'Nowadays, national literature doesn’t mean much: the age of world literature is beginning, and everybody should contribute to hasten its advent.’ This was Goethe, of course, talking to Eckermann in 1827; and these are Marx and Engels, twenty years later, in 1848: ‘National one-sidedness and narrow-mindedness become more and more impossible, and from the many national and local literatures, a world literature arises.’ Weltliteratur: this is what Goethe and Marx have in mind. Not ‘comparative’, but world literature: the Chinese novel that Goethe was reading at the time of that exchange, or the bourgeoisie of the Manifesto, which has ‘given a cosmopolitan character to production and consumption in every country’. Well, let me put it very simply: comparative literature has not lived up to these beginnings. It’s been a much more modest intellectual enterprise, fundamentally limited to Western Europe, and mostly revolving around the river Rhine (German philologists working on French literature). Not much more.
This is my own intellectual formation, and scientific work always has limits. But limits change, and I think it’s time we returned to that old ambition of Weltliteratur: after all, the literature around us is now unmistakably a planetary system. The question is not really what we should do—the question is how. What does it mean, studying world literature? How do we do it? I work on West European narrative between 1790 and 1930, and already feel like a charlatan outside of Britain or France. World literature?
Many people have read more and better than I have, of course, but still, we are talking of hundreds of languages and literatures here. Reading ‘more’ seems hardly to be the solution. Especially because we’ve just started rediscovering what Margaret Cohen calls the ‘great unread’. ‘I work on West European narrative, etc. . . .’ Not really, I work on its canonical fraction, which is not even one per cent of published literature. And again, some people have read more, but the point is that there are thirty thousand nineteenth-century British novels out there, forty, fifty, sixty thousand—no one really knows, no one has read them, no one ever will. And then there are French novels, Chinese, Argentinian, American . . .
Reading ‘more’ is always a good thing, but not the solution. 
Perhaps it’s too much, tackling the world and the unread at the same time. But I actually think that it’s our greatest chance, because the sheer enormity of the task makes it clear that world literature cannot be literature, bigger; what we are already doing, just more of it. It has to be different. The categories have to be different. ‘It is not the “actual” interconnection of “things”’, Max Weber wrote, ‘but the conceptual interconnection of problems which define the scope of the various sciences. A new “science” emerges where a new problem is pursued by a new method.’  That’s the point: world literature is not an object, it’s a problem, and a problem that asks for a new critical method: and no one has ever found a method by just reading more texts. That’s not how theories come into being; they need a leap, a wager—a hypothesis, to get started.
World literature: one and unequal
I will borrow this initial hypothesis from the world-system school of economic history, for which international capitalism is a system that is simultaneously one, and unequal: with a core, and a periphery (and a semiperiphery) that are bound together in a relationship of growing inequality. One, and unequal: one literature (Weltliteratur, singular, as in Goethe and Marx), or perhaps, better, one world literary system (of inter-related literatures); but a system which is different from what Goethe and Marx had hoped for, because it’s profoundly unequal. ‘Foreign debt is as inevitable in Brazilian letters as in any other field’, writes Roberto Schwarz in a splendid essay on ‘The Importing of the Novel to Brazil’: ‘it’s not simply an easily dispensable part of the work in which it appears, but a complex feature of it’;  and Itamar Even-Zohar, reflecting on Hebrew literature: ‘Interference [is] a relationship between literatures, whereby a . . . source literature may become a source of direct or indirect loans [Importing of the novel, direct and indirect loans, foreign debt: see how economic metaphors have been subterraneously at work in literary history]—a source of loans for . . . a target literature . . . There is no symmetry in literary interference. A target literature is, more often than not, interfered with by a source literature which completely ignores it.’ 
This is what one and unequal means: the destiny of a culture (usually a culture of the periphery, as Montserrat Iglesias Santos has specified)  is intersected and altered by another culture (from the core) that ‘completely ignores it’. A familiar scenario, this asymmetry in international power—and later I will say more about Schwarz’s ‘foreign debt’ as a complex literary feature. Right now, let me spell out the consequences of taking an explanatory matrix from social history and applying it to literary history.
Writing about comparative social history, Marc Bloch once coined a lovely ‘slogan’, as he himself called it: ‘years of analysis for a day of synthesis’;  and if you read Braudel or Wallerstein you immediately see what Bloch had in mind. The text which is strictly Wallerstein’s, his ‘day of synthesis’, occupies one third of a page, one fourth, maybe half; the rest are quotations (fourteen hundred, in the first volume of The Modern World-System). Years of analysis; other people’s analysis, which Wallerstein’s page synthesizes into a system.
