It is true, we should concede at the outset, that Buddhism is not a “historical religion” in the sense that applies to the monotheistic faiths of the Semitic world. What the Buddha discovered in his experience of enlightenment, so the Buddhist scriptures tell us, was not any new revelation irrupting into the world for the first time, but a timeless truth about the nature of reality, identical to the truth discovered by all other enlightened sages before him. Thus, even the central event of Buddhist mythology —the Buddha’s experience of illumination under the Bodhi tree —is not viewed as a decisive historical event in the sense that Jews, Christians, or Muslims might use to speak of the receipt of the ten commandments by Moses, the incarnation of God in the person of Jesus, or the revelation of the Qur’an to Muhammad. Indeed, it is a central contention of virtually all schools of Buddhism that the Buddha’s experience is by definition repeatable and is accessible (at least in theory) to all human beings.

The agenda in the present study, then—for to raise such a question without facing it oneself would be illegitimate —has been to demonstrate that we will miss (and misunderstand) a great deal of what has gone on in Buddhist history if we assume that this tradition has been indifferent to historical change. On the contrary, it is my contention that the question of “what time it is” has mattered, and at times has mattered very much, to a substantial proportion of Buddhist believers. A major objective of this study has been to demonstrate that there is considerable evidence in the Buddhist canonical literature itself that, far from being concerned only with “timeless realities,” the Buddhist tradition has often paid careful attention to the transitory realities of this earth.

[T]his acute sense of the inability of ordinary human beings living in the Final Age to emulate their spiritual predecessors has led to what would be described, in a Christian context, as “dispensationalism”: that is, the idea that while certain teachings and practices may have been appropriate in an earlier age, we now find ourselves in a completely different era (or “dispensation”) in which a wholly new spiritual repertoire is called for. Thus the arguments set forth by the Japanese Buddhist teacher Hōnen (1133-1212) in favor of discarding all Buddhist scriptures other than those concerning the “original vow” of the Buddha Amitābha (Jpn. Amida) were based on the idea that a fundamentally new age was now in effect. Likewise, while his compatriot Nichiren (1222-1282) argued that his advocacy of chanting the daimoku was fully in accord with the intention of the Buddha Śākyamuni, he was also well aware that this constituted a radically new practice in the eyes of his fellow Buddhists, and he argued for its legitimacy precisely on the basis of such “dispensationalism.”

Thus, while in South, Southeast, and Inner Asia (including Tibet) the threat of the decline and ultimate demise of the Dharma served largely to elicit conformity with the existing tradition and to reinforce the importance of preserving whatever elements of the Dharma still remain, in East Asia a long series of Buddhist leaders —from the Pure Land teacher Tao-ch’o in 6th-century China to the Kamakura reformers of 13th-centuryJapan —found in the idea of mo-fa an incentive to innovation, often leading to the formulation of new religious ideas and practices of striking creativity.

The idea of decline, then, is clearly multivalent and has served a number of seemingly contradictory purposes in Buddhist religious history. Its significance in any given time and place—or for any given individual —will be influenced by a great number of factors, one of the most important of which is the presence, or absence, of a concept of mo-fa. The task of living within a prolonged period of the “Final Dharma” is quite different from that of facing the imminent demise of the Dharma as a whole, and it is hardly surprising that these two prospects should have evoked such different responses.

Our search for the origins of the three-period system has brought us, then, to the following conclusions. The notions of saddharma and saddhanna-pratirūpaka were well established in Indian Mahāyāna literature by the middle of the 2nd century CE at the latest. Meanwhile, the term Paścimakāle (“latter age”) had also entered Indian Buddhist literature, likewise in a Mahāyāna context, as a reference to the period following the death of the historical Buddha. Most often the latter term was used in contexts in which the Buddha was described as recommending the acceptance, preservation, dissemination, and so forth of the sūtra in question during the time after his death, sometimes in conjunction with a discussion of the difficulties that might attend those who do so. By the latter half of the 3rd century CE Buddhist scriptures containing all of these terms were being translated into Chinese by Dharmarakṣa, who appears to have introduced the two vital terms hsiang-fa (as a translation of saddharma-Pratirūpaka) and mo-shih (as a translation of Paścimakāle) into Chinese Buddhist discourse.

