Reading the Mahāvaṃsa: The Literary Aims of a Theravāda Buddhist History

AS HISTORIES, the vaṃsas are explicitly concerned with linking the “good people” of the textual community to the Buddha through both narratives and the actual presence of relics. The Pāli Mahāvaṃsa is at once historical and literary—the former because it was written in a particular cultural and temporal moment about other events in the deep recesses of the collective imagination or inherited cultural memory of its community of production; the latter because it employs devices such as metaphor and plot development to tell that story to its audience. But the categories “history” and “the literary” are far from mutually exclusive; a rethinking of the Pāli Mahāvaṃsa reveals how very dependent on narrative and “the literary” the early Buddhist textual community in Sri Lanka was in formatting and arguing its own “history.”¹

Hayden White has extensively analyzed the structure and processes of history writing and examined how the rhetorical choices made by the historian in the presentation of his or her material prefigure the range of audience responses, much like the intentional fabrication concocted by the literary author.² Dominick LaCapra extends the discussion of the intentionally fictive dimension of history writing and interpretation to claim that in narrative “worklike” elements engage the audience in imaginative and interpretive acts that move beyond mere “documentary” representations of the past.³ Paul Ricoeur, too, has extensively examined the aesthetic and ethical effects of narrative, whether classed as historical or fictive in nature.⁴

In her argument for renewed appreciation for a vibrant and productive Buddhist literary culture in medieval Tamil areas, Anne Monius argues that certain “literary works” are agents of religious community formation and imagination: “Ornate poetic narrative, even poetic theory itself, draws attention to various kinds of cultural practices in the absence of archaeological or inscriptional evidence and offers fresh insight into the long and complex historical processes of debate, selection, transmission, and recreation that constitute religious community.”⁵ This assertion may appear at first the converse of Gregory Schopen’s recommendation for the Buddhologist (and, more broadly, the historian of religion) to eschew the “Protestant presuppositions” inherent in the study of religious communities, whereby the text is privileged over concrete sources of historical information that can be excavated, literally, through archaeology and epigraphy.⁶ But in fact both Monius and Schopen call for a new way of reading those textual, “historical” sources. In the absence of definitive sources from material culture to employ as a corrective to the overly textual orientation of the historian of religion, and in the face of the continued (or new) emphasis on the centrality of the Mahāvaṃsa in Sri Lankan Theravāda Buddhism, rethinking Buddhist histories such as the Pāli Mahāvaṃsa in explicitly literary terms becomes a fruitful corrective to the common “documenting history” or “establishing a political charter” readings that are more typically performed. My method for reading does not undermine the utility of prior readings but rather extends the possible interpretations of the Mahāvaṃsa to include the kind of personally and communally transformative literature that creates and sustains a religious, textual community.

Reading the Mahāvaṃsa as a literary work rather than exclusively as a historical source requires us to change our expectations regarding the value of the text or the possible effect it has had or will have. We simply cannot be sure of the veracity or objectivity of any of the claims in the Mahāvaṃsa when it is read as a source for social history. We have no concrete evidence outside of our texts, except for brief epigraphical donative inscriptions, to confirm or corroborate the timeline of events—the “history”—presented therein.⁷ Thus, we should approach these sources with the hermeneutic of suspicion;⁸ these texts are ideological in nature, not documentary resources through which a reasonable facsimile of the events and circumstances of history may be constructed. They may be arguments, agendas, and ideological treatises or self-conscious attempts at rendering good local literature into Pāli for particular religious or aesthetic aims or both. We may never know the precise sociohistorical situation of the textual communities responsible for the Pāli Dīpavaṃsa and Mahāvaṃsa, but I suspect we can know something about their literary concerns. I contend that one of the primary concerns for the fourth- and fifth-century Mahāvihāran monks responsible for compiling these two vaṃsas was to provide a vehicle for the continued presence and proximity of the Buddha through evocative, transformative literature—a narrative unfolding through the course of time as recounted in the text.

The primary key employed to unlock the hidden meanings of the Mahāvaṃsa and for fleshing out the scanty data available about its historical context is its medieval commentary Vaṃsatthappakāsinī. The exact date of composition of this text is unknown. The earliest it could have been written would have been after the reign of Dāṭhopatissa II (659–667 C.E.).¹⁰ It is referred to in other texts by the thirteenth century. Whereas Mahāvaṃsa editor and translator Wilhelm Geiger dated the Vaṃsatthappakāsinī to 1000–1250, based on his assumption that the text references the tenth-century Bodhivaṃsa in its present form,¹¹ G. P. Malalasekera dates it much earlier, citing the fluid nature of texts prior to critical editions. He suggests that it was written in much closer temporal proximity to the Mahāvaṃsa, believing that there is evidence in the text to suggest that the “original sources” for the Mahāvaṃsa “were still being studied” while the commentary was being written. As Malalasekera writes, “Since Mahānāma [the author-compiler of the Mahāvaṃsa] is generally believed to have lived in the sixth century [C.E.], it would not, I feel, be too early to assign the author of the MṬ [the Vaṃsatthappakāsinī] to about two or three centuries later, the eighth or ninth century. This would also allow sufficient time for variant readings to develop in the [Mahāvaṃsa] text.”¹² Whether we follow Geiger or Malalasekera’s dating, we know that the Vaṃsatthappakasinī was written at least a few centuries after the Mahāvaṃsa and thus represents the concerns of an entirely different community than that which produced the Mahāvaṃsa. The Vaṃsatthappakasinī follows the narrative of the Mahāvaṃsa quite closely, although it chooses to elaborate on some sections far more than others. Both early translators, George Turnour and Wilhelm Geiger, used the commentary to read and translate the Mahāvaṃsa. The commentary does elucidate otherwise hard-to-render passages of the Mahāvaṃsa, but it nonetheless represents later concerns and values used to interpret the text and not necessarily the concerns and values of the Mahāvaṃsa itself.

Likewise, reading the later section of the Mahāvaṃsa, or what Geiger called the Cūlavaṃsa, alters the overall effect of reading the Mahāvaṃsa.¹³ Each time the Mahāvaṃsa was continued, the concerns of the new interpretive community were inserted into the text; and even if a past section was not altered, the document as a whole was nevertheless affected. The later section further reorients the text as it brings the history of the island to the modern period; its inclusion in a reading of the Mahāvaṃsa results in a sense of this text as history. The first part of this later section begins in the fourth century with the arrival of the tooth relic, a key event in the history of Theravāda Buddhism, and then spans centuries to culminate in a narration of the twelfth-century reign of the influential king Parākramabāhu I.¹⁴ The compilation of this section is claimed by the monk Dhammakitti in the twelfth century. The material may have gradually accumulated, making Dhammakitti only one hand among many making editorial decisions, but the ultimate form that the first part of this later section takes certainly reflects the concerns of the twelfth-century Mahāvihāran community. The next installment continues through the arrival of the British in 1815, which segues into the more recent interpretations (and, in fact, to the continuation of the Mahāvaṃsa to the present day). Here in this book we are singularly concerned with the first section of the Mahāvaṃsa that concludes with the thirty-seventh chapter, entitled “King Mahāsena,” and ends with the simple subscript, “Mahāvaṃso niṭṭhhito” (The Mahāvaṃsa has ended).

The Mahāvaṃsa is thus layered with accretions, with each new textual community subjecting it to a new layer of interpretation. Any stratum we might isolate to undergo historicizing scrutiny bears the marks of the interpretive community of the era in which it was created.¹⁵ As R. G. Collingwood observes, “The historical past is the world of ideas which the present evidence creates in the present. In historical inference we do not move from our present world to a past world; the movement in experience is always a movement within a present world of ideas.”¹⁶ Of course, textual communities through time are hardly as distinct and bound as layers of an onion, with each new understanding of a text revealed as we peel back intervening interpretations. Nor are they necessarily monolithic, although a singular, unified interpretive vision is the general idea the vaṃsas intend to provoke. The text of the Mahāvaṃsa acts as the central cohesive unit for such a textual community at the time of its production, but later communities of interpretation are in constant evolution, especially as they begin to produce commentarial material that attaches successive and sometimes disparate layers of meaning to the original cohesive text.

The modern interpreter must use his or her imagination and a study of history to arrive at any understanding of the community. Even the most thorough of investigations will result in conjecture at best, not “fact.” As Brian Stock says,

Jonathan Walters concurs with this situated reading of the textual communities when he recommends that we read Pāli texts as “actions within the sociohistorical circumstances of their production rather than as passive transmitters of neutral information.”¹⁸ If we pay attention to the active, constructive, worklike elements of the text, whether explicitly rendered as messages to an intended audience (such as “good people”) or subtly conveyed through rhetorical devices, we may be able to better understand how the texts have been interpreted through the centuries. We may also discover that we have no recourse to “actual history” within these texts; they may indicate much about the textual community responsible for their production, but even this information comes to us in a highly manufactured form. All written histories play with the persuasive powers of text and demand that the reader-hearer follow along the trajectories they set, accepting even the most fanciful of situations in service of maintaining a narrative arc that leads to the present.

