It is instructive to review Stephen Owen's two most recent books together, not merely on the grounds of their shared authorship and year of publication. Though they treat different moments in Chinese poetic history, they share a single critical and historical perspective that, if embraced by scholars, as Owen urges, would redirect the discipline's customary discourse about traditional Chinese literature and raise its level of sophistication in the process.

The subject of Owen's Classical Poetry is primarily pentasyllabic verse: the anonymous “old poems,” pieces in Li Ling's corpus, and much of the poetry attributed to writers of the early third century. It is a body of informal poetry—Owen's useful term is “low register”—that enjoyed popularity among the elite in that era. The word “making” in Owen's title has two senses, signaling two parallel stories in this study: the poets' process of composing low register poetry, on the one hand, and the culture's process of preserving and canonizing it as “classical,” on the other. The result is a major departure from the standard chronological narrative of Chinese classical poetry, which is occupied with the development of genres and distinctions between authors. Such a model requires the poems to be attributable to particular periods and poets, but classical poetry, observes Owen, is “a kind of shared poetic practice of uncertain origin” (p. 8), and much of what survives comes to us without author or date attached. To account for this poetry without forcing the anonymous poems to fit into a “historical network that confers meaning,” Owen proposes a synchronic approach to poetry: instead of “genre,” “author,” or “poem,” Owen focuses on shared “themes, topics, sequences of exposition, templates”—the “range of verbal habits” by which the poetry was generated (p. 9).

Thus, in chapters 2–6, Owen compellingly illustrates that classical poetry was in many ways “one poetry,” “created from a shared poetic repertoire and shared compositional procedures” (p. 3), by analyzing the “grammar” of early poetry and “mapping” two dominant and related themes (immortality and feasting) in poems assigned to different periods and poets, including anonymous ones. In contrast, Owen does give a historical account of China's early medieval writing of literary history. In chapter 1, he examines the evidence for textual conservation practices in the Southern dynasties, the processes by which pre-Jian'an poetry was reproduced and transmitted, to establish that “the textual medium through which we have received this material has been selected and sometimes radically reshaped [in the fifth and early sixth century particularly] to produce our standard account of the origins of classical poetry” (p. 3).

Possibly I would single out chapter 2 as exemplifying most vividly the liberating power of treating early poetry “as ‘one poetry’ in varying realizations, as if synchronic” (p. 9), though no chapter is without its rewarding insights. There, Owen shows that “problem” passages and poems often become comprehensible once they are read “within sets of other poems of the same kind” (p. 10). He does so by analyzing various segments that seem out of sequence or even entirely out of place in certain yuefu poems and by examining several bird images that seem to scholars misapplied within the context of poems across different category boundaries (yuefu, anonymous old poems, and informal poetry of Jian'an poets). In each instance, Owen demonstrates persuasively that these are traces of shared “poetic material … open for multiple variations,” components and formulae that have been expanded or contracted, realized in different registers, applied singly or linked with others (pp. 105–6). In short, Owen gives us means to account for poetry that has long been characterized too simply according to the lockstep procedures of conventional biographical and chronological literary history.

Owen sets himself a greater challenge in his other book. Late Tang represents the final installment in his comprehensive history of Tang poetry, which started with The Poetry of the Early T'ang in 1977 (New Haven, Conn.: Yale University Press), followed by The Great Age of Chinese Poetry: The High T'ang in 1980 (New Haven, Conn.: Yale University Press) and, in a different vein, The End of the Chinese “Middle Ages”: Essays in Mid-Tang Literary Culture in 1996 (Stanford, Calif.: Stanford University Press). But naturally, as Owen acknowledges, developments in the field of literary studies and changes over the years in his own critical interests have led him to conceive of literary history quite differently than he did in earlier decades, despite there being “some constancy of questions” that allow him “to call it by the same name” (p. 14). “Literary history, as an enterprise,” he has come to appreciate, “is a function of the particularities of the period it represents”—“not a unified discipline that focuses on different ‘objects’ but [one] reconfigured by the objects it pretends to describe” (p. 14). Having presumably written Classical Poetry alongside Late Tang, moreover, Owen is doubly aware of the partial and “mediated” nature of what survives to represent Late Tang poetry to us. “We have many tantalizing texts that remind us—by their very limitations—of a large and more diverse world of poetic production in the Late Tang, material that has largely been lost,” he cautions (p. 12).

