in the early sixteenth century, and especially after the entry of English and Dutch companies into the highly competitive spice trade at the turn of the seventeenth century, that a wider body of European writings about the region began to build up, including a significant quantity of specifically scientific literature in both Dutch and English. In this literature, considerable attention was given to the question of poison.

Although the medieval accounts all seem to locate the poison tree in Java, the bulk of seventeenth-century enquiry was focused on vegetable poisons associated with Makassar, a port on the southwestern peninsula of Sulawesi. As Carey (2003) points out, the attention paid to poisons from Makassar was largely due to “the importance of the Makassar sultanate in the period and its role in international trade” (2003: 518). Southwestern Sulawesi had long been a hub of maritime trade, and in the seventeenth century the Gowa Sultanate of Makassar was the most influential indigenous polity of eastern Indonesia. This state of affairs led, inevitably, to conflict with the Dutch, who were then ascendant in the neighbouring “spice islands” of Maluku, and in the entrepôts of the north Java coast, having largely ousted their Portuguese and English rivals. The conflict culminated in the Dutch conquest of Makassar in 1667. During their various clashes with Makassarese forces, the Dutch certainly encountered weapons primed with vegetable toxins, and reports of poisoned darts and daggers appear to have generated interest amongst the scientists of Europe. In London, the Royal Society paid considerable attention to “Makassar poison” in the 1660s (Carey, 2003: 533). This attention, however, largely focused on the mechanics of the poison’s deployment and the nature of its effects, typically through attempts to prepare and deploy the occasional dried samples that reached Europe from Indonesia in what passed at the time for laboratory conditions. Certain key terms — particularly the Javanese word upas and its Malay/Makassarese equivalent, ipoh — were already in use amongst European scientists, and the source of the toxin was widely understood to be a tree. Fantastical stories appear to have been generally absent.

Curiously, it was only in the works of writers with first-hand experience of Indonesia that the more fantastical elements of the upas myth began to emerge. Of these, one of the first to be made publicly available, in a German scholarly journal in 1685 (Carey, 2003: 529), was penned by Admiral Cornelis Speelman, who had led the Dutch fleet which conquered Makassar, and later served as governor-general for the Dutch East India Company — a dramatic indication of the link between scholarship and political power. Speelman’s informants were the subjugated nobles of Makassar, who were apparently reluctant to divulge information about the poison. Indeed, Speelman reports that they were actually bound by oath not to discuss its source. However, some did agree to talk and revealed that the best poison was produced in the Toraja highlands. Speelman’s account is interesting in that it introduces a glimmer of indigenous agency into the construction of the upas myth: Makassarese nobles asserting control over knowledge of what was, quite literally, a secret weapon. Perhaps the most significant element of the Speelman account, however, is that his informants told him that “the soil in which the trees affording the poison grow, for a great space around about produces no grass nor any other vegetable growth” (qtd. in Yule and Burnell, 1996/1886: 956).

By the time Speelman’s account was published, another Dutch scholar was already conducting his own investigations, though his work would not be made publicly available until the following century. Georg Everhard Rumpf — better known as Rumphius — was a German-born botanist who produced a vast catalogue of Indonesian plant species, the Herbarium Amboinense. Rumphius’s account of the upas tree — which he also locates near Makassar — stands out as a fantastical inclusion in what is otherwise a solidly sober gazetteer of Indonesian botany:

Neither grass nor leaves grow under this tree, nor for a stone’s throw around it, or any other trees, and the soil remains barren, russet, as if scorched. And under the strongest kinds one will find the telltale sign of bird feathers, for the air around the tree is so tainted that if birds want to rest themselves on the branches, they soon become dizzy and fall down dead… Everything perishes that is touched by its wind, and all animals avoid passing by, or the birds to fly over it. (qtd. in Dove, 2013: 32)

Rumphius’s account was almost certainly based in part on stories furnished by local informants — again, introducing a hint of indigenous agency in the development of the myth. His own exaggerations were likely initially prompted by local exaggerations which may well have been deliberate and considered, a means of retaining control over the production of a valuable substance. As Dove has it, “His fantastic accounts of the poison tree can be seen as a co-production of European fear and native guile” (2013: 41).

