I have been wondering about a long-running controversy in Indian textual studies as to the relative merits of analyzing a given textual tradition either ‘into layers’ or ‘as a whole’. I am sure that these two things mean little to someone not in the business of analyzing pre-modern texts, but it is an important issue for our project as, more and more, scholars are identifying with one or the other camp. Beyond this, the issue of how scholars approach religious texts is, in particular, an important one. Questions concerning the historical origin and development of the canonical texts of Christians, Muslims, Jews, Hindus and Buddhists, to name only the most prominent religious traditions, are both contentious and important (and not just to academics). Before I begin my ruminations, I suspect it might be useful for me to say precisely what I mean by analyzing sources ‘into layers’ (by which I refer to textual criticism) or ‘as a whole’ (which I call textual analysis).
The analysis of texts into ‘layers’ involves the examination of the available manuscripts of a given textual tradition to see which parts of it came first and which later. There are quite complex ‘rules’ about how to compare manuscripts with one another, but the basic principle is simple: one finds similarities and differences between manuscripts and consider why they might be there. This is the main business of textual criticism. The practice of textual criticism is also often extended to include educated guesswork as to the development of a given text before we have direct manuscript evidence for it (this is very important in Indian Studies, where there is little manuscript evidence before the latter part of the first millennium CE). This sort of guesswork is often based on the idea that one style of writing (or speaking) might be earlier than another one, or that the presence or absence of words borrowed from other languages is significant. Scholars also look for relationships between different texts (especially if we have more reliable dates for some of them). In Indian Studies, for example, there are a considerable number of Chinese versions of Buddhist texts for which we have reliable dates. This can help us, even if only tentatively, to say something about the date of non-Buddhist sources that have something in common with the Buddhist ones. Scholars also make use of corroborating historical data, such as literature or material culture. They even use their own ideas about the order of development of ideology and practices in India to suggest the presence, or absence, of ‘layers’ of composition in a given text. This might be described as the use of educated guesswork from one area to inform educated guesswork in another! Whatever the precise method, the goal is a constant: the re-construction of the history of the development of a given text.
The analysis of a text ‘as a whole’ involves a consideration of its structure and content in order to say interesting things about its nature and probable function. It relies heavily on the close and careful reading of a given source in order to say something about the key ideas and themes that circulate within it. Such readings pay close attention to the words and language that are used, and to any features of a text that appear to provide an insight into the logic of its composition (and use). This sort of close-reading tends to be accompanied by a less dogged focus on dates, and the order of development of ideas and practices, and more on the characterization of textual ‘concerns’. These analyses are often informed by broader academic interests in the reconstruction of such things as gender, personhood or the exercise of power. Suchlike analyses are often subsequently historically contextualized on the basis of the idea of a ‘fit’ between the ‘concerns’ of the text and the ‘concerns’ of a given period. Scholars also use a wide variety of literary and material cultural sources in order to corroborate their ‘readings’ of a given text (and its historical context). This is often no less of an example of the use of one form of guesswork to inform another than we saw in the practice of textual criticism. Whatever the precise method employed, the goal of textual analysis is a constant: the reconstruction of the purpose of a text on the basis of its design (whether or not this purpose was consciously known to its authors and subsequent users or not).
Black and Geen exemplify this second ‘holistic’ camp in their introduction to their edited volume ‘The Character of Character in Early South Asian Religious Narratives’ (see Naomi’s summary of this volume in another post for the full bibliographical details). They comment:
Although we draw upon both philological and historical studies, as well as other trajectories in western scholarship, we have taken a more synchronic approach, accepting an individual text as it now exists and taking it as a unified whole. (p.9)
Such a comment conceals some knotty problems for the person engaging with pre-modern Indian texts. To take a text ‘as it now exists’ is often, for the ‘synchronic’ reader, to take up a text that has already received the attentions of someone practicing the discipline of ‘textual criticism’ (this is true for a large number of the texts cited in the bibliographies of Black and Geen’s volume). In fact, this is invariably the case unless one is studying a specific manuscript (or an unedited printed edition). As if to emphasize this contradiction, Black and Geen go on to say:
…we merely suggest that a preoccupation with textual layers often results in a glossing over of the creative and deliberate ways by which early South Asian narratives have been composed, compiled and edited. (p. 9)
Now, this is an extraordinary claim and one which says more about the current division between those who read in ‘layers’ and those who read in ‘wholes’ than anything else; the very means by which we know anything about how a text has been ‘compiled and edited’ (if not always composed) is through textual criticism. What is, in fact, being said here is that a synchronous reading reserves the right to postulate an overarching textual logic on the basis of observable features of the text regardless of how it has been edited by others. I am not saying that this is necessarily wrong (that depends on the particular text, the evidence available for it and the quality of the editor, not to mention the quality of the ‘synchronic’ reading), but I am saying that there is no need to repudiate textual criticism if this is one’s agenda. Black and Geen’s volume is of the highest quality and my comments in no way detract from the rigor of their overarching analyses nor of the quality of the individual contributions. My concern is the idea that we have to ‘choose’ between reading in terms of ‘layers’ and ‘wholes’. One might also mention the fact that we are ‘turning away’ from the very means by which the majority of texts that we use in pre-modern Indian Studies were made. More than this, I am concerned that we are being inconsistent in our treatment of the role of inference in the two ‘types’ of textual study; the person who addresses issues of textual development and change over time is increasingly branded a reactionary and a crank if they posit a development based upon, but moving beyond, the available evidence, while the ‘synchronist’ is granted considerably more freedom and is labelled progressive for doing much the same in relation to textual content. This is not a scientific stance. For our project, I hope that Naomi and I can rest between the two ‘camps’ and indulge in the luxury (and necessity) of speculation about both ‘layers’ and ‘wholes’.