Bhante Sujato recently alerted me to the fact that the word saṅghāṭi – normally understood to mean “outer robe” – is not used consistently in this sense in the EBTs. I had noticed some unusual usages of this word myself, but had dismissed it as the typical variety in and inconsistency of language. But when two people notice the same thing, there are probably good grounds to have a closer look at what is going on, which is what this little write-up is about. For anyone without a special interest in the meaning of saṅghāṭi, the remainder of this essay is bound to be mind-numbingly boring. Should you still be interested in the conclusion, however, I would suggest going straight to the last few paragraphs.
The earliest word that consistently means “robe” in the EBTs seems to be cīvara. This word is used regularly in the pātimokkha rules. By way of contrast, there is nothing in the rules about antaravāsakas (lower robes), uttarāsaṅgas (upper robes), or indeed saṅghāṭis in its usual sense of “outer robe.”
Cīvara also means robe-cloth in general, and it is not entirely clear to me whether this usage existed from the very beginning. In bhikkhu nissaggiya pācittiya 3 cīvara is used in the sense of “robe-cloth,” not just “robe,” and this would seem to suggest that it has meant “robe-cloth” all along. On the other hand, the apparent difference in usage between rule and vibhaṅga in bhikkhu nissaggiya pācittiya 1 (and a number of other rules) – where cīvara seems to mean “robe” in the rule, but “robe-cloth” in the vibhaṅga – could mean that “robe-cloth” is a later development. But the evidence from bhikkhu nissaggiya pācittiya 3 must probably be given the most weight, since it is all internal to the rule, which is part of the earliest stratum of the EBTs. So cīvara probably meant both “robe” and “robe-cloth” from the beginning, but with an emphasis on “robe.”
The pātimokkha also contains another word for robe, sāṭika, as in bhikkhu nissaggiya pācittiya 24, bhikkhu pācittiya 90, and bhikkhunī pācittiya 22. (Although in bhikkhu nissaggiya pācittiya 24 rule we have the compound sāṭikacīvara, which probably shows that sāṭika is not quite synonymous with cīvara.)
In the pātimokkha, in bhikkhu nissaggiya pācittiya 2 and 29, we also find the expression ticīvara, “the three robes,” but nowhere in the pātimokkha are they given individual names, and so this seems to be a later development. (With the partial exception of saṅghāṭi, as we shall see below.) We do, however, find antara and uttara, “inner and upper,” as in santaruttara in bhikkhu nissaggiya pācittiya 7. These are perhaps the earliest individual names for two of the three robes. We also find the words nivāsessāmi and pārupissāmi in the sekhiya rules, also suggesting a distinction between lower and upper robes. Much of this variety in terminology is no doubt due to a gradual movement towards a standardised vocabulary.
The use of saṅghāṭi in the vinaya
Although the three robes are not individually named in the pātimokkha rules, we do find these names in the vibhaṅga material. In bhikkhu nissaggiya pācittiya 20, for instance, we have:
At that time Venerable Upananda the Sakyan had become skilled at making robes. He made a saṅghāṭi of rags, well-dyed and beautifully executed, and he wore it.
Tena kho pana samayena āyasmā upanando sakyaputto paṭṭo hoti cīvarakammaṃ kātuṃ. So paṭapilotikānaṃ saṅghāṭiṃ karitvā surattaṃ suparikammakataṃ katvā pārupi.
According to the continuation of this story, Upananda exchanges his robe for the robe (saṭa) worn (pārupitvā) by a wanderer, and the wanderer puts on (pārupitvā) the saṅghāti he has been given by Upananda. During the whole story, which goes on for quite a bit longer, everyone refers to the saṅghāṭi as if it is a clearly distinguishable garment, not just a technical term for a robe that happens to be the “outer” one. In all of this it seems the word saṅghāṭi is effectively equivalent to an uttarāsaṅga, “an upper robe,” but one made of rags.
In the vinītavatthu (the case stories) to bhikkhu pārājika 4 we have:
As I was coming down from Mount Vulture Peak, I saw a monk flying through the air. His saṅghāṭi was ablaze and burning, as were his bowl, his waistband, and his body.
Idhāhaṃ, āvuso, gijjhakūṭā pabbatā orohanto addasaṃ bhikkhuṃ vehāsaṃ gacchantaṃ. Tassa saṅghāṭipi ādittā sampajjalitā sajotibhūtā, pattopi āditto sampajjalito sajotibhūto, kāyabandhanampi ādittaṃ sampajjalitaṃ sajotibhūtaṃ, kāyopi āditto sampajjalito sajotibhūto.
Here saṅghāṭi could mean “outer robe,” but it is more natural to read it as “upper robe” (uttarāsaṅga), which is the normal robe worn by bhikkhus on the upper part of the body.