Now, if we take this model seriously, the study of world literature will somehow have to reproduce this ‘page’—which is to say: this relationship between analysis and synthesis—for the literary field. But in that case, literary history will quickly become very different from what it is now: it will become ‘second hand’: a patchwork of other people’s research, without a single direct textual reading. Still ambitious, and actually even more so than before (world literature!); but the ambition is now directly proportional to the distance from the text: the more ambitious the project, the greater must the distance be.
The United States is the country of close reading, so I don’t expect this idea to be particularly popular. But the trouble with close reading (in all of its incarnations, from the new criticism to deconstruction) is that it necessarily depends on an extremely small canon. This may have become an unconscious and invisible premiss by now, but it is an iron one nonetheless: you invest so much in individual texts only if you think that very few of them really matter. Otherwise, it doesn’t make sense. And if you want to look beyond the canon (and of course, world literature will do so: it would be absurd if it didn’t!) close reading will not do it. It’s not designed to do it, it’s designed to do the opposite. At bottom, it’s a theological exercise—very solemn treatment of very few texts taken very seriously—whereas what we really need is a little pact with the devil: we know how to read texts, now let’s learn how not to read them. Distant reading: where distance, let me repeat it, is a condition of knowledge: it allows you to focus on units that are much smaller or much larger than the text: devices, themes, tropes—or genres and systems. And if, between the very small and the very large, the text itself disappears, well, it is one of those cases when one can justifiably say, Less is more. If we want to understand the system in its entirety, we must accept losing something. We always pay a price for theoretical knowledge: reality is infinitely rich; concepts are abstract, are poor. But it’s precisely this ‘poverty’ that makes it possible to handle them, and therefore to know. This is why less is actually more. 
The Western European novel: rule or exception?
Let me give you an example of the conjunction of distant reading and world literature. An example, not a model; and of course my example, based on the field I know (elsewhere, things may be very different). A few years ago, introducing Kojin Karatani’s Origins of Modern Japanese Literature, Fredric Jameson noticed that in the take-off of the modern Japanese novel, ‘the raw material of Japanese social experience and the abstract formal patterns of Western novel construction cannot always be welded together seamlessly’; and he referred in this respect to Masao Miyoshi’sAccomplices of Silence, and Meenakshi Mukherjee’s Realism and Reality (a study of the early Indian novel).  And it’s true, these books return quite often to the complicated ‘problems’ (Mukherjee’s term) arising from the encounter of western form and Japanese or Indian reality.
Now, that the same configuration should occur in such different cultures as India and Japan—this was curious; and it became even more curious when I realized that Roberto Schwarz had independently discovered very much the same pattern in Brazil. So, eventually, I started using these pieces of evidence to reflect on the relationship between markets and forms; and then, without really knowing what I was doing, began to treat Jameson’s insight as if it were—one should always be cautious with these claims, but there is really no other way to say it—as if it were a law of literary evolution: in cultures that belong to the periphery of the literary system (which means: almost all cultures, inside and outside Europe), the modern novel first arises not as an autonomous development but as a compromise between a western formal influence (usually French or English) and local materials.
This first idea expanded into a little cluster of laws,  and it was all very interesting, but . . . it was still just an idea; a conjecture that had to be tested, possibly on a large scale, and so I decided to follow the wave of diffusion of the modern novel (roughly: from 1750 to 1950) in the pages of literary history. Gasperetti and Goscilo on late eighteenth-century Eastern Europe;  Toschi and Martí-López on early nineteenth-century Southern Europe;  Franco and Sommer on mid-century Latin America;  Frieden on the Yiddish novels of the 1860s;  Moosa, Said and Allen on the Arabic novels of the 1870s;  Evin and Parla on the Turkish novels of the same years; Anderson on the Filipino Noli Me Tangere, of 1887; Zhao and Wang on turn-of-the-century Qing fiction;  Obiechina, Irele and Quayson on West African novels between the 1920s and the 1950s (plus of course Karatani, Miyoshi, Mukherjee, Even-Zohar and Schwarz). Four continents, two hundred years, over twenty independent critical studies, and they all agreed: when a culture starts moving towards the modern novel, it’s always as a compromise between foreign form and local materials. Jameson’s ‘law’ had passed the test—the first test, anyway.  &  And actually more than that: it had completely reversed the received historical explanation of these matters: because if the compromise between the foreign and the local is so ubiquitous, then those independent paths that are usually taken to be the rule of the rise of the novel (the Spanish, the French, and especially the British case)—well, they’re not the rule at all, they’re the exception. They come first, yes, but they’re not at all typical. The ‘typical’ rise of the novel is Krasicki, Kemal, Rizal, Maran—not Defoe.
