One of the seemingly most useful sources for reconstructing the biography of Sakyamuni is the traditions and lore associated with the rainy season retreats. As we have seen though in Article 5 ("Locations of Sakyamuni's Residence during the Rainy Season: A Comparative Study of the Early Buddhist Scriptural Sources and Later Traditions"), these are not based on the early Buddhist sources. It is thus necessary that we verify the documentary value of these traditions, and perhaps the best way of doing so is to clarify the evidence upon which they are based. Since it is unlikely that they completely ignore the information contained in the early scriptural sources, it can be conjectured that they grew up, built on some kind of information found in the early scriptures. What kind of information contributed to shaping these traditions? This article constitutes one line of research exploring this issue.

The traditions concerning the years of the rainy season retreats tells us that they were all held in Sāvatthi from either the twentieth or the twenty-fifth year after the enlightenment, other than the last, which was held in Veḷuva, on the outskirts of Vesālī. In the early scriptural sources, it is from either the twentieth or the twenty-fifth year after the enlightenment that Ānanda began acting as the Buddha's attendant. If these two coinciding reports are somehow connected, Sakayamuni spent the rainy season only in Sāvatthi from the time Ānanda became his attendant. Conversely, the twenty or twenty-five years following the enlightenment when the rainy season retreat was spent in a number of places, beginning with Bārāṇasi Isipatana, corresponds to the period when Ānanda was not the Buddha's attendant.

Various later Pali commentaries (aṭṭhakathā) as well as works such as the Mahāprajnāpāramitopadeśa (Da zhidu lun, Treatise on the Perfection of Great Wisdom) list the names of men who were the Buddha's attendants before Ānanda, such as Nāgasamāla (Nāgapāla), Nāgita, Upavāna, Sunakkhatta, Cunda samaṇuddesa, Sāgata, Rādha and Meghiya. Can the information in the scriptures where they appear not be closely related to the development of traditions concerning the rainy season retreats? In other words, we can hypothesize that places where the Buddha resided in those sources in which attendants other than Ānanda appear must have been the rainy season retreats in the early years after the enlightenment that appear in the retreat traditions, whether or not they are given as rainy season retreats in the scriptural sources.

Evidence for this hypothesis may be gleaned by studying the places of residence of the Buddha when each of the attendants who preceded Ānanda made an appearance in the scriptures as attendants, and their correspondence with places that feature in the retreat traditions. That is the purpose of this article, and our conclusions are set out below.

Places where pre-Ānanda attendants appear are as follows. Nāgasamāla (Nāgapāla) appears at Vesālī (fifth retreat) and the Maṅkula peak (sixth retreat). Sunakkhatta is associated with Vesālī (fifth retreat). Sāgata is connected with a number of places, including Kosambī (ninth retreat), Ceti (tenth retreat), the Vulture Peak above Veluvana (second to fourth retreats), and Suṃsumāragira (eighth retreat). Rādha appears at the Maṅkula peak (sixth retreat). Meghiya appears with Sakyamuni at the Cālikā peak (thirteenth and eighteenth retreats). When Nāgita appears as an attendant, it is in Vesālī.

The connections between pre-Ānanda attendant traditions, the deeds of those attendants in the scriptural traditions, and traditions concerning rainy season retreats are varied. The relationship between places where the Buddha dwelt with his attendants before the time of Ānanda and place names in the rainy season retreat traditions show a disparity between the Northern and Southern Transmissions in their traditions about the locations of rainy season retreats and about attendants, and there is a considerable lack of uniformity too between the two transmissions in such descriptions in the early scriptural sources. It is difficult to unravel what is connected with what, and what is based on what else, in the development of the lore and traditions. We should however consider the following points.

(1) There are cases where the appearance of pre-Ānanda attendants can be confirmed at places appearing as locations of rainy season residence in the traditions, even where there are no reports concerning Sakyamuni's rainy season residence. The Cālikā peak appears in the traditional accounts of rainy season retreats in both the Southern and Northern Transmissions. Since, though, there is no mention of Sakyamuni residing there during the rainy season in the early scriptural sources, as far as Meghiya's deeds at the Cālikā (Cāliya) peak are concerned, the rainy season retreat traditions provide convincing materials for connecting them with items related to the appearance of pre-Ānanda attendants rather than with items about Sakyamuni's rainy season retreats in the early scriptural sources. The deeds of Sāgata at Kosambī and Ceti are contained in part in materials where Ānanda appears, but since Ceti is mentioned in the rainy season retreat traditions contained in the Sengqieluochasuo jijing in the Chinese Avadana but not in items about Sakyamuni's rainy season retreat in the early scriptural sources, this too can be regarded as material showing the linkage between pre-Ānanda attendants and rainy season retreat traditions.