In bhikkhunī pārājika 8 we find the following:
If, for the purpose of indulging in inappropriate sexual conduct, a lustful nun consents to a lustful man holding her hand or consents to him holding the edge of her saṅghāṭi or …
Yā pana bhikkhunī avassutā avassutassa purisapuggalassa hatthaggahaṇaṃ vā sādiyeyya, saṅghāṭikaṇṇaggahaṇaṃ vā sādiyeyya … etassa asaddhammassa paṭisevanatthāya …
It would make little sense if this offence should be restricted to outer robes, and saṅghāṭi here almost certainly refers to upper robes in general, that is, both saṅghāṭis and uttarāsaṅgas. Indeed the vibhaṅga expands the meaning to include any robe:
If, for the purpose of indulging in inappropriate sexual conduct, she consents to him holding her lower robe or upper robe, she commits a serious offense.
Saṅghāṭikaṇṇaggahaṇaṃ vā sādiyeyyāti etassa asaddhammassa paṭisevanatthāya nivatthaṃ vā pārutaṃ vā gahaṇaṃ sādiyati, āpatti thullaccayassa.
At bhikkhunī pācittiya 24 we have a similar situation:
If a nun does not move her saṅghāṭi(s) for more than five days, she commits an offense entailing confession.
Yā pana bhikkhunī pañcāhikaṃ saṅghāṭicāraṃ atikkāmeyya, pācittiyanti.
This is then explained in the vibhaṅga as follows:
If she does not wear or sun her five robes for five days, then, when the fifth day has passed, she commits an offense entailing confession.
Pañcamaṃ divasaṃ pañca cīvarāni neva nivāseti na pārupati na otāpeti, pañcamaṃ divasaṃ atikkāmeti, āpatti pācittiyassa.
So again, saṅghāṭi is here a general term for all robes. By the way, as for bhikkhunī pārājika 8, there is no reason to think that saṅghāṭi here refers exclusively to “rag-robes.” It seems that saṅghāṭi needs to be qualified as paṭapilotikā to acquire the specific meaning of “patch-work robe” or “rag-robe.”
At bhikkhunī pācittiya 96 we find the following awkward situation:
At that time a nun entered a village for almsfood without wearing her vest. While she was walking along a street a whirlwind lifted up her saṅghāṭi. People shouted out, ‘She has beautiful breasts and stomach!’”
Tena kho pana samayena aññatarā bhikkhunī asaṅkaccikā gāmaṃ piṇḍāya pāvisi. Tassā rathikāya vātamaṇḍalikā saṅghāṭiyo ukkhipiṃsu. Manussā ukkuṭṭhiṃ akaṃsu – ‘sundarā ayyāya thanudarā’ti.
Saṅghāṭi must here be equivalent to uttarāsaṅga, “upper robe.”
The above usage of saṅghāṭi in three pātimokkha rules suggests quite strongly that this word was used distinctly from or earlier than either uttarāsaṅga or antaravāsaka. If this is correct, then saṅghāṭi must have an additional and earlier usage separate from its use as one of the three robes.
Moving on to the Mahāvagga, in the Mahākhandhaka we find the following:
He should also attend on his preceptor in the water. When he has bathed, he should be the first to come out of the water. He should dry his own limbs and put on his lower robe. He should then wipe the water off his preceptor’s limbs, and he should give him his lower robe and then his saṅghāṭi.
Udakepi upajjhāyassa parikammaṃ kātabbaṃ. Nahātena paṭhamataraṃ uttaritvā attano gattaṃ vodakaṃ katvā nivāsetvā upajjhāyassa gattato udakaṃ pamajjitabbaṃ, nivāsanaṃ dātabbaṃ, saṅghāṭi dātabbā.
Again, we find saṅghāṭi where we would normally have expected uttarāsaṅga. Here we also find the word nivāsana (lower robe) where we might expect antaravāsaka. This combination of nivāsana with saṅghāṭi is common in the Khandhakas.
