Unwritable Dwellings/Unsettled Texts:
Robert Louis Stevenson’s In the South Seas and the Vailima House
On the 6 April, 1889, during a six month stay in Hawai'i, Robert Louis Stevenson provided the following description of his living circumstances, in a letter to Adelaide Boodle:
[t]he buildings stand in three groups by the edge of the beach. [.] The first is a small house [...] All about the walls [are] our South Sea curiosities, war clubs, idols, pearl shells, stone axes, etc; [...] The next group of buildings is ramshackle and quite dark; [...] you go in, and find photography, tubs of water, negatives steeping, a tap, and a chair and an ink-bottle, [...] right opposite the studio door you have observed a third little house [...] herein [you find] one Squire busy writing to yourself [...] The house is not bare; it has been inhabited by Kanakas and (you know what children they are!) the bare wood walls are pasted over with pages from the Graphic, Harper's Weekly, etc. [...]; on the panels [...] a sheet of paper is pinned up, and covered with writing. I cull a few plums.
"A duck-hammock for each person.
A patent organ like the commandant's at Tai-o-hae.
Cheap and bad cigars for presents.
Permanganate of potass.
Liniment for the head and sulphur.
What do you think this is? Simply life in the South Seas foreshortened. 
The buildings themselves, by their contents and their construction, stand in metaphorical relation to the extent to which the Pacific had been textualized in 1889. Stevenson's shorthand Pacific is a tableau of textualizing activity. The first house, decorated like a museum with various South Sea curios and paraphernalia, represents a constant flow of recontextualizations, of object, language, and people, and refers to the way in which modes of display and the performance of culture can themselves be considered a kind of textualization. The second house, which contains photographic and writing equipment, the apparatus of representation, signifies the prerogative of the writer to represent, to produce a record of what he has observed.
It is the third and last house, however, that yields the most intriguing instance of text-building. With its walls covered by a great many cuttings and advertisements from magazines, and lists of articles of trade, it most fully represents the written Pacific, in which the trade in words written about the Pacific and the encroachment of material culture overlap. The house is, in effect, a built text. It has experienced multiple over-writings. The reference to 'life in the South Seas foreshortened' carries a multiple resonance; it bears a proleptic reference to Stevenson's death in Samoa in 1894; the corrupted condition of the house itself also resonates with what Patrick Brantlinger has called an 'extinction discourse' which, from the late eighteenth century onwards, preoccupied much of the writing about other, supposedly primitive cultures, and which described, 'the inevitable disappearance of some or all primitive races'. The perception that, in accordance with the precepts of social Darwinism, a process of natural selection was acting among many of the island cultures of the Pacific, was common currency by the time of Stevenson's visit. In addition, each house is a reference to the difficulties of remaining settled in the Pacific, against which the final resonance, the desire to annotate the Pacific for the sake of posterity, can be read.
The purpose of this article is to consider Stevenson's writing about his Pacific experiences, and to explore the tensions inherent between what appear to be commensurate urges to write and to settle. In doing so, it will look at the encounters described by Stevenson as mediated by dual modes of production, that is, building and writing, that are similarly engaged in the production and assertion of presence. In this sense, the article considers the way in which buildings can be made to signify, and the way texts are constructed. Both function to demonstrate the fact that the writer/builder has occupied a place, has genuinely 'been there'.  They are monuments of presence, indicating the ability of the authors to represent themselves.
The practice of building a house to mark or commemorate presence has a traceable precedence in the post-exploration Pacific. In 1837, on Vaitahu in the Marquesas Islands, the missionary John Stallworthy witnessed the departure of his fellow missionaries, the Rodgersons, and was left 'alone on the beach. To celebrate his lonely survival, Stallworthy built a new house, spending weeks on the construction of its walls and paths and fence'. It is significant that Stallworthy's house is built to signify a singular presence-he is the sole occupant of the beach. This kind of singular, settling presence implicitly refers to a corresponding absence, in this case both of other settlers and, more ominously, of any indigenous presence. The assertion of presence is thus associated with the absolute control of context.
To negotiate the convergence of building and writing, this article will invoke the notion of the 'text-builder', from Alton Becker's call for a '"[...] new philologist-a specialist in contextual relations-in all areas of knowledge in which text-building.is a central activity"'.  The term 'text-builder' expresses the convergence of writing and building, and encapsulates the convergence of understanding culture as space and culture as knowledge: both concepts are occupied by the text-builder. Thus, through what it tells us about presence, building and settling (or dwelling) are intertwined, problematically, with writing. As the practice of dwelling is reduced to writing, it implies the centrality of certain activities to the performance of text-building: the necessary ability to exert control over context; and also the possibility of resistance: that others apart from the text-builder will engage in simultaneous textualizing practises that contradict or actively oppose the practices of the text-builder.
Rather than as a spur to his imagination, and to the production of further tales of adventure and romance such as Kidnapped and Treasure Island, Stevenson imagined his Pacific travels as a period of research out of which he would compose a comprehensive survey of the cultures he had encountered. In a letter to Sidney Colvin, written as he approached Samoa on 2 December, 1889, Stevenson described his grand design for his book, next to which he predicted there would be 'few better books now extant on this globe', and which he proposed to call The South Seas: 'it is a rather large title, but not many people have seen more of them than I; perhaps no-one: certainly no-one capable of using the material'.  As the title suggests, it was intended to be a definitive work, divided into eight sections, dealing first with general issues, ' [o]f schooners, islands, and maroons',  and following with a section on each of the Pacific regions Stevenson had visited: The Marquesas; the Tuamotu Archipelago; Tahiti; Hawai'i; The Gilberts; and Samoa, 'which I have not yet reached'.  As Vanessa Smith has asserted, 'Stevenson's travels were envisaged as a book'. This forms one of the principal starting points of my own argument: namely, that his intention from the beginning was to textualize his experiences. What is crucial to my argument here, is the manner in which Stevenson foregrounds his role in collating and synthesising the material he has gathered; that is, in constructing the text. The 'I', his own presence, is central, and essential, and I want therefore to argue that what Stevenson's South Sea book was intended to represent was a kind of monument to his role as author-to his ability to gain access to other cultures and experience an encounter that was unlimited in depth, but sacrificed nothing in the exchange.