Experiments with history
See the beauty of distant reading plus world literature: they go against the grain of national historiography. And they do so in the form of an experiment. You define a unit of analysis (like here, the formal compromise),  and then follow its metamorphoses in a variety of environments —until, ideally, all of literary history becomes a long chain of related experiments: a ‘dialogue between fact and fancy’, as Peter Medawar calls it: ‘between what could be true, and what is in fact the case’.  Apt words for this research, in the course of which, as I was reading my fellow historians, it became clear that the encounter of western forms and local reality did indeed produce everywhere a structural compromise—as the law predicted—but also, that the compromise itself was taking rather different forms. At times, especially in the second half of the nineteenth century and in Asia, it tended to be very unstable:  an ‘impossible programme’, as Miyoshi says of Japan. At other times it was not so: at the beginning and at the end of the wave, for instance (Poland, Italy and Spain at one extreme; and West Africa on the other), historians describe novels that had, certainly, their own problems—but not problems arising from the clash of irreconcilable elements.
I hadn’t expected such a spectrum of outcomes, so at first I was taken aback, and only later realized that this was probably the most valuable finding of them all, because it showed that world literature was indeed a system—but a system of variations. The system was one, not uniform. The pressure from the Anglo-French core tried to make it uniform, but it could never fully erase the reality of difference. (See here, by the way, how the study of world literature is—inevitably—a study of the struggle for symbolic hegemony across the world.) The system was one, not uniform. And, retrospectively, of course it had to be like this: if after 1750 the novel arises just about everywhere as a compromise between West European patterns and local reality—well, local reality was different in the various places, just as western influence was also very uneven: much stronger in Southern Europe around 1800, to return to my example, than in West Africa around 1940. The forces in play kept changing, and so did the compromise that resulted from their interaction. And this, incidentally, opens a fantastic field of inquiry for comparative morphology (the systematic study of how forms vary in space and time, which is also the only reason to keep the adjective ‘comparative’ in comparative literature): but comparative morphology is a complex issue, that deserves its own paper.
Forms as abstracts of social relationships
Let me now add a few words on that term ‘compromise’—by which I mean something a little different from what Jameson had in mind in his introduction to Karatani. For him, the relationship is fundamentally a binary one: ‘the abstract formal patterns of Western novel construction’ and ‘the raw material of Japanese social experience’: form and content, basically.  For me, it’s more of a triangle: foreign form, local material—and local form. Simplifying somewhat: foreign plot; local characters; and then, local narrative voice: and it’s precisely in this third dimension that these novels seem to be most unstable—most uneasy, as Zhao says of the late Qing narrator. Which makes sense: the narrator is the pole of comment, of explanation, of evaluation, and when foreign ‘formal patterns’ (or actual foreign presence, for that matter) make characters behave in strange ways (like Bunzo, or Ibarra, or Bràs Cubas), then of course comment becomes uneasy—garrulous, erratic, rudderless.
‘Interferences’, Even-Zohar calls them: powerful literatures making life hard for the others—making structure hard. And Schwarz: ‘a part of the original historical conditions reappears as a sociological form . . . In this sense, forms are the abstract of specific social relationships.’  Yes, and in our case the historical conditions reappear as a sort of ‘crack’ in the form; as a faultline running between story and discourse, world and worldview: the world goes in the strange direction dictated by an outside power; the worldview tries to make sense of it, and is thrown off balance all the time. Like Rizal’s voice (oscillating between Catholic melodrama and Enlightenment sarcasm),  or Futabatei’s (caught between Bunzo’s ‘Russian’ behaviour, and the Japanese audience inscribed in the text), or Zhao’s hypertrophic narrator, who has completely lost control of the plot, but still tries to dominate it at all costs. This is what Schwarz meant with that ‘foreign debt’ that becomes a ‘complex feature’ of the text: the foreign presence ‘interferes’ with the very utterance of the novel. The one-and-unequal literary system is not just an external network here, it doesn’t remain outside the text: it’s embedded well into its form.