As to why such an otherwise rigorous scientific recorder of the natural world should have admitted this flight of fantasy to his account, Dove and Carpenter argue that, as the Dutch were beginning properly to consolidate their colonial project in Indonesia in the seventeenth century, the idea of a dangerous entity hidden in the inner forests spoke to their anxieties:

[The] upas tree represented fears associated with European dreams of wealth based on control of the spice trade. In particular, fear of upas reflected the threat to the colonial project of uncontrolled plants and people in the interior, mountainous areas of the archipelago. (2005: 184)

This is certainly a compelling idea. It does not, however, explain the ultimate iteration of the upas myth, which was to emerge a full century later, when the Dutch colonial project had passed its tenuous, anxiety-inducing early stages and when overarching Orientalist knowledge of the archipelago was rapidly being constructed.

A fantastical culmination: The 1783 Foersch account

In December 1783, British readers of the London Magazine were treated to a sensational story. The piece was attributed to J. N. Foersch, having been translated “from the original Dutch” by a Mr Heydinger, “Formerly a German bookseller near Temple-Bar” (1783: 512; emphasis in original). ² Foersch claimed to have served as a surgeon during the previous decade at Batavia, the headquarters of the Dutch East Indies (today the Indonesian capital, Jakarta). During his time in Java, he wrote, he had received various reports of what he called the “Bohon-Upas” — “Bohon” being a corruption of the Malay word for tree, pohon. Having obtained official sanction from the Dutch governor-general, and bearing a letter of introduction to a “Malayan priest” appointed by the “Emperor” of Java to guard the realm of the upas tree, he claimed that he had travelled into the mountains to procure his gripping eyewitness account:

The Bohon-Upas is situated in the island of Java, about twenty-seven leagues from Batavia, fourteen from Soura-Charta [Surakarta], the seat of the Emperor, and between eighteen and twenty leagues from Tiukjoe [Yogyakarta], the present residence of the Sultan of Java. It is further surrounded on all sides by a circle of high hills and mountains, and the country round it, to the distance of ten or twelve miles from the tree, is intirely [sic] barren. Not a tree, not a shrub, nor even the least plant or grass is to be seen. I have made the tour all around this dangerous spot, at about eighteen miles distant from the centre, and I found the aspect of the country on all sides equally dreary. The easiest ascent of the hills, is from that part where the old ecclesiastic dwells. From his house the criminals are sent for the poison, into which the points of all warlike instruments are dipped. It is of high value, and produces a considerable revenue to the Emperor. (513)

According to Foersch, criminals condemned at the court of Surakarta — the traditional seat of the major Javanese kingdom of Mataram, which had recently been partitioned between Surakarta and Yogyakarta, with, as Foersch correctly described, a susuhunan or “emperor” reigning from the former and a sultan installed at the latter — were routinely used to obtain the poison. Having been blessed by the “old ecclesiastic” and fitted out with goggles and projective leather armour, the condemned men were despatched into the wasteland from upwind. So noxious was the air around the tree that fewer than one in ten made it back alive — though Foersch supposedly stayed long enough at the edge of the toxic zone to interview several survivors and thus provide compelling details of the tree itself and its immediate surrounds.

Foersch also claimed to have witnessed the execution at Surakarta by means of upas poison of 13 royal concubines “convicted of infidelity to the Emperor’s bed”. Each was tied to a stake and stabbed with a poison-tipped lancet:

My astonishment was raised to the highest degree, when I beheld the sudden effects of that poison, for in about five minutes after they were lanced, they were taken with a tremor, attended with a subsultus tentinum, after which they died in the greatest agonies, crying out to God and Mahomet for mercy. In sixteen minutes by my watch, which I held in my hand, all the criminals were no more. Some hours after their death I observed their bodies full of livid spots, much like those of the Petechia. Their faces swelled, their colour changed to a kind of blue, their eyes looked yellow, &c. &c. (516)

