BAGA, an Old Iranian term for “god,” sometimes designating a specific god.
Baga- is attested in early and late Iranian with two meanings (1) agent noun “distributor,” glossed by Parsi Sanskrit vibhaktar- and (2) noun of action and result “portion.” Beside the noun, the verb bag- “give or receive portions” is in frequent use: bag-, present baj- and baxš-, participle baxta-, and nominal derivative bāga-. The cognate words are cited in full in Dictionary of Khotan Saka, p. 300: Avestan baž-, baxta-, bāga-; Sogdian βaxš-, βxt-, βγn-; Zoroastrian Pahlavi baxš-, baxt; Khotan Saka būṣṣ-, būta-; Old Indian (Sanskrit) bhájatiṃbhaktá-, bhága-, bhāgá, traced in Indo-European in J. Pokorny, Indogermanisches etymologisches Wörterbuch, Bern and Munich, 1959, p. 107 s.v. bhag-.
The agent noun baga- “distributor” was adopted as a name for “god” and “gods.” In Old Persian baga- occurs beside maθišta bagānām “greatest of gods” for Ahuramazdā. It renders the Akkadian Semitic ilu “god.”
In the Avesta the word yazata- “worshipful” was preferred, the later Zor. Pahl. yazd and Ossetic izäd, in Greek script isdi- in the name Isdigerdēs, Yazdkirt, Armenian Yazkert. The other word of divinity, dai- “shine,” is retained once in the Avestan word dyaoš “from the sky” (genitive from dyau-) and in daiva- of the Avestan phrase daēvāiš-ca mašyāiš-ca “with gods and men.” Normally this word for the ancient gods was developed to “devils” in the Zoroastrian tradition. The North Iranian Khotan Saka also had developed dyūva- “devil” from daiva. A trace of the older “god” can be seen in the name Dēv-dād “created by god.” Xerxes in a Persepolis inscription reported daivadānam viyakanam “I destroyed the house of devils” (XPh 37).
In the later Zoroastrian books Pahlavi and Pazand baga- became baγ (bkˈ, bγˈ, Pāzand baγ) and occurs associated with yazata- in the phrase yazdān baγān “gods” (bkʾn, bγʾn). A lost nask “book” of the Avesta (a word preserved in its earlier non-technical sense in Caucasian Georgian naskʾ-v-i “knot,” verbal naskʾ-va) was called Baγ (bkˈ) and Baγān. The Avestan baγō.baxta- can be rendered either “bestowed as a portion” or “distributed by the gods,” the second meaning can be supported by the compound (Dēnkard, ed. Madan, p. 603.6) baγ-tāšīt “created by gods.” The Aramaic ALḤA “god” was interpreted by baγ; it is used in the Ḥājjīābād inscription and in the inscription of Šāpūr I, Parthian line 1 mzdyzn ALḤA, *mazdēẓn baγ, Greek masdaasnēs theos. The king Ḵosrow I is called im baγ “his present Majesty” (Dēnkard, p. 413.9). Šāpūr son of Ohrmazd is quoted in Škand-gumānīk Vičār (ed. P. J. de Menasce, Fribourg, 1945, pp. 118-19, 10.70) ōi baγ šāhpūr ī šāhą šāh i hōrməzdq “the god Šāpūr king of kings son of Ohrmazd.” The baγ yazdkart is cited in Dēnkard, p. 949, 21. The word baγa- gave Sogdian βγ (in Arabic script this β- was expressed by the letter f with three dots on it, which in normal Arabic could be replaced by f). The Buddhist Sogdian employed βγ- of the Buddha P 2.1137-45 βγʾn βγtm pwty “the most godlike of gods Buddha.” The Khotan Saka had preferred gyastānu gyastä “god of gods” from yazata-. Sogdian had the plural βγʾyšt, *βaγišt, gen. plural βγʾn, *βaγān, and βγʾstʾn, *βaγastān for “paradise.”
The Kušan bagapouro (from baga-puθra-) is represented in Turfan Parthian bγpwhr and New Persian faγfūr, and found in Armenian čen-bakour, the Chinese imperial title ṭʿien-tsï “son of heaven,” to which the Old Indian word was deva-putra-. Manichean and Christian texts use this same word for “son of god.”
