The editorial of the first number of George Woodberry’s Journal of Comparative Literature claimed for comparatists a ‘new citizenship in the rising state which – the obscurer or brighter dream of all great scholars from Plato to Goethe – is without frontiers or race or force, but there reason is supreme’ (Woodberry 3–4). The coming ‘unity of mankind’ and ‘upbuilding of the international state in a homogeneous civilization’ were to be achieved by identifying the universals in literature and culture. After a solid basis had been achieved in European languages, Woodberry foresaw ‘the approaching exploitation of the old literatures of the Orient, which is the next great event in the literary history of the world’ (8). Optimism about the cosmopolitan future would be crushed by two world wars (already, its precarity could be discerned wherever gunboats made the law), but the position of ‘the old literatures of the Orient’, always on the threshold of being integrated into a cultural world order defined by others, remains.

Theorists of ‘world literature’ often refer back to Goethe’s 1837 statement that ‘national literature means little now, and the era of world literature is at hand’ (Eckermann, 325, translated in Damrosch, Melas, and Buthelezi, 22), or to Marx and Engels saying in the 1848 Communist Manifesto that ‘the bourgeoisie has through its exploitation of the world market given a cosmopolitan character to production and consumption in every country’, so that ‘from the numerous national and local literatures, there arises a world literature’ (Marx and Engels, 36). In this understanding, the ‘world’ of ‘world literature’ is a side effect of globalization, understood as a network of political and economic relations set in motion by Columbus’s voyages and the subsequent implantation of European colonies around the world. Whether this is an accurate representation of global economic history is debatable. Those who have in their sights a longer period speak of a world economy centred on Asia and the Near East, disrupted for only about 400 of the last 4000 years by the European newcomers.

Others take ‘world’ to stand for one pole of the axis ‘general–particular’. It is world literature as opposed to Sulawesi, Russian, Qatari literatures. In this view the act of exalting world literature amounts to downgrading whatever is local, specific, untranslatable, tied to its place and time of origin. One criticism of such accounts of ‘world literature’ champions the particular – but in general, on principle. Another emphasizes the provinciality endemic to supposedly global descriptions, their failure to live up to the universal mission they proclaim.

I take from the existence of rival accounts of globalization the lesson that it matters how we specify the ‘world’ in ‘world literature’. What is a world? I will propose, in a relatively abstract fashion relying on Niklas Luhmann, that a world is specified by the attainable horizon of communication for those who live in it. Thus, for example, although intelligent beings may dwell elsewhere in the universe, they are not yet part of our world. Although the figurines buried on the Cycladic islands are known to us, and so belong to our world, the societies that produced them are not well enough understood to be part of our world, though this might change with as yet unforeseen discoveries. By defining ‘world’ in this way, we open up the term to allow for many overlapping or possibly mutually unrelated worlds. ‘World literature’ is thus the literature of that world which is available to the persons who call it so.

Now one of the limits of ‘world literature’ as usually conceived comes into view – a limit that is usually taken as a sign of its relevance. The ‘world’ imagined by contemporary globalization narratives is a world whose history leads up to us, the audience for the product. But that world is not the only world. To pluralize the noun ‘world’, there is no need to call on science fiction or hypothesize additional dimensions. Human history is crammed with worlds indifferent to our present condition, worlds from which our world is practically unimaginable. But one ought not reject world literature with a mere ‘no’; it is always better to oppose something with something else. As an experimental corrective to several of the biases that inflect current conceptions of world literature, I propose that comparatists set themselves the task of considering the historical relationships among the languages, literatures and cultures of Asia before 1800.

Asia, because it includes several written cultures that had built themselves up in world-defining terms, accumulated massive textual archives and translated each other; pre-1800, because in that period the interference from Europe was relatively inconsequential. A comparative literary history of Asia must reach far back in time, to a period before the nations of today originated. It will demonstrate the cosmopolitan character of a single world region before the current dynamics of globalization got underway.