As a hoax — and it was, of course, a hoax — Foersch’s account was a supremely sophisticated piece of work. For all that the story was obviously fantastical, its presentation featured a series of strategies which served to establish a tone of authority and veracity, and careful anticipations of the charges of fabrication which might be levelled against the author. Foersch begins by anticipating potential incredulity: previous accounts of the upas have been “so tinctured with the marvellous, that the whole narration has been supposed to be an ingenious fiction by the generality of readers” (512; emphasis in original). Foersch declares that he himself “long doubted the existence of this tree, until a stricter enquiry convinced me of my error”. His account, he states, is a relation of “simple, unadorned facts of which I have been an eye-witness. My readers may depend upon the fidelity of this account” (512). By acknowledging the unbelievable nature of his subject and positioning himself at the outset of his enquiry as a former doubter, Foersch issues a compelling invitation: by means of strict enquiry he has convinced himself of the poison tree’s existence; his readers ought, then, to join him in accepting the truth of his narrative; to continue to disbelieve is an “error”.

Repeated foregrounding of supposed eyewitness testimony, inclusion of exhaustive detail, and “the power of describing ‘circumstance’, or what we might call situatedness, in vivid fashion” have long been key rhetorical strategies in establishing the apparent veracity of accounts by travellers and explorers (Carey, 2016: 6), and Foersch deploys all these elements in his text. The verisimilitude of the account is further propped by specific references to Foersch’s qualities as a professional scientific observer, such as the cool-headed use of a stopwatch and the noting of specific medical symptoms during the gruesome execution. He was evidently familiar with the scientific language of the day, and had a clear knowledge of the kind of operations conducted by the denizens of the Royal Society, for he also included detailed accounts of his own experiments with the poison — on three unfortunate puppies, a cat, and a chicken. These experiments featured further precise use of a stopwatch (“none of these animals survived above thirteen minutes”) and extensive post-mortem details:

I opened the body [of the third puppy], and found the stomach very much inflamed, as the intestines were in some parts, but not so much as the stomach. There was a small quantity of coagulated blood in the stomach, but I could discover no orifice from which it could have issued, and, therefore, supposed it to have been squeezed out of the lungs, by the animal’s straining while it was vomiting. (516)

Significantly, this passage — which, of itself, contains nothing fantastical, and indeed sounds like the genuine report of a post-mortem examination — is placed after the account of the tree itself. The repetition of similarly detailed observational language, and the reiteration of Foersch’s construction of himself as a qualified scientific observer, serves to bolster the credibility of the earlier, more obviously fantastical sections of the article.

Again anticipating doubts amongst his audience, Foersch even included an explanation for the previous absence of detailed accounts of the upas tree, despite the fact that the Dutch had been installed in Java for the best part of two centuries: “the object of most travellers to that part of the world consists more in commercial pursuits than in the study of Natural History and the advancement of sciences” (517). The elaborate and careful construction of the article meant that considerable resistance would be required on the part of readers not to be convinced by Foersch’s statement towards the end of the article that “the few facts I have related here will be considered as a certain proof of the existence of this pernicious tree, and its penetrating effects” (517).

Like John Mandeville four centuries earlier, Foersch appears to have drawn on several sources to construct his hoax. The first and most obvious is Rumphius’s account, which had, significantly, only been made publicly available in the mid-eighteenth century (Carey, 2003: 530). Besides his use of textual sources, it seems reasonably clear that the author was genuinely familiar with Java, for the political and circumstantial details are largely accurate. Indeed, it may be conjectured that the environs of the upas tree in Foersch’s narrative — a barren upland, suffused with noxious gases and surrounded by high ridges — could have been inspired by the numerous active volcanic craters that pock Java’s central chain of mountains. ³ Yule and Burnell suggest that at least one of the distances from Batavia, Surakarta, and Yogyakarta mentioned for the location of the tree in the Foersch account may be a transcription error, as they come nowhere close to an actual intersection anywhere in Java (1996/1886: n953). However, a journey of “twenty-seven leagues from Batavia” would bring a traveller to the vicinity of the large crater of Gunung Papandayan, which had suffered a major eruption in 1772, shortly before Foersch claimed to have been based in Java. Even the figure of the “old ecclesiastic” appointed by the emperor to guard the poisoned realm has a possible real-world parallel in the figure of the mystic guardian, still appointed by the Yogyakarta court in the twenty-first century to dwell on the high slopes of the sacred Merapi volcano.