Turfan Parthian had bg, *baγ with plural baγān and adjective baγānīγ. Turfan Persian had developed -y- from -γ- in by, *bay, plural baʾān (bʾʾn) and adjective bayānīγ. This bai is found in New Persian bē-doxt, bē-loft from *baga-duxtā “daughter of god” for the planet Venus. About A.D. 800 Iranian baγ is present in Kirkut (Qiṛγiz) bai, written by the Chinese sign buâi (Mathews, Chinese-English Dictionary, no. 4991, modern pei) cited in JA, 1950, p. 298. The Kirkut adopted the Turkish language only in the 7th-8th centuries.
Khotan Saka has beʾga-, baʾga-, and be- in the title beʾgarakä, baʾgarakä, and berakä for Turkish bägräk. From Turkish, in proper names Ossetic has bī “prince,” as in Baras-bī “tiger, prince.”
Turkish took the title bäg, beg, and later bey; in early Turkish in Brāhmī script it was bhek, and in Byzantine Greek the Khazar title was written mékh. The form bägräk occurs in a Turfan Manichean source. In Arabic script the Turkish title was bag, baḵ, and bāk.
The sacred fire called farn-baγ “the distributor (god) of farn” (farn was equated with Aramaic GDHǰ “fortune”) was located at Kārīān in Fārs. The Pāzand read the name as farō-bag and in Armenian it probably gave hour-bak.
In Kushan inscriptions this baga- is vaka- and vvaga-in the name Vaka-mihira and Vvaga-mihira “the god Miθra.” The v- and vv- are an attempt to express the fricative β. The Kushan bakana-pati- “temple official” is βγn-pt in Sogdian. Armenian has bagin (q.v.) “shrine, altar” from Parthian.
Choresmian from Toprak-kala has βγy for the sixteenth day of the month corresponding to mtr “Miθra,” as Bīrūnī gave fyγ, *βaγ, from βaγi for the same day.
The word baga- occurs in many proper names in ancient Persia. In Greek these names are written with Baga- and Mega-. The name bg-srw, *baga-srava- occurs in an Aramaic papyrus from Egypt.
Two toponyms contain the word baga- in Persia, others are in Armenia. The rock of Behistūn, Bīsotūn is on tò Bagístanon óros. The Bundahišn cites Baγ-dāt ī baγān-dāt “Baghdad which is created by the god(s)” (TD2, p. 205.12).
In the sense of “portion” Gathic Avestan has baga-, beside later baγa- and baγā-, which is glossed by Zor. Pahl. baγ, bažišn, and baxtārīh.
The derivative baxta- “distributed, allotted” also is frequent later. The Avestan phrase in Vištāsp yašt 38, baxta-ca nivaxta-ca, is explained to mean pat nēvakīh ut frazandān “possessed of good things and children” where “possessed” is Pahlavi baxtadār (baxtiyār). The baxt “fortune” takes the form of a varrak ī vazurg “great ram” mounted behind Artaxšaθr (Ardašīr) in his flight from Artapān (Ardavān).
The noun derivative bāga- is also important in connection with baga-. A later form is attested in Khotan Saka haṃbāya- (sing. haṃbā, plural haṃbāya) “portion” corresponding to Turfan Parthian ʾmbʾg, *ambāγ, Turfan Persian hmbʾw, *hambāv, “rival,” Zor. Pahl. hambāγ, Pāzand hambāi, hambāe “companion.” Through connection with “possessions” the Iranian bāy gave Turkish bai “rich,” whence Mongol bayan “rich.” In Ossetic the Turkish bai- occurs in the personal family name Bai-tuγantä “rich” and “bird of prey.” As a “portion of land” New Persian has bāγ “garden” and Pāzand has bāγastą i vahəšt “garden of Paradise,” that is Garō’amān, and vahišt is in Zor. Pahlavi bō’astān “garden as place of perfumes.” From Alanian (that is older Ossetic) Caucasian Čečen has (with many other words of Alanian origin) the word bai “meadow, lawn,” plural beš, with Inguš bai, plural bäš (the plural suffix is -aš and -s-).
This bāy “rich” is also found in the Finno-Ugrian Ostyak way.
