The ’go ba'i lha – usually translated with “personal protective deities” – are often approached as an integral part of Tibetan popular or folk religion.Footnote 1 These gods are usually said to be born with an individual, to reside in his or her body, and to protect various facets of his or her existence. While there are varying numbers of gods belonging to the group, depending on the source, the most common formulation includes five, e.g. pho lha, mo lha, zhang lha, srog lha, and dgra lha. As for the etymology of ’go ba'i lha, while “protective deities” is the dominant translation, it remains a highly communicative and contextual gloss of ’go ba, whose attested meanings do not include “to protect”. This communis opinio is probably informed by an alternative name for the group, mgon pa'i lha. The present contribution queries the category of ’go ba'i lha by offering a new analysis of the verb ’go based on attestations in Old Tibetan texts from Dunhuang from the ninth and tenth centuries. In doing so, the article not only analyses the meaning of the verb ’go, and proposes a new etymology of ’go ba'i lha, but also touches on the manner in which Tibetans have imagined the relationship between humans and divinities within their rituals and cosmologies.
A note on the ’go ba'i lha lnga
While this enquiry is devoted principally to the meaning of the verb ’go, it has implications for the ’go ba'i lha lnga, usually translated with “five personal protective deities”. Since this phrase and this group of deities provide the most well-known context for the appearance of ’go ba, it will suffice to offer a few prefatory remarks on them. The ’go ba'i lha lnga have been the subject of numerous studies, and are mentioned in most surveys of Tibetan folk religion. To summarize briefly what has been written about these deities, we should begin with the fact that they do not always appear as a pentad. Giuseppe Tucci (Reference Tucci and Samuel1980: 193) observes that they “are classified in a quite different manner” and occur in groups of 5, 7, 13, and 21. The deities are said to be born with the individual, and to reside in his or her body. The group of five is the most dominant classification, but even within this pentad the identities of the five deities vary. Nebesky-Wojkowicz (Reference Nebesky-Wojkowicz1956: 327–8) made this point in his influential, albeit brief, treatment of the topic. In two of his sources,Footnote 2 the group consists of the mo lha, srog gi lha, pho lha, yul lha, and dgra lha, but in another sourceFootnote 3 it consists of the pha lha, ma lha, zhang lha, dgra lha, and srog lha. Also, depending on the source, these deities are said to reside at the crown of the head, on the shoulders, in the armpits, and at the heart, with the particular location of each god once again varying (Stein Reference Stein1962: 187; Vinkovics Reference Vinkovics and Kelényi2003: 90; Berounský Reference Berounský and Pakhomov2007: 331–2).Footnote 4
As for the ritual and cosmological relevance of this group of gods, the ’go ba'i lha lnga play a role in Tibetan marriage when a bride moves into her husband's home and must undergo a ritual in which her personal and household deities are detached so that those of her husband and his house can be attached to her (Shastri Reference Shastri and Kvaerne1994: 760; Karmay Reference Karmay, Karmay and Watt2007: 161). This is referred to as “loosening the gods” (lha bkrol) and “attaching the gods” (lha ’dogs). The ’go ba'i lha also come into play in tantric rituals for harming or killing one's enemies. In order to make one's victim most vulnerable to attack, one must first carry out rituals to remove his or her ’go ba'i lha lnga. This is evident in the fascinating rkyal ’bud ritual studied by Amy Heller, but it is equally a feature of several other tantric practices of “liberation” (Heller Reference Heller, Aziz and Kapstein1985: 262–3).
Previous translations of ’go ba'i lha
There are various translations for the word ’go, many of which are most easily apparent from how scholars have translated ’go ba'i lha or ’go ba'i lha lnga. The examples that follow are representative of the different choices that translators have made vis-a-vìs the word ’go, but it should be noted that these do not come from reflective studies of the word, and most seem to rely on the scholarly consensus that ’go ba'i lha means “personal protective deities”. It is clear, too, that some scholars take ’go for a noun, while others read it as a verb. For our purposes, the few instances of footnotes justifying a scholar's translation are of most use, in that they make explicit how ’go is understood in a given translation.
Scholars have generally opted for one of four main translations for ’go ba'i lha lnga. The most dominant of these is “personal protective deities” or some variant; a more philologically grounded translation is “sticking deities” who are “like a shadow”; an early curiosity is “five superior demi-gods”; and a translation that takes ’go for a noun is “head gods”. The most influential translation comes from Nebesky-Wojkowicz's brief treatment of the ’go ba'i lha lnga, which he translates with “five guardian-deities of each person” (Reference Nebesky-Wojkowicz1956: 264). In his survey, Tibetan Civilization, Rolf Stein (Reference Stein1962: 187) translates them with “dieux protecteurs”, which was also the preferred choice of Ariane Macdonald (Reference Macdonald1971: 300–1). Tucci (Reference Tucci and Samuel1980: 193), in The Religions of Tibet, translates ’go ba'i lha with “guiding deities” and “protective gods”. Todd Gibson echoes this in a short appendix to his 1991 PhD thesis, where he translates ’go ba'i lha lnga with “five protective spirits” (Gibson Reference Gibson1991: 256). Geoffrey Samuel, surveying previous remarks largely by Nebesky-Wojkowicz, Stein, and Tucci, prefers “personal protective deities” and “five guiding deities” (Samuel Reference Samuel1993: 187, 438). Gyurme Dorje (Reference Dorje2001: 296, 303–4), in his translation of the White Beryl, translates ’go lha with “life-long companion gods”. Dung dkar Blo bzang ’phrin las further ratifies this with his gloss, “srung byed kyi lha” (Dung dkar 2002: 630). Gergely Orosz (Reference Orosz and Kelényi2003: 22) supports this same view with his translation, “five protecting gods”, as does Judit Vinkovics (Reference Vinkovics and Kelényi2003: 90–1) with her “five personal protective deities”. Sarah Harding translates ’go bar byed pa'i lha lnga, a variant of ’go ba'i lha lnga, with “five patron gods”, and reports Ringu Tulku's comment that these are “five gods that stick with one like a scent (’go)”Footnote 5 (Harding Reference Harding2003: 128, and 309 n. 22).