Who exactly J. N. Foersch was has been a point of some debate down the decades. The compilers of Hobson-Jobson suspected that the author of the 1783 article may have been the same pseudonymous hoaxer behind a similarly fantastical account of a Ceylonese anaconda which had appeared 15 years earlier (Yule and Burnell, 1996/1886: 954). Other anecdotal reports suggested that the hoax may have been the work of George Steevens, sometime assistant to Samuel Johnson. However, the author’s apparent first-hand knowledge of Java makes this unlikely, and John Bastin (1985) has convincingly demonstrated that not only did a “J. N. Foersch” exist and serve as a “Surgeon Third Class” at Semarang (rather than Batavia) in the 1770s, but that he was definitively present in London around the time that his article appeared. Most remarkably, Bastin’s investigations reveal that Foersch — who may, like Rumphius, have been born in Germany — was an acquaintance of one of the foremost botanists of the day, Sir Joseph Banks (Bastin, 1985: 38).

Of course, none of this gives any insight into Foersch’s motivations for concocting his elaborate hoax. That said, Bastin’s information that he intended to publish an “Account of the manners & customs” of Indonesia and had already collected subscribers by the time his article appeared suggests that he may simply have been looking to generate publicity with a dash of sensationalism (Bastin, 1985: 39). But Foersch’s article in the London Magazine does not stand alone; it was part of a European discourse that had first emerged 400 years earlier. In considering this, a remarkable pattern becomes apparent. Jordanus’s account appeared in the fourteenth century at a time when exotic stories of cynocephali (dog-headed men) and suchlike were very popular, and when Indonesia was almost entirely unknown to Europeans. And yet his account of the upas tree is only moderately fantastical — a tree emitting a potentially deadly fragrance. By the seventeenth century, when cynocephali and their ilk had generally been consigned to the realms of fiction and when first-hand European accounts of Indonesia were readily available, the myth had become considerably more elaborate and more fantastical: the tree was now capable of felling birds from the sky. But it was only in the late eighteenth century, deep within the era of the Enlightenment, with rigorous scientific enquiry and technological advancement going on all around, and with Indonesia increasingly falling under centralized Dutch control, that the upas myth reached its sustained fantastical culmination.

The myth’s increasing elaboration and ever-more fantastical nature ran absolutely contrary to the development of empirically-grounded European awareness of Indonesia, from the late medieval period to the eighteenth century. At each stage it was read as fact by at least some of its audience. I will consider the question of why this might have happened in the final section of this article; but first it is necessary to examine the two divergent streams of response to Foersch’s account.

“A degree of impudence scarcely to be believed”: Scientific responses to Foersch

The reaction of the scientific community in Europe to Foersch’s article was decidedly sceptical from the very outset. In Java itself, the Batavian Society of Arts and Sciences instigated an investigation of the upas which eventually produced a journal article, published in 1789, containing a direct refutation of the Foersch account and specific evidence of its fraudulence (Bastin, 1985: 32). Various other Dutch, French, and English accounts followed, most significantly an 1810 paper by Jean-Baptiste Louis Théodore Leschenault de la Tour, a respected French botanist who had spent three years in Java after disembarking from a passing maritime expedition during a bout of ill health. De la Tour not only conducted various experiments regarding the effects of upas poison; he also travelled in person into the forests of Banyuwangi in the far east of Java and there identified the source species. De la Tour’s account contained a direct and scathing rebuttal of Foersch’s article:

[Foersch] has endeavoured to mislead Europe with a degree of impudence scarcely to be believed or forgiven. After having made a collection of absurd stories, to which he has added his own inventions, he, on his return to Europe, gave, as an eye-witness, a narrative, accompanied by all those minute and circumstantial details, which are generally the seal of truth, and which prevent a man being accused of falsehood, unless he is held in the most profound contempt. (Stockdale, 2003/1811: 330)

The tone here gives some indication of the pronounced anger directed towards Foersch by members of the European scientific community around the turn of the nineteenth century.