(H. W. Bailey)
The old IE. term for the heavenly gods, the cognate of OInd. devá-, Lat. deus, etc., was generally devalued in Iranian to “false or evil god” and ultimately “demon” (Av. daēva-, Old Pers. daiva-, Mid. Pers. dēw, etc.). In the sense “god,” this word was replaced either by yazata- “(a being) worthy of worship” (Av. yazata-, Mid. Pers. yaz(a)d, NPers. īzad, etc.) or by baga-.
It is not likely that OIr. baga- derives from a word meaning “god” in IE., despite the occurrence of bogŭ “god” in the Slavonic languages (probably as a result of Iranian influence). Etymologically, baga- “god” belongs to the verbal root bag “to distribute, allot” and may be equated with OInd. bhága-, a divine epithet probably meaning “dispenser; generous one,” beside which there exists also a second noun bhága- “distribution, allotted portion” equivalent to Av. baγa- “portion, lot,” Slavonic *bogŭ in derivatives such as ubogŭ “poor” (from “*portionless”). It is this latter word which is personified in the Rigveda as the minor deity Bhaga (one of the Ādityas, a group headed by Varuṇa and Mitra). (See P. Thieme in B. Schlerath, ed., Zarathustra, Darmstadt, 1970, p. 401.)
Av. baga- occurs only once in the Gāθās, in a much-debated passage (Y. 32.8) where it possibly means “portion” or “distribution.” In the later Avestan texts baγa- occurs occasionally as a title of divinities (Ahura Mazdā, Māh, Mithra), but it is debatable whether it should still be understood in the earlier sense of “dispenser” (cf. OInd. bhága-) or already as “god.” The development of the latter meaning is securely attested in Old Persian, where baga-, translated by words meaning “god” in Babylonian and Elamite, is the only generic term for the divinities worshipped by the Achaemenids.
There is no certain trace of baga- in the Mid. Ir. Saka dialects, where forms derived from yazata- (Khot. gyasta-, Tumshuqese jezda-) are used instead, while Islamic Choresmian has adopted Ar. ʾllh for “God” (but cf. Chor. (‘)βγ(y)k “doll” from *baga-ka-). All other attested Mid. Ir. languages retain forms derived from baga- in the sense “god”: Mid. Pers. bay (plur. bayān, baʾān), Parth. baγ, Bactr. bago, Sogd. βγ-. All these forms are used as honorific titles, not only of gods but also of kings and other men of high rank, in which case the translation “lord” is usually appropriate, e.g., Bactr. i bago šao Kanēški “the lord king Kanishka.” In Zoroastrian Pahl. im bay “this lord” and ōy bay “that lord” mean “His (present) Majesty” and “His late Majesty” respectively. A peculiar use of the plur. form to refer to the monarch is found in the Sasanian inscriptions (and perhaps in the Sogd. “Ancient letters,” see W. B. Henning, “Soghdisch βγʾʾn-,” ZDMG 90, 1936, pp. 197-99), e.g., Inscr. Parth. LKM ʾLḤYN (išmāh baγān) “Your Majesty.” A derivation of the Turk. lordly title bäg, later bey, from Iranian is superficially attractive but quite uncertain (see G. Doerfer, Türkische und mongolische Elemente im Neupersischen II, Wiesbaden, 1965, pp. 402-06).
It is probable that baga- “god” sometimes designates a specific deity as “the god” parexcellence. Various attempts have been made to identify “the god” so referred to, as Ahura Mazdā, as Mithra, or as an Iranian equivalent of the Vedic Varuṇa (see M. Boyce, “Varuna the Baga,” further identifying Varuṇa with the Av. Apąm Napāt [q.v.]), but no basis has ever been stated for the assumption that baga- “the god” in a personal name such as *Baga-dāta- “given by (the) god” must refer to the same divinity at all periods and in all parts of the Iranian world. However, strong evidence that Baga was sometimes used as a by-name of Mithra is provided by calendrical data first exploited by J. Marquart (Markwart), Untersuchungen zur Geschichte von Eran I, Göttingen, 1896, pp. 63-65; II, Leipzig, 1905, pp. 129, 132-34 (see also V. A. Livshits, “The Khwarezmian Calendar and the Eras of Ancient Chorasmia,” Acta Antiqua Academiae Scientiarum Hungaricae 16, 1968, pp. 444-46, for Chor. data). Thus, the seventh month of the year is named after Mithra in Mid. Pers. Mihr māh, Arm. Mehekan, early Chor. mtr, but after Baga in Old Pers. *Bāgayādiya (attested in El. transcription) and Bāgayādi(š) (q.v.), Sogd. βγkʾn(c) (cf. the similar formations used to name the ninth month: Mid. Pers. Ādur māh, Arm. Mehekan, early Chor. ʾtrw, Old Pers. Āçiyādiya (q.v.) and *Āçiyādi(š)). Likewise, the sixteenth day of the month is known as Mihr rōz in Mid. Pers., mtr in early Chor., myš in the Sogd. calendar given by Bīrūnī, etc., but as fyγ (for *βyγ) in Bīrūnī’s Chor. calendar, βγy-rwc in a Sogd. document from Mount Mug, bgy in a Chor. inscription.