The current map of Asia will not be much help in designing the comparative project, because the nations, languages and peoples that are familiar to us have emerged alongside the very processes of globalization that serve as justificatory meta-narrative to the usual ‘world literature’ investigation. The ‘nation’ cannot be taken for granted as the unit of comparison for the further reason that the preconceptions of national literary histories tend to make us think that the past has value as prediction of the present, and thus that the useful part of the past is that which corresponds to the now. National histories offer a cumulative, teleological and thus sanitized image of the past. An inclusive comparative project must include para- and pre-national narratives, exactly what national histories typically repress or neutralize through incorporation. The shadow of the nation-state too easily dominates history-writing, for the good reason that historians have always written on behalf of rulers, dynasties and states. Archaeology is needed to supplement history’s selective vision. Large areas of China, for instance, had a long existence as separate kingdoms more or less attached to the orbits of the regional powers before coming under the sway of the Qing Empire in the seventeenth and eighteenth centuries. The ‘Central Regions’ ancestral to Chinese civilization were at one time just a small corridor of shared culture. Competitors to the ‘Central Regions’ have left traces within and without Chinese mainstream culture. Linguistic and archaeological evidence shows the mobility of ancient populations, whether journeying along the caravan routes of the Silk Road, transhuming and raiding in the northern grasslands, practising mountain agriculture in the warmer zones, or engaging in long- and short-range trade in the archipelagos of the southeast. Tibetan speakers once occupied an area far larger than present-day Tibet. Buddhist preachers and pilgrims carried much more than the Four Noble Truths when they spread the doctrine to new areas. Industrial or military techniques, the arts of administration, musical modes and the like circulated despite barriers of language.

It is no surprise that China plays a large role in the cultural history of Asia, and this is in no small measure due to the kinds of creativity that the Chinese dedicated to ‘letters’, as we may call them. The term wen or wenzhang in Chinese, it is generally known, refers not only to ‘literature’ but to principles of organization applied to family life, political order, visual artefacts, morals, ideas about the natural world and so forth. It comes close to being, in some texts, a near synonym for ‘Chineseness’ or ‘Chinese values’. Obviously an investigation intended to be cosmopolitan cannot perpetuate such a sense of cultural preference, but the meanings of wenzhang do indicate that the scope of this investigation cannot be restricted to ‘the literary’ as we know it today in the West. And indeed what is transmitted through and as wenzhang is far broader than lyric, imaginative or narrative writing. The effect of letters in this broad sense accounts for much of what we call culture.

An internally comparative history of Chinese literature could be built from case studies of the encounter of Chinese and non-Chinese (or as-yet-non-Chinese) around the subject of ‘letters’ and language, at different moments in history. The emergence of China was a process contemporaneous with the diffusion of Chinese writing. Let us first consider how that process appears to modern observers situated within the ‘event horizon’ (Hawking, 92), so to speak, of the Chinese written language:

China has always had a clear and stable center, even if its margins have at times been vague and unstable. […] Unlike Europe, Chinese political borders and cultural spaces spread from the centers to the margins. Bracketing the most ancient three dynasties for the moment, ever since Qin and Han times we have enjoyed, as the saying goes, ‘standardized cart tracks, standardized writing system, and standardized conduct.’ The European notion that ‘nations are new phenomena characteristic of the later stages of human history’ does not hold as well here, where a nation coalesced much earlier under the influence of language, ethics, customs and the political system. (Ge 21, 23)