From 1811 the Anglophone world had further opportunity to receive reputable scientific accounts of upas. During August of that year British forces invaded Dutch territories in Java. The invasion was a far-flung episode of the global Napoleonic conflict (the Netherlands had been seized by France in 1794, making their overseas possessions de facto enemy territory for Britain), and it led to a five-year interregnum during which a nascent British East Indies was ruled by an appointed lieutenant-governor, Thomas Stamford Raffles, better known today for his later role in the founding of a British settlement at Singapore. Raffles had serious scientific interests of his own, and he requested a report on the upas from Thomas Horsfield, an enthusiastic naturalist and unlikely American expatriate, long resident in the royal Javanese city of Surakarta. Horsfield delivered his “Essay on the Oopas or Poison Tree of Java” in March 1812. Like de la Tour — with whom he had communicated — Horsfield had travelled to Banyuwangi where he encountered the upas-producing tree in its natural environment — namely, surrounded by other vegetation and animal life, and emitting no noxious fumes whatsoever. Horsfield also recorded that the tree was locally known as ancar (or antshar or antiar) (1812: 4), and from this, its scientific name — Antiaris toxicaria — was developed.

As well as investigating the source of the poison, Horsfield embarked on a particularly extended series of experiments on its effects, sacrificing large numbers of dogs and chickens, as well as sundry other species — from water buffalos to flying squirrels — in the name of science, and recording their death pangs in vivid detail with stopwatch to hand. The similarities between Horsfield’s accounts of the effects on dogs and Foersch’s own description of its use on puppies suggest that the hoaxer may well have had real experience with upas poison during his time in Java, or at very least had encountered a reliable first-hand account of such an experiment. Horsfield’s report also included the now near-ubiquitous direct and angry refutation of Foersch (1812: 2–4).

By this stage any slender benefit of the doubt that Foersch might ever have enjoyed amongst serious scientists had been roundly obliterated. But the reaction to his article elsewhere had been rather different. For the more strictly literary — as opposed to scientific — world of Europe, the fantastical account of the upas tree had proved irresistible.

“The Hydra-Tree of death”: Popular and literary responses

The London Magazine appears to have been a respectable generalist publication aimed at educated readers — a similar market to that later served by Blackwood’s Magazine. Had it been otherwise, perhaps Foersch’s article would have had less impact. As it was, the tale of the upas tree quickly proliferated, reprinted in various other publications and translated into French and Russian (Gustafson, 1960: 103). It sometimes appeared with inaccurate attributions, which probably served to further its ubiquity and its impact (Bastin, 1985: 38).

It is impossible to tell what proportion of its original readers took the piece at face value. The editor who first published it apparently accepted its essential truth, while retaining at least a hint of scepticism judging by an editorial note appended to it in the London Magazine:

With regard to the principal parts of the relation, there can be no doubt. The existence of the tree, and the noxious powers of its gums and vapours, are certain. For the story of the thirteen concubines, however, we should not choose to be responsible. (512; emphasis in original)

Among the readers who did take serious notice was Erasmus Darwin, grandfather of Charles. Darwin was both a serious botanist and a poet — interests which he combined in The Botanic Garden, a two-part epic poem published in 1791 which blended Linnaean taxonomy with lurid anthropomorphism. Crucially, the poem referenced reputable scientific texts as its source material, and these featured in the extensive endnotes. Amongst the species that appeared in the poem was the upas tree:

Fierce in dread silence on the blasted heath

Fell Upas sits, the Hydra-Tree of death. (1825/1791: 168)

Darwin’s source was evidently Foersch and, indeed, he included the entire article in the endnotes for this section of the poem (1825/1791: 191). ⁴ In doing so he gave the account both a literary and a scientific gloss, and this literary–scientific fusion appears to have influenced a number of other writers. Upas featured in Robert Southey’s 1801 epic, Thalaba the Destroyer (Paterson-Morgan, 2016: 41), which was almost certainly inspired by the blended Darwin–Foersch account.