A different interpretation of these facts was proposed by W. B. Henning, “A Sogdian God,” BSOAS 28, 1965, pp. 242-54. Chiefly on the basis of an etymology of Sogd. βγʾny-pš-ktʾkw “wedding” as “making of a Baga-union” and a reference to βγy, apparently as a god distinct from my’rʾ “Mithra,” in a Sogd. marriage contract, he identifies Baga with the Vedic Bhaga: both in India and in Iran, B(h)aga would be a god with a “special interest in marriage” and closely associated (but not identical) with Mit(h)ra. This theory has been supported with onomastic data by P. Gignoux, “Le dieu Baga en Iran,” Acta Antiqua Academiae Scientiarum Hungaricae 25, 1977 (1980), pp. 119-27, but it has also been strongly challenged, see especially A. Dietz, “Baga and Miθra in Sogdiana,” in Ētudes mithriaques, Acta Iranica 17, Tehran and Liège, 1978, pp. 111-14, and S. Zimmer, “Iran. baga—ein Gottesname?” MSS 43, 1984, pp. 187-215, and cannot be regarded as proven.
The article by S. Zimmer just cited includes a useful discussion of most problems concerning baga-, while the Old and Mid. Ir. usage is surveyed by M. Boyce, “Varuna the Baga,” in Monumentum Georg Morgenstierne, Acta Iranica 21, Leiden, 1981, pp. 59-73, esp. 61-65.
On the relationship between OIr. baga- and OInd. bhága-, Slavonic bogŭ see M. Mayrhofer, Kurzgefasstes etymologisches Wörterbuch des Altindischen II, Heidelberg, 1963, pp. 457-59.
Regarding H. W. Bailey’s suggested reading and interpretation of Khot. *vvūvayau as a royal title derived from *baga-yauna- (e.g., in The Culture of the Sakas in Ancient Iranian Khotan, Col. Lect. Ser. 1, Delmar, 1982, p. 11) see R. E. Emmerick in Studies in the Vocabulary of Khotanese II, ed. R. E. Emmerick and P.O. Skjærvø, Vienna, 1987, pp. 132-34.
Baga- as a name element occurs in names of towns, e.g., Bagadāta-, modern Baghdad “given by baga” or “established by baga,” *Bagadāna- “temple (?) of (the) god(s),” Old Armenian Bagaran, a town and a village (see Hübschmann), with derivative suffix -ka- Bactrian bagolaggo “altar” (or proper name “The Sanctuary”), today Baḡlān, name of the region (see Henning), Bagastāna- in Media “place of the gods,” so named after the nearby Bīsotūn carvings and inscriptions, Bag(e)is in Lydia; rivers, e.g., Bagrades, today Nābendrūd, Bagossala, probably modern Bug, Ukraine; and mountains, e.g., tò Bagístanon óros at Bagastāna, see above, Bagôon óros between Areia and Drangiana (see Pauly-Wissowa, II, 2, cols. 2765-74; S. I, col. 237). Most of the names, however, are personal names and exclusively men’s names. All three Indo-European name types are attested, namely, one-stem-names (“short names”), derivatives (mostly hypocoristica), and compound names. Short names are seldom (e.g., Baga inscription of Priene, 1st cent. A.D.; perhaps Vaga, “Indoscythian” stratega inscription, ca. 20 A.D.). The hypocoristica (e.g., Bagaios, Bagauka-) have been treated in many articles by R. Schmitt (see Schmitt, 1982 and “Indogermanische Chronik”). Of compound names, Bagabigna- and Bagahuxša- are attested in the Old Persian inscriptions (see IPNB), but more than 60 other Old Iranian names are found from the sixth or even eighth century B.C. in Aramaic, Armenian, Babylonian, Egyptian, Elamite, Greek, Hebrew, Indo-Arian, Latin, Lydian, and Lykian tradition, see Schmitt, 1982 with references, also to the studies on the phonetic shape of Old Iranian baga- in these languages respectively. The Middle Iranian names are only partially published, see Gignoux, 1980a, 1980b, and forthcoming parts of Iranisches Personennamenbuch. At present, four baga names are known from epigraphical Middle Persian: ʾtwrbg, bgʾtwr, bgʾwhrm[zd], plnbgy (this itself part of compounds), and one from Sogdian: βγyfrn. On the linguistic classification and interpretation of these compound names, see Schmitt, 1981, and Zimmer, 1984.