Hayden White reflects on the tension between the reader’s present activities and the past as presented by the text: “The historically real, the past real, is that to which it can be referred only by way of an artifact that is textual in nature. The indexical, iconic, and symbolic notions of language, and therefore of texts, obscure the nature of this indirect referentiality and hold out the possibility of (feign) direct referentiality, create the illusion that there is a past out there that is directly reflected in the texts.”¹⁹ The Mahāvaṃsa in particular requires a new reading, one that pays attention to the material of the text and how it may be a window onto the particular concerns of the textual community of its production, so often assumed to be the fifth- or sixth-century manifestation of the Mahāvihāra. But it also begs to be read as a literary text, full of self-proclaimed good poetry and emotionally evocative material, which happens to find a “historical,” chronologically organized narrative the most effective framework for transmission.

Because much of the vaṃsa literature chronicles the successive reigns of kings, it is commonly read through a political lens. Any attempt to simplify the explanation of the complex relationship between the saṅgha and the ruling power leads to oversimplification and to the bias toward a politicized reading of the Mahāvaṃsa. Heinz Bechert writes:

There are several problems with this argument, most obviously the anachronistic use of the term state.²³ Nonetheless, the parallel vaṃsas (lineages) of the saṅgha and the king are present in both the Dīpavaṃsa and the Mahāvaṃsa. Ample fodder exists for the interpretation of the lineages as “a model of and a model for the Sinhalese Buddhist identity…. The resultant stereotypes are one element in the Sinhalese Buddhist world view.”²⁴

Richard Gombrich asserts that “the Mahāvaṃsa is the charter of Sinhalese Buddhist nationalism,”²⁵ but framing the Mahāvaṃsa as a nationalist charter is an anachronistic and self-serving interpretation and a politically legitimizing usurpation of a multivalent literary text. By stating that the vaṃsas constitute “the sacred history of a people destined with a sacred mission, namely, to maintain the purity of the Dhamma in a world of impermanence and self-seeking,” Bardwell Smith comes closer to the sort of interpretation I advocate.²⁶ “Sacred history,” or Heilsgeschichte, operates beyond the mere reportage of events; it is a theological and ideological claim.²⁷ In the Mahāvaṃsa, we read such a claim in the proem, and it is repeated in the stock phrase at the end of each chapter that no matter how banal the subject may be, it was “made for the anxious thrill and serene satisfaction of good people.” However, key words in Smith’s argument, such as sacred, destined, and purity, oversimplify the impact of this text and add a layer of interpretation because of his word choice. “Purity” and “destined” are concepts never explicitly used in the text in reference to the dhamma, but the text’s modern interpreters nevertheless read them as implicit.²⁸

Smith’s assessment is also employed by scholars working in other fields to further their own arguments even though they may not have read the Mahāvaṃsa. For example, political scientist David Little concludes that “the chronicles have a mythical aspect. Historical facts are embellished in various ways to lay the foundations for the ‘charter’ of an ideal social order.”²⁹ Why privilege the “historical” as the basis upon which “mythological” embellishments accrue? Couldn’t the Mahāvaṃsa just as easily be read as a mythological text to which historical “facts” are added to suggest the idea of charter? After all, the text begins not with a “historical” event but in the mythical realm, with the Buddha’s three visits to the island.

Alice Greenwald, too, categorizes the Mahāvaṃsa as a charter, but in a broader sense than a specifically political charter. She almost claims the converse of Little’s statement: “[A] [c]harter is conceived of in the Malinowskian sense as that which, through the medium of myth (or, in this case, religious historiography), both engenders and substantiates a cultural self-consciousness.”³⁰ The Dīpavaṃsa and Mahāvaṃsa are charters in this latter sense insofar as they both generate and validate a “cultural self-consciousness” by and for the textual communities that compiled them. They are the texts that form the fulcrum for the textual community, the reason for its very existence.³¹

I disagree that the vaṃsas are charters for contemporary political purposes. I understand the Mahāvaṃsa in particular as a charter for an internal conversion experience or a deepening of faith for the audience, individually or collectively. After all, this text was compiled by and for monastics and became available to a secular audience only because it had been preserved in monastic settings. If its proem is to be believed, it certainly did not set out to chart new territory or political agendas for the twentieth- or twenty-first-century Sinhalese; the narrative focus is on the past and the particular sentiments generated when one reads or hears it. The Mahāvaṃsa does claim superiority over what it deems to be lesser expressions of the same material, but this claim is made in literary (and not overtly political) terms.

One could argue, however, that the vaṃsas do chart the religious and ethical aspirations of communities in the future. In that sense, they are as much about the future as about the past, but only insofar as they reach out to the future reader and implicate that reader in participating in the text’s religious dimension.³² For example, we find in Mahāvaṃsa XXXII an adaptation of the standard formulation the Buddha reveals at the end of every Jātaka story (“At that time, I was so-and-so,” “Ānanda was so-and-so,” etc.). This formula comes to be anticipated by the reader of the Jātakas, who becomes increasingly aware of the interwoven relationships that transcend individual births. It is this “group karma” encountered in the paradigmatic Jātakas that is mimicked in the Mahāvaṃsa, but not with reference to the past birth of the Buddha. It instead appears in the context of the future—namely, the anticipated birth of Duṭṭhagāmaṇī in the time of the Buddha Metteyya, the future Buddha. In the scene of Duṭṭhagāmaṇī’s death, replete with a chariot from Tusita heaven to carry him away, circling the Mahāthūpa three times in the air, we see the narrator conclude the story with the observation that “Great King Duṭṭhagāmaṇī, he who is worthy of the name king, will be the first disciple of the blessed Metteyya; (Metteyya’s) father (will be) the king’s father, his mother the mother, (Duṭṭhagāmaṇī’s) younger brother Saddhātissa will be his second disciple. But the son of the king Sālirājakumāro will be the son of the blessed Metteyya.”³³ If this particular portion of text is working as a charter, it is a religious—not political—one.

Another problem with the classification of the Mahāvaṃsa as a political charter is that its proem explicitly declares its religious function—namely, to arouse saṃvega (awe, religious agitation) and then pasāda (serene joy, peace). The primary aim for a political charter typically is not to arouse specific emotional or ethical effects in a reading audience. The proem makes it clear that the Mahāvaṃsa is a charter only insofar as it circumscribes a certain textual community (of “good people,” sujana), a community that it seeks to create.

Thus we can see that the term charter does not easily apply to the Mahāvaṃsa, in spite of its common use in politicized readings performed for specific political agendas. The modern reception and categorization of the Mahāvaṃsa as a charter nonetheless has had an impact on Sri Lankan Theravāda communities of the present. H. L. Seneviratne refers to the dominant Sinhalese Buddhist understanding of the island’s history as the “Mahāvaṃsa view,”³⁴ and Tessa Bartholomeusz refers to it as the “Mahāvaṃsa mentality,”³⁵ a bellicose one at that.

The Mahāvaṃsa is best read by suspending the type of reading that has been informed by state-building, identity-reifying political rhetoric. Although the document attests to the struggles for power through the ages, jumping straight to that conclusion (of preauthorized, even mandated and sacralized Sinhala hegemony) overlooks the work of the text. Eschewing recent rhetoric that espouses the “Mahāvaṃsa view” or the “Mahāvaṃsa mentality” as a singularly hegemonic political ploy, we open ourselves to the text’s more literary function and to what may have been at stake for the community that inspired the fifth-century composer.

To fully appreciate the literary qualities of the vaṃsas, however, we must situate these works, in particular the Mahāvaṃsa, in the community responsible for generating them and for whom the literary maneuvers were crafted. Herein lies a methodological dilemma that has plagued (but not impeded) all prior research into the historical era responsible for the earliest Pāli vaṃsas. The primary source for the historical information reported in the plethora of secondary sources is the Mahāvaṃsa itself. The object of our literary study is thus also the main informant on the social history of the time. It is certainly possible that the text can reveal significant information about the period in question, especially when one reads between the lines and accepts the chronology and implicit argument of the text’s author-compiler. But there remains an inescapable tension between what we think we can be certain about and how we have derived what we know from the text that is the object of our study. This inscrutable tension between methods and genres must inform a new evaluation of the Mahāvaṃsa.