The upas as a feature in popular discourse then received a considerable boost when the public received news of the 1811 British expedition to Java. Taking advantage of the new flurry of interest in the region amongst British readers, and before word of the expedition’s success had even reached London, the publisher John Joseph Stockdale hurried a book into print, comprised of reprinted and translated material from existing sources. It was titled Sketches Civil and Military of the Island of Java and its Immediate Dependencies: Comprising Interesting Details of Batavia and Authentic Particulars of the Celebrated Poison-tree. Stockdale evidently understood the sensational value of the upas, and he included the complete text of Foersch’s article in the book. He was clearly aware that the account had been roundly debunked by this stage, indicating as much in a footnote which described the article as a “notorious fabrication” (2003/1811: n311). Later in the book he also included a translation of de la Tour’s account, and an excerpt from a paper on the effects of upas poison presented to the Royal Society earlier that year. These were presumably the “authentic particulars” of the book’s lengthy subtitle, but the fact that they appeared after the Foersch account — the unreliability of which was signalled only by that easily overlooked footnote — is perhaps significant. Stockdale was, in essence, presenting two divergent versions of the upas tree to his readers, the fantastical and the strictly scientific, and allowing them to choose which they preferred. Judging by the flurry of literary activity in the subsequent decade, it was the former which had the greater impact.

On 18 November 1811, a pantomime based on the Foersch account titled The Poison Tree, or Harlequin in Java premiered in London (Cohen, 2009: 89). Whether its author, Jane Scott, had first encountered the story in Stockdale’s book is unclear; if so, she had worked very quickly, for the first edition had only appeared on 24 September. Other, more serious, literary renderings soon followed. Upas à la Foersch–Darwin featured in Canto IV of Byron’s Childe Harolde’s Pilgrimage (1812–1818), and in Coleridge’s 1813 play, Remorse. Another play based extensively on the Foersch narrative, The Law of Java by George Colman the Younger, appeared in London in 1822. “Upas” was also by now well on its way to becoming a “customary metaphor”. ⁵

Of course, these were all strictly literary works, and there would likely have been at least some tacit understanding amongst both their authors and their audiences that the image of the upas they deployed was indeed a myth. But what is key is that two distinct versions of the story were now at large: that which featured empirical detail of a real tree growing in the forests of Banyuwangi, and that which featured dramatic fantasy. These two versions had been explicitly placed together by Stockdale, offering a clear choice to readers. Those who were not scientists generally appear to have chosen the latter version.

This idea of an explicit choice is underscored in one of the later literary renderings of the upas myth, penned in 1828 by the Russian poet Alexander Pushkin. Usually titled “The Poison Tree” in English translation, in the original it was called “Ancar”. Attempting to decipher the precise antecedents of the poem, and suggesting Darwin as the principal source, Richard Gustafson (1960) acknowledges that the Russian title is a “problem” for his thesis, given that the term ancar does not appear in either The Botanic Garden or the original Foersch article. However, he then appears to dismiss the issue, suggesting that Pushkin “may even have just heard” the term ancar (1960: 103). However, Gustafson misses a considerable significance here. Ancar and its variations, based on the authentic Javanese word for the tree most closely associated with upas poison, appear for the first time in the accounts of de la Tour and Horsfield — both firmly scientific texts, and both containing explicit refutations of Foersch. The word subsequently occurs in similarly scientific texts, but does not seem to appear in any of the other literary or popular accounts of upas. Pushkin, then, must have encountered at least one scientific counter-narrative, as well as The Botanic Garden or some other literary source. He chose the fantastical version, probably because it offered greater literary possibilities, and perhaps also because it seemed a more fitting representation of Southeast Asia in the popular European imagination of the 1820s.

Pulling in different directions: The upas myth and diverging strands within Orientalism

The five-year British interregnum in Java (which ended with the restoration of Dutch colonial rule in 1816) produced two significant works of Orientalist scholarship. The first, and the best known today, was that penned by the lieutenant-governor himself: Thomas Stamford Raffles’ hulking two-volume The History of Java, which was published in 1817. The second, appearing in 1820, was the three-volume History of the Indian Archipelago by John Crawfurd, another key figure in the erstwhile British administration of Java. Typically of such “histories” of the time, these books contained not only historical accounts of the territories they aimed to cover, but also a wide array of economic, ethnographic, botanical, zoological, and geographical information (the obvious precedent for both books was The History of Sumatra by William Marsden, published the year after Foersch’s article appeared in the London Magazine).