Ringu Tulku's comment brings us to another etymology of ’go ba'i lha lnga. Daniel Berounský translates ’go ba'i lha lnga with “five sticking deities” who are characterized as being “stuck to the body of the person as a shade” (Berounský Reference Berounský and Pakhomov2007: 331). Namkhai Norbu also makes use of this latter simile in his book Drung, De'u and Bon. There, Norbu translates ’go ba'i lha lnga with “five deities of the individual”, and describes them as following the individual “like a shadow” (Norbu Reference Norbu, Clemente and Lukianowicz1997: 65, 66). A shadow is also mentioned in the Rgya bod tshig mdzod chen mo’s definition of ’go ba'i lha: “protecting gods or class of non-human beings that accompany each person's body like a shadow” (mi so sor lus dang grib ma ltar ’grogs nas skyob pa'i lha'am mi min gras; Zhang et al. Reference Zhang1998: 498).
Another translation of ’go ba'i lha lnga appears to have begun with Sarat Chandra Das, whose dictionary entry defines them as “five superior demigods” (Das Reference Das1902: 291). Roerich followed this rendering in his dictionary, but otherwise it has been largely ignored (de Roerich Reference de Roerich, Parfionovich and Dylykova1983–93: 194).
A further strand of translation trades on the meaning of ’go ba as “chief”, “beginning”, and as a variant for “head” (mgo). We can cite here Françoise Pommaret's translation of ’go ba'i lha lnga with “the five head gods” (Pommaret Reference Pommaret, Blondeau and Steinkellner1996: 41), which she justifies in a footnote as a play on words incorporating “various meanings of ’go head, but also chief and principal”. Samten Karmay follows a similar tack, translating ’go ba'i lha lnga with “five gods of the head”, and noting also that within Tibetan culture the head is the most sacred part of the body and is to be kept undefiled (Karmay Reference Karmay, Karmay and Watt2007: 160–2). Francis Lim, in his monograph on the Lang tang region of Nepal, also favours the “five head gods” for translating ’go ba'i lha lnga (Lim Reference Lim2008: 95, n. 4).
As is already clear from this brief and by no means exhaustive survey, the most attractive choice for translators and lexicographers has been to gloss ’go ba in ’go ba'i lha lnga with “protective” or a close synonym. Beyond the excellent contextual reason, namely that these are gods born with the individual, and reside in the body and protect the individual and his or her fortune, the texts themselves also support this gloss. Berounský, for example, cites a passage in the Mi'u rigs bzhi lha sel, a text published in Karmay and Nagano's collection The Call of the Blue Cuckoo, where mgon pa “protector” stands in the place of ’go ba, i.e. the group is referred to as the mgon pa'i lha lnga, or “five protecting deities” (Berounský Reference Berounský and Pakhomov2007: 334–5; Karmay and Nagano Reference Karmay and Nagano2002: 1–33). This is not an isolated occurrence.
The verb ’go
Translations such as “companion”, “stuck”, and “head gods” are understandable given the attested meanings of ’go. Turning to dictionaries, Jäschke (Reference Jäschke1881: 95) defines ’go as “to stain, to lose colour; to dirty, sully oneself”; “to infect, with a disease”. These definitions are repeated in Das (Reference Das1902: 291) and de Roerich (Reference de Roerich, Parfionovich and Dylykova1983: 194), the latter in dependence upon the former. Johannes Nobel, in his dictionary to the Suvarṇaprabhāsottama-Sūtra, similarly has ’go as “pf. gos, beschmutzen” and “anstecken (mit einer Krankheit)”. Nobel also offers a note that one should compare ’go with another verb, bsgo (Nobel Reference Nobel1950: 33). In the Mahāvyutpatti, two entries within the section on the names of illnesses include ’go ba. Number 9526 is nad ’go ba = saṁcāravyādhiḥ, meaning “infectious disease” and number 9555 is ’go ba'i nad = upasargaḥ, also meaning “infectious disease” (Sakaki Reference Sakaki Ryōzaburō1916: 609, 610). These meanings are also covered in the Rgya bod tshig mdzod chen mo: gos pa/ ’go ba// shan zhugs pa dang/ sbags pa'am ’byar ba/…’go ba'i na tsha ’gos pa/…skyon ’gos pa/…gos pa'i nyes pa/ (Zhang et al. Reference Zhang1998: 498). Most recently, the Wörterbuch der tibetischen Schriftsprache ratifies the understanding of ’go ba'i lha as designating protective deities: “Bez. von Schutzgottheiten” (Franke et al. Reference Franke, Hartmann and Höllmann2005: 373). For ’go, it lists two meanings: first, “Anfang, Erstes, Kopfteil; vgl. mgo”; and second, echoing Nobel, “pf. ’gos anstecken [mit einer Krankheit]”.
The attested meanings of the verb ’go are helpfully summarized in Nathan Hill's dictionary of verbs as attested in the Tibetan grammatical tradition: “To stain, lose color, dirty sully. To be infected; to be covered. For one thing to stick to or contaminate another thing. 1. To catch (an illness), to be infected. 2. To be stained, to be sullied, to have a stain or substance stick to oneself” (Hill Reference Hill2010a: 49). Citing Betty Shefts Chang (Reference Chang1971: vii.18), who picked up Nobel's suggestion, ’go is listed as the resultative form of sgo, and the latter's entry reads: “(obsolete) To say; to speak, (mostly) to bid, to order. … 1. To make one thing to stick to or contaminate another thing. 2. To teach to another” (Hill Reference Hill2010a: 64).