Around the same time the notion of two periods in the history of the Buddhist religion (found in such Indian texts as the Lotus Sūtra) was becoming well established in Chinese Buddhist circles. In light of this twofold time system some Chinese Buddhists began to interpret the term mo-shih, which would naturally be understood in Chinese as “final age,” as the name of a third such period. Based on this understanding, certain Buddhist translators (of whom the most influential, and probably the earliest, was Kumārajīva) began to use the term mo-fa as an occasional substitute for mo-shih, thus bringing the latter into greater symmetry with its “predecessors,” the periods of cheng-fa and hsiang-fa, respectively. Having thus entered the scriptural corpus, the term mo-fa took on a life of its own, and Chinese commentators undertook with enthusiasm the task of describing the nature and duration of this anticipated third period. That they chose for its duration the quintessentially Chinese figure of 10,000 years (with its underlying implication of “an eternity”) demonstrates that they were free from any constraints encountered in Indian documents, for while mo-shih (Skt. paścimakāle) is often described as comprising the “latter five hundred years,” the newly coined term mo-fa was subject to no such restrictions. Likewise it reveals their profound sense of optimism (or, at the very least, of wishful thinking), for in assigning to this newly created period of mo-fa a duration of 10,000 years these Chinese commentators expressed the hope that Śākyamuni’s teachings would last forever, albeit in a reduced and less accessible form.

In light of the scenario outlined above it is no longer surprising that explicit references to a three-period system first appear not in sūtra and śāstra literature translated from Indian originals, but in 6th-century Chinese scholastic texts. For if the developed notion of three periods in the history of the Dharma indeed arose in the context of reflection on the meaning of the term mo-fa —itself a Chinese “apocryphal word” —we should expect this notion to have emerged well after the first appearances of this expression in Chinese Buddhist literature, which took place around the beginning of the 5th century CE. And this is precisely what we find, for it is Nan-yüeh Hui-ssu (515-577), best known as the teacher of T’ien-t’ai Chih-i, who is credited with having been the first to set forth in writing a three-period system based on eras of cheng-fa, hsiang-fa, and mo-fa, respectively, in a work completed in 558 CE.

But the absence of any direct evidence for the existence of a three-period system in the literature of Indian Buddhism has not stifled attempts by modern scholars (in particular, by modern Japanese scholars, in whose own religious traditions this system continues to hold a central place) to find evidence for its origins in India.

An examination of the contexts in which Skt. Paścimakāle and its Tibetan and Chinese counterparts are found in the Buddhist sūtra literature provides further evidence that this phrase was originally intended to be read in the sense of a “latter” or “future” age, not as a “final age” in the superlative sense. Most telling is the fact that this “latter age” is never contrasted with any earlier period other than the lifetime of the Buddha himself, where such a contrast regularly takes the …”after the death of the Tathāgata, in the latter time, in the latter period, in the latter five hundred years.” Just as the expression saddharma-pratirūpaka regularly refers simply to the duration of the Dharma after the parinirvāṇa of the Buddha, so in these instances paścimakāle (and its more extended equivalent) seems to refer simply to the period following the Buddha’s death.

Most interesting from the point of view of the historian of Buddhism, however, is the evident agenda that seems to have led to the use of this expression in sūtras produced by Mahāyāna writers. For the expression is used, in the vast majority of cases, in contexts where the Buddha is described as recommending “this sūtra” (i.e., the Mahāyāna sūtra in which the expression appears) for circulation among Buddhist believers after his death. The expression Paścimakāle serves, in other words, as a kind of “Good Housekeeping Seal of Approval,” certifying the sūtra in question for acceptance and dissemination in the latter (i.e., the post-Śākyamuni) age.

In light of this evidence it is particularly interesting to recall where the expression mo-shih does, and does not, appear in the Chinese translation literature. It does not appear (according to the Taishō index) anywhere in the Āgamas (i.e., the non-Mahāyāna sūtras) or in the abhidharma literature, and only two occurrences are registered in the vinaya texts. The overwhelming majority of appearances of the term are found in the Mahāyāna sūtras, with another sizeable group of occurrences in the mi-chiao (tantric) literature. The term seems to have been most popular, in other words, in the Mahāyāna sūtras, where—by stating explicitly that the Buddha promised vast merits to those who would accept and transmit the sūtra in question after his death —it served as a certification of their legitimacy.

This evidence concerning the motive for the use of the term, together with the fact that this “latter age” is never explicitly contrasted with any earlier period in the history of Buddhism (other than the lifetime of the Buddha himself), brings us to a rather unexpected conclusion: that the expression “in the latter age” (Skt. paścimakāle) was originally introduced into Buddhist discourse simply as a reference to the time after the death of the Buddha. Just as in the case of saddharma-pratirūpaka discussed above, it would seem that the idea of a “latter time” originally implied no periodization whatsoever within the lifetime of the Dharma after the Buddha’s Parinirvāṇa, but simply referred to this era as a whole.

It is noteworthy that a substantial proportion of the traditions in this category [external causes for the decline of Buddhism] are found in Mahāyāna texts, while those in the internal causes category are found almost exclusively in Nikāya Buddhist works. It is tempting to draw doctrinal conclusions from this asymmetry; yet the more important distinction, perhaps, is that the Mahāyāna sūtras in fact almost never predict the actual demise of the Dharma, except where they have inherited (and maintained) certain pre-Mahāyāna accounts. Even in these cases the Mahāyāna texts frequently modify these accounts, indicating that the Dharma will only appear to die out, but will not actually do so.