In the medieval West, “histories” come in several forms, one of which is the chronicle, a form that adds depth and detail to the basic temporal sequence of the annals. As Hayden White explains,

Structurally, then, the Mahāvaṃsa and Dīpavaṃsa could be classified as chronicles because, at least in the case of the Dīpavaṃsa, they do trail off into the compiler’s present. Yet the designation chronicle is complicated; it may attenuate a fuller comprehension of the text and unhelpfully inform the modern reader about what to pay attention to and what to mentally edit upon its encounter. This designation divests the first story cycle of its narrative strength. The very designation of these texts as “chronicles” in the minds of their interpreters, who have been infected by modern standards and methods of historiography, seems to have triggered a certain mode of interpretation that consciously overlooks the ethical dimension of the act of reading or receiving the text and that edits out the historically challenging aspects of the narrative (such as the miracles surrounding the relics and the nāgas). The problem exists in calling the text a chronicle and thus exporting medieval European expectations (or, more appropriately, modern scholars’ and historiographers’ notions) about what a chronicle conveys.

The structure and content of the first chapters of both the Mahāvaṃsa and the Dīpavaṃsa seem to be closely related to the genre of Sanskrit purāṇa.³⁷ Both open with the construction of a mythic landscape in a timeless time, through which the narratives of the lineages of kings are to be interpreted. Acknowledging the assertively literary qualities of the Mahāvaṃsa and based on his reading of the five elements (pañcalakṣaṇa) of purāṇa writing, Steven Collins concludes that the Dīpavaṃsa and Mahāvaṃsa have much in common with Indic purāṇa literature. As he suggests, “We see the Pāli chronicles in this perspective as a part of the literary genre of the purāṇa in the widest sense, listing the genealogy and deeds of the lineage of the Buddha and his heritage.”³⁸ Sequence, paramparā (lineage), legitimacy, and security in being connected to authoritative sources (whether the Buddha or the deities of the purāṇas) are themes that tie the vaṃsas and purāṇas together. Considering the vaṃsas as purāṇa is by no means an original observation by Collins; in fact, Buddhaghosa himself refers to the Dīpavaṃsa as “the Purāṇa.”³⁹ But just as the term history can never fully translate the term purāṇa, so I contend that it is also an insufficient translation of vaṃsa—and yet many scholars translate it this very way.

Stephen Berkwitz argues that because the term chronicle is a similarly insufficient translation of vaṃsa, we should call this genre of text “Buddhist histories.” Following arguments by Stephen Greenblatt, Keith Jenkins, and Hayden White, he concludes that “historical writings are ‘fictions’ in the sense of being made or fashioned, things shaped by human imagination and the available resources of narration while still referring to actual events believed to have taken place apart from its description.” Recognizing that Buddhist histories, with their abundant stories of miracles and fanciful landscapes and characters, may challenge a historian’s reading—even one attuned to the artfully manufactured quality of history—Berkwitz argues that in comparison to modern Western historiography, “Buddhist vaṃsas are more transparent with their ‘fictive’ qualities and revel in the creative potency that is ascribed to their narratives.”⁴⁰

The term history, of course, comes with its own semantic baggage. A proper history assumes a firsthand witness as source. Classicist Carolyn Dewald discusses the origins of the classification “history” as a titular creation by Herodotus as he sought, a full forty years after events had transpired, to uncover the reasons why the Greeks, not the Persians, had won the great war in 481 B.C.E.:

[History] was a noun formed out of a very old Indo-European stem vid- meaning sight, or knowledge gained through sight. We have its cousin words through Germanic Old English in wit, or wisdom, and through Latin, another cousin language, we have vision, and video. In Greek very early, before Homer, the digamma or w-sound was lost, and the stem came in as the verb idein, to see, and in its noun of agency, a (w)istor, or histor, a witness, a man who knows things because he has seen them. In the Iliad, a text written three hundred years or so before Herodotus, a histor is in one passage a judge of a horse race, waiting at the end goalpost to declare the winner, and in another passage he is one of a group of judges gathered to weigh the evidence and determine the outcome of a manslaughter trial.⁴¹

The vaṃsas’ chroniclers were not alive to witness the Buddha’s three visits, the landing of Vijaya, the coming of Mahinda, and the rise of Duṭṭhagāmaṇī. In fact, the narrative telos for the Dīpavaṃsa is the description of the end of King Mahāsena’s reign, a fact that has been interpreted to mean that the writer of the Dīpavaṃsa had thus reached his (or her) contemporary period. But the first cycle of the subsequently written Mahāvaṃsa ends with this same event, which suggests that it is simply a retelling of the content of the Dīpavaṃsa (whether the Dīpavaṃsa itself was the source or not). Given the commonly accepted dating of the Mahāvaṃsa as about one hundred years after the Dīpavaṃsa, why wouldn’t the Mahāvaṃsa compiler have had the confidence to continue the narrative to his own time? (After all, he asserts in the proem such confidence in his poetic ability!) One possibility is that the two texts are, in fact, more contemporaneous than evidence suggests. The evidence that separates them, however, is compelling. Buddhaghosa mentions the Dīpavaṃsa by name but does not mention the Mahāvaṃsa. This seems to set Buddhaghosa’s period of production between the two documents, unless Buddhaghosa was residing with monks who were preserving the lineage and claims of the Dīpavaṃsa, and the Mahāvaṃsa was being written by another faction within the Mahāvihāra. The contentious, assertive proem of the Mahāvaṃsa would then make sense; the Dīpavaṃsa represents a lesser work by a lesser contemporaneous community. Could the assertions of the Mahāvaṃsa author have been an attempt by a separate faction of the Mahāvihāra brethren to curry the court’s favor or to lay claim to the authentic paramparā (lineage) of the Buddha through Mahinda by utilizing the successful rhetorical strategies employed by the compiler of the Dīpavaṃsa?

Although previous attempts at dating the texts have attempted to do so on stylistic ground alone, I am unconvinced that the crude form of the Dīpavaṃsa necessitates an earlier composition date.⁴² After all, poor poets exist in every generation. Several scholars have suggested that the clumsiness of the poetry was due to a lack of confidence on the author’s part in the new idiom, Pāli. It is true that there are several awkward passages in the text, but this awkwardness could have easily been due to the compiler’s aspirations to preserve variant traditions in the text. Of course, as Walpola Rahula asserts, it may also be simply due to multiple authors: “The rugged nature of its language and style, its grammatical peculiarities, its many repetitions and the absence of any plan or scheme in its narrative convince the reader that the Dīpavaṃsa is not the continuous work of one individual, but a heterogeneous collection of material like ballads of some unskilled versifiers who lived at different periods in different parts of the Island.”⁴³ Whereas the Dīpavaṃsa may have been about collating and documenting important stories of lineage (and especially connections to the Buddha through the monks and nuns and to Asoka’s court through the coming of Mahinda to Sri Lanka), the Mahāvaṃsa is about provoking an emotional response in primed readers.

According to previous scholarship, the Mahāvaṃsa springs from Sri Lanka of the fifth century, specifically from the Mahāvihāra monastic complex in Anurādhapura, which was at the time the seat of the monarchy and the center of political and cultural production. All scholars are in agreement that the textual community responsible for the Mahāvaṃsa inherited a set of source material contained in the Sinhalese (or Eḷu, Old Sinhalese) Sīhaḷaṭṭhakathā. Although no such text is extant today, several references found in later texts attest to its prominence and influence at that time. The Mahāvaṃsa proem refers to an older version of the Mahāvaṃsa “made by the ancients” (porāṇehi kato), which remains unnamed there. The medieval commentary Vaṃsatthappakāsinī explains that this source text was written in Sinhalese prose interspersed with Pāli verse.⁴⁴ It also states that monks of the Mahāvihāra were responsible for writing the aṭṭhakathā.⁴⁵ Malalasekera comes to several conclusions about this early compilation of texts:

It is generally agreed that there was preserved in the Mahāvihāra in Anurādhapura an old commentary to the Tipiṭaka, which was generally referred to as the Aṭṭhakathā. This work was written in Sinhalese prose, interspersed with Pāli verses. Oldenberg has suggested that the Sīhalaṭṭhakathā-Mahāvaṃsa…formed a historical introduction to the dogmatic part of this Aṭṭhakathā of the Mahāvihāra fraternity, much in the same way as the historical introduction to Buddhaghosa’s Samantapāsādikā. Geiger disagrees with this view and expresses the opinion that the “Mahāvaṃsa of the ancients” was an independent chronicle, which the dwellers of the Mahāvihāra carried down to Mahāsena, and certainly would have continued further had not the Mahāvihāra been destroyed and depopulated and their peaceable work disturbed by the violence of Mahāsena…. It is quite probable that originally this chronicle formed a sort of historical introduction to the Canonical commentary and dealt only with the 13 subjects mentioned in the proem of the Dīpavaṃsa. Later, however, it burst its framework by the incorporation of a great deal of extraneous information and developed into an independent compilation, whose custodians were the Mahāvihāra monks.⁴⁶