In one of Rolf Stein's earliest articles, on a pebble divination text of unknown provenance, but coming perhaps from Ladakh, we find the verb ’go in a context similar to that which we will explore below. Here ’go appears three times in an auspicious prognosis associated with the goddess Lha sman dkar mo: bsrungs ma ’go te nyin gyi bya ra mtshan gyi mi rtse byas / lha sman bzang mos mgo te…klu sman ’go bas klu gtor btang /. Stein translates this with “Sous la surveillance de ce génie tutélaire, faites les veillées de jour et de nuit! Sous la présidence de la Bonne Médecine des Dieux … Mais si vous êtes protégé par la (déesse) Médecine des Génies Aquatiques (klu, nāga) offrez un torma pour Génies Aquatique (klu)!” (Stein Reference Stein1939: 308). Stein thus translates the verb ’go first with “surveillance” and then with “protégé”. Mgo, which is here a verb, Stein translates with “présidence”. This is almost certainly the same verb, as Stein observes in a long introductory note on orthography where he uses mgo ~’go as an example of the alternation of m and ’ prefixes in this text (Stein Reference Stein1939: 302). In each case, the action or attitude is performed by a deity, which is precisely the context we will now explore in Old Tibetan divinatory and ritual texts in order to attempt to clarify the meaning of the verb ’go with recourse to the earliest available textual witnesses.
The verb ’go in Old Tibetan
Many of the previous analyses of the meaning and function of ’go ba in the context of ’go ba'i lha lnga are well thought out, and take into account various social facts. It is not the intention of this article to question their characterizations of the connotations of ’go ba'i lha lnga, and of various Tibetan understandings of them. They are, functionally, “personal protective deities”, and only a pedant would fault a translator for using this contextual gloss. Rather, the aim here is to subject the verb ’go to a lexicological analysis in order to elucidate its meaning and syntax based on examples in Old Tibetan (OT) texts where I have encountered it. This will make it easier to assess proposed etymologies linking ’go with “protect”, “head”, “beginning”, and “to be stuck”, and with analyses that pair it with the verb sgo.
Reading widely in OT ritual texts, including several texts concerning dice divination, one repeatedly encounters the verb ’go in contexts describing, in a formulaic manner, the proper and improper relationships between humans and gods. I collate these here, with my translations, before offering a semantic and syntactic analysis. I exclude the many other examples in which ’go appears in a context where it clearly means “chief” or “beginning”, where it is a variant for the noun “head” (mgo), or where it is part of a person's personal or family name.
The translations are provisional, and they do not exhaustively treat the philological problems posed by these passages when these are peripheral to the matter at hand. For the sake of brevity, I have sometimes skipped over text in my translations, advertising this with ellipses. ’Go and another pivotal verb, ’che(s), are left untranslated in the presentation of the passages, but definitions are proposed in the analysis.
Over two-thirds of the attestations given here occur in dice divination texts, with which we begin. Nine of these attestations are found in the same text, ITJ 738, and these occur in the same context, at the beginning of the evaluation that follows the cryptic, sometimes versified omen.Footnote 6 These typically issue from the mouths of named deities, and address the supplicant in the second person as “you, o human!” (myi khyod). There are surely other instances of ’go in OT documents that I have overlooked. The following 20 passages, however, should constitute a representative sample, and, to the extent that they agree on a specific meaning of the verb, one would expect this to hold true for further attestations not included here.
All but one of these twenty examples come from Dunhuang, the one exception coming from Mīrān. Tibetan dice divination texts, according to Tsuguhito Takeuchi, largely date from the period of the Tibetan Empire, and the Tibetan occupation of Dunhuang (786–848), with one exception being the tenth-century divination codex ITJ 739 (Takeuchi Reference Takeuchi, Takeuchi and Hayashi2012: 9). None of the texts has been subjected to a detailed study of its codicological, palaeographical, orthographical, and linguistic features, however, so we should be circumspect in making assumptions about their dates.Footnote 7 It is preferable at present to leave the matter vague, and to state that the following passages date generally to the ninth and tenth centuries.
Passages
1. ITJ 738, 1v13–14
khyIm phya dang srog phya la btab na / / myI khyod la lhad pal Footnote 8 bzang po cIg ’go zhong Footnote 9 ’ches zhIng ’dug pas / bgyegs sil Footnote 10 la lha [mchod] cIg / cI bsam ba bzhIn du grub par ’ong ’o / / mo ’dI cI la btab kyang bzang / /
“If you've cast [this divination] for household fortune or life-force fortune, then human, an excellent sacred god ’go-s and ’ches-s you. Clear away the bgegs [demons] and worship the gods! Whatever you wish for will be accomplished. This divination is good for whatsoever you have cast it.”
2. ITJ 738, 1v19–21
/ / khyIm phya dang srog phya la btab na / / lha bkur phong pas / / lha ’go zhing ’dug / / snyIng dang sbyan ba thong na rang la pha[-] gdon lan ma mchIs / / gnyen byas na bu tsa gzhin / / khang kyim byas na kha dro / ’dron po la btab na ’ong / / bor lag [byung?] na rnyed / / mo ’dI cI la btab kyang bzang / /
“If you've cast [this divination] for household fortune or life-force fortune, then though you have renounced respecting the gods, the gods ’go you … This divination is good for whatsoever you have cast it.”
3. ITJ 738, 1v84–85
khyIm phya dang srog phya la btab na lha dpal ’che ste ’go / / dgra phya la btab na / dgra myed srId phya la btab na srId phyin / / don gnyer na grub / / ’dron po la btab na ’ong / / mo ’di cI la btab kyang bzang / /
“If you've cast [this divination] for household fortune or life-force fortune, the sacred gods ’che and ’go you. If you've cast it for enemy fortune, there shall be no enemy. If you've cast it for livelihood fortune, livelihood will be provided. If you've cast it for a certain matter, it shall be accomplished. If you've cast it for a visitor, s/he will come. This divination is good for whatsoever you have cast it.”