That is how it looks for an observer who knows what would eventually become the ‘centre’ and what would be its ‘shifting margins’. But at the time when China was coalescing out of many competing states with some degree of common culture, no one could have known as much. To look at the situation without retrospective wisdom, we should say rather that in the pre-imperial period, the languages and writing systems covering the Chinese cultural zone were not uniform. People from regions far from the ‘Central Territory’, from Qin or Chu for example (areas centred in present-day Shaanxi and Hubei provinces, respectively), spoke and wrote ‘strangely’, according to documents penned by and for the public of the Central States. Scribes from different areas used (again an anachronism) ‘variant’ characters even when copying down texts familiar to the later possessors of the tradition. The unification of script by the First Emperor just after 221 BCE caused Chinese to ‘enjoy’, as Field and Qin put it, the benefits of standardization. Those whose languages could be conveyed by the official writing system, and who figured as protagonists in the history written in that system, were deemed to belong to the culture and descent-group we now know as ‘Chinese’. Just as, in earlier eras, royal genealogies were periodically re-engineered to permit the accommodation of previously alien groups, so regional modes of speech became ‘dialects’ of Chinese script. Sparsely documented exotica – place names, bits of lexicology – suggest that Sinification involved winnowing out of the non-standard elements. Peoples affected by Chinese conquest or influence sometimes adopted Chinese script as an administrative language, dooming to muteness their earlier spoken idioms, or eventually nativized it, resulting in idiomatic or diglossic forms of writing as in Vietnam, Korea and Japan. Some languages were notated in several different families of script simultaneously. A variety of acculturational styles are exhibited by the different kinds of response to Chinese script on the frontiers of the Chinese culture zone.

Buddhism, one of the main cultural binding agents of Asia, entered China through Central Asian contacts and long appeared as a challenge to the language and values of the Chinese literate elite. Its early linguistic vehicles were multiply alien: Parthian intermediaries retelling sermons originally delivered in Sanskrit or Pali, whose words were taken down in rough Chinese and subsequently polished by trained scribes. Through the establishment of monastic translation committees and the work of a few exceptional polyglot interpreters compiling a corpus of tens of thousands of texts, Chinese became the dominant medium of Buddhist teaching in Asia. (Sanskrit and Tibetan texts circulated in more specialized circuits.) The introduction of Buddhism via Central Asia in around 150 CE presented the Chinese with the novelty of an advanced culture that did not have the same beginnings as their own. Foreign languages and scripts – Indic, Sogdian, Parthian, Khotanese – were sometimes assumed to be continuous with the language of the birds and beasts, sometimes treated as magical incantations. Some regions were notable for their multilingualism: the caves of Dunhuang, first excavated in 1900, provide astonishing direct evidence of Silk Road linguistic, ethnic and religious diversity. Many of the languages spoken by tenth-century Central Asians no longer exist, or have become identified with cultures and religions far different from the ones known to Dunhuang travellers.

One international ‘literary world’ that has received recent attention is the ‘Sinosphere’ of ‘Sinophone’ or ‘Sinographic’ communication, designating China and those of its neighbouring states that adopted and adapted Chinese-character writing: Korea, Vietnam, Japan. In pre-modern times this literacy transcended national borders – among elites, concepts and models enshrined in books such as the Chinese Classics could circulate without hindrance. But the Sinographic model of international culture closely mirrors a particular state narrative, that of the Chinese Empire with its surrounding tributary states. To cast aside Eurocentrism and adopt Sinocentrism is not much of a gain for cosmopolitanism. The vast Chinese tradition, with its almost credible self-sufficiency, never saturated even the space of present-day China. To the groups that adopted Chinese writing and Chinese texts, we must contrast the neighbours whose models for writing came from elsewhere, from India or the Near East, or whose verbal artistry was realized in the medium of speech. After all, Sino-Vietnamese, Sino-Japanese and Sino-Korean literature can be (and often is) included as a chapter in Chinese literary histories. The element of difference integral to comparison enters more starkly when Manchu, Mongolian, Tibetan, Tangut, Sogdian, Khotanese, Burmese, Thai, Cambodian, Miao, Yi and other sources are put in dialogue with Chinese and with each other. Indeed, even under dynastic rule the importance of understanding alien languages and cultures was recognized. Although for many centuries the norm in East Asia was for kings to claim autocratic powers bestowed on them by heaven, one fourth-century Chinese monarch is recorded as proposing that someone who is able to ‘penetrate the customs of regions so remote that nine relays of translators are required to communicate with them’ (遠通殊方九譯之俗) deserves state support. (He says this, incidentally, in response to a Buddhist cleric’s challenge to his law-giving authority: an extraordinary concession for a ruler to make.). Although the model generated within China and prominent in Confucian moral writing and history is that of a powerful centre whose progressively less powerful margins are often in need of central recalibration, in practice the inhabitants of marginal zones experienced the call of many competing centres, cultural, political and religious. Thus, the Chinese literate class was not deaf to the call of the foreign – most particularly to that of the dharma, but also to the ‘nomad flute’ (Rorex and Fong) or to the ‘mountain songs’ (Ōki and Santangelo) of incompletely Sinicized regions.