Trees, waves and cultural history
Forms are the abstract of social relationships: so, formal analysis is in its own modest way an analysis of power. (That’s why comparative morphology is such a fascinating field: studying how forms vary, you discover how symbolic power varies from place to place.) And indeed, sociological formalism has always been my interpretive method, and I think that it’s particularly appropriate for world literature . . . But, unfortunately, at this point I must stop, because my competence stops. Once it became clear that the key variable of the experiment was the narrator’s voice, well, a genuine formal analysis was off limits for me, because it required a linguistic competence that I couldn’t even dream of (French, English, Spanish, Russian, Japanese, Chinese and Portuguese, just for the core of the argument). And probably, no matter what the object of analysis is, there will always be a point where the study of world literature must yield to the specialist of the national literature, in a sort of cosmic and inevitable division of labour. Inevitable not just for practical reasons, but for theoretical ones. This is a large issue, but let me at least sketch its outline.
When historians have analysed culture on a world scale (or on a large scale anyway), they have tended to use two basic cognitive metaphors: the tree and the wave. The tree, the phylogenetic tree derived from Darwin, was the tool of comparative philology: language families branching off from each other—Slavo-Germanic from Aryan-Greco-Italo-Celtic, then Balto-Slavic from Germanic, then Lithuanian from Slavic. And this kind of tree allowed comparative philology to solve that great puzzle which was also perhaps the first world system of culture: Indo-European: a family of languages spreading from India to Ireland (and perhaps not just languages, a common cultural repertoire, too: but here the evidence is notoriously shakier). The other metaphor, the wave, was also used in historical linguistics (as in Schmidt’s ‘wave hypothesis’, that explained certain overlaps among languages), but it played a role in many other fields as well: the study of technological diffusion, for instance, or the fantastic interdisciplinary theory of the ‘wave of advance’ by Cavalli-Sforza and Ammerman (a geneticist and an archaeologist), which explains how agriculture spread from the fertile crescent in the Middle East towards the North-West and then throughout Europe.
Now, trees and waves are both metaphors—but except for this, they have absolutely nothing in common. The tree describes the passage from unity to diversity: one tree, with many branches: from Indo-European, to dozens of different languages. The wave is the opposite: it observes uniformity engulfing an initial diversity: Hollywood films conquering one market after another (or English swallowing language after language). Trees need geographical discontinuity (in order to branch off from each other, languages must first be separated in space, just like animal species); waves dislike barriers, and thrive on geographical continuity (from the viewpoint of a wave, the ideal world is a pond). Trees and branches are what nation-states cling to; waves are what markets do. And so on. Nothing in common, between the two metaphors. But—they both work. Cultural history is made of trees and waves—the wave of agricultural advance supporting the tree of Indo-European languages, which is then swept by new waves of linguistic and cultural contact . . . And as world culture oscillates between the two mechanisms, its products are inevitably composite ones. Compromises, as in Jameson’s law. That’s why the law works: because it intuitively captures the intersection of the two mechanisms. Think of the modern novel: certainly a wave (and I’ve actually called it a wave a few times)—but a wave that runs into the branches of local traditions,  and is always significantly transformed by them.
This, then, is the basis for the division of labour between national and world literature: national literature, for people who see trees; world literature, for people who see waves. Division of labour . . . and challenge; because both metaphors work, yes, but that doesn’t mean that they work equally well. The products of cultural history are always composite ones: but which is the dominant mechanism in their composition? The internal, or the external one? The nation or the world? The tree or the wave? There is no way to settle this controversy once and for all—fortunately: because comparatists need controversy. They have always been too shy in the presence of national literatures, too diplomatic: as if one had English, American, German literature—and then, next door, a sort of little parallel universe where comparatists studied a second set of literatures, trying not to disturb the first set. No; the universe is the same, the literatures are the same, we just look at them from a different viewpoint; and you become a comparatist for a very simple reason: because you are convinced that that viewpoint is better. It has greater explanatory power; it’s conceptually more elegant; it avoids that ugly ‘one-sidedness and narrow-mindedness’; whatever. The point is that there is no other justification for the study of world literature (and for the existence of departments of comparative literature) but this: to be a thorn in the side, a permanent intellectual challenge to national literatures—especially the local literature. If comparative literature is not this, it’s nothing. Nothing. ‘Don’t delude yourself’, writes Stendhal of his favourite character: ‘for you, there is no middle road.’ The same is true for us.