4. ITJ 738, 3v57–61
khyim [phya dang] srog phya la btab na / / ma myes gyI lha bzang po cIg yod gIs / / ’go zhing ’che bar ’ong bas / / don po dang gnum lha pa yod na / / gsol cIg / / ma dang bu gnyIs phrad phar ’ong ba dang ’dra ste / / skad snyan te phrad par 'ong / / gnyen byas na srId yod / / grog phya la btab na grog che / / gsol shags byas na gnang / / tshong byas na khe phyin / / ’dron po la btab na [’ong] / / [mo ’dI] cI la btab kyang bzang / /
“If you've cast [this divination] for household fortune or life-force fortune, since there is the excellent god of your maternal ancestors, it will ’go and ’che you … This divination is good for whatsoever you have cast it.”
5. ITJ 738, 3v62–66
mo ’dI nI khyIm phya dang srog phya la btab na / / dpal lha bzang po myi thams shad la ’go ba dang ’dra ste / / myI khyod kyang shing bzang po [lo?] ma [---] la [rten?] bcas na / grIb ma che ste bsIl ba la phan pa dang ’dra ste / / rI mthon po la kharu byung na / snang yal che / / rje btsan ba la[-]kyabs [bdzal?] na / / kha gzes myI tshugs / / lha bzang po la mchod na [phug (/phyug)] zhing mtha’ myI ’grI / / don gnyer na grub lag byung na rnyed / / ’dron po la btab na nye zho myed ste ’ong / / gnyen byas na bzang / / mo ’di [ci] la btab kyang bzang rab ’o /
“If you've cast this divination for household fortune or life-force fortune, since the excellent sacred gods seem to ’go ba all people, to you also, human … this is like an excellent tree, whose leaves offer extensive cool and beneficial shade … This divination is very good for whatsoever you have cast it.”
6. ITJ 738, 3v74–79
khyIm phya dang srog phya la btab na / myI khyod [sngun] nyon mongs zhIg ’dug la / / lha bzang po zhig ’go zhIng ’dug ’o / / mye ngan ched po la skal mas nas ’dug pa las / gnam gi lhas char phab nas / skom rnyed pa dang ’dra / / srId myed nas phongs las / ma lha chen mos srId byIn ba'I ngo yod / / rje blas zhus na bka’ lung snyan pa thos pa'I ngo / / tshong zhig byas na khe phyIn / / ’dron pho cig la btab na phyIn te ’ong / / nad pa la btab na stsI sman myI dgos / / don gnyer na grub / / shIs phya la btab na bzang / / mo ’dI cI la btab kyang bzang ’o / / /
“If you've cast [this divination] for household fortune or life-force fortune, you man, who were previously afflicted, an excellent god ’go-s you. It is as if the gods of the sky cast down the rain, and a great suffering that had been present for ages were quenched. It is the omen of the great mother goddess giving livelihood to one who has no livelihood and is poor. It is the omen of hearing a pleasing pronouncement when one has applied for official duty. If you go trading, you will make a profit. If you've cast it for a visitor s/he will come. If you've cast this for a sick person, s/he won't need a medical elixir. If it is for a specific matter, it will be accomplished. If you've cast it for shIs Footnote 11 fortune, it is good. This divination is very good for whatsoever you have cast it.”
7. ITJ 738, 3v115–118
mo ’dI nI khyIm phya dang srog phya la btab na / / lhad Footnote 12 dpal bzang po ’am sman dkar mo zhIg yod pas / / de la mchod ’phras Footnote 13 legs par byas na / / khyed la ’go zhing ’dug pas / rIgs Footnote 14 bshor na sod / rje blas zhus na gnang / / don gnyer na grub / ’dron po la btab na ’ong / / nad pa la btab na sos / gsol shags byas na gnang / / bor lag byung na rnyed / / tshong byas na khe phyIn / / mo ’dI [ci] la btab kyang bzang /
“If you've cast this divination for household fortune or life-force fortune, since there is an excellent sacred god or a white sman goddess, if you offer well to [the god or goddess], s/he will ’go you. If you go hunting, you will have success/kill [your quarry]. If you request official duty, it shall be granted. If you undertake some matter, it will be accomplished. If you've cast this for a visitor, s/he will come. If you've cast it for a sick person s/he will recover. If you've cast if for a legal dispute, it will be granted. If it is for something lost, you will find it. If you go trading, you will make a profit. This divination is very good for whatsoever you have cast it.”
8. ITJ 738, 3v130–131
mo ’dI nI khyIm phya dang srog phya la btab na / / khyed sngun gI pha myes lha bzang po cIg yod zIn / / da lta yang lha de cung thag rIng zhIng ’dug pa ltar byed gyIs / / lha de legs par gsol na / khyed la ’go zhing ’ches phar ’ong / /
“If you've cast this divination for household fortune or life-force fortune, there was previously the excellent god of your paternal ancestors, and now, although that god seems as if it were a long way off, if you offer well to that god, it will ’go and ’ches you.”
9. ITJ 738, 3v156–158
mo ’dI nI khyIm phya dang srog phya la btab na / / lha cIg ’go zhIng ’dug pas / / lha de ma yIn na myI khyod ltogs Footnote 15 te yang de'u shI las lta des bskyabs zhIng ’dug ’o / / cI bya na yang lha la gsol ba thob la gyIs dang / bya bar rung bar ’ong / /
“If you've cast this divination for household fortune or life-force fortune, a god ’go-s you. If it is not that god, [and] you, o human, turn your back, de'u shi las lta,Footnote 16 it does protect you. In whatever you will undertake, pray to the gods and do it, and that which you undertake will turn out well.”