This project was characterised rather disdainfully by Fanny Stevenson as 'a sort of scientific and historical impersonal thing, comparing the different languages (of which he knows nothing, really) and the different peoples [.] suppose Herman Melville had given us his theories as to the Polynesian language and the probable good or evil results of the missionary influence instead of Omoo and Typee'.  Contrary to both Stevenson's aspirations and his wife's apprehensions, however, scientific travel writing in the nineteenth century was not a wholly neutral and objective exercise; as Nigel Leask has observed, achieving a balance 'between itinerary narrative and scientific disquisitions [.] was seen as a major problem for eighteenth- and nineteenth-century scientific travel writing'.  Stevenson arrived in the Pacific in the wake of a wave of literature about other cultures in which the scientific and the performative converged in the process of transcribing experience. European travellers who preceded Stevenson to the Pacific were in the main missionaries and sailors who had jumped ship. Several of the latter produced narrative accounts of their time spent among the indigenous islanders, describing their immersion in the local culture as a mixture of personal narrative and observation of the strange customs and practices they witnessed. It was this kind of 'beachcomber' narrative, such as those produced by William Dampier, Jean Cabri and George Vason,  which provided Herman Melville with a model for Typee (1846), a text that so gripped Stevenson's imagination. 
The habit of texts, such as those by Cabri and Dampier, of 'crossing' different narrative modes, mixing accounts of personal adventure with descriptions of hitherto undescribed peoples, demonstrates that the practice of writing encounter in the Pacific was not carried out in isolation from the subject. Rather, 'the transformation of the trajectory of the journey into that of a narrative [...] involved other kinds of cultural and linguistic crossings and contacts'.  These crossings-of cultural and physical boundaries, locate the role of the writer of encounter, such as Stevenson was planning to undertake, in a mode of reciprocal exchange.
The requirement to engage in reciprocal exchange had been recognised long before Stevenson set out his plans for his book of the South Seas . In her 1838 treatise on ethnographic practice, How to Observe , Harriet Martineau advocated that,
the observer must have sympathy; and his sympathy must be untrammelled and unreserved. [.] Nothing was ever more true than that "as face answers to face in water, so is the heart of man". To the traveller there are two meanings in this wise saying, both worthy of his best attention. It means that the action of the heart will meet a corresponding action, and that the nature of the heart will meet a corresponding nature. 
For Martineau, the traveller's sympathy is vital if the other is to be called into a reciprocal exchange; the performance, therefore, of his own presence determines the outcome of the encounter. As Mary Louise Pratt has argued, in relation to another piece of travel writing by a Scottish author, Mungo Park 's Travels in the Interior Districts of Africa, of 1799, '[r]eciprocity, [...] is the dynamic that above all organises Park's human-centred, interactive narrative. It is present sometimes as a reality achieved, but always as a goal of desire, a value'.  This was a value shared by Stevenson; in a letter to Sidney Colvin, sent from Honolulu on 2 April, 1889, Stevenson recommended two Tahitians who were to visit Europe, and advised Colvin on how to meet them: 'A hint: if you want to put these people at their ease, talk of our Middle Ages, Highland Clans, etc.- then they no longer feel foreign -this has been my highly successful principle'.  However, as Stevenson was to discover, '[w]ritings from the beach [...] were subject to the risks of encounter'. 
Greg Dening has described one such example; Peter Heywood, one of the Bounty mutineers, responded to overtures from his Tahitian hosts that he be tattooed with a great readiness. In a letter to his mother describing his reasons for becoming tattooed he states, 'it was my constant endeavour to acquiesce in any little custom which I thought would be agreeable to them, though painful in the process, provided I gained by it their friendship and esteem'.  Heywood discovered himself living amongst an excessively amenable host, whose generosity inspired him to make a gesture in kind. As a demonstration of reciprocity, Heywood thus 'let them make him like themselves,' discovering 'joy in being possessed', as Dening has put it. As the example of Heywood demonstrates, the 'crossings' that followed the initial crossing of the beach involved a degree of partiality-that is, a sacrifice of presence. This corresponds with Greg Dening's definition of the beach as a site of loss, the space that is initially crossed in instances of cultural interaction. He famously described the exchange that occurs as parties negotiate the beach: 'things come across the beach partially, without their fuller meaning'.  Those that cross the beach represent their own culture in ways that are partial or incomplete, and their experience of exchange allows only a partial immersion in the host culture. In contrast to Dening's notion of 'partial', I will refer to Stevenson's ideal imagined encounter as a 'complete' encounter, in which the assertion of presence through the text is not infringed upon or compromised-the phrase is intended to suggest an encounter which does not concede to the partial condition of the beach. One of the principle tensions in In the South Seas, the book which Stevenson did eventually produce from his Pacific travels, is that between his desire for a fully reciprocal encounter-to properly cross the beach and engage with his subject-and the anxiety produced by the partial condition that inevitably followed. Where this partial condition threatened the centrality and authority of Stevenson's authorial performance, textual crisis ensued.
In line with the concern for building and writing, this article will examine Stevenson's production of two monuments to his own presence in the Pacific; In the South Seas, which recorded his experiences of touring Polynesia, in the yacht Casco from June 1888 to January 1889, and in the schooner Equator for June 1889 to December of the same year; and also the house he built near Apia, in Samoa, having resolved to settle there in September 1890. This article will consider three instances in which Stevenson's encounters are mediated by or centred around a building or buildings: in the Marquesas, where he investigated indigenous dwelling places; in the Gilbert Islands, where the King of Apemama built a temporary house for Stevenson; and in Samoa, where he elected to settle and build himself a house. In doing so, it will look to consider three aspects of tension relating to presence in In the South Seas: that is, the anxiety of the unwritable subject; the destabilizing influence of a proficient indigenous textualizing presence; and the extent to which the pursuit of a 'complete' encounter, which does not admit the partial nature of the beach, is frustrated by building and writing.