Scholars such as Hermann Oldenberg, Wilhelm Geiger, E. W. Adikaram, G. C. Mendis, and G. P. Malalasekera are in relative agreement that there were texts (porāṇa, aṭṭhakathā) in circulation (oral or written) that formed the source material for the Dīpavaṃsa and Mahāvaṃsa.⁴⁷ Oldenberg considers the porāṇa to be a historical preface to the aṭṭhakathā (or to Sīhalaṭṭhakathā-Mahāvaṃsa).⁴⁸ Geiger seems to consider the aṭṭhakathā and porāṇa as “one and the same.” Adikaram and Malalasekera seem to be in agreement that the porāṇa refer in a general way to “teachers of old.”⁴⁹

Although the Vaṃsatthappakāsinī states that the Mahāvaṃsa (there named Padyapadoruvaṃsa) presents a faultless version of the Sīhālaṭṭhakathā-Mahāvaṃsa, which was riddled with dosa (defects), it also states that “they are not really dosa, but they make the work difficult to understand and remember. It is only these ‘difficulties’ that the Mhv. [Mahāvaṃsa] had tried to avoid.”⁵⁰ But this source, too, is dated centuries after the Dīpavaṃsa or Mahāvaṃsa would have been compiled. We cannot be certain of the exact nature, provenance, or even the existence of the source texts for the Dīpavaṃsa and Mahāvaṃsa, but we can at least question why the authors of both texts would have explicitly claimed in their proems that they were revising older materials if such source texts did not exist (Dīpavaṃsa I.4; Mahāvaṃsa I.2). Reading the Dīpavaṃsa and Mahāvaṃsa together, we can be fairly certain that each text followed some manifestation of an urtext rather closely because whole story cycles are repeated (sometimes practically verbatim) in both texts.

Both texts begin with the ultimate legitimating persona, the Buddha of this (our) era. We can be fairly certain of one thing: in the Buddhist milieu during Gotama Buddha’s life, there was a central authority that could ensure, to some degree, the continuity and coherence of the religious order known as the saṅgha. Following Max Weber’s understanding of the “routinization of charisma,”⁵¹ we can assume that in the social life of the saṅgha in the period immediately following the Buddha’s corporeal death (parinibbāṇa), considerable attention was paid to the systematization of institutional resources and policies to ensure the longevity of the Buddha’s vision for his community. Much is said in the canonical sources to explain the necessity to preserve the sermons (suttas) and the rules of the order (vinaya). Maintaining continuity with the Buddha’s own saṅgha became an imperative of the developing and expanding Buddhist community, so much so that new techniques to appeal to the legitimacy and authority of the Buddha were utilized by all communities, ancient and nascent, seeking legitimization.⁵² Authenticity became paramount in claims of legitimacy, especially in light of variant readings of literature and various traditions that all claimed a spot in the Buddha’s paramparā. In this context, the designation “Theravāda” (Doctrine of the Elders) can be seen and evaluated as carrying with it an argument about legitimacy.

This is not the place to delve into the morass of scholarship and speculation on early Nikāya Buddhism (that is, the period after the Buddha’s parinibbāṇa when several variant traditions and communities of interpretation arose). Nonetheless, we may want to destabilize the familiar yet problematic categorization of a monolithic Theravāda as the sole form of early Buddhism and the most legitimate heir to the Buddha’s sāsana, even though this is the position that the Pāli vaṃsa texts vociferously argue. Indeed, this destabilization should penetrate even the designation “Theravāda” itself; disparate vihāras (monastic complexes) competing for limited sources of patronage show that the Theravāda then, as now, was hardly a monolithic institution in spite of its periodical councils and purifications and otherwise convincing rhetoric.

The Dīpavaṃsa and Mahāvaṃsa imagine that the island itself was primed to receive the dhamma. A timeline is constructed in the text to support this argument: the Buddha visits, and upon his parinibbāṇa Vijaya arrives. These two events mark the beginnings of the lineages, both religious and royal. The Dīpavaṃsa and Mahāvaṃsa maintain that Buddhism was brought to the island by the missionary monk Mahinda during the reign of Devānaṃpiyatissa (247–207 B.C.E.). Could it be that this event is portrayed in the early vaṃsas for rhetorical force and to legitimize a particular strain or inheritance of the Mahāvihāra? K. M. de Silva suggests that given the degree of trade and traffic between Laṅkā and mainland India, early Buddhist influences are likely to have been imported to the island before Mahinda’s mission: “Once again the Mahāvaṃsa’s account of events conceals as much as it reveals, and what it hides in this instance is the probability that Buddhists and Buddhism came to the island much earlier than that.”⁵³ The monastic institution granted by the Sinhalese king to this proselytizing monk from India was the Mahāvihāra, so the text claims, thus establishing—at least for the purposes of the narrative—a named community with a definitive beginning. Considering that it is this very named community that preserves and recasts the narrative through the Dīpavaṃsa and Mahāvaṃsa, we may think critically about just what the term vaṃsa in this context means, especially given Collins’s definition: “Bamboo grows by sending out one, and only one, shoot: unlike our concept of a genealogical tree, therefore, a vaṃsa genealogy allows one legitimate successor at a time. Thus the term not only describes a line of transmission, but at the same time ascribes to the members of the vaṃsa a specific status and authority as legitimate heirs of that transmission.”⁵⁴ Thus, we may wonder if the vaṃsa genealogies of the competing monastic institutions, the Abhayagiri and the Jetavana, might have contained similar originary tales highlighting the central position of their own lineage. We cannot know for sure; the resurgence of the Mahāvihāra after a period of persecution resulted in reciprocal persecution of variant sects and the burning of these vihāras’ texts by the queen’s command.⁵⁵ When writing alone was not powerful enough to ensure hegemony, physical force was used.

If the timeline of the series of events in the Dīpavaṃsa and Mahāvaṃsa is to be believed, the Mahāvihāra itself was not established until Mahinda brought Buddhism to Laṅkā during Devānaṃpiyatissa’s reign, which would put the founding of this school in the third century B.C.E. Even if, as many have speculated, Buddhism was a known entity on the island prior to Mahinda’s arrival, the designation of actual space for the establishment of the saṅgha marks the institutional commencement. The very identity and self-understanding of the Mahāvihāra community responsible for writing the history is contingent upon direct, fundamental association with Mahinda. What is important to consider here is the representation of that founding—in the Dīpavaṃsa and Mahāvaṃsa, the Mahāvihāra is understood to be the legitimate heir to the sāsana and the king (here Devānaṃpiyatissa, the first convert) its patron. The founding of the Mahāvihāra also comprises the third discreet narrative cycle in the litany of events recounted in the text, the first being the Buddha’s three visits and the second being the arrival of Vijaya.⁵⁶

Following the chronological temporal structure, the Mahāvaṃsa urges us on to the narrative about Duṭṭhagāmaṇī. According to the text, the reign of the famed Duṭṭhagāmaṇī (101–77 B.C.E.) resulted in a massive building campaign for the Mahāvihāra. As we will see, so much of the Mahāvaṃsa is devoted to the epic narration of this king’s deeds that its story cycle can be referred to as the “Duṭṭhagāmaṇī epic,” and narratively it can stand apart from the rest of the vaṃsa. In the Mahāvaṃsa, the “Duṭṭhagāmaṇī epic” is expanded significantly from the abbreviated form found in the Dīpavaṃsa—the Duṭṭhagāmaṇī cycle constitutes ten of the thirty-seven chapters that make up the first section of the Mahāvaṃsa, more than one-quarter of the entire text. And yet this king is virtually absent in the Dīpavaṃsa account.

How each text represents this king’s reign could be a clue regarding its textual community’s motivating concerns. That the Dīpavaṃsa barely mentions Duṭṭhagāmaṇī by name may suggest that the Dīpavaṃsa is, as several scholars have argued, a faithful, documentary reconstruction of previous source materials, an assemblage of various texts, many of which were likely in oral circulation, and its aim was to set down and preserve as much as possible of those earlier texts. The reign of this king was secondary in importance to the spiritual lineage that garners greater focus in the Dīpavaṃsa. By contrast, the aim of the Mahāvaṃsa appears not to record the story but to amplify it and develop it poetically. The historical hero-king Duṭṭhagāmaṇī provides the Mahāvaṃsa with an ideal protagonist and the opportunity for a good, inspiring story.⁵⁷ A narrative about a king is the medium for many of the composer’s most effective poetic flourishes. Kings are certainly important in this text, but Duṭṭhagāmaṇī here is depicted as just one loyal Buddhist subject. What really takes center stage are the emotions that his story evokes.

Buddhist histories and the Theravāda Buddhists’ historicizing impulse have garnered much scholarly attention.⁵⁸ George Turnour’s translation of the Mahāvaṃsa in 1837 caused excitement among the early Indologists because the chronologies of kings in this text were exposed in a Rosetta Stone fashion to help James Prinsep in his translation of Asokan epigraphy.⁵⁹ The early Orientalists took express pleasure in the apparent and relative historical veracity of the Mahāvaṃsa, with its “ring of truth” and its attention to the presentation of chronologies of kings that most approximated the standard Western formulation of history that the scholars had always hoped to find in the South Asian context. The Mahāvaṃsa even provides dates that correspond with known epigraphical references. Early academic treatments of this text as real, authoritative, and trustworthy history, unlike the overly whimsical and blatantly devotional purāṇas of mainland India, subsequently informed the way they were translated, read, and used as documentary texts. In hindsight, we see that the early scholars’ interests, priorities, and biases determined how these scholars evaluated their source material, and we see the translation of the term vaṃsa as “history” reifying those same interests, priorities, and biases.