10. ITJ 739, 7v3–5
lha gnyan po ni ’go zhing tshad zin pas ni rang skyid / da ltar khyod mnga bas mthar bar ’ong ste bzang rab bo /
“The powerful gods ’go you and are as delighted as can be. Now what you possess will become complete. [A] very good [divination].”
11. PT 1052, r40–42
de d[-]d [-]n Footnote 17 / lha dpal ’go [ste] skyid cing bde bar 'ong / [-o] / bton lan [myi (/mye)] [---] mo ’dI jI la btab gyang bzango / /
“…the sacred gods ’go [you], and [you] will become delighted and happy … For whatsoever you cast this divination, it is good.”
12. PT 1052, r230–233
mo ’di jI la btab gyang lha dpal che Footnote 18 / nas dpal ’go zhing bla ma ’is skyabs byed pa yod pa skyId cing phyugs te bzango /
“For whatsoever you have cast this divination, there are the sacred gods that che, and the sacred [gods] that ’go, and the bla ma that acts as a refuge. They are delighted and prosperous; it is good.”
13. S.155, v51–55
ngam le [mgyungs] ke'o zhal nas myI khyod la lhad Footnote 19 dpal ’goste dgra dang g.yag la lan [ka dang] phrad na yang [ ± 2] ’ong ngo / blon por bcug na srid du ’ong ngo khyod la ’phrul gyI dum [-] [nad] cig byin byas [lge rmyis] kyang mkhas ba'i rigs / zla ba tshes shIg yul phyogs su rI mthong [ba'I] khar zhugs [shang sgol?] cig dang ji bka’ ba dang jI ’dod pa dang phrad do / rtags su ni thogs so [ ± 2] thab cing rmyi'Is cha'o // Footnote 20
“From the mouth of Ngam le mgyungs ke'o: ‘you, o human, the glorious gods have ’gos-ed you, and even if you should meet with the retribution of yaks and enemies … will happen. If you are appointed councillor, you shall gain power … you will meet with whatever you desire …’.”Footnote 21
Moving on from dice divination texts, we also find ’go used in a relevant context in a prayer, a ritual text, an apocalyptic text, Buddhist polemics, and (Confucian) sayings.
14. ITJ 751, 36r4 (Prayers of De ga g.yu tshal)
gtsug lag khang thams chad tshul bzhin mdzad pas nI / sku bla rIng rdzI dang / gnam sa'i lha klu / thams chad kyang ’go zhing dgyes / /
“…By [Emperor Khri Gtsug lde brtsan's] correctly founding all the temples, the tutelary deities and ring rdzi and all the gods and spirits of heaven and earth ’go and are delighted.”Footnote 22
15. PT 1194, 21–22 (Non-Buddhist funeral rite)
pha mas lasu lha mchod mchod na mgosu lha ma ’goste dmu dag Footnote 23 dkar po gyang du bchado
“… when his parents … were offering to the gods, the required gods did not ’gos, and the white dmu cord was indeed cut.”
16. PT 126, ll. 47–48; also found in PT 992, 12r8–9 (“A Summary/Sūtra of the sacred Bhikkṣu's Teachings for Future Generations”; ’Phrul kyi byig shus phyi ma la bstan pa'i mdo)
bde ba'i tshe na lha myi dran / / na ba'i tshe na / lha la ltos Footnote 24 / sdig to myi la lha myi ’go /
“When you are happy, you don't pay mind to the gods. When you are sick, you look to the gods. [You] are wicked. The gods do not ’go men [like you].”Footnote 25
17. PT 992, 15r3–4 (“A summary/sūtra of the sacred Bhikkṣu's teachings for future generations”)
myi ’dzangs pa ni ma bskos par myi ’jug la / / lha gnyen po ni myi mchod par myi ’go / /
“A wise man, unappointed, cannot be of use, and a powerful god, unpropitiated, does not ’go.”
18. PT 1058b, v13–14 (a text on morality, possibly Confucian)
skyes bu dam pa ma bcug par myi rten / / lha gnyen po ma mchod par myi ’go / /
“A true person, unappointed, cannot be of use. A powerful god, unpropitiated, does not ’go.”
19. ITJ 733, ll. 1–3 (Apocalyptic text)
[g]nam du bro[s] nas / yul dang r[i] st[o]ng par gyur to / bar bar du yul sa dang chos ma nor par rjed na / [---] stong pa [la] rjed pa’ nI / jI la yang myi phan te’ / ’greng myI ’o cog gtham zer na’ lha ’dI sngun lta[r] ’go’ la myi drag na ji nyes shes zer bar ’ong ngo’ / / ’ung nas chos dang tshe ngan pa la babste
… escaped to the heavens, and the lands and the mountains became empty. Gradually, when they unerringly worshipped the customs (chos) and yul sa … worshipping empty [lands and mountains] nothing would be benefitted. All of the people had a saying: “What a shame if this god is not strong enough to ’go us, as it did before!” And then came the time of evil customs and the evil age.
20. Tak 490 (M.I. xii.5/ Or. 15000/378; fragment from Miran)
$/:/ blon skyes bzang la/ / lha ’go'i[…] laM ’tsal zhing mchis/ / rgyal b[zang…] nas/ / [ ± 7] btang b[…] bdag gi [ ± 4] dus dusu gdab […] Footnote 26
A [letter from] Lha ’go to councillor Skyes bzang …Footnote 27
Semantic analysis
To summarize, ’go is performed by gods for humans, usually following offerings that humans have made to the gods. The gods' act of ’go ba can also be withheld from undeserving supplicants. ’Go is often paired with the verb ’che(s), and it is sometimes accompanied by delight, either of the gods, the humans, or both.