In the South Seas opens with Stevenson's decision, in 1890, to make his home on Samoa: 'the axes of my black boys are already clearing the foundations of my future house; and I must learn to address readers from the uttermost parts of the sea'.  The projects of narration and settlement are linked in the creation of a text; here, then, we find at the outset of Stevenson's text a dual undertaking: to assert presence through building, and to write the Pacific. Vanessa Smith has described the problem of writing about Pacific encounters experienced by Stevenson, by drawing a contrast between Walter Benjamin's storyteller and Claude Levi-Strauss's bricoleur, which also is instructive in considering the adjacent concern of building. In his essay 'The Storyteller', Benjamin describes storytelling as 'an artisan form of communication'.  He associates the storyteller with the medieval craftsman; as Vanessa Smith has said, Benjamin 'links the production of narrative with a context of licensed workmanship'.  It therefore describes a conjunction of building and writing anticipated by Stevenson; the construction of a narrative, or of a dwelling structure, is ordered, controlled, and fitted to a particular design, by a single figure. In addition, Benjamin also describes storytelling as a bridge for contact: it is 'the ability to exchange experiences. [...] The storyteller takes what he tells from experience-his own and that reported by others. And he in turn makes it the experience of those who are listening to his tale'.  Located at the centre of reciprocal experience, the storyteller does not concede any remainder, or element of the partial, in his craft.
Stevenson similarly aspired to avoiding the partial in his constructions. In a letter to Charles Baxter, dated 6 September 1888 while the Casco was sailing to the Tuamotu archipelago , Stevenson goes so far as to suggest that the experience of the beach goes hand in hand with a compulsion to write:
Excuse me if I write little: when I am at sea, it gives me a headache; when I am in port, I have my diary crying, "Give, give". I shall have a fine book of travels, I feel sure: and will tell you more of the South Seas after very few months than any other writer has done-except perhaps Herman Melville. 
This passage is of interest as it provides an insight into the manner in which Stevenson conceived of the space of the beach at the outset of his travels. Rather than the space of loss described by Dening, it is when he reaches the beach that Stevenson feels the pressure to transcribe his experiences-that is, it was on the beach he most fully engaged in his role as author. His confidence is high that in his authorial role he will achieve a kind of narrative such as that described by Benjamin, one that will resist the partial nature of encounter.
However, in contrast to Benjamin's model of artisanal text production, Smith has posited the beachcomber narratives that preceded Stevenson, 'producing objects and narratives that are makeshift rather than crafted, from materials to hand at the periphery of empire',  as the archetype of the written encounter in the Pacific. In How to Observe, Martineau has also conceded the inevitably partial nature of the encounter:
[The traveller] does not command the expanse; he is furnished, at best, with no more than a sample of the people; and whether they be indeed a sample, must remain a conjecture which he has no means of verifying. He converses, more or less, with, perhaps, one man in ten thousand of those he sees; and of those few with whom he converses, no two are alike in powers and in training, or perfectly agree in their views [.]. 
The partial nature of narratives of encounter constantly undermines Stevenson's efforts to assert an authentic presence through his role as author. It is at this point that the crisis of In the South Seas is located: in the possibility of authentic presence set against the distance imposed by transcription. Significantly, Benjamin imagined his storyteller operating in an oral mode, in contrast to which, '[t]he novelist isolates himself'.  Reciprocal exchange is an oral transaction, to which the practice of writing and transcription is an impediment. David Richards has noted the relationship, applied by Derrida, between speech and literacy, and social presence: 'For Derrida, the "ethic of speech is the delusion of presence mastered", "the image of a community present to itself, without difference. Writing is [...] a condition of social inauthenticity"'.  Despite his privileging of authorial authority, reciprocity was important to Stevenson in his encounters with the indigenous population, as a means of overcoming the problems of difference. For Stevenson, the desire for reciprocity is oriented around a desire to achieve some form of authentic (spoken) communication. Yet his function as a writer stifles any possibility of this, reinforcing the difference between Stevenson and his subject; rather than reciprocal, dialogic encounter, writing introduces a set of codifying, representational power-relations that centre on the performance of presence. Whereas his aspirations are similar to those of Benjamin's storyteller, his participation in the technology of writing critically undermines his endeavours. Furthermore, where he meets resistance to his efforts to write encounter, Stevenson's text struggles against the threat of the effacement of authorial presence; the anxiety that, if what he encounters is unwritable, then he will be made to occupy a position of absence, continually disturbs Stevenson.
The problem of unwritability confronts Stevenson in his first pacific landfall, at Nuku Hiva in the Marquesas. As he approaches the shore he anticipates his forthcoming encounter in terms of a complete immersion: 'the anchor plunged. It was a small sound, a great event; my soul went down with these mooring whence no windlass may extract it nor any diver fish it up'.  There is also, however, a hint of anxiety; 'it was with something perhaps of the same anxious pleasure as thrilled the bosom of discoverers that we drew near these problematic shores'.  The unfamiliarity of the 'problematic shores' arouses his anxiety about how to write what he sees:
The suffusion of vague hues deceived the eye; the shadows of clouds were confounded with the articulations of the mountain; and the isle and its unsubstantial canopy rose and shimmered before us like a single mass. [.] Somewhere, in that pale phantasmagoria of cliff and cloud, our haven lay concealed. 
His first encounter with the Marquesan inhabitants compounds this; they 'swarmed' the Casco, offering 'no word of welcome', such that Stevenson fears they are 'beyond the reach of articulate communication'.  He describes the experience of writing his journal in his cabin as his Marquesan guests 'squatted cross-legged on the floor, and regarded me in silence with embarrassing eyes'.  Edmond has described this moment as 'a classic example of the observer observed',  and drawn a comparison with the incident in Typee where Tommo first encounters the Typee: 'One of them in particular, who appeared to be the highest in rank, placed himself directly facing me; looking at me with a rigidity of aspect under which I absolutely quailed. He never once opened his lips, but maintained his severe expression of countenance, without turning his face aside for a single moment. Never before had I been subjected to so strange and steady a glance; it revealed nothing of the mind of the savage, but it appeared to be reading my own'.  In terms of the performance of presence, the text-based logic of encounter should locate the Marquesan silence as absence; silence becomes proof of inarticulacy, rather than a reluctance to speak, and the Marquesans are thus figured as inauthentic, lacking the facility that will provide Stevenson with proof that he has encountered the authentic, and which he can textualize. Yet this is subverted, as the author is subjected to a reversal of gaze. Frank McLynn has explained Stevenson's mistake lay in his ignorance of local custom:
At this stage RLS knew nothing about Polynesian etiquette, which required a host to sit as a gesture of reconciliation; to remain standing was an insult. Stevenson unwittingly gave offence by continuing to write instead of sitting on the floor with his guests, and did not, as expected, clap his hands as a signal for the Marquesans to initiate a welcoming ceremony in song and dance. 
For all his extensive research and desire to engage with the indigenous population, writing remains the obstacle. This experience reinforces his anxiety that the subject will prove unwritable: 'my new fellow-creatures sat before me dumb like images. Methought, in my travels, all human relation was to be excluded; and when I returned home (for in those days I still projected my return) I should have but dipped into a picture-book without a text'. 
In order to write, therefore, it becomes apparent that Stevenson must limit the alienation his has so far encountered, particularly in terms of language. He declares, '[t]he impediment of tongues was one I particularly over-estimated',  and offers proof in a comparative table of various Polynesian equivalents of house and love:
House Love Tahitian FARE AROHA New Zealand WHARE Samoan FALE TALOFA Manihiki FALE ALOHA Hawai'ian HALE ALOHA Marquesan HA'E KAOHE
Stevenson is discovering a means to cope with the problematics of textualizing his subject; by suggesting a degree of similarity between different languages, he hopes to eliminate unwritable diversity. To further manage the unfamiliar, he refers to a context of comparison:
It was perhaps more important that I had enjoyed in my youth some knowledge of our Scots folk of the Highlands and the Islands . [.] points of similarity between a South Sea people and some of my own folk at home ran in my head in the islands; and [.] [i]t was so that I hit upon a means of communication which I recommend to travellers. When I desired any detail of savage custom, or of superstitious belief, I cast back in the story of my fathers, and fished for what I wanted with some trait of equal barbarism: Michael Scott, Lord Derwenter's head, the second-sight, the Water Kelpie,--each of these I have found to be a killing bait. The black bull's head of Stirling procured me the legend of Rahero; and what I knew of the Cluny Macphersons, or the Appin Stewarts, enabled me to learn, and helped me to understand, about the Tevas of Tahiti. The native was no longer ashamed, his sense of kinship grew warmer, and his lips were opened. 
Reference to Scots culture is intended to engender a sense of intimacy between Stevenson and his indigenous subject, by bringing the unfamiliar within the sphere of the familiar. Intimacy is necessary for the mode of sharing that constitutes a reciprocal encounter. He notes a similarity between Scots and Polynesian pronunciation-'common to both tongues [is] the trick of dropping medial consonants' -and it is significant that, as with the language table, Stevenson focuses his approach to comparison through the problem of how to write the Marquesan houses. Gaston Bachelard has described the house as a 'domain of intimacy', without which 'man would be a dispersed being'.  Stevenson's references to Scots culture, and especially Scots houses, are therefore a strategy to counter-act the threatened dispersalizing consequences, upon his text, of the unwritable.
Familiarity, or intimacy, with a Highland turf house allows Stevenson to 'read' its appearance:
In Scotland wood is rare, and with materials so rude as turf and stone the very hope of neatness is excluded. And in Scotland it is cold. Shelter and a hearth are needs so pressing that a man looks not beyond; [...] he has not an appetite for more. Or if for something more, then something higher; a fine school of poetry and song arose in these rough shelters. 
In contrast, he says of the Marquesan paepae-hae: 'The same word, as we have seen, represents in many tongues of Polynesia , with scarce a shade of difference, the abode of man. But although the word be the same, the structure itself continually varies'.  Because he lacks intimacy with Marquesan structures, they remain as unwritable as their inhabitants. Rather than provide a basis for reciprocal narrative, Stevenson's references to his own culture merely counterbalance the sense of displacement he feels when confronted with what he cannot write, and which resists incorporation into his text. The implication is that the unreadable paepae-hae will therefore resist his efforts to write it.
The comparison with Highland cultures also, however, provokes Stevenson's anxiety about his capacity to write his encounter. Stevenson was well aware of what T.M. Devine has called 'The Disintegration of Clanship'  during the eighteenth century: 'Not much beyond a century has passed since they were in the same convulsive and transitory state as the Marquesas of to-day'.  By invoking a deteriorated culture as the basis of his efforts to engender a reciprocal relationship with the Marquesas islanders, he demonstrates an anxious undercurrent of feeling that his subject is in a similar process of terminal dissolution, rapidly disappearing before his eyes, and before he can write it up properly. Here we can clearly see the influence of social Darwinism, and Adam Smith's stadial theory of culture, on Stevenson's text.  He describes his impression of a prevailing sense of fatalism in Marquesan culture: 'The thought of death [...] is uppermost in the mind of the Marquesan [...] death reaps them with both hands',  and the people are to his eyes visibly waning from European disease. Stevenson sees everywhere the threat of deracination, such as an encounter with a young Marquesan mother:
[...] she began with smiling sadness, and looking on me out of melancholy eyes, to lament the decease of her own people. "Ici pas de Kanaques ", said she; and taking the baby from her breast, she held it out to me with both her hands. "Tenez- a little baby like this; then dead. All the Kanaques die. Then no more". The smile, and this instancing by the girl-mother of her own tiny flesh and blood affected me strangely; they spoke of so tranquil a despair. 
The scene does indeed affect Stevenson strangely. Edmond has described this as evidence of Stevenson's pronounced 'use of the pathetic fallacy', superimposing his own waning health upon his environment: 'a dying narrator confronts a Polynesian world represented in similarly terminal decline'. What is most interesting here is the dying narrator, and it is my contention that in Stevenson's rendering of the scene, it is a short step from the dying narrator and dying subject, to the death of narrative.
Stevenson's imagination translates the Marquesan predicament to a European context: 'in a perspective of centuries I saw their case as ours, death coming like a tide, and the day already numbered when there should be no more Beretani, and no more of any race whatever, and (what oddly touched me) no more literary works and no more readers'.  He goes on to describe the devastating consequences of an outbreak of smallpox among the Hapaa: 'in less than a year two survivors, a man and a woman, fled from that newly-created solitude. A similar Adam and Eve may some day wither among new races, the tragic residue of Britain '.  He imagines a reversal of the creation narrative, thus figuring the end of narration.
Confronted with the depopulation of indigenous society, Stevenson is also confronted with anxiety that he will be unable to dwell. He has hinted that he considers dwelling a tenuous state, one that might endure for a time but will always end in departure. Although when describing his decision to build his Samoan house, he declares that, '[f]ew men who come to the islands leave them; they grow grey where they alighted',  he simultaneously, when recounting his first Marquesan landfall, retains a sense of this partial condition, which he articulates by recourse to indigenous culture: '"The coral waxes, the palm grows, but man departs", says the sad Tahitian proverb; but they are all three, so long as they endure, co-haunters of the beach'. Significantly, whereas he appears to be appropriating a native voice in order to describe the incompleteness of exchange, he is actually borrowing a phrase with a complex, intertextual history. The proverb in question had already been included in, amongst other sources, William Ellis's Polynesian Researches (1816) and Herman Melville's Omoo (1847), both of which Stevenson was aware of. Thus, in quoting it himself he is textualizing his own account, in such a way that demonstrates the difficulty of settling with conviction. His settler's confidence is undermined by the new visitor's sense that man's dwelling is inevitably temporary and limited in depth. Stevenson refers to this proverb just before his account of the moment of arrival; its inclusion and his brief exposition introduce an undertone of anxiety to his first recollections of the Marquesans; his insight into the partial condition is intimately informed by a sensibility of his own potentially inauthentic condition.
The perception of imminent depopulation increases the urgency of Stevenson's textualizing project; but he realises that if the subject proves to be unwritable then the culture and the text will die also. What most affects him about the possible demise of the Marquesans is the end of their potential writability. The loss of his subject would create a debilitating absence at the centre of Stevenson's 'travels envisaged as a book'. If the subject is permanently unwritable there exists no further possibility of dwelling. Stevenson even makes an association between depopulation of the island and the subsequent decay of their 'empty paepaes. When a native habitation is deserted, the superstructure-pandanus thatch, wattle, unstable tropical timber-speedily rots, and is speedily scattered by the wind. Only the stones of the terrace endure; nor can any ruin, cairn, or standing stone, or vitrified fort present a more stern appearance of antiquity'.  This passage suggests the relief of the land when those who dwell upon it leave, and despite Stevenson's rationalising, social Darwinist assertion that depopulation is due to the rapid pace of social change, brought on by exposure to European culture-'change of habit is bloodier than a bombardment' -there is a sense here that he shares in the Marquesans' perceived fatalism. This is the fear against which Stevenson's twin urges to write and to settle constantly struggle, and which is compounded by the paradoxical condition of the writer. Writing, for the writer, is his condition of being, and is both an inducement and a barrier to integration with the host culture. His presence is a consequence of his desire to write, yet it also prevents him from fully integrating, serving as a constant reminder of his outsider status. Inability to integrate fully means an inability to settle; the partial condition of writing encounter is to alienate the writer from his material, thus writing, or at least textualizing, alienates dwelling.
Stevenson landed at Apemama, in the Gilbert Islands , on 1 September 1889, where he met an islander as complicit in textualizing strategies as himself-the Gilbertese king, Tembinok', whose text-creating facility was demonstrated by his voracious collecting. Stevenson wrote:
He is greedy of things new and foreign. House after house, chest after chest, in the palace precinct, is already crammed with clocks, musical boxes, blue spectacles, umbrellas, knitted waistcoats, bolts of stuff, tools, rifles, fowling-pieces, medicines, European foods, sewing-machines, and, what is more extraordinary, stoves. 
Susan Stewart has described the process of collecting as, 'the reframing of objects within a world of attention and manipulation of context': 
[I]ts function is not the restoration of context of origin but rather the creation of a new context, a context standing in a metaphorical, rather than a contiguous, relation to the world of everyday life. Yet unlike many forms of art, the collection is not representational. The collection presents a hermetic world: to have a representative collection is to have both the minimum and the complete number of elements necessary for an autonomous world-a world which is both full and singular, which has banished repetition and achieved authority. 
The collector engages in similar processes as the text-builder, appropriating and recontextualizing his materials, pursuing an ideal of representation that is unattainable. Neither the collector nor the text-builder can have absolute authority over their subject, making it a necessary condition of possession in both instances that the subject be removed from its original context and placed in a new, boundaried context, which is available for control.
Stevenson's response to this fabulous assemblage of mundane artefacts is to catalogue it. However, this is not an encounter with the unwritable, as in the Marquesas, but with the already-written. Under Tembinok', Apemama is a highly textualized environment. Speaking in relation to the toy, Stewart has described the relationship between the collection and the text: 'The toy is the physical embodiment of the fiction: it is a device for fantasy, a point of beginning for narrative. The toy opens an interior world [...] To toy with something is to manipulate it, to try it out within sets of contexts, none of which is determinative'.  Tembinok's collection, secured in a series of specially constructed houses, 'toys' with meaning and context in an open-ended, narrative process, and can thus be figured as a text. Tembinok's collection is a response to the recontextualization of the Pacific. Stevenson notes that Apemama is the only island remaining on whose beaches a white man may not embark as he pleases, whereas '[t]he white man is everywhere else, building his houses. '  Apemama, by contrast, 'is a closed island, lying there in the sea with closed doors'.  Tembinok's hermetic collection is an expression of his ability to impose his authority on the context in which material culture encroached upon his islands. Stevenson notes that Tembinok' made him swear that he would, 'give his [the king's] subjects no liquor or money (both of which they are forbidden to possess) and no tobacco, which they were to receive only from the royal hand'.  As O.H.K. Spate has said, the influx of nineteenth-century materialism made over the Pacific as 'a European artefact'.  By appropriating the artefacts of materialism and manipulating their context-arranging them within his collection and acting as the sole authority determining what may and may not be received and exchanged in his kingdom-Tembinok' demonstrates his mastery of context. As Spate has said, '"The Pacific" is a European artefact: not so the peoples who inhabit its shores and islands-rather, they are artificers'. 
Tembinok' further demonstrates his mastery of context in the terms by which he allows the Stevenson party to dwell on the island: 'We were to choose a site, and the king should there build us a town. His people should work for us, but the king only was to give them orders. [...][H]e was to come to meals with us when so inclined'.  These terms include a prohibition on interfering with island trade, which the king's sole preserve; and a reference to his rigid control of discourse. Stevenson describes him as 'an admirer of silence in the island; he broods over it like a great ear; has spies who report daily; and had rather his subjects sang than talked'.  As the Equator is unloaded, Stevenson describes the curious scene of the construction of their compound: 'It was singular to spy, far off through the coco stems, the silent oncoming of the maniap' , at first (it seemed) swimming spontaneously in the air-but on a nearer view betraying under the eaves many score of moving naked legs'.  Tembinok', rather than build a settlement from raw materials, has confiscated the houses (maniap's) of his subjects and ordered them to be transported to what was to become Equator Town, recontextualizing the already-made. Stevenson later encounters one of the dispossessed population; while walking on the beach he realises that, '[n]ear by there had been a house before our coming, which was now transported and figured for the moment in Equator Town'. 
Tembinok' is able to dictate context; thus he is able to recontextualize Stevenson to suit his purposes. He builds a fence around the perimeter of the compound, and places the whole under tapu, manufacturing a segregated space in which foreigners can be observed free from the intrusion of everyday island life: 'we sat in the midst in a transparent privacy, seeing, seen, but unapproachable, like bees in a glass hive'.  Stevenson is 'the subject of a constant study'  by the king. His purpose is to collect information pertinent to his rule:
The quality of his white visitors puzzled and concerned him; he would bring up name after name, and ask if its bearer were a "big chiep", or even a "chiep" at all [...] He was struck to learn that our classes were distinguishable by their speech, and that certain words (for instance) were tapu on the quarter-deck of a man-o-war; and he begged in consequence that we should watch and correct him on the point. [...] We were showing him the magic-lantern; a slide of Windsor castle was put in, and I told him there was the "outch" of Victoreea. "How many pathom he high?" he asked, and I was dumb before him. It was the builder, the indefatigable architect of palaces, that spoke; collector though he was, he did not collect useless information; and all his questions had a purpose. 
It is significant that Tembinok's motives leave Stevenson speechless; within the text-based logic of presence, he is therefore fixed in a category of absence. McLynn notes that, as the Stevensons departed Apemama, the king 'told him how much he had learned about the world from white men, whom he called his "books"';  thus in the confrontation with a textualizing indigenous presence, Stevenson himself is made a text.
Stevenson's final stop was in Samoa, in December 1889, where he made his decision to settle, and commenced the difficult task of revising his journals into a publishable form. Rod Edmond has described Stevenson's Pacific output as, 'the product of complex transactions between an already extensively textualized Pacific and his own distinctive experience as a traveller-writer who was to become a settler'.  This complex mediation of genre and vocation reflects the difficult formal history of what became In the South Seas. Whilst staying at Saranac with his family in the winter of 1887-1888, Stevenson was approached by Samuel McClure to produce a series of fifty articles, or letters, describing the journey he was to take in the South Pacific. For these he would receive '£20 per letter in England , and a further $200 each in the USA',  which would provide the necessary finance to undertake the voyage. However, Neil Rennie has noted that the journal kept by Stevenson, from which he would extract material for his letters, exhibits a remarkable similarity to the published text of In the South Seas, suggesting, 'Stevenson was not so much writing a journal as drafting an account for publication'.  He imagined a book on a large scale, encompassing the rich and diverse heritages of language, myth and folklore from the whole of the South Seas. It was in Samoa, where he was also engaged in building himself a house, that Stevenson set about building his text from the various materials he had accumulated during his travels.
On 2 December, 1889, Stevenson wrote to Colvin of his design for his book, consisting of great inter-island comparisons of the like that had not previously been attempted. He finally reached Samoa on 7 December, and by January had decided upon a site near to the capital Apia on which to build his permanent Pacific residence. However, by this time his grand design had already begun to fragment, as a letter to Baxter written on 28 December illustrates: Stevenson had begun to envisage a separate volume dealing with Samoan history and politics, which became A Footnote to History, in addition to his 'big South Sea Book: a devil of a big one, and full of the finest sport'. 
Already he was exhibiting uncertainty regarding how his material should be used, although his confidence that he would be capable of using it well was undiminished. But his grand design contradicted his original commission for a series of letters based on his experiences in the Pacific. He was faced with several different pressures acting at once upon the material he had gathered between 1888 and 1889: his ambition to create a definitive work, and the demands of publishers and family, that narratives of personal experience should not be subordinate to geological inquiry and etymological theorising. Stevenson's cruise on the Janet Nicholl in the middle of 1890 was 'chiefly occupied with revising earlier material, rather than with [.] day-to-day accounts';  apart from a series of four letters describing his visit to Penrhyn, none of which were included in In the South Seas,  Stevenson laid down his notebook in order to devote his energies to the organisation of his material. Yet these energies were divided between the construction of his book and the letters. By 19 July, 1890, he was able to send McClure, as promised, an instalment of fifteen finished letters worked up from the Marquesas material, but with the following qualification: 'what you are to receive is not so much a certain number of letters, as a certain number of chapters in my book. The two things are identical but not co-terminous'. He left it to McClure 'to choose out of the one what is most suitable for the other,' and promised the remaining letters by 'at the latest before Easter '91.'  Although McClure secured syndication of the letters with the New York Sun, and Black and White in London, his response confirmed that Stevenson was facing a crisis of form, when he complained in early 1891, 'the letters did not come as letters are suppose to come. They were not a correspondence from the South Seas, they were not dated and .in no way...fulfil[ed] the definition of the word "letter" as used in newspaper correspondence'. These fifteen chapters were privately printed in November, in a copyright edition, under the title The South Seas: A Record of Three Cruises.  However, this partial printing, which included only the Marquesas material, emphasised the partial nature of Stevenson's written encounter.
By this time Stevenson was facing the prospect that his material was going to slip away from him, and it is significant that his increasingly apparent difficulty in marshalling his material coincided with the gradual transformation of his sense of his presence in the Pacific: first as a visitor, but later as a settler. Roger Swearingen notes that it was between 31 August and 25 October, 1889, while staying at Apemama, that Stevenson first 'began working on The South Seas as distinct from his day-to-day accounts',  and this coincided with his initial decision to remain in the Pacific. He arrived with his family on the estate he was to call Vailima in September 1890, and it was in the midst of the clearing and laying down of his settlement that he realised the problem handling such 'a mass of stuff' as he had gathered on his first trips was 'architectural-creative-[how] to get this stuff jointed and moving',  as he explained in a letter to Colvin in November. The crisis of form was compounded by the fact that his materials would not cohere, and it is evident from his correspondence that, as he sought simultaneously to build his Samoan house and build his text of the South Seas, that the satisfaction he found in the former was increasingly lacking in the latter. He complained to Colvin in March 1891, that:
Today I have not weeded, I have written instead from six till eleven, from twelve till two [.]; a damned Letter is written for the third time; I dread to read it, for I dare not give it a fourth chance-unless it be very bad indeed. [.] On the whole I prefer the massacre of weeds. 
The intransigence of the South Sea material contrasted greatly with the marvellous construction of the 'new two-storied Vailima residence':
[.] it was an imposing sight. [.] Painted peacock-blue, with a red-iron roof, Vailima was equipped with shutters against gales and gauze-screened windows to keep out mosquitoes, cockchafes, moths, beetles and other insects. There were no doors, only hanging mats or curtains, but a verandah ran along the whole north side of the house, both upstairs and downstairs. [.] [Stevenson's] study was a small room off the library-actually an enclosed bit of the north-west corner of the upper verandah. 
In another letter to Colvin, in late April 1891, Stevenson was able to declare, 'Our old house is now demolished; it is to be rebuilt on a new site'.  The Stevensons had originally occupied a small cottage on the Vailima site, built for them by H. J. Moors, a local trader, while their larger house was being built. Thus he experienced far greater satisfaction in the redrafting of his settlement than he did of his text.
By the end of the year he had acquired a household of as many as nineteen Polynesian servants, in addition to his retinue of family and hangers-on. He had filled his mansion with fine objects from his family home: '[f]rom the households of Herriot Row and Skerryvore came Mahogany and rosewood furniture, chests of silver and linen, mirrors, a piano, a Rodin sculpture, wine glasses, decanters and, above all, books'.  As he progressively assumed the role of settler, building his household and developing his interest in island politics, Stevenson felt himself entering into another role that came to him via the myths of his native Scotland, which he had frequently drawn upon to assist his understanding of the Polynesian myths and traditions he encountered on his tours: McLynn has noted that, as his household became established, Stevenson came to regard 'himself as an old-style patriarch, like Cameron of Lochiel or Cluny MacPherson, with quasi-familial duties towards his extended family of retainers'.  As already demonstrated by his descriptions of Scots and Polynesian houses, Stevenson's familiarity with Scottish folklore provided him with a basis for approaching the complexities of Polynesian society, and thus for many of the encounters he intended to write; yet as he increasingly began to occupy a role styled after the figures in Scottish legend, rather than use them as a lens through which to view the difference he encountered in the Polynesians, he found the use of this lens to be increasingly beyond his reach. The more he came to regard himself as resembling the old Highland patriarchs, with a retinue of servants, the further removed he became from a truly dialogic encounter. In other words, the more settled he became, the greater the difficulty he experienced in making his material cohere. By May 1891 he was writing to Colvin with excited 'sketch plan[s] of the present state of our empire', while at the same time complaining of the painful progress of the commissioned letters: '[d]ays and days of unprofitable stubbing and digging, and the result still poor as literature'. Building represented an impediment to writing encounter, just as did writing to the encounter itself, at least of the reciprocal, engaged kind Stevenson wanted to experience.
The combined pressure to produce the letters, and the threatening unwritability of his subject matter which resisted his efforts to make it cohere, presented too great a burden for the material or its author to bear, and by the end of 1891 Stevenson's great South Sea book had all but collapsed. As early as April he had threatened to 'simply make a book of it by the pruning knife'  once the commitment for the letters had been fulfilled, of which thirty seven of the promised fifty eventually appeared in syndication. It was not until 1894, following the proposed publication of an 'Edinburgh Edition' of Stevenson's collected works, that he re-engaged with his South Sea material to make selections for inclusion in the collection, which was interrupted by Stevenson's death in December. In the South Seas appeared in 1896, assembled from the letters by Colvin, whom McLynn condemns for his inclination to 'suppress, censor, distort, mangle, and bowdlerise' Stevenson's material.  Partly due to Colvin's intervention, and partly to the condition of the material Stevenson left behind, the assembled text fell far short of Stevenson's original grand design; yet it remains, as Vanessa Smith has written, 'a text precisely about how it might be possible to write the Pacific islands'.  Although he produced a range of discerning work relating to the Pacific, including history, fiction, and balladry, his attempts to write the Pacific he himself encountered, as he journeyed first on the Casco and then on the Equator, and later as he settled on Samoa, were fraught with anxiety, and often frustrated. Yet they were also prone to considerable insight into the complexities of textualization and settlement. In September 1891, as his project for a South Sea book was collapsing, Stevenson wrote to Colvin to defend the impersonal tone he had adopted:
As far as telling you where I went or when, or anything about Honolulu , I would rather die; that is plain and fair. How can anybody care when and how I left Honolulu ? [.] If ever I put in any such detail, it is because it leads into something or serves as a transition. To tell for its own sake, never! 
Stevenson struggled, throughout the composition of his material, to convert his own experiences into a comprehensive survey of Pacific culture in the islands he visited. Yet although his encounters were frequently subordinated to the impulse to represent them, that is, to writing, it may also be said that writing for Stevenson was subordinate to the encounter; it was a means, however flawed, of achieving a 'transition' that would imply, despite the pressure to textualize, the possibility of engaging in a reciprocal encounter.
The Letters of Robert Louis Stevenson vol. 6 August 1887-September 1890 eds. Bradford A. Booth, Ernest Mehew (New Haven and London: Yale University Press, 1995) pp.279-281. I would like to acknowlege the support of the University of Leeds, where the work for this article was carried out. The article has been taken from my forthcoming book, Unsettled Narratives: The Pacific Writings of Stevenson, Ellis, Melville and London (London and New York: Routledge 2007). RETURN TO TEXT
 Brantlinger begins his analysis of extinction narratives by quoting from Charles Darwin's The Descent of Man : 'When civilised nations come into contact with barbarians the struggle is short, except where a deadly climate gives its aid to the native race'. Patrick Brantlinger, Dark Vanishing: Discourse on the Extinction of Primitive Races, 1800-1930 (Ithaca and London: Cornell University Press, 2003) p.1. The notion of a 'fatal impact', pursuant to the encounter between indigenous and Western races, was rapidly appropriated by Manichean imperialism; as Rod Edmond has observed, at the end of the nineteenth century 'evolutionary explanations of the origins of the races' were 'applied to human societies for confirmation of Anglo-Saxon superiority'. Rod Edmond, Representing the South Pacific: Colonial Discourse from Cook to Gaugain (Cambridge: Cambridge University Press, 1997) p.156. RETURN TO TEXT
Edmond, 1997 p.190. For more examples of the importance of building in the assertion of incursive presence, see John Williams, A Narrative of Missionary Enterprises in the South Sea Islands (London: J. Snow, 1838); and William Ellis, Polynesian Researches (London; Henry G. Bohn, 1853). RETURN TO TEXT
5 Quoted in Clifford Geertz, 'Blurred Genres: The Refiguration of Social Thought', in Local Knowledge: Further Essays in Interpretive Anthropology (London: Fontana Press, 1983) pp.19-35 (p.32). RETURN TO TEXT
 Nigel Leask, 'Darwin 's "Second Sun": Alexander von Humbolt and the Genesis of The Voyage of the Beagle ', in Literature, Science, Psychoanalysis, 1830-1970: Essays in Honour of Gillian Beer eds. Helen Small and Trudi Tate (Oxford: Oxford University Press, 2003) pp.13-36 (p.15). RETURN TO TEXT
 For extracts of these kinds of Pacific narratives, see Exploration and Exchange: A South Seas Anthology 1680-1900 eds. Jonathan Lamb, Vanessa Smith, Nicholas Thomas (Chicago and London: University of Chicago Press, 2000) RETURN TO TEXT
 In the case of missionary writing, texts of empirical observation were frequently produced to promote a particular sectarian agenda. Missionary texts were credited with considerable scientific value among students of the Pacific-Darwin described William Ellis's Polynesian Researches (1816) 'as one of the main sources of his understanding of Polynesian cultures' (Edmond, 1997 p.104)-but as Christopher Herbert has asserted, 'the idea of atonement for origins stamps itself powerfully upon early experiments in ethnography. [...] Missionaries were attempting to gather authenticated empirical proof of the proposition that unredeemed human nature is a horrifying mass of lust and wickedness'. Christopher Herbert, Culture and Anomie: Ethnographic Imagination in the Nineteenth Century (Chicago and London: University of Chicago Press, 1991) p.159. Bernard Smith has also described how the missionary endeavour to provide, in their accounts, examples of indigenous art was intended to promote 'an active dislike in Europe for the figural arts of Pacific peoples. For the missionaries the question was inextricably bound up with idol worship'. Bernard Smith, European Vision and the South Pacific, 1768-1850 (Oxford: Clarendon Press, 1960) p.245. RETURN TO TEXT
 'Introduction' in Exploration and Exchange: A South Seas Anthology 1680-1900 eds. Jonathan Lamb, Vanessa Smith, Nicholas Thomas (Chicago and London: University of Chicago Press, 2000) p.xxi. RETURN TO TEXT
 Adam Smith asserted that, '[t]here are four distinct stages which mankind passes through. First the age of Hunters; second, the age of Shepherds; third, the age of Agriculture; fourth, the age of Commerce'. Adam Smith, Extract from 'MSS Notes of his Lectures on Jurisprudence, 6 vols, 24 December 1762-12 April 1763 (Glasgow University Library, MS Gen. 94/I), Vol. I, entry for 24 December 1762', in The Origins of the Scottish Enlightenment 1707-1776 ed. Jane Rendall (London: MacMillan Press, 1978) pp.141-143 (p.141). David Richards has summarised the influence of Smith's stadial theory: 'The stadial theorists' conviction that societies developed according to the regulating principles of history uncovered by the categorisations of ethnography was an immensely powerful philosophical construction which profoundly shaped the representation of other peoples. Since the European 'past' could be seen in the lives of contemporary primitives who lived nonetheless in a 'previous' epoch of human development, travel to these peoples involved not only a geographical journey, but a voyage in time'. Richards, 1994 p.125. RETURN TO TEXT