(2) We might expect that the scriptures of the Southern Transmission would match the commentaries of that transmission, and that the scriptures of the Northern Transmission would match the traditions of that transmission, but there are cases where this is not necessarily so. The appearance of Nāgasamāla (Nāgapāla) and Rādha on the Maṅkula peak can be confirmed only in the early scriptural sources of the Northern Transmission; since the Maṅkula peak is not mentioned at all in the Pali Canon, if we explain its occurrence in the rainy season retreat traditions in the commentaries (aṭṭhakathā) in terms of the appearance of the pre-Ānanda attendants, then it must have come about through items in the early scriptural sources of the Northern Transmission. And though Rādha is not mentioned anywhere in the Pali Canon as an attendant of the Buddha, he appears as such in traditions about attendants in the AN.-A. (Burmese edition). This shows a discrepancy between the attendant traditions and the scriptures. Again, it is only in the Mūla-Sarvāstivādin Vinaya that there is a connection between Sāgata and Suṃsumāragira. If we explain the fact of Suṃsumāragira's appearance in the rainy season retreat traditions of the Southern Transmission in terms of Sāgata, then there must have been Northern Transmission influence on the rainy season retreat traditions in the aṭṭhakathā. The reverse can also be found. If we seek the origins of the inclusion of Vesālī in the rainy season retreat traditions through the appearance of Nāgita, since Nāgita appears as an attendant only in the Pali scriptures, we must look to them for how Vesālī came to be listed in the rainy season retreat traditions of the Sengqieluochasuo jijing. However, because we can seek the basis for Vesālī appearing in these traditions through other attendants, we do not need to consider this possibility. Incidentally, Nāgita does not appear as an attendant in the traditions of the Northern Transmission. The above suggests that it might be possible to show the influence of traditions of the Northern Transmission on the traditions in the aṭṭhakathā. There is no convincing evidence for the reverse. Also, since the Maṅkula peak is not found as a place name in Sakyamuni's rainy season retreat episodes recorded in the early scriptural sources, it may be regarded as convincing evidence for a connection between the pre-Ānanda attendants and the rainy season retreat traditions.

(3) There is no particular need to seek among pre-Ānanda attendants in the rainy season retreat traditions to confirm places like Rajagāha and Vesālī, both of which are recorded in the scriptures as places of the Buddha's rainy season residence. Nevertheless, their appearance there provides a reason for considering the event to have occurred in the early post-Enlightenment period. In this connection we may mention the fact that references to Sunakkhatta are concentrated on Vesālī, that Sāgata appears in Rajagāha in the account of Soṇa-kolovisa, and that Kassapa appears as an attendant only in the Pinimujing.

(4) Through the above examination, we may conclude that one base of the rainy season retreat traditions is the pre-Ānanda attendants. Conversely, we can say confidently that there is little possibility that the records in the early scriptural sources of Sakyamuni's rainy season retreats that begin with the formula "At one time the Buddha was spending the rainy season at a certain place with a large number of bhikkhus" were the basis for the rainy season retreat traditions.

(5) The rainy season retreat traditions comprise not just locations but information about the year they took place. We can suppose that there is the possibility of some kind of connection between the order in which the names of pre-Ānanda attendants are listed and the reports about the years given in the rainy season retreat traditions, but there is no uniformity about the traditional lists of attendants, and so it is difficult to make a clear judgement.

We have discussed the rainy season retreat traditions as they appear in the aṭṭhakathā and the Sengqieluochasuo jijing, and elucidated the origins of those traditions according to the above. The remaining locations are listed below. We omit however the first and last locations, Bārāṇasī and Veḷuva.

Though there is a possibility that Guishenjie in the Sengqieluochasuo jijing is the same as Suṃsumāragira, we cannot be completely sure, but we must leave aside that question for now. The legend that a rainy season retreat was held in the Heaven of the 33 gods is widely known today, but there is no mention of this in the Pali scriptures. It was first introduced in the aṭṭhakathā of the Southern Transmission, as part of the legends associated with the life of the Buddha. Separate study is needed regarding why it was placed in that particular position, in the early post-Enlightenment period, before Sāvatthī.

The location of Jetavana as the fourteenth retreat corresponds to the first year of the propagation of Buddhism towards Sāvatthī. In the early scriptural sources, Jetavana is said to have been established for the Buddha to spend his first rainy season retreat in Sāvatthī. Why this was said to have been the fourteenth retreat requires further consideration at another time.

Names of other rainy season retreats reported in the aṭṭhakathā are, as far as present research has been able to confirm, Pārileyyaka, Nālā brāhmaṇagāma, Verañjā, Kapilavatthu, Āḷavī and Rajagāha. There are problems about the chronology concerning Kapilavatthu and Rajagāha, as well as Pārileyyaka and Verañjā. However, as has been presented in Article 5 of Monograph 6, it is clear that the episodes regarding Sakyamuni's rainy season retreats in the early scriptural sources have a solid basis. Remaining are Nālā brāhmaṇagāma and Āḷavī, which, though appearing in the rainy season retreat traditions, are not recorded as locations of the Buddha's retreat in the early scriptural sources. This question should be considered on a separate occasion.