We have seen how the depiction of these texts as “histories” happened in the scholarly, external environment of the texts. Do these arguments have internal corollaries? How do the Pāli Dīpavaṃsa and Mahāvaṃsa frame themselves historically? Walters argues that they construct and participate in a sequence of “dispensations” of the Buddha that challenge any standard historical conception of time and at the same time pay particular attention to chronologies, events, and dates, even dating specifically each of the Buddha’s three “prehistorical” visits to the island: “‘History’—thought about the past in the then-present—proceeded in pre-colonial Sri Lanka within an episteme (to borrow M. Foucault’s useful term) consisting of a temporal scheme and anthropology quite foreign to modern sensibilities, one that leads me to consider the Vaṃsas as ‘successions’ of the Buddha’s presence rather than as mere ‘chronicles’ of events.”⁶⁰ This argument is very useful in destabilizing the dominant method of reading and using these texts because I do see the vaṃsas as narratives working to bring the Buddha (and his relics) into proximity with each successive future community.

Here I draw a distinction between the religious work of Buddhist histories and the rather overdetermined semantic field of the terms history and historiography in contemporary scholarship. By “religious work,” I refer to the explicitly contemplative and transforming effects that the act of reading or hearing the text is supposed to engender. There simply is no systematic concept of history at work for the compilers of the Dīpavaṃsa and the Mahāvaṃsa, and so it would be anachronistic at best to attribute these texts to a historical mode of writing. This does not mean that the saṅgha was unconcerned with historical matters.

What, then, is an acceptable understanding of “history” by which we might be amenable to calling the vaṃsas “Buddhist histories?” Jan Nattier provocatively suggests that “for most of the Buddhists for most of the time, the question of history—of where we are in the cosmos and of how much time remains in the world as we know it—has been of central, not peripheral, importance.”⁶¹ History in the vaṃsas is a temporally structured framework that orients the narrative. Relative proximity in the chronology is thus more important than the actual dates of events, and the religious implications or specifically Buddhist elements and interpretations of events expressed bear more weight than the chronologies themselves.

The “religiosity” expressed in the Mahāvaṃsa is thus perceived in the way the temporal frame works on the reader. The operative chronology carrying the narrative of the text is synchronized with the biography of the Buddha and with the reign of the Buddhist monarch Asoka. This synchronization can be interpreted as a powerful rhetorical device employed for a specific effect—namely, the construction of a connection between the textual community of the Mahāvaṃsa’s compiler and the legitimating, charismatic presence of the Buddha in the past. B. G. Gokhale considers the Theravāda understanding and use of what can be called history, but he, too, connected what is meant by history to what is perceived to be theology. He cites the Buddha as having among his many achievements an aversion to “low talk” (tiracchānakathā), which includes “tales of kings and their proconsuls, wars and heroes, cities and countryside, comely maidens and awesome bandits, items commonly regarded as the stuff out of which history is constructed.” Gokhale argues that in spite of this presumed aversion to all things historical, the growing “need for hagiological literature” in early Theravāda practice catalyzed the Buddhist return to history from the Buddha’s seemingly antihistoricist stance. He also claims that the “growth of the hagiological spirit is intimately related to the development of Buddhism as a religion.”⁶² At the time Gokhale made these claims, over fifty years ago, the debate about whether Buddhism is more properly characterized as a philosophy or a religion was raging, but we can nonetheless see some relevance to our understanding of the Mahāvaṃsa, in particular the accounts of the Buddha’s three visits to the island, as working in religious ways. But the predominant method of reading of the Mahāvaṃsa is driven by overattentiveness to the political chronologies it contains.⁶³

Malalasekera echoes many scholars’ assessment when he notes that the Pāli vaṃsas’ “chronology is admirably accurate…and they possess an honesty of intention and an accuracy of chronological record” absent in the Indian texts such as the purāṇas. But he also says they “manifest the same credulity and superstition, the same exaggeration in description, the same adulation of kings and princes, as is met with in the annals and religious history of [other] nations,” which, for his reading, does not detract from their historical value.⁶⁴ For Malalasekera, the whimsical aspect of the vaṃsas—that is, the rampant miracle stories and presence of the nāgas—does little to diminish the texts’ historicity, and much like the exaggerations in other (e.g., European) chronicles, these elements can be ignored in the process of reading for history. His perspective is but one angle of the seemingly universal position among the early scholars that (in Geiger’s terms) the “kernel” or “germ” of true history may be culled from the texts, while the more legendary, miraculous, and fantastic material can be downplayed.⁶⁵ Yet the mythical aspects are present throughout the text, not just in the first stratum, and so there is no easy division between what is primarily provocative narrative and what is historically sound. Instead of ignoring the “exaggerations,” my project here has been to consider the work that the presence and, in fact, the dominance of such mythical elements does to the text’s hearer-reader.

Just as the presence of mythical elements challenges a conventional historical reading of the vaṃsas and points to poetic and religious concerns, so might the choice of narrative subject (and poetic elements) reveal something about their time of production. It is commonly accepted that the Mahāvaṃsa and Dīpavaṃsa were written in times of duress. If one reads carefully and considers the narrative of the tumultuous time in which the Dīpavaṃsa was composed, the argument is not that the Mahāvihāra is the legitimate Buddhist sect but rather merely the first in a line of successive vihāras, with the Abhayagiri and other sects described as splinter groups from the main branch that is the Mahāvihāra.⁶⁶ As Rahula understands the history, “originally” all bhikkhus from all over the island “owed ecclesiastical allegiance” to the Mahāvihāra.⁶⁷ Rahula comes to his conclusion by consulting a Mahāvihāran text, one that envisions all rival schools as mere branches of the one authoritative, great Mahāvihāra. In spite of this, Rahula provocatively muses about the designation “Mahāvihāra,” “What was meant by the term Mahāvihāra? Was it only a place geographically defined or an institution?”⁶⁸

In this brief comment, we see just how little has been definitively resolved about the history of the textual communities of the Dīpavaṃsa and Mahāvaṃsa. It may be that “Mahāvihāra” was an umbrella name meant to designate a general Buddhist monastic complex, not a distinct institution or a particular sectarian orientation. It may be that the rival institutions were less disparate than we assume they were, at least when times were good and resources plentiful⁶⁹—that it was in times of duress, with concomitant competition for limited resources, that contention arose. Indeed, many have suggested that it is periods of contention that present the catalyst for literary production in general.⁷⁰

If we follow the timeline of the Mahāvihāra presented in the Mahāvaṃsa, the first real schism occurred during the reign of Vaṭṭagāmaṇī in the first century B.C.E.,⁷¹ when the Abhayagiri separated itself from the Mahāvihāra, presumably over practical rather than doctrinal differences. It was also during his reign (29–17 B.C.E.) that the Pāli Tipiṭaka was purportedly committed to written form.⁷² Collins notes, “The Pāli canon, like most other religious Canons, was produced in a context of dispute, here sectarian monastic rivalries. King Vaṭṭagāmaṇī supported the rivals of the Mahāvihāran monks, those of the recently founded Abhayagiri monastery.”⁷³ E. W. Adikaram argues that the shifting political currents were responsible for the impetus to write down the canon rather than the period of famine or general political strife that is usually considered the cause for the surge in textual production. He suggests that the Tipiṭaka was written by the Mahāvihāra monks in Matale and thus removed from the contested site of the capital Anurādhapura at the time when the Abhayagiri was founded through the support of King Vaṭṭagāmaṇī.⁷⁴

Could it be that the Dīpavaṃsa was written at a time when there were immediate concerns for survival and the authenticity and legitimacy of the Mahāvihāra were being challenged? The Dīpavaṃsa would then have been a plea for recognition and maybe even continued support, but not a claim for outright superiority vis-à-vis the other monastic institutions or the king himself.