Both Raffles and Crawfurd covered the topic of upas, and both felt compelled to issue fresh condemnations of Foersch. For Raffles, his account had been a “gross imposition practiced on the people of Europe” (1817: n50). For Crawfurd, subsequent scientific investigations of the poison and its source had demonstrated “the egregious mendacity of the man who propagated the fable respecting it, which has obtained currency in Europe, and the extraordinary credulity of those who listened to his extravagant fiction” (1820: 471). It is striking that both writers drew attention to Foersch in such stridently condemnatory terms. But why, almost four decades after his article first appeared, and with its every detail long since dismissed by serious scientists, did Foersch have such capacity to annoy Raffles and Crawfurd?

Raffles and Crawfurd were exemplars of the model of Orientalist scholarship which was then emerging, and which would remain associated with European colonialism in Asia throughout the coming century. Their activities in Java featured attempts at military, political, and economic control of the island, and a concerted scholarly endeavour to obtain all available information about the place. This amounted to what Farish A. Noor calls a “militarised commercial-academic enterprise” (2016: 75), in which the scholarship could not be divorced from the other activities: an attempt to gain control, politically, militarily, and intellectually. In this it was essential that their scholarly works conveyed an impression of comprehensiveness and accuracy, which Raffles’ work in particular, with its enormous scope and vast array of footnotes and appendices, certainly achieves. “Knowledge is power”, Raffles once wrote, “and in the intercourse between the enlightened and ignorant nations the former must and will be the rulers” (Raffles, 2013/1835: 478).

The postcolonial understanding of “Orientalism”, developed by Edward W. Said in his hugely influential 1978 book of that name, recognizes endeavours such as those of Raffles and imaginative and literary European renderings of “the East” as part of a single overarching discourse, intended to be a “western style for dominating, restructuring, and having authority over the Orient” (2003/1978: 3). But the obvious annoyance prompted in Raffles and Crawfurd by Foersch — whose account they presumably knew still to be enjoying a vigorous afterlife in the popular and literary imagination of Europe at the time their own scholarly works appeared — reveals a clear tension between the militarized commercial–academic urge comprehensively to know, and thus to control, a place such as Java, and the enduring attraction of fantastical stories, the potential credibility of which amongst a European audience relied entirely on their supposed location being unknown and uncontrolled. Judging by the responses to Foersch’s account of the upas, for readers, writers, and theatre-goers in early nineteenth-century Europe, the most attractive vision of Indonesia was that of an unknown, uncontrolled, and mysterious land where the wondrous could still be encountered. This attraction appears to long predate the emergence of Orientalism as a concerted political project, as we can see if we return to Foersch’s fourteenth-century precedent as arch-hoaxer on the topic of Southeast Asia.

In the Travels of Sir John Mandeville, the section on Sumatra and Java plagiarized from Odoric represents the end of the material loosely connected to geographical reality. But Mandeville then continues his itinerary eastwards through a litany of imaginary islands, peopled with all manner of fantastical grotesques. There are men with “heads like dogs”, two-headed geese, cyclops, headless men with faces growing out of their chests, and all manner of other people “of evil customs” (Moseley, 2005/1983). That this section with its mass of individual islands actually tallies with the archipelagic reality of Indonesia beyond Sumatra and Java is probably a coincidence, for Mandeville seems principally to be using this blank space as a dumping ground for all the tall tales that he was unable to accommodate elsewhere. Some of the mythical figures he places here are drawn from earlier medieval “wonder books” such as Wonders of the East; others can be traced as far back as Herodotus.