Of the divination prognoses, passages 7 and 8 describe a reciprocal relationship between humans and gods: the humans make offerings to the gods, and the gods then ’go them or ’go and ’ches them.Footnote 28 While passages 14 to 19 contrast with 1 to 13 in that they describe an inauspicious situation in which the gods do not protect humans, they also lay bare, in its rupture, the characterization of the ideal relationship between humans and gods. This dynamic of exchange, or of the conditionality of the gods' act of ’go ba, is strongly reinforced in particular by passages 15, 16, 17, 18, and possibly also 19. Most striking in this regard is the repetition (in both the “Summary/Sūtra of the sacred Bhikkṣu's Teachings for Future Generations” and in the moral sayings) of the phrase “a powerful god, unpropitiated, does not ’go” (lha gnyen po ma mchod par myi ’go).
We observe that all of the examples from divination prognoses are from good (bzang) or very good (bzang rab) prognoses. The surrounding context, where I have given it, also confirms that ’go has a positive and auspicious value. The direct or indirect result of the gods' performing ’go, generally in recompense for offerings, is invariably good: in passage 1 it states for example that “whatever you wish for will be accomplished”. In passages 1, 3, 4, 8, and probably also 12, ’go is paired with the verb ’che(s). In passages 10, 11, 12, 13, and 14 it is associated with delight and happiness, either of the god(s) or of the human supplicant or divination client.
In terms of the broader picture of interaction between humans and gods in the context of divination, there are other verbs and phrases that are also stereotypical. Without venturing too deeply into the semantic fields of divination, we can observe, for example, that humans “revere” (rjed), “honour” (bkur), “make obeisance” (phyag ’tshal), “petition” (gsol), “offer gifts” (yon gsol), and “make offerings” (mchod) to gods. Gods often “show consideration” (lhas thugs dgongs mdzad), “protect” (srung or skyob) and are pleased (dgyes, thugs dgyes or dga’) or displeased (myi dgyes or myi dga’). A further attitude is expressed by the verb ’phangs, when a god is repulsed by the humans and abandons them. This meaning was well appreciated by Ariane Macdonald, who already remarked in the context of the divination text PT 1047 on “l'expression ’phang-ba ou myi-phang-ba, qui caractérise l'hostilité ou la faveur des Sku-bla à l’égard des hommes…” (Macdonald Reference Macdonald1971: 301). In this context we can observe that ’go is essentially the antonym of ’phangs.Footnote 29
The conditional reciprocity of the relationship between humans and gods is also very clear from a reading of OT dice divination texts. To choose but one example, this is the prognosis in ITJ 740 for the dice roll 2 – 1 – 4:
@@ @ @@@@ $ :/ / myI khyod chaste ’gro na lha la phyag ’tshol cIg dang lhas thugs dgongs mdzade dgra dang g.yag la song na yang dgra thub g.yag sod / tshong bya na tshong rgyal / lha la phyag ma ’tshal na ’dI kun myI ’byor gyIs lha la phyag ’tshol cIg / dang mo bzango /; ITJ 740, ll. 158–162.
“You, o human – if you set out and go, make obeisance to the gods, and the gods will show you consideration. If you are going against yaks and enemies, you'll defeat the enemies and kill the yaks. If you go trading, you'll prevail at trade. If you don't make obeisance to the gods, you won't enjoy any of this, so make obeisance to the gods! A good divination.”
The various roles of the gods within the omens and prognoses of Tibetan divination are too complex to describe in detail here.Footnote 30 For the present, it suffices to emphasize reciprocity and conditionality as the operative principles governing human–god relations around the verb ’go.
Syntactic analysis
These passages also establish the normative syntax of ’go as absolutive–oblique, with the agent or theme marked in the absolutive (Ø) and the patient or beneficiary marked in the oblique (OBL), i.e. “god Ø human OBL V-’go,” or, in reverse order, “human OBL god Ø V-’go.” This is clear from the following examples:Footnote 31
5. dpal lha bzang po myi thams shad la ’go ba dang ’dra ste
sacred god good Ø human all OBL V-’go NOM SOC like SEM
13. myI khyod la lhad dpal ’goste
human you OBL god sacred Ø V-’gos SEM
16. sdig to myi la lha myi ’go /
wicked SFP human OBL god Ø neg. V-’go
1. myI khyod la lha dpal bzang po cIg ’go zhing ’ches
human you OBL god sacred good one Ø V-’go COOR V-’ches
6. myI khyod [sngun] nyon mongs zhIg ’dug la / / lha bzang po zhig ’go zhIng ’dug ’o
human you before afflictions one V-sit OBL god good one Ø V-’go COOR V-sit SFP
8. khyed la ’go zhing ’ches phar ’ong
you (pl.) OBL V-’go COOR V-’ches NOM-TERM V-come
All of the examples display a clear syntax in which the agent (or theme) of the verb ’go is marked in the absolutive. This is particularly clear in the following examples:
2. lha ’go zhing ’dug /
god Ø V-’go COOR V-sit
3. lha dpal ’che ste ’go
god sacred Ø V-’che SEM V-’go
9. lha cIg ’go zhIng ’dug pas
god one Ø V-’go COOR V-sit NOM-ERG
10. lha gnyan po ni ’go
god powerful Ø ISO V-’go
14. gnam sa'i lha klu / thams chad kyang ’go zhing dgyes /
sky earth GEN god serpent-spirit all CON V-’go COOR V-delight
The possible counter-examples where one finds an ergative following the agent or theme of the verb ’go are in passages 4 and 7.