An even more interesting passage is found just a few lines earlier:
If the preceptor wants to enter the village, the student should give him a lower robe and receive the one heʼs wearing. He should give him a belt. He should put the saṅghāṭis together, overlapping each other and edge-to-edge, and then give them to him. … If the preceptor wants an attendant, the student should put on his lower robe evenly all around, covering the navel and the knees. He should put on a belt. Putting the saṅghāṭis together, overlapping each other and edge-to-edge, he should put them on and fasten the toggle …
Sace upajjhāyo gāmaṃ pavisitukāmo hoti, nivāsanaṃ dātabbaṃ, paṭinivāsanaṃ paṭiggahetabbaṃ, kāyabandhanaṃ dātabbaṃ, saguṇaṃ katvā saṅghāṭiyo dātabbā … Sace upajjhāyo pacchāsamaṇaṃ ākaṅkhati, timaṇḍalaṃ paṭicchādentena parimaṇḍalaṃ nivāsetvā kāyabandhanaṃ bandhitvā saguṇaṃ katvā saṅghāṭiyo pārupitvā gaṇṭhikaṃ paṭimuñcitvā …
Here we have saṅghāti in the plural, that is, it would seem to encompass both the saṅghāti and the uttarāsaṅga. To me this suggests that uttarāsaṅga is either a later term or that it was used differently in the earliest period, otherwise I can see no reason why it would not be used here.
In the Cammakkhandhaka we find:
That bad lay follower then slaughtered that calf, skinned it, and gave the skin to the bad monk. The monk hid the skin in his saṅghāṭi and left. The mother-cow, longing for her calf, followed closely behind him. When the monks asked him why the cow was following him, he replied he did not know. But his saṅghāṭi was smeared with blood, and so the monks said, ‘What happened to your saṅghāṭi?’
Atha kho so pāpupāsako taṃ vacchakaṃ vadhitvā cammaṃ vidhunitvā tassa pāpabhikkhuno pādāsi. Atha kho so pāpabhikkhu taṃ cammaṃ saṅghāṭiyā paṭicchādetvā agamāsi. Atha kho sā gāvī vacchagiddhinī taṃ pāpabhikkhuṃ piṭṭhito piṭṭhito anubandhi. Bhikkhū evamāhaṃsu – “kissa tyāyaṃ, āvuso, gāvī piṭṭhito piṭṭhito anubandhī”ti? “Ahampi kho, āvuso, na jānāmi kena myāyaṃ gāvī piṭṭhito piṭṭhito anubandhī”ti. Tena kho pana samayena tassa pāpabhikkhuno saṅghāṭi lohitena makkhitā hoti. Bhikkhū evamāhaṃsu – “ayaṃ pana te, āvuso, saṅghāṭi kiṃ kata”ti?
Here the saṅghāṭi seem to take the place the uttarāsaṅga had above, that is, it seems to be used to make a bundle (although the word “bundle” is not actually used in this case).
There are further examples of how saṅghāti is used as generic term for upper robe in the Cūlavagga. Here are a couple of examples:
(1) When alms have been given, they should cover the bowl with the saṅghāṭi and leave carefully without hurrying.
Bhikkhāya dinnāya saṅghāṭiyā pattaṃ paṭicchādetvā sādhukaṃ ataramānena nivattitabbaṃ.
(2) At that time Venerable Ānanda had dressed in light saṅghāṭis and then entered the village for alms.
Tena kho pana samayena āyasmā ānando lahukā saṅghāṭiyo pārupitvā gāmaṃ piṇḍāya pāvisi.
In the latter case we again find saṅghāti used in the plural, presumably referring to both the saṅghāti and the uttarāsaṅga.
Turning briefly to the suttas, saṅghāṭi again only seems to be used for monastics, including those of other sects. We also find the word saṅghāṭika, referring to a person who wears a saṅghāṭi:
I do not say that the recluse’s status comes about in a saṅghāṭika through the mere wearing of the saṅghāṭi, nor in a naked ascetic through mere nakedness, nor in one caked with dust and dirt through being caked with dust and dirt, nor in a ritualistic bather through mere ritualistic bathing, nor in a tree-root dweller through mere dwelling at the root of a tree, nor in an open-air dweller through mere dwelling in the open air, nor in a practitioner of continuous standing through mere continuous standing, nor in a taker of food at stated intervals through mere taking of food at stated intervals, nor in a reciter of incantations through mere recitation of incantations; nor do I say that the recluse’s status comes about in a matted-hair ascetic through mere wearing of the hair matted.
Nāhaṃ, bhikkhave, saṅghāṭikassa saṅghāṭidhāraṇamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, acelakassa acelakamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, rajojallikassa rajojallikamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, udakorohakassa udakorohaṇamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, rukkhamūlikassa rukkhamūlikamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, abbhokāsikassa abbhokāsikamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, ubbhaṭṭhakassa ubbhaṭṭhakamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, pariyāyabhattikassa pariyāyabhattikamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, mantajjhāyakassa mantajjhāyakamattena sāmaññaṃ vadāmi. Nāhaṃ, bhikkhave, jaṭilakassa jaṭādhāraṇamattena sāmaññaṃ vadāmi.
The context here suggests that the saṅghāṭi was an ascetic type of garment, perhaps a rag-robe. Although the actual translation renders saṅghāṭi here as a “patchwork-cloak,” there is no obvious reason why “patchwork” should be used.