Mary B. Campbell has noted that in European discourse, eastwards is “historically the direction in which the chiefest among the marvels moved” (1988: 51). As empirically-grounded information of the more westerly regions of what Europe saw as “the East” became more abundant, the perceived location of the most fantastical images tended to retreat from the frontier of the “known”. Thus, the cynocephali had shifted from Libya in Herodotus’s time, via India, to somewhere east of Sumatra in Mandeville’s era. The section of the world in reality occupied by Indonesia could serve as a cache for the fantastic, and that fantastic could remain at least partly credible to Europeans, precisely because it was entirely unknown.

What may not have been apparent to European writers of Mandeville’s era was that in Indonesia the easternmost limits of available imaginative territory had been reached (the Pacific Ocean excepted). From now on, the chiefest among the marvels had nowhere else to go. Given this, the increasing elaboration of the upas tree myth in the centuries that followed can be viewed as a distillation of the attraction that the fantastical clearly continued to hold for the Europeans. As cynocephali and other wonders succumbed to the impact of science, exploration, and increasing colonial involvement in Indonesia, the poison tree continued to grow, becoming ever more fantastical as a last surviving focus of a Mandevillean urge to maintain a wonder-filled unknown.

The vision of the upas recorded by Speelman and Rumphius in the seventeenth century may well in part have been a reflection of Dutch anxieties about the security, and perhaps even the legitimacy, of the colonial project in Indonesia, as Dove and Carpenter (2005) argue. A century and a half later, Raffles and Crawfurd clearly had no such anxieties. What made them anxious was the fact that Foersch’s vision of the upas was still at large and possibly still enjoying some lingering credibility in Europe, undermining their own project of presenting Indonesia as totally known, and thus totally controlled.

Saidian interpretations of European discourses about the “mysterious Orient” may suggest that such mystery and difference was constructed specifically as a pretext for political–military–academic incursions, which would overcome the mystery and so “tame” the previously unknown space. But the anxiety that the upas myth prompted amongst men such as Raffles around the turn of the nineteenth century, suggests that by this stage the function of such a discourse was in fact to preserve a titillating air of mystery rather than to overcome it, to place a certain imaginative Oriental territory beyond the control of demystifying scholars and intruding soldier–politicians.

Obviously, the visions of Indonesia which both Mandeville and Foersch, and Raffles and Crawfurd, sought to convey were equally European constructions. Neither took any account of “the concrete detail of human experience” (Said, 1981: xxxi) amongst the indigenous inhabitants on the ground (though interestingly, if we accept Dove’s argument, it could be said that the locals had a more active input in the construction of the fantastical vision). But the fact that the two visions are directly contradictory is significant.

In the four decades since the publication of Orientalism, Said’s theories have been endlessly debated. Some detractors — such as Robert Irwin (2006; 2013) — have attempted a wholesale refutation on grounds of both minutiae and overall substance, while those who accept and work with Said’s basic premises have often sought to add nuance to the view of “Orientalism” as a monolith. As Ziauddin Sardar notes: “It would be more fruitful to see Orientalism as a whole series of discourses, changing, adapting to historic, scholarly and literary trends, but interconnected by a series of coherent common features” (1999: 55). But how exactly those multiple discourses might intersect and diverge, interact and counteract, perhaps requires more attention. This is where an examination of the history of the upas myth has proven valuable. It reveals a marked divergence within the discourse around a single artefact at the turn of the nineteenth century, as a highly developed commercial–military–academic Orientalism emerged. These divergent strands do not simply demonstrate nuances or even contradictions within Orientalism; they are directly oppositional, acting against one another, with the one (the imaginative legacy of Foersch) a direct challenge to the ambitions of the agents of the other (the knowledge-as-power project of Raffles).

Attempting to identify other examples of such oppositional strands within Orientalism, around specific myths, artefacts, or imaginative territories, would be a valuable area of further research. Charting the continuing trajectories of those diverging strands and the tensions between them through the nineteenth century and beyond would also be a worthy undertaking, for the traces of those tensions can arguably still be discerned today, in the attempts at Western military and economic — and, for that matter, academic — interventions in the modern “Orient” on the one hand, and on the other in the eternal quest of travel writers, documentary-makers, and tourists for surviving fragments of “mystery” in places such as Indonesia.