Whatever mo-shih (that is, Paścimakāle) may have meant in these texts, it was clearly not a third period occurring after an era of hsiang-fa.

We must return, therefore, to the contexts in which this expression is used to determine its original meaning. Here our task is made easier by the fact that, in the overwhelming majority of cases, the context is quite consistent: the term occurs in the context of a discussion of the merit to be derived from the acceptance, maintenance, and distribution of a given sūtra—that is, the sūtra in which the passage in question appears. A typical example is found in one of the earliest Mahāyāna scriptures, the Lotus Sūtra:

One should always reverentially salute him with joined hands, as if he were the Chief of Jinas or the Self-born, he who in the fearful latter time (Skt. Paścimakāle [sic]) upholds this sūtra [i.e., the Lotus Sūtra of the Extinct [Buddha].

Those who accept the legitimacy of the Lotus Sūtra and uphold its teachings after the death of the Buddha, in other words, are worthy of the reverence of the world.

When we combine this chronological information with the fact that in a significant number of cases Ch. mo-fa corresponds to Skt. Paścimakāle (i.e., to an expression that is more regularly translated into Chinese as mo-shih), we may find the key to our puzzle: namely, that mo-fa originated simply was a variant of mo-shih, introduced by a Chinese writer already familiar with the periodization scheme built on sequential eras of cheng-fa and hsiang-fa, respectively. Viewed in this context it would have required only a small leap of inference to conclude that mo-shih (which implies, unlike its Sanskrit counterpart, not merely a “latter” but a final period) was meant as a reference to a discrete third period in the history of the Dharma, which could be expressed more clearly (or at least more symmetrically) by the term mo-fa. Once mo-fa had gained some currency—a process that must have been stimulated by the use of the term in some of the translations of Kumārajīva, whose works are among the most influential ever produced in China—the die was cast, and it fell to the lot of Chinese Buddhist scholastic writers to expound on the nature and duration of this supposed third period. In light of this scenario, it should no longer surprise us that the first extended discussion of the three-period time scheme appears well after the introduction of the term mo-fa in the works of Kumārajīva and others, and that such discussions appear not in translated sūtras, but in the works of Chinese commentators themselves.

If this line of reasoning is correct we should no longer view the term mo-fa as a Chinese translation of an Indian Buddhist term, but rather as a stylistic variant of mo-shih (itself a genuine translation of Skt. Paścimakāle …). The term mo-fa subsequently took on a life of its own, stimulating seemingly endless commentarial reflections in East Asia. Mo-fa is thus a Chinese “apocryphal word”: a term created in China, with no identifiable Indian antecedent.

If we have had little difficulty in locating examples of the use of the term saddharma-pratirūpaka (and of the two-part timetable with which it eventually came to be associated) in Buddhist scriptures composed in India, matters are altogether different when we come to the third dispensation in the history of the Buddhist religion, known in East Asian sources as mo-fa (Jpn. mappō). Though the term itself appears in a number of sūtras translated into Chinese from Indian originals, it is not at all clear what Sanskrit Buddhist term —if any—can properly be described as its antecedent. Moreover, though the idea of a three-part timetable of cheng-fa, hsiang-fa, and mo-fa is so ubiquitous in East Asian Buddhist writings that much of the history of Buddhism in this region would be incomprehensible without it, it has proved singularly difficult to find examples of such a three-part scenario in any Buddhist source of certifiably Indian origin.

If mo-fa cannot be viewed simply as a translation of a well-known Sanskrit technical term, we must turn directly to the Chinese Buddhist literature in our attempt to determine its significance. Restricting our inquiry at this point to those texts translated from Indian originals (both in order to focus on the point of entry of the term into Chinese Buddhist usage and to continue our attempt to establish its proper Indian antecedent, if any), we will begin by tabulating the occurrences of the term in the first seventeen volumes of the Taishō canon —that is, in the Agama, Avadāna, and Mahāyāna sūtra literature. …

In sharp contrast to what we might expect, the above list shows not a multitude of texts in which hsiang-fa and mo-fa appear in conjunction, but quite the opposite. With only a handful of exceptions … either hsiang-fa or mo-fa may appear in a given text, but not both.

[W]e should also take note of another important fact: that is, that in the entirety of the first seventeen volumes of the Taishō canon, comprising 847 separate scriptures and totaling over 16,000 pages of printed text (that is, approximately 25,296,000 Chinese characters), only 22 individual works containing the term mo-fa are registered in the Taishō index. Moreover, in virtually all of these cases the term appears only once in a given text, rather than being used repeatedly and serving as a major topic of discussion in its own right. While there are undoubtedly other occurrences of the term that have escaped the notice of the indexers, the overall trend is quite clear: this expression is as rare in the canonical sūtra literature as it is ubiquitous in the East Asian commentaries.