 I address the problem of the great unread in a companion piece to this article, ‘The Slaughterhouse of Literature’, forthcoming in a special issue of Modern Language Quarterly on ‘Formalism and Literary History’, spring 2000.
 Max Weber, ‘Objectivity in Social Science and Social Policy’  in The Methodology of the Social Sciences, New York 1949, p. 68.
 Roberto Schwarz, ‘The Importing of the Novel to Brazil and Its Contradictions in the Work of Roberto Alencar’  in Misplaced Ideas, London 1992, p. 50.
 Itamar Even-Zohar, ‘Laws of Literary Interference’ in Poetics Today, 1990, pp. 54, 62.
 Montserrat Iglesias Santos, ‘El sistema literario: teoría empírica y teoría de los polisistemas’, in Dario Villanueva (ed.), Avances en teoría de la literatura, Santiago de Compostela 1994, p. 339: ‘It is important to emphasize that interferences occur most often at the periphery of the system.’
 Marc Bloch, ‘Pour une histoire comparée des sociétés européennes’, Revue de synthèse historique, 1928.
 Or to quote Weber again: ‘concepts are primarily analytical instruments for the intellectual mastery of empirical data’. (‘Objectivity in Social Science and Social Policy’, p. 106.) Inevitably, the larger the field one wants to study, the greater the need for abstract ‘instruments’ capable of mastering empirical reality.
 Fredric Jameson, ‘In the Mirror of Alternate Modernities’, in Karatani Kojin, Origins of Modern Japanese Literature, Durham–London 1993, p. xiii.
 I have begun to sketch them out in the last chapter of the Atlas of the European Novel 1800–1900 (Verso: London 1998), and this is more or less how they sound: second, the formal compromise is usually prepared by a massive wave of West European translations; third, the compromise itself is generally unstable (Miyoshi has a great image for this: the ‘impossible programme’ of Japanese novels); but fourth, in those rare instances when the impossible programme succeeds, we have genuine formal revolutions.
 ‘Given the history of its formative stage, it is no surprise that the early Russian novel contains a host of conventions popularized in French and British literature’, writes David Gasperetti in The Rise of the Russian Novel (De Kalb 1998, p. 5). And Helena Goscilo, in her ‘Introduction’ to Krasicki’s Adventures of Mr. Nicholas Wisdom: ‘The Adventures is read most fruitfully in the context of the West European literature on which it drew heavily for inspiration.’ (Ignacy Krasicki, The Adventures of Mr. Nicholas Wisdom, Evanston 1992, p. xv.)
 ‘There was a demand for foreign products, and production had to comply’, explains Luca Toschi speaking of the Italian narrative market around 1800 (‘Alle origini della narrativa di romanzo in Italia’, in Massimo Saltafuso (ed.), Il viaggio del narrare, Florence 1989, p. 19). A generation later, in Spain, ‘readers are not interested in the originality of the Spanish novel; their only desire is that it would adhere to those foreign models with which they have become familiar’: and so, concludes Elisa Martí-López, one may well say that between 1800 and 1850 ‘the Spanish novel is being written in France’ (Elisa Martí-López, ‘La orfandad de la novela española: política editorial y creación literaria a mediados del siglo XIX’, Bulletin Hispanique, 1997).
 ‘Obviously, lofty ambitions were not enough. All too often the nineteenth century Spanish-American novel is clumsy and inept, with a plot derived at second hand from the contemporary European Romantic novel.’ (Jean Franco, Spanish-American Literature, Cambridge 1969, p. 56.) ‘If heroes and heroines in mid-nineteenth century Latin American novels were passionately desiring one another across traditional lines . . . those passions might not have prospered a generation earlier. In fact, modernizing lovers were learning how to dream their erotic fantasies by reading the European romances they hoped to realize.’ (Doris Sommer, Foundational Fictions: The National Romances of Latin America, Berkeley–Los Angeles 1991, pp. 31–2.)
 ‘Yiddish writers parodied—appropriated, incorporated, and modified—diverse elements from European novels and stories.’ (Ken Frieden, Classic Yiddish Fiction, Albany 1995, p. x.)