Walters considers the Dīpavaṃsa nothing short of a “plea for survival.”⁷⁵ The Dīpavaṃsa is certainly motivated on some level by its compiler’s (or compilers’) operative ideology, and for that reason alone it cannot be considered a course for purely historical reflection. The only evidence we have for the strain on the textual community comes from within the text itself, although there is external, epigraphical evidence to corroborate its claims that “[s]upport for the ecumenical Abhayagiri monks had reached a fevered pitch at the end of the third century, under Mahāsena (274–301).”⁷⁶ The Dīpavaṃsa was likely written in the fourth century C.E., just on the heels of the rise in the Abhayagiri’s popularity. In a footnote, Walters considers the various epigraphical testimonies to the cosmopolitan and prolific support of the Abhayagiri, citing donation inscriptions from the first and second centuries.⁷⁷ As Collins notes, a monastic institution mired in “a context of dispute” becomes a likely site for textual productivity, and duress is a significant catalyst for textual production.⁷⁸ He thus echoes Adikaram’s observations not of the vaṃsa production proper but of two other periods of Pāli literary activity: “It is worthy of notice that the two most important events, namely, the writing down of the Pāli texts at Āloka-vihāra and the translation of the Commentaries into Pāli, both took place during the reigns of kings who were not favorably disposed towards the Mahāvihāra and who actively helped the opposing camp, the Abhayagirivihāra.”⁷⁹

Why did these times of duress provoke this particular kind of literary productivity—namely, textual production in the translocal language Pāli over the vernacular Sinhalese? Why would the response to challenging circumstances provoke such translocal aspirations? Perhaps it was all part of the broader agenda of an upstart vihāra to recenter the Buddhist world, and making claims about the centrality of Laṅkā in the translocal language of Pāli was a tactic to better convince readers-hearers “out there” of the center “here.” Yet these assertions seem audacious. Wouldn’t the claims have been more believable coming from a community in a period of growth and prosperity, not from a beleaguered minority voice? The use of Pāli for the composition of these texts may have been to gain greater sway in global Buddhist concerns, but another motivation lies closer to home within the textual community itself. By composing the texts in Pāli, the producers made a statement about the authenticity of their religious vision and of the claims made manifest in the proem and explained through metaphor and nāga. Pāli, after all, was claimed to be the mūla-bhāsa, or the “root language,” the Māgadhan dialect spoken by the Buddha himself. The coming of the Buddha, his relics, Vijaya, and the narratives of the kings of Laṅkā all take on a different hue when heard in the Pāli, the language connected with the canonical suttas and thus with the Buddha’s enduring voice. The intended effect of reading the vaṃsas was an emotionally transformative one, and perhaps composing them in Pāli to make the hearer work hard (and most likely to work through an interpreter, one of the key features of Brian Stock’s “textual community,” for the right interpretation) made the effect even sweeter an attainment.

During the reign of Vohārika Tissa (269–291), the controversial Vaitulyavāda sect (following Mahāyāna elements) was suppressed, and there was also a “purification” of the saṅgha. But perhaps most curious for our reading of the two Mahāvihāran literary productions in question is that the reign of Mahāsena saw an accelerated rise of interest in and outright royal support of certain Mahāyāna ideas in circulation at the time. According to the later section of the Mahāvaṃsa and the Vaṃsatthappakāsinī, Mahāsena oversaw the destruction of some of the Mahāvihāra buildings, and, in a move concretely symbolic of the shifting royal favoritism, instructed that bricks and building materials from the desecrated Mahāvihāra structures be used to enlarge the rival Abhayagiri complex. It was during this same king’s reign that a third contender monastic institution, the Jetavana, was built, and the Mahāvihāra itself was deserted.

For the later communities reading and interpreting this “historical” narrative, the designation “Mahāvihāra” signifies much. It cannot be coincidental that the narrative chronicles of both the Dīpavaṃsa and the first section of the Mahāvaṃsa end with the reign of the Lambakaṇṇa king Mahāsena (334–361). Although the Dīpavaṃsa and Mahāvaṃsa represent this king’s rule as a politically stable period (the Mahāvaṃsa even suggests that he was a good king on balance), later texts and epigraphical sources suggest that religious tension among competing institutions was also at a high during his reign. The Abhayagiri seceded from the Mahāvihāra during the reign of Vaṭṭagāmaṇī Abhaya (de Silva gives the date as around 103 B.C.E.⁸⁰), and it appears that there were frequent confrontations and arguments between the two institutions on monastic as well as doctrinal issues. From the vaṃsas themselves, it seems that by this time the Mahāvihāra had by no means secured the position of authoritative hegemony it would come to occupy centuries later. Collins reiterates the common assessment that “during the first millennium the Mahāvihāra was a minority tradition alongside the Abhayagiri and Jetavana monastic fraternities.”⁸¹ Thus, the “documentary” content within the vaṃsas reveals a competitive environment among three “minorities” struggling for resources, support, and legitimacy as well as the possibility that the vaṃsa production was a tool for one fraternity’s claims of authority.

This interpretation becomes yet more evident when we read between the lines, turning our attention away from the documentary “evidence” suggested in the text and toward the underlying narrative. The patterns of alternating patronage and withdrawal of royal support may have been a narrative ploy intended to heighten the audience’s sensitivity and response. The Mahāvaṃsa attributes much agency to the Coḷa Saṅghamitta, a resentful South Indian monk with allegiance to the Abhayagiri monastery. King Goṭhābhaya (249–262) had continued the policy of Vohārika Tissa (209–231) to suppress the Madhyamaka-influenced Mahāyāna movement known as the Vetullavāda (Sanskrit vaitulyavāda), which had become favored by the Abhayagiri institution. Goṭhābhaya deported sixty monks of the Abhayagirivihāra to India; Saṅghamitta was the disciple of one of these displaced (and likely indignant) Sri Lankan monks. We can assume that Saṅghamitta was groomed in a particular way and that he brought his animosity toward the Mahāvihāra with him to Sri Lanka.⁸² Saṅghamitta then curried favor with Goṭhābhaya’s successor, Jeṭṭhatissa I (263–273), and became the steward for Mahāsena, the intended next in line. Mahāsena’s feelings of intolerance for the Mahāvihāra, influenced by his adviser Saṅghamitta, spurred his campaign of destruction of the Mahāvihāra. Its buildings were razed, its monks forced into exile, and its valued possessions pillaged and claimed by the Abhayagiri (Mahāvaṃsa XXXVI.11–25, XXXVII).⁸³ Assuming the Dīpavaṃsa was in fact written at the time the events of its narrative cease—that is, during the reign of Mahāsena—then it was certainly written under a particular kind of duress: uncertainty about the king’s future patronage and active persecution, including exile to India. Thus, it appears that the Dīpavaṃsa may have been composed as an assertive claim to legitimacy in the face of favored contenders. As we have seen, however, it was not written de novo; the Dīpavaṃsa drew its content from the native Sinhala aṭṭhakathā literature.⁸⁴ If the Dīpavaṃsa represents earlier, originally Sinhalese source material in its claims to legitimacy, it is a conscious rhetorical strategy.

The nine years the Mahāvihāra monks were in exile should have obliterated their lineage; they had no support from the king, no residence for the monks, and no sense of stability in which to attend to the religious office of the saṅgha. Continuity, the theme par excellence, was seemingly interrupted. The later section of the Mahāvaṃsa attempts to recast these years, explaining that monks were in actuality left behind (in hiding) to protect the Mahāvihāran monastery grounds and artifacts for the continuity of the community. Walters offers an enticing and plausible reason why the lineage did not fade into oblivion in the nine years of exile. He suggests that “if we read the [Mahāvaṃsa] account of Mahāsena’s reign as saying that only the Mahāvihāran monks were expelled from the monastery—if the monks had indeed been destitute (and denied access to their libraries) for nine years—then we could explain why the task of pleading for mercy should have fallen upon the remaining nuns.”⁸⁵

Malalasekera cites Hugh Nevill’s unpublished manuscript catalog at length because Nevill presents a compelling argument that the nuns’ community may have been responsible for the production of the Dīpavaṃsa: “It [the Dīpavaṃsa] can scarcely be a record of the Theravāda fraternity of the Mahāvihāra, because in the very reign in which it was put forward by royal patronage (Dhātusena’s) Mahānāma set about to supersede it by his Mahā-vaṃsa. It certainly is not a record of the Dhammaruci sect of the Abhayagiri community, because it passes over the history of that wealthy, royal foundation with a well-calculated but short notice that could offend no one. But it dilates on a third society, the community of Theravādin nuns.” Nevill provocatively asserts that the extensive list of the nuns in Dīpavaṃsa XVIII should rightly have been preceded by a corresponding list of monks but that composition and circulation in a nuns’ institution would explain why the monks’ section would have been omitted. He also suggests that the compilation of the Mahāvaṃsa was necessary because Mahānāma was “jealous of the nuns,”⁸⁶ a less than satisfactory reason for which there is no evidence, drawing attention to the completely conjectural nature of his observations.

This issue of authorship may also explain the reason for the coarseness of the Pāli in the Dīpavaṃsa. Nevill names the possible compiler of the Dīpavaṃsa based on the nuns’ lineage presented in Dīpavaṃsa XVIII, claiming that two historian nuns from India, Sivalā and Mahāruhā, revised the compiled traditions and stories from Princess Mahilā and “very probably formed [them] into the unpolished almost aboriginal Pāli we now possess, to which additions were made by Nāgamittā, and later by Sanhā and Samuddā.”⁸⁷ Nevill’s conclusion that the nuns’ community was responsible for the Dīpavaṃsa leaves open the possibility that the two vaṃsas may have been more contemporaneous than we would think on the basis of stylistic grounds and that the Mahāvaṃsa represents another monastic community’s reaction to the Dīpavaṃsa at the time the Dīpavaṃsa was produced.