Some cases of inter-Asian cross-cultural adaptation are well known in the secondary literature. Ji Xianlin has observed that the animal fables of the Panchatantra are found migrating in the company of Buddhist texts from India to China, where they encounter another line of animal stories collected in the Zhuangzi (c. 340 BCE). Victor Mair has described how the Mahāmaudgalyāyana narrative – multiply adapted, from sutra to story to ritual opera – uses the Buddhist tale of a descent into hell in order to propose a reconciliation between monastic and familial-piety systems of value. Plot lines from late imperial Chinese fiction were borrowed and renarrated according to the expectations of Korean, Japanese and Vietnamese readers, becoming the basis for major works in the respective national canons (see Pastreich and Huỳnh). Although elements of content from Chinese literature readily crossed over into neighbouring languages and cultures, the reverse path was much steeper. It was not until the twentieth century that anyone thought to translate The Tale of Genji or Songs of Flying Dragons into Chinese. Poems in Chinese characters written by Koreans, Japanese or Vietnamese were, however, often included in Chinese imperial anthologies with no special framing or apparent need for interpretation. Such are the fruits of diglossia. The pre-modern Asian literary space is therefore not an undifferentiated one with texts rolling about like so many unconstrained billiard balls. Rather, it has its filters, barriers and occasional connecting tunnels. Here, Chinese characters and Chinese genres have maximum portability. Mahayana Buddhism, indeed, adopted Chinese as its second (and, some would say, dominant) language. Other writing systems travel in more constrained channels. The same is true of genres. The originally Tibetan epic of King Gesar of Ling, recomposed in Mongolian, Turkic, Manchu and other languages, is found in sung and written form in all in the lands surrounding China on three sides, but is unknown in Chinese tradition itself. This seems to indicate not only an indifference to Tibetan cultural achievement outside of religion, but hostility to military epic as a genre, to which Chinese terrain may have been inhospitable by the early acquisition of writing and through a behavioural code based on hierarchy and civility. Or if we consider the ritualized, formulaic language of oral epic to be a kind of writing, then the Gesar epic’s invisibility in China is merely another case of the Chinese character pushing its rivals to the margins. These examples of generic and thematic fluidity – and the barriers opposed to such fluidity – teach us how to read the map of pan-Asian pre-modern cultures. To take it as a system of coordinates is insufficient: like a navigational chart, it indicates currents, tides, shoals, beacons and other features that permit or hinder passage.

Where Woodberry saw in Weltliteratur the realization of a dream dreamt ‘from Plato to Goethe’, the empirical charting of worlds defined by the passage, or not, of Asian letters offers a different lesson. Asian literary history before European influence is a series of world orders, each hierarchical, each exalting some principle derived from debates that perturbed the communities of the so-called ‘Axial Age’. It offers not one but many alternatives to current thinking about what constitutes global culture. Merely to describe the ways in which it varies from the models now in circulation amounts to a challenge. Let us replace ‘the exotic charm of another system of thought […] the stark impossibility of thinking that’ (Foucault, xv) with the incontrovertible evidence that some did. And others didn’t.