 Matti Moosa quotes the novelist Yahya Haqqi: ‘there is no harm in admitting that the modern story came to us from the West. Those who laid down its foundations were persons influenced by European literature, particularly French literature. Although masterpieces of English literature were translated into Arabic, French literature was the fountain of our story.’ (Matti Moosa, The Origins of Modern Arabic Fiction , 2nd ed. 1997, p. 93.) For Edward Said, ‘at some point writers in Arabic became aware of European novels and began to write works like them’ (Edward Said, Beginnings , New York 1985, p. 81). And Roger Allen: ‘In more literary terms, increasing contacts with Western literatures led to translations of works of European fiction into Arabic, followed by their adaptation and imitation, and culminating in the appearance of an indigenous tradition of modern fiction in Arabic.’ (Roger Allen, The Arabic Novel, Syracuse 1995, p. 12.)
 ‘The first novels in Turkey were written by members of the new intelligentsia, trained in government service and well-exposed to French literature’, writes Ahmet O. Evin (Origins and Development of the Turkish Novel, Minneapolis 1983, p. 10); and Jale Parla: ‘the early Turkish novelists combined the traditional narrative forms with the examples of the western novel’ (‘Desiring Tellers, Fugitive Tales: Don Quixote Rides Again, This Time in Istanbul’, forthcoming).
 ‘The narrative dislocation of the sequential order of events is perhaps the most outstanding impression late Qing writers received when they read or translated Western fiction. At first, they tried to tidy up the sequence of the events back into their pre-narrated order. When such tidying was not feasible during translation, an apologetic note would be inserted . . . Paradoxically, when he alters rather than follows the original, the translator does not feel it necessary to add an apologetic note.’ (Henry Y. H. Zhao, The Uneasy Narrator: Chinese Fiction from the Traditional to the Modern, Oxford 1995, p. 150.) ‘Late Qing writers enthusiastically renewed their heritage with the help of foreign models’, writes David Der-wei Wang: ‘I see the late Qing as the beginning of the Chinese literary “modern” because writers’ pursuit of novelty was no longer contained within indigenously defined barriers but was inextricably defined by the multilingual, crosscultural trafficking of ideas, technologies, and powers in the wake of nineteenth-century Western expansionism.’ (Fin-de-siècle Splendor: Repressed Modernities of Late Qing Fiction, 1849–1911, Stanford 1997, pp. 5, 19.)
 ‘One essential factor shaping West African novels by indigenous writers was the fact that they appeared after the novels on Africa written by non-Africans . . . the foreign novels embody elements which indigenous writers had to react against when they set out to write.’ (Emmanuel Obiechina, Culture, Tradition and Society in the West African Novel, Cambridge 1975, p. 17.) ‘The first Dahomean novel, Doguicimi . . . is interesting as an experiment in recasting the oral literature of Africa within the form of a French novel.’ (Abiola Irele, The African Experience in Literature and Ideology, Bloomington 1990, p. 147.) ‘It was the rationality of realism that seemed adequate to the task of forging a national identity at the conjuncture of global realities . . . the rationalism of realism dispersed in texts as varied as newspapers, Onitsha market literature, and in the earliest titles of the African Writers Series that dominated the discourses of the period.’ (Ato Quayson, Strategic Transformations in Nigerian Writing, Bloomington 1997, p. 162.)
 In the seminar where I first presented this ‘second-hand’ criticism, Sarah Golstein asked a very good, Candide-like question: You decide to rely on another critic. Fine. But what if he’s wrong? My reply: If he’s wrong, you are wrong too, and you soon know, because you don’t find any corroboration—you don’t find Goscilo, Martí-López, Sommer, Evin, Zhao, Irele . . . And it’s not just that you don’t find positive corroboration; sooner or later you find all sorts of facts you cannot explain, and your hypothesis is falsified, in Popper’s famous formulation, and you must throw it away. Fortunately, this hasn’t been the case so far, and Jameson’s insight still stands.
 OK, I confess, in order to test the conjecture I actually did read some of these ‘first novels’ in the end (Krasicki’s Adventures of Mr. Nicholas Wisdom, Abramowitsch’s Little Man, Rizal’s Noli Me Tangere, Futabatei’s Ukigumo, René Maran’s Batouala, Paul Hazoumé’s Doguicimi). This kind of ‘reading’, however, no longer produces interpretations but merely tests them: it’s not the beginning of the critical enterprise, but its appendix. And then, here you don’t really read the text anymore, but rather through the text, looking for your unit of analysis. The task is constrained from the start; it’s a reading without freedom.