4. lha bzang po cIg yod gIs / / ’go zhing ’che bar ’ong
god good one V-exist ERG V-’go COOR V-’ches NOM-TERM V-come
7. / lha dpal bzang po ’am sman dkar mo zhIg yod pas / /
god sacred good OPT goddess white one V-exist NOM-ERG
de la mchod ’phras legs par byas na / / khyed la ’go zhing ’dug pas
that OBL V-offer fruit good NOM-TERM V-did INE you (pl.) OBL V-’go COOR V-sit NOM-ERG
These are clearly “ergatives of reason” and not “ergatives of the agent”, and their use as such is evident from my translations in the passages section above.
Morphology
As for the morphology of the verb ’go, it appears in the form ’go in the present (e.g. ’go zhing ’che bar ’ong in passage 5) and future (e.g. myi la lha myi ’go in passage 16) stems. There are two possible occurrences of a past stem ’gos, with the negation ma, in passages 13 and 14 in the phrases lhad dpal ’goste and mgosu lha ma ’goste. There is no attested example of the imperative stem.
In passages 2, 6, 7, and 9, the verb ’go is followed by the perfective auxiliary zhing ’dug, indicating an ongoing state or action, and the same auxiliary follows ’go zhing ’ches in passage 1.
Proposed definition
As a preliminary to suggesting a definition of ’go, and placing it in the context of other Tibetan verbs, let us consider briefly its frequent companion, the verb ’che(s). This verb appears in close connection with ’go in passages 1, 3, 4, 8, and 12. In passages 1, 4, and 8, the two verbs are linked by the coordination particle:
1. ’go zhing ’ches;
4. ’go zhing ’che bar ’ong;
8. ’go zhing ’ches phar ’ong.
Verbs so linked are almost always performed by the same agent or theme. They are often equal partners, and typically synonyms, which is why the coordination particle is most often translated in English with “and”. Verbs linked by the coordination particle can also be temporally subordinate, where one action comes after the other. Alternatively, one verb can qualify the other, in which case one often uses a gerundive or adverbial translation, e.g. “’go-ing, he ’che(s)-s” (Hahn Reference Hahn1985).
In passage 3, the two verbs are linked by the semi-final particle, which, by contrast with the coordination particle, less often subordinates one verb to another, temporally or otherwise. The semi-final particle also allows for a change of agent or theme in a way that the coordination particle typically does not. Passage 3 reads lha dpal ’che ste ’go. Here the order is reversed, with ’che in the subordinate position.
As for the meaning of ’che(s), it does not correspond perfectly to the verb ’che as given in the dictionaries, whose meanings are “to attest, to promise; to affirm” (Hill Reference Hill2010a: 92). This latter's morphology is given as: present ’che; perfect ’ches/bces/ches; future ’che/bce; and imperative ches (Hill Reference Hill2010a: 92). Our verb ’che(s) is rather closer, both semantically and syntactically, to the verb gces, which follows an absolutive–oblique syntax and means “to have attachment, to love”; “to feel sympathy or pity for”; and “to value, cherish, prize, respect” (Hill Reference Hill2010a: 80). Without venturing any further hypothesis about the morphology of ’che(s), I will tentatively propose that it also means something like “to feel sympathy or pity for”. These sentiments, we should note, are particularly fitting for the relationship described between gods and humans in our texts.
If we accept that the verb ’che(s) means something like “to feel sympathy or pity for”, then this further narrows the semantic field for the verb ’go, which is almost certainly a synonym or near synonym. Context can helpfully rule out many of the attested meanings of the verb ’go within the Tibetan grammatical tradition and within available dictionaries and lexicographical resources. Among these, “to stain”, “to sully”, “to stick”, and “to infect” have no obvious relevance to the context and possible meaning of OT ’go in the above passages. We can also rule out the nouns “head” and “chief” as relevant translations for what is demonstrably a verb. We are left to consider similar verbs to which ’go might be related, as a variant or otherwise. Here one might think of the “word group” posited by Géza Uray that includes the verbs sko “to settle, establish”, ’god “to design, to establish”, and ’khod “to dwell, to be established” (Uray Reference Uray1972: 19, n. 3). This can be quickly ruled out on syntactic and morphological grounds: they are largely ergative – absolutive verbs, and one cannot easily account for the loss of the d suffix or the s prefix in trying to link these verbs to ’go. The same is true of the verb bgo, “to divide, allot”, which we find a god performing in a passage from the divination text ITJ 739: “good sman lha Footnote 32 allotted it such that in the end it will turn out happily and well” (sman lha ni bzang mo zhig gis bgo bas mthar skyid par cing bzang bar ’ong ngo; ITJ 739, 10r5–7). More appropriate syntactically is go, “to hear, to understand”, a single stem non-controllable verb with an absolutive–oblique syntax. Morphologically, however, ’go has at least two stem forms, so it is obviously not a “variant” of go. Semantically, it also has more strongly positive connotations than are customarily carried by the verb go.
We are left with the supposition that OT ’go in these passages is related to the verb ’kho, “to wish, to want, to think useful, serviceable, necessary” (Hill Reference Hill2010a: 24–5). As a voiced counterpart to ’kho, no doubt derived from the same root, ’go would mean something like “to be favourable [towards sm.]”.Footnote 33 As a non-controllable verb of affect, “to be favourable” agrees well with the syntax, fits the context in each of the above passages, and forms a coherent phrase when paired with ’che(s), i.e. “to feel sympathy for and to be favourable [towards]”.Footnote 34
Conclusion
In the foregoing I have proposed that the twenty passages above, drawn mostly from OT ritual texts, attest to the existence of an OT verb ’go meaning “to be favourable [towards]”. A larger question, and one that touches on historical processes of religious, cultural, and linguistic change, concerns the relevance of this verb to the group of deities known as the ’go ba'i lha lnga.