The use of uttarāsaṅga and antaravāsaka in the vinaya
Before we can draw any definite conclusions from the above, we need a more comprehensive picture of how the other names for individual robes, especially uttarāsaṅga (“upper robe”) and antaravāsaka (“lower robe”), are used. Uttarāsaṅga is used throughout the vinaya (probably over a 100 times) in the standard phrase ekaṃsaṃ uttarāsaṅgaṃ karitvā (“having put the uttarāsaṅga over one shoulder”). Apart from this it is used only rarely. We find it twice to define ticīvara in the vibhaṅga to bhikkhu nissaggiya pācittiya 2 and bhikkhu nissaggiya pācittiya 29. And it is found in the origin stories to bhikkhu nissaggiya pācittiya 5, bhikkhu nissaggiya pācittiya 16, and bhikkhu pācittiya 84, where the uttarāsaṅga is used to make a bundle for meat, wool, and jewellery, respectively. In the Besajja-khandhaka, the lay follower Suppiya wraps her thigh in her uttarāsaṅga after cutting off some of her own flesh to feed the monks!
Turning to the suttas, uttarāsaṅga is actually found in a number of places. However, it does not seem to refer to monastic robes in particular, since it is used also used for lay people. This contrasts with the use of saṅghāṭi which always seems to be used for monastic robes, whether Buddhist or otherwise.
The above usages in the vinaya suggest to me that uttarāsaṅga was used only when there was a need to distinguish it from the lower robe. But since it is not a term used exclusively of monastics, saṅghāti seems to have been the preferred term when no such distinction was required.
As for the antaravāsaka, it must refer to the lower robe, as can be seen from the vinītavatthu to pārājika 4:
As she was paying respect, she held up his antaravāsaka and took his penis into her mouth.
Sā vandantī antaravāsakaṃ ukkhipitvā mukhena aṅgajātaṃ aggahesi.
(I apologise for the rather explicit nature of this, but such is the monastic vinaya.)
And at bhikkhunī saṅghādisesa 3 we have:
Wherever, after covering the three circles, the antaravāsaka gets wet when the nun is crossing, this is called a ‘river.’
Nadī nāma timaṇḍalaṃ paṭicchādetvā yattha katthaci uttarantiyā bhikkhuniyā antaravāsako temiyati.
Apart from this I cannot find anything of much interest about the antaravāsaka in the vinaya. And the word does not seem to occur at all in the four Nikāyas. Nivāsana, the other word for “lower robe,” occurs twice in the four Nikāyas (AN 3.39 and MN 54), and it seems to be the earlier of the two.
The above suggest to me the following evolution in usage. In the very earliest times the robes were simply known as cīvara and they were distinguished as antara (inner) and uttara (outer). (It seems fairly obvious that antaravāsaka is derived from antara, whereas uttarasaṅga probably existed as an independent word from the beginning.) In fact, it may well be that antara and uttara are not names as such, but just markers of the position of the robe, used as a short-hand and succinct description in the “legalese” of the pātimokkha.
Next, judging from the pātimokkha rules, we find saṅghāti used to refer to any sort of simple monastic robe, including the lower robe. This suggestion is reinforced by the almost complete absence of any word for the lower robe in the earliest strata of text (four Nikāyas and pātimokkha). This usage of the word saṅghāti seems to go back to the earliest times.
At some stage saṅghāṭi became a generic term for “upper robe,” as distinct from the lower robe (at first called nivāsana, only later antaravāsaka), with no linguistic distinction being made yet between the upper robe and outer robe. Eventually, however, this evolves further into what became the three-fold description of robes into antaravāsaka, uttarasaṅga and saṅghāti, with antaravāsaka becoming the new term for nivāsana (“lower robe”) and uttarasaṅga (“upper robe”) taking over one of the functions of saṅghāti. From this point on saṅghāti exclusively refers to “outer robe.”
It follows from this that the word saṅghāti has a triple usage. It was first used as a generic term for any monastic robe. It then became paired with the nivāsana to refer to any upper robe. Only later did it acquire the specific meaning of “outer robe.” As an additional observation, I don’t think a definite case can be made that the earlier usage of saṅghāti refers specifically to “rag-robes,” or “patchwork robes.” However, since the saṅghāti seems to refer specifically to monastic robes, it seems likely that it was frequently made of rags. This could account for its regular description as a patchwork robe in the suttas.
So the question is: to what extent should this triple usage of saṅghāti be reflected in translation? Ideally I would suggest we should use “monastic robe” for its earliest usage, then “upper robe,” and finally “outer robe.” But is this going to work?