The main problem with this theory, again, is a lack of corroborating evidence aside from the artful reading of the Dīpavaṃsa. Especially problematic is Nevill’s attribution of the editions of the Dīpavaṃsa to individually named nuns even though there is no evidence to support such a claim. The text certainly pays much attention to the position of nuns, which should grab any reader’s attention. But could this, too, have been a rhetorical device utilized by the compiler? The careful attention paid to the nuns’ lineage is nowhere evident in the Mahāvaṃsa, where even the story of the nun Saṅghamittā seems presented in a reduced form compared to its appearance in the Dīpavaṃsa.

The Dīpavaṃsa and Mahāvaṃsa do share a common conclusion: the death of King Mahāsena. Geiger suggests that the Dīpavaṃsa ends with Mahāsena more or less by accident: “It also happens that the fact that [the Dīpavaṃsa] finished with Mahāsena was not founded on any definite plan, but depended rather on an unfortunate occurrence. If the Mahāvihāra had not been destroyed at that time and depopulated, then the list of kings would without a doubt have been continued further. According to my idea the ‘Mahāvaṃsa of the ancients’ was an independent chronicle, which the monks of the Mahāvihāra carried on to Mahāsena, and certainly would have continued still further had they not been disturbed in their peaceable work in some violent manner.”⁸⁸ But Walters makes a persuasive argument that the Dīpavaṃsa’s temporal end and explicit polemical stance were in fact calculated and intentional. He suggests that the Dīpavaṃsa “evaluates Mahāsena’s action, implicitly suggesting that it violates the truths of both calculable and incalculable history, and concludes its narrative with him because that is what the entire presentation was leading up to: a demonstration that the Mahāvihāra is worth continuing because it embodies these truths.”⁸⁹ Here Walters addresses the chronology as serving the narrative. There is a rhetorical need for a Mahāsena for the text to argue against, which serves as the culmination of the text itself. The Mahāvaṃsa can then be understood to have shaped the narrative to result in something. The text is not so much documenting the past as it is constructing an argument.

The Dīpavaṃsa could have been created by nuns, might represent the written version of orally performed stories, and might be deficient in the criteria of good literature (“here too long, there too short”) due to its primary intention to preserve variant authoritative stories. Regardless, it was most likely written in contentious times and thus represents a “plea for survival” in the midst of competition and an uncertain future. Walters concludes that this palpable “plea for survival” may also indicate the “newcomer” status of the Dīpavaṃsa compilers: “This reworking of the history of Buddhist orders and their textual canons, which dismisses the claims of the Mahāsāṃghikas and Sarvāstivādins by appropriating Aśoka for the Theravādin cause, hoped to secure a place for the Theravāda at the Mahāvihāra, a Buddhist disciplinary order that was probably a newcomer on the global Buddhist scene.”⁹¹ Perhaps the Dīpavaṃsa was not a reassertion or a defense of the Mahāvihāra as legitimate heir of the Buddha’s sāsana but in fact an argument for the Mahāvihāra’s nascence in the midst of several competing traditions.

We can piece together some information based on the firsthand witnessing by the Chinese monk Faxian, who visited Sri Lanka in the early years of the fifth century, and nearly all scholars utilize his documentation. But can we be certain of the accuracy or unbiased nature of this monk’s impressions? He was, after all, not a disinterested observer, but a temporary resident of the Abhayagiri. Our suspicion aside, he gives the impression that by the time of his fifth-century visit there were three distinct institutions: “According to Fa Hien, at the time he visited Ceylon in the 5th century, there were 5,000 monks at the Abhayagiri, 3,000 at the Mahāvihāra and 2,000 at Mihintalee. Further he says that there were about 60,000 monks who got their food from their common stores, and that the king, besides, prepared elsewhere in the city a common supply of food for five or six thousand more.”⁹² If this were the case, it is very interesting to consider the reasoning behind Buddhaghosa’s choice of the Mahāvihāra for his commentarial work. Buddhaghosa’s teacher supposedly sent him to Sri Lanka because of the deplorable state of the commentarial literature in mainland India. Buddhaghosa would have gone to the institution closest in alliance to his home vihāra. The Abhayagiri at this time, according to the Mahāvaṃsa, was leaning toward the Mahāyāna.

In its conclusion, the Vaṃsatthappakāsiṇī⁹³ asserts an individual author of the Mahāvaṃsa, Mahānāma, who is likely the fifth-century thera (monk) composer from the Mahāvihāra monastery in Anurādhapura. This is the only reference to the compiler’s purported name; it is given nowhere in the Mahāvaṃsa itself. Geiger notes: “It [the Vaṃsatthappakāsiṇī] says of him [Mahānāma] that he lived in the monk’s cell built by the general Dīghasanda. Dīghasanda was, according to [Māhavaṃsa] 15, 230, a leader of the army of Devānaṃpiyatissa, and the cell founded by him, which bore after him the name Dīghasandasenāpatipariveṇa, belonged to the Mahāvihāra.”⁹⁴ The provenance of the Māhavaṃsa is thus set in the context of the Mahāvihāra, but that provenance is not attributed until perhaps five or six centuries after the Māhavaṃsa’s composition. It is because of this named author that the date of the text is assumed to be fifth century; the running historical narrative (that is, the first narrative that ends with the reign of Mahāsena) leaves off at an earlier date. This endpoint is the same in the Dīpavaṃsa, which lends credibility to the theory that the Mahāvaṃsa is either a direct reorganization and retelling of the Dīpavaṃsa as we know it or that both texts are working from the same urtext and are envisioning themselves to be better, more sophisticated renditions of the same narrative. In either case, there is an undeniable relationship between the Dīpavaṃsa and the Mahāvaṃsa even if we cannot sufficiently or precisely date either text’s composition.

Along this line contrary to Geiger, Collins, in his own reading of the later commentary Vaṃsatthappakāsinī, sees the reference to the “Mahāvaṃsa of the ancients” as pointing not to the Dīpavaṃsa but to early Sinhalese accounts.⁹⁶ This has come to be the more commonly accepted conclusion among scholars. With the information we have, we cannot make a conclusive argument that this phrase in the Mahāvaṃsa is not actually referring to the Dīpavaṃsa. Even if they are not directly related, reading the proem of the Mahāvaṃsa as indicative of a certain view against older source materials, including the likes of the Dīpavaṃsa and perhaps, as Geiger wishfully concluded, even the Dīpavaṃsa itself, is helpful to highlight the intentionally literary turn made by the Mahāvaṃsa.

There is undeniably plenty of historical information provided within these two texts, particularly in what has become the accepted chronology of events that is the subject of the narrative of the Dīpavaṃsa and the Mahāvaṃsa. Scholars write about the community of the early Pāli vaṃsa production by leaning on prior interpreters of the later extension of the Mahāvaṃsa, the very source of the information about the Mahāvihāra from the fourth and fifth century on (the time of production of the Dīpavaṃsa and Mahāvaṃsa, respectively). Most scholars agree to a fifth-century date for the Mahāvaṃsa, but some, such as G. C. Mendis, consider that “it was probably composed about the sixth century A.D.”⁹⁷ This discrepancy in dating follows other chronological anomalies in the text itself. In a footnote, Gombrich explains: “Like Rahula, I am keeping to the chronology established by Geiger (1912). According to Mendis (1947), Duṭṭhagāmaṇī’s rule should be dated 161–137 BCE and all Sinhala dates from then on til [sic] the end of the fourth century be moved back 60 years. The problem is that somewhere between Devānaṃpiya Tissa and the late eleventh century the Mahāvaṃsa put in 60 regnal years too many. Geiger thinks it is in the fifth century, at the very beginning of the Cūlavaṃsa; Mendis argues that it is between Devānaṃpiya Tissa and Duṭṭhagāmaṇī. I find Mendis’ arguments plausible but not conclusive.”⁹⁸

In other words, we are unsure of the dates that have been fixed for the composition of the Mahāvaṃsa. With the date of the text itself unsettled, how sure can we be of the other information we glean from the text? Can we even be certain of what precisely was meant by the designation “Mahāvihāra” at this early juncture? The bolstering force of the saṅgha reforms (purification) by the twelfth-century king Parākramabāhu I has forevermore colored our understanding of the Mahāvihāra community. He unified the three main, competing sects, the Mahāvihāra, Abhayagiri, and Jetavana, under the sole leadership of the Mahāvihāra, which spurred the exportation of the unified Theravāda to other countries and thus secured a relatively unchallenged position of orthodoxy for the Mahāvihāra once and for all. Of course, what we know about Parākramabāhu’s royal activity comes from the literary sources, such as the extension of the Mahāvaṃsa.