Such repeated reminders that “the record is skewed in odd ways by vagaries of taste and accidents” (p. 40) call for a literary history that must perform something of a balancing act. “Rather than making generalizations about the period as a whole,” he tells us, “I have tried to be as historically specific as possible, always paying attention to the way Tang poetic legacy was textually preserved” (p. 11). Owen nevertheless does offer generalizations about the period as whole, and many more about the authors and groupings of poems that he treats. And, as in his earlier books, these generalizations are frequently keenly perceptive, even invigorating, for so often Owen puts his finger on just the right details and captures their effects in just the right words. About Late Tang poetry generally, Owen observes an intense and persistent retrospective gaze: it is a period of “poetic conservatism” (p. 14), he remarks, when poets looked to the past for inspiration so intently that its poetry strikes us as “traumatically ossified” (p. 7). This fixation on continuity as opposed to innovation in the last eighty years of the Tang warrants Owen's decision to make 860 rather than 907 the endpoint of his discussion (860 is a “convenience rather than the mark of a moment of change,” p. 7), but also it generates the book's recurring metacritical argument, the challenge to recognize that the evidence for this period's conservatism “is not only part of a history but shapes the very notion of history” (p. 14). “If we wish to uncover the relationship between the history of poetry and the larger sense of ‘history,’ ” Owen suggests, “we may find it not in changes in poetry but in poetry's refusal to change, in its fine couplets, its absorption in pleasures both poetic and sensual” (p. 7).

When Owen turns to capture the distinguishing characteristics of individual poets, he provides still greater service to readers by mapping the poetic landscape of the period with memorable technical and literary-critical markers. “Tang poetry tended to parataxis, with each line a distinct predicate,” he notes, but “Li He's parataxis was more extreme: there was more distance between the lines, creating the effect of verbal fragments that only partially cohered,” giving his verse a “dreamlike” quality (p. 163). Cao Tang's “eroticized and romanticized poetry” comprises “images of ninth-century romance culture and fantasies about palace life … transferred to the immortal world,” but because Cao Tang also “adds the technical vocabulary of Daoism and the immortals' perspective on time,” he ultimately “underscor[es] the difference between the immortal and mortal worlds” (pp. 320–21). The appeal of Li Shangyin's “hermetic poems,” to cite a final example, “lies in the strong contrast between the density and difficulty of the inner couplets and the relatively directness of the closure, often marked by an urgency of address or some emotional coloring” (p. 389).

Such characterizations as these have much potential utility. Yet I would return our attention to the balancing act of Owen's literary history to locate his books' most important contribution. Granted, this balancing act is necessitated by a paradox. If we take to heart Owen's reminders of the fluidity of texts over time, the accidents of taste and luck that account for the transmission of some literary texts but not others, and the skepticism we must ever maintain about traditional attributions of poems to particular poets and time periods, then we want to know, at what point are we warranted to have confidence in generalizations about texts, poets, and periods? In Classical Poetry, discussing a poem attributed to Cao Zhi, Owen makes a point of stressing that the received version's source text depended on a “manuscript tradition” that “belonged to an age of textual fluidity,” and he asserts, “The text as we have it may be exactly as Cao Zhi composed it, but on some level that ceases to matter. We have one possible ‘rendition’ ” (pp. 142–43). Elsewhere, however, Owen does not tell us how he knows where to draw the line on taking this perspective. Yet the line does get drawn—in Owen's casual reference to “the term Cao Cao used” in a poem, for example (p. 168), and in comments such as this one: “Some of the Cao Pi attributions are questionable, but there is no reason this poem could not have been composed by him” (p. 193).

Likewise in Late Tang, Owen cites the example of “Yang Yi's [974–1020] painstaking reconstruction of Li Shangyin's collected poems” as “a sober reminder of how our understanding of many other poets' works has probably been distorted by their preservation only in manuscript anthologies, which were meant to please late ninth- and tenth-century taste” (p. 338). But there is another obvious implication of Yang Yi's “reconstruction” efforts, conducted well over a century after Li's death. Owen summarizes Yang Yi's report that he had started with “just over 100 poems” but eventually gathered more than four hundred, which represents “just over two-thirds the size of the current collection” (p. 337). Surely we should have suspicions about every stage of this process. How many of the poems gathered by Yang Yi, and how many added to them since, are “authentically Li's”—as opposed to poems composed by others out of the same repertoire of shared components and formulae, perhaps written in imitation of Li Shangyin but perhaps merely discerned by later generations as poems he must have written? And how similar are the texts of these poems today to what they were a millennium ago? How much were they adjusted, by multiple hands over decades or centuries, to make them seem more like poems that Li Shangyin must have written?

In short, textual fluidity and dubious attributions threaten to stop us cold, unless means can be found either to determine where fluidity and doubt end and reliability begins, or to redirect our practices as literary historians and critics to accommodate, even exploit, the evidence of textual fluidity and dubious attributions. It may seem strange to praise Owen's recent books for arguing a perspective that would seem to undermine other of his arguments, but it is to his credit that he does not shy from doing so. He has thereby posed both himself and us with a challenge that promises to be inconvenient, even distressing, but at last liberating if it helps to push the field in the new directions toward which it is fitfully groping.