Contrary to the opinion of Gignoux, 1980a and b, these names do not provide any evidence for the existence of an Iranian deity “Baga.” See the detailed analysis of the arguments in Zimmer, 1984. Old Iranian baga-, from Indo-Iranian *bhaga-, means “allotment, distribution,” later also “god” [but see also baga i, above], and in Middle Iranian bag/bay “god, lord” as a mode of address of gods and noble men. Nowhere in Iranian is this word a name. Only in India did the concept of “allotment, distribution” become the Āditya Bhaga.
Mayrhofer’s suggestion (Etymological Sanskrit Dictionary II, p. 457) that baga- in Iranian names replaces the old *daiva- whose meaning had been changed to “evil deity” by Zarathustra, is quite attractive, but hard to prove.
The Old Persian month name Bāgayādi- (q.v.) is interpreted by R. Schmitt as “(month of) worship of the god,” against Henning’s suggestion “(month of) fertilizing (?) the fields” (comparing Middle and Modern Persian bāḡ “garden” [q.v.]). This interpretation is supported by the parallel formation Açiyādiya/*Açiyādiš (Elamite) “(month of) fire-worship” (cf. the Middle Persian month Ādur, Armenian Ahekan; see i, above). This could imply a special worship of Miθra (?) in one month, and of Ātar in the other, perhaps because of some major feasts celebrated in these months.
Ph. Gignoux, “Le dieux Baga en Iran,” Acta Antiqua Academiae Scientiarum Hungaricae 25, 1977 , pp. 119-27 (1980a).
Idem, “Les noms propres en moyen-perse épigraphique,” Pad nām ī yazdān, Université de la Sorbonne Nouvelle, Travaux de l’Institut d’Ētudes Iraniennes 9, 1979 , pp. 35-100 (1980b).
W. B. Henning, “Surkh Kotal,” BSOAS 18, 1956, pp. 366-67, Acta Iranica 15, 1977, pp. 503-04.
W. Hinz, Altiranisches Sprachgut der Nebenüberlieferungen, Wiesbaden, 1975.
H. Hübschmann, “Die Altarmenischen Ortsnamen,” Indogermanische Forschungen 16, 1904, pp. 410-11.
M. Mayrhofer, A Concise Etymological Sanskrit Dictionary (Kurzgefasstes Etymologisches Wörterbuch des Altindischen), Heidelberg, 1965-80. Idem, Onomastica Persepolitana, Vienna, 1973, esp. 8.184-8.244.
M. Mayrhofer and R. Schmitt, eds., Iranisches Personennamenbuch, Vienna, 1977ff. (IPNB). R. Schmitt, “Altpersisch-Forschung in den Siebzigerjahren,” Kratylos 25, 1980 , pp. 1-66.
Idem, “Indogermanische Chronik, chap. V c 3,” Die Sprache, Vienna, twice yearly.
Idem, review of Gignoux, 1980b, Studia Iranica 10, 1981, pp. 154-59.
St. Zimmer, “Iran. baga- —ein Gottesname?” MSS 43, 1984, pp. 187-215 (with further references).
R. Zwanziger, Studien zur Nebenüberlieferung iranischer Personennamen in den griechischen Inschriften Kleinasiens, Ph.D. dissertation, Vienna, 1973.