There are in fact some features in the above passages concerning gods showing or withholding favour that are redolent of descriptions of the ’go ba'i lha lnga. Passage 5, for example, uses the simile of a shade tree to describe the roles of the gods in showing favour. This recalls what seems to have become a standard simile applied to the ’go ba'i lha lnga, mentioned by Norbu and Berounský and enshrined in the Rgya bod tshig mdzod chen mo. In passages 4 and 8 the gods who show favour are linked to one's paternal or maternal ancestors (pha myes; ma myes), respectively, a point that recalls two members of the ’go ba'i lha lnga, the pha lha/pho lha and the ma lha/mo lha. Indeed passage 6 even refers to a ma lha, albeit a ma lha chen mo, which need not be the same as the ma lha/mo lha in the context of the ’go ba'i lha lnga. Looking further afield, one finds other members of the pentad, e.g. yul lha and dgra lha, in other divination prognoses and in other OT texts. These similarities, while intriguing, are far from sufficient to confirm the existence of the group of ’go ba'i lha lnga at the time of our texts; more conservatively, one can conclude that some of the ingredients for the tradition were already present, and leave it at that.
Though we do not find the phrase in the searchable OT lexicon, we can state that if it did occur, a * ’go ba'i lha in this early Tibetan context would be a “god who shows favour”, and whom one supplicates with offerings. This relationship of exchange between humans and gods is easily the clearest point that comes out of the passages analysed here. It is also one that was very much under threat from the introduction of Buddhist cosmological ideas that would contribute to the disappearance of dice divination traditions that involved omens issuing from the mouths of non-Buddhist deities. Indeed we can detect in passage 16 a Buddhist attempt to moralize this human–divine exchange by claiming that the gods will not show favour to the wicked (sdig to myi la lha myi ’go). The model of reciprocal exchange most certainly lived on, but it was increasingly directed towards Buddhist deities, and the idea of the gods' favour was pushed somewhat – though by no means entirely – into the background. A sense of exchange, for instance, and of a personal connection with a deity, inform the relationship between tantric adept and tutelary deity (yi dam). Here an initiate finds his or her yi dam through ritual and even divinatory means (e.g. tossing a flower into the mandala) that reveal a pre-existing dispositional and karmic connection. It is no stretch to observe a parallel with the gods of one's paternal or maternal ancestors in passages 4 and 8 who, though seemingly in abeyance, can be invoked through the reactivation of the human–god relationship, such that they will once again show their favour. The difference is that it is an ancestral, lineal connection on the one hand, and a karmic connection on the other. Of course there are more fundamental differences in the nature of the relationship: most crucially, there is no sense whatsoever that one can identify with, unite with, or in any sense seek to become a god who shows favour.
Despite these similarities, and the sense that the model of human–divine interaction in the passages translated above could be construed in a very loose way as a forerunner to the relationship between a tantric devotee and a yi dam deity, it is clear that this earlier model declined in stature with regard to Tibetan rituals and cosmologies while the latter, tantric model grew in importance. It may not be possible to pinpoint the date at which this transition took place, or to date the emergence of the ’go ba'i lha as a group of “personal protective deities”, but this has not been an obstacle to scholarly speculation. On an abstract level, Tucci posits that the ’go ba'i lha had their origin as “protective souls, before they changed into protective gods on whom depend the bodily integrity of the individual” (Tucci Reference Tucci and Samuel1980: 193; see also Samuel Reference Samuel1993: 438–9). Looking for a more specific temporal origin of the ’go ba'i lha, Todd Gibson extends Ariane Macdonald's articulation of the problem by proposing that a singular, royal deity, the sku bla, gave way to a plural, non-royal group of deities, the ’go ba'i lha lnga. Gibson speculates that the group of ’go ba'i lha may have originated in the eleventh century following the vacuum left by the collapse of the monarchy (Gibson Reference Gibson1991: 260). The musings of Tucci and Samuel, along with the matter of the relation of the ’go ba'i lha to “external souls” as expressed in the Tale of Dri gum btsan po, the Gesar Epic, and so forth, hinges on how scholars imagine pre-Buddhist beliefs and rituals in Tibet. This is a thorny question that I will not entertain here. As to the hypotheses of Macdonald and Gibson, these rest on problematic definitions of sku bla, and we can also note that our material does not directly involve the gods of kings, but rather those of divination clients (who are sometimes, but by no means always kings). A putative transition from “gods who show favour” (lha … myi la ’go) to “personal protective deities” (’go ba'i lha) has the advantage of using more or less the same terminology and referring to an equally wide sector of society. While this does not falsify Macdonald's and Gibson's speculations, it does render them overly complicated by comparison.
Another point that should be made here is that the old meaning of the verb was not totally lost. We saw it persist, for example, in a few lines in the dice divination text studied by Rolf Stein. In most such cases, people (that is, scholars, and probably also diviners and clients) have reinterpreted or misunderstood the word. This is not to discount the possibility that its OT meaning persisted in small geographic areas or in specific genres or ritual traditions such as divination. By and large, however, the old meaning was reanalysed, and the historical developments briefly sketched above offer a plausible scenario for why traditional understandings of ’go ba'i lha have been impressionistic (they are “like a shadow”), glossological (they are “head gods”), or reliant on contextual glosses (they are mgon pa'i lha). If we assume that the tradition of the ’go ba'i lha ultimately originated in a context like that of the above passages, as a “god who shows favour” in return for offerings, we can speculate that this meaning was forgotten when this model of human–divine interaction was deposed by competing Buddhist models, which also overlaid the older practices with their own revised language. This sort of reinterpretation of words whose archaic meanings have been lost is very common, and is highly productive of folk etymologies and of accompanying narratives and myths. It is also a treasure trove for the sort of wordplay so valued by Tibetans, who trade on pun and polysemy to enliven their poetics and to poke fun at one another and at their traditions. In these ways, it appears that the “gods who show favour” gave way to five or more “shadowy”, “sticky”, or “protecting” “head gods”.