The narrative practices employed within the text—including and perhaps even privileging the mythological elements so readily dismissed by modern historical methods as flights of fancy or, at best, elements that contain merely a germ of the truth—actually prefigure the modern study and interpretation of the text as a work of history by downplaying the centrality of the mythical mores of the text and overinterpreting what are already overdetermined characters (such as the yakkhas and nāgas). Far more than the chronological structure, these literary elements—first the reading instructions given in the proem, then the subtle employment of metaphor and other poetic devices, and, finally and most importantly, the saturated landscape populated by the nāgas—are employed for a lasting effect on the primed audience.

By arguing that the narrative is not to be read as linearly historical, I do not tread entirely new methodological grounds. My approach to the “history” that we glean from our sources is informed by Stephen Berkwitz’s warnings regarding the facts or context behind the text: “Attempts to arrive at the historical context—as if a single ‘context’ ever exists—of Buddhist histories results in a problem directly related to the fictive reconstruction of historical narratives. The particular context upon which an interpretation can be based is, like the plot of histories, something that is as much invented as found. Any historical context that is presumed to exist behind and prior to a given text constitutes an abstract image of the ‘real past,’ a past that is itself created and synthesized by the historian out of readings from other texts.”⁹⁹ In other words, what is read as historical “data” within the Mahāvaṃsa is itself a part of an inventive, imaginative literary project. The corroborating “evidence” gleaned from other texts, such as the Vaṃsatthappakāsinī and the Mahāvaṃsa extensions (Cūlavaṃsa), is also a part of an ongoing literary project and can hardly provide an objective assessment of the historical context of the Mahāvaṃsa. In his consideration of the late-thirteenth-century Sinhala Thūpavaṃsa, Berkwitz points to the need to break the hold of the so-called purely historical interpretation of Buddhist history in favor of framing and strategically reading these texts as literature: “The choice to frame Buddhist history as a form of religious narrative used to accomplish religiously defined and sanctioned goals instead of a sequence of real, external events from the past recasts our study of historical texts in Sri Lankan Buddhism. Textuality and language replace historical fidelity as the objects of inquiry in Buddhist historiography.”¹⁰⁰ Although his argument refers specifically to the vernacular Buddhist histories of the late medieval period and to the Sinhala Thūpavaṃsa in particular, I believe this method may be profitably applied to the reading of the Dīpavaṃsa and the Mahāvaṃsa as well.¹⁰¹ The Mahāvaṃsa itself defies temporal fixity even in the midst of its finely tuned attention to chronologies and dates. The patterns, icons, and emotional qualities it invokes defy temporal bounds and speak persuasively and practically to Theravāda Buddhist communities in the past, present, and future. Reading the Mahāvaṃsa as a religiously conceived and utilized text, one that begins with the Buddha himself, rather than as a history of the acts of kings allows us to begin to see how persuasive the text is and how it does not merely document history but works on its audience to provoke a response. By “religiously conceived,” I mean that the compiler or compilers of the Mahāvaṃsa worked within a monastic context to inscribe the vaṃsa or lineage of his or their monastic institution while simultaneously shaping his or their text to inspire a powerful sense of location, community, and faith in the audience. We see how the Mahāvaṃsa builds an argument about its own efficacy and transformative powers vis-à-vis the primed community of reception. Lest we forget the text’s express purpose, the line that concludes each chapter of the Mahāvaṃsa—“Sujanappasādasaṃvegatthāya kate mahāvaṃse” (Made for the anxious thrill and serene satisfaction of good people)—reminds us, even in the midst of the chronology, that this text is a religious one, as I discuss further in the conclusion.

The historically situated concerns of the Dīpavaṃsa and Mahāvaṃsa textual communities may have shaped the way their respective stories unfold in terms of both structure and rhetorical details. Following the story too closely, assuming that the texts are documenting what exists around them rather than participating in a constructive, literary maneuver, obfuscates the work that each text is doing to construct its community. Walters attempts to marry the dominant narrative of writing in a time of political upheaval with the more specialized literary tactics at work in the Mahāvaṃsa as compared to those at work in the Dīpavaṃsa. He characterizes the salient reasons for the composition of the Dīpavaṃsa as an argument for legitimacy in a time of contention:

The monks and nuns of the Mahāvihāra produced the [Dīpavaṃsa], the first Pāli text of Sri Lanka, as a challenge to the privileged position enjoyed by the residents of the Abhayagirivihāra in the fourth century. More than that, by claiming to be the descendants of an originary Buddhist order uncorrupted with later Great Vehicle practices, it also posed a challenge to the “idealist” Buddhists who were coming into prominence in the Gupta-Vākāṭaka imperial formation in response to the “realist” Buddhists who had prevailed when imperial rulers recognized theirs as superior to other ways of life. The complex author of the [Dīpavaṃsa] composed that text as part of its fight for survival (in the wake of Mahāsena’s disastrous reign).¹⁰³

Just as the authors or compilers of the Dīpavaṃsa and Mahāvaṃsa represented stories within the framework of a historical trajectory, so we organize data culled from these texts according to our own most salient frameworks of interpretation. Seneviratne reminds us that “history, therefore, can no longer be fact—not because facts are only the first step in creative interpretation—but because the kind of facts it seeks, namely facts about the past, are inherently unestablishable, and, therefore, those facts with which we endow the past are imaginings based on a present enclosed in its own processes of creating attitude, perspective and value.”¹⁰⁵ This statement is as true for our contemporary reading of “historical” texts as it was for the compilers of the Dīpavaṃsa and Mahāvaṃsa. Seneviratne continues to place himself and his fellow contributors on a scale of historical interpretations, one that I believe extends into history beyond the reflective act of twentieth-century scholarship looking at ancient texts, wherein texts are constantly renegotiating and rearticulating, forging arguments with and against each other, and isolating particularly salient stories. Awareness of both texts’ literary aims changes the type of reading we perform on them. It is within the vaṃsas that we see the “world wishes” articulating and constructing rather than documenting the worlds in which the texts were composed. Seneviratne continues, thinking about current events in Sri Lanka and how we can see “the present as shaping the past”: “Newly centralized and unified polities invent or interpret myths to justify their exclusive rights to habitation in a given locality and to subordinate or exclude those defined as others. Accounts of the past are embellished and interpreted through the perspective of present-day ethnic and other group identities, aspirations, values, interests and so forth.”¹⁰⁶

Turning to the sense of urgency conveyed in the Mahāvaṃsa, I am reminded of the volume of recent scholarship on the “Mahāvaṃsa mentality,” the postcolonial phenomenon of employing the rhetoric and vision of the Mahāvaṃsa to bolster nationalistic, hegemonically Sinhalese Buddhist ideas about what Sri Lanka is. Briefly, this “Mahāvaṃsa mentality” is susceptible to deconstruction on the basis of its anachronistic conception of statehood.¹⁰⁷ Too often scholars are seduced by the efficacy and interpretability of the chronological structure of the Mahāvaṃsa. Indeed, there are recognizable patterns of dissolution and reformation, power struggles and oppositions happening in the text.¹⁰⁸ Consider, for example, what Bardwell Smith has to say about postcolonial Sinhalese Buddhist identity in his essay “Sinhalese Buddhism and the Dilemmas of Reinterpretation” and how it might be fruitful to apply the same deductive logic to the scenario articulated in the proems: “The manner in which a community in times of profound crisis reaches into the past to discover its identity and redefine itself in relationship to this past, even to perceive distinction and uniqueness over against it, has considerable psychological and social importance.”¹⁰⁹

I am inclined to add “ethical importance” here, both for the individual reader-hearer and for the community of interpreters. In the case of our Pāli vaṃsas, the same technology of calling forth the past to help define the present is performed in two distinct ways. First, both vaṃsas begin with a foray into the deep and legitimating past, the Buddha himself, even though chronologically the Buddha would have lived centuries prior to either text’s composition. Second, the Mahāvaṃsa’s proem distances itself from what it calls “the Mahāvaṃsa of the ancients,” commonly understood to be either the Dīpavaṃsa itself or the urtext (aṭṭhakathā) from which both the Dīpavaṃsa and the Mahāvaṃsa cull their material.

Paying attention to the circumstances—social, institutional, political, and religious—in which the two earliest extant Sri Lankan vaṃsas were produced is obviously an important way to understand the literary devices we have encountered in these texts. But we must be wary that just as our own situatedness affects the way we structure our understanding of the historical time period whence these texts sprang, their fourth- and fifth-century situation also informs the way they have been conceived. Both vaṃsas collect previous information; neither claims to be an original work, and we must remember to read them as products of a literary world in which originality was not the point. But both of them present the past in ways that are influenced by their present. Fresh reading practices help us encounter Buddhist histories on literary terms.¹¹⁰