 For practical purposes, the larger the geographical space one wants to study, the smaller should the unit of analysis be: a concept (in our case), a device, a trope, a limited narrative unit—something like this. In a follow-up paper, I hope to sketch out the diffusion of stylistic ‘seriousness’ (Auerbach’s keyword in Mimesis) in nineteenth- and twentieth-century novels.
 How to set up a reliable sample—that is to say, what series of national literatures and individual novels provide a satisfactory test of a theory’s predictions—is of course quite a complex issue. In this preliminary sketch, my sample (and its justification) leave much to be desired.
 Scientific research ‘begins as a story about a Possible World’, Medawar goes on, ‘and ends by being, as nearly as we can make it, a story about real life.’ His words are quoted by James Bird in The Changing World of Geography, Oxford 1993, p. 5. Bird himself offers a very elegant version of the experimental model.
 Aside from Miyoshi and Karatani (for Japan), Mukherjee (for India), and Schwarz (for Brazil), the compositional paradoxes and the instability of the formal compromise are often mentioned in the literature on the Turkish, Chinese and Arabic novel. Discussing Namik Kemal’s Intibah, Ahmet Evin points out how ‘the merger of the two themes, one based on the traditional family life and the other on the yearnings of a prostitute, constitute the first attempt in Turkish fiction to achieve a type of psychological dimension observed in European novels within a thematic framework based on Turkish life. However, due both to the incompatibility of the themes and to the difference in the degree of emphasis placed on each, the unity of the novel is blemished. The structural defects of Intibah are symptomatic of the differences between the methodology and concerns of the Turkish literary tradition on the one hand and those of the European novel on the other.’ (Ahmet O. Evin, Origins and Development of the Turkish Novel, p. 68; emphasis mine.) Jale Parla’s evaluation of the Tanzimat period sounds a similar note: ‘behind the inclination towards renovation stood a dominant and dominating Ottoman ideology that recast the new ideas into a mould fit for the Ottoman society. The mould, however, was supposed to hold two different epistemologies that rested on irreconcilable axioms. It was inevitable that this mould would crack and literature, in one way or another, reflects the cracks.’ (‘Desiring Tellers, Fugitive Tales: Don Quixote Rides Again, This Time in Istanbul’, emphasis mine.) In his discussion of the 1913 novel Zaynab, by Husayn Haykal, Roger Allen echoes Schwarz and Mukherjee (‘it is all too easy to point to the problems of psychological fallacy here, as Hamid, the student in Cairo acquainted with Western works on liberty and justice such as those of John Stuart Mill and Herbert Spencer, proceeds to discuss the question of marriage in Egyptian society on such a lofty plane with his parents, who have always lived deep in the Egyptian countryside’: The Arabic Novel, p. 34; emphasis mine). Henry Zhao emphasizes from his very title—The Uneasy Narrator: and see the splendid discussion of uneasiness that opens the book—the complications generated by the encounter of western plots and Chinese narrative: ‘A salient feature of late Qing fiction’, he writes, ‘is the greater frequency of narrative intrusions than in any previous period of Chinese vernacular fiction . . . The huge amount of directions trying to explain the newly adopted techniques betrays the narrator’s uneasiness about the instability of his status . . . the narrator feels the threat of interpretive diversification . . . moral commentaries become more tendentious to make the judgments unequivocal’, and at times the drift towards narratorial overkill is so overpowering that a writer may sacrifice narrative suspense ‘to show that he is morally impeccable’ (The Uneasy Narrator, pp. 69–71).
 In some cases, even translations of European novels went through all sorts of incredible somersaults. In Japan, in 1880, Tsubouchi’s translation of The Bride of Lammermoor appeared under the title Shumpu jowa [Spring breeze love story], and Tsubouchi himself ‘was not beyond excising the original text when the material proved inappropriate for his audience, or converting Scott’s imagery into expressions corresponding more closely to the language of traditional Japanese literature’ (Marleigh Grayer Ryan, ‘Commentary’ to Futabatei Shimei’s Ukigumo, New York 1967, pp. 41–2). In the Arabic world, writes Matti Moosa, ‘in many instances the translators of Western fiction took extensive and sometimes unwarranted liberties with the original text of a work. Yaqub Sarruf not only changed the title of Scott’s Talisman to Qalb al-Asad wa Salah al-Din (The Lion Heart and Saladin), but also admitted that he had taken the liberty of omitting, adding, and changing parts of this romance to suit what he believed to be his audience’s taste . . . Other translators changed the titles and the names of the characters and the contents, in order, they claimed, to make the translated work more acceptable to their readers and more consistent with the native literary tradition.’ (The Origins of Modern Arabic Fiction, p. 106.) The same general pattern holds for late Qing literature, where ‘translations were almost without exception tampered with . . . the most serious way of tampering was to paraphrase the whole novel to make it a story with Chinese characters and Chinese background . . . Almost all of these translations suffered from abridgement . . . Western novels became sketchy and speedy, and looked more like Chinese traditional fiction.’ (Henry Zhao, The Uneasy Narrator, p. 229.)
 Why this difference? Probably, because in Southern Europe the wave of French translations encountered a local reality (and local narrative traditions) that weren’t that different after all, and as a consequence, the composition of foreign form and local material proved easy. In West Africa, the opposite situation: although the novelists themselves had been influenced by Western literature, the wave of translations had been much weaker than elsewhere, and local narrative conventions were for their part extremely different from European ones (just think of orality); as the desire for the ‘foreign technology’ was relatively bland—and further discouraged, of course, by the anti-colonial politics of the 1950s—local conventions could play their role relatively undisturbed. Obiechina and Quayson emphasize the polemical relationship of early West African novels vis-à-vis European narrative: ‘The most noticeable difference between novels by native West Africans and those by non-natives using the West African setting, is the important position which the representation of oral tradition is given by the first, and its almost total absence in the second.’ (Emmanuel Obiechina, Culture, Tradition and Society in the West African Novel, p. 25.) ‘Continuity in the literary strategic formation we have identified is best defined in terms of the continuing affirmation of mythopeia rather than of realism for the definition of identity . . . That this derives from a conceptual opposition to what is perceived as a Western form of realism is difficult to doubt. It is even pertinent to note in this regard that in the work of major African writers such as Achebe, Armah, and Ngugi, the movement of their work has been from protocols of realist representation to those of mythopeic experimentation.’ (Ato Quayson, Strategic Transformations in Nigerian Writing, p. 164.)
 The same point is made in a great article by António Cándido: ‘We [Latin American literatures] never create original expressive forms or basic expressive techniques, in the sense that we mean by romanticism, on the level of literary movements; the psychological novel, on the level of genres; free indirect style, on that of writing . . . the various nativisms never rejected the use of the imported literary forms . . . what was demanded was the choice of new themes, of different sentiments. (‘Literature and Underdevelopment’, in César Fernández Moreno, Julio Ortega, Ivan A. Shulman (eds), Latin America in Its Literature, New York 1980, pp. 272–3.)
 ‘The Importing of the Novel To Brazil’, p. 53.
 Rizal’s solution, or lack thereof, is probably also related to his extraordinarily wide social spectrum (Noli Me Tangere, among other things, is the text that inspired Benedict Anderson to link the novel and the nation-state): in a nation with no independence, an ill-defined ruling class, no common language and hundreds of disparate characters, it’s hard to speak ‘for the whole’, and the narrator’s voice cracks under the effort.
 In a few lucky cases, the structural weakness may turn into a strength, as in Schwarz’s interpretation of Machado, where the ‘volatility’ of the narrator becomes ‘the stylization of the behaviour of the Brazilian ruling class’: not a flaw any longer, but the very point of the novel: ‘Everything in Machado de Assis’s novels is coloured by the volatility—used and abused in different degrees—of their narrators. The critics usually look at it from the point of view of literary technique or of the author’s humour. There are great advantages in seeing it as the stylization of the behaviour of the Brazilian ruling class. Instead of seeking disinterestedness, and the confidence provided by impartiality, Machado’s narrator shows off his impudence, in a gamut which runs from cheap gibes, to literary exhibitionism, and even to critical acts.’ (Roberto Schwarz, ‘The Poor Old Woman and Her Portraitist’ , in Misplaced Ideas, p. 94.)
 ‘Grafting processes’, Miyoshi calls them; Schwarz speaks of ‘the implantation of the novel, and of its realist strand in particular’, and Wang of ‘transplanting Western narrative typologies’. And indeed, Belinsky had already described Russian literature as ‘a transplanted rather than indigenous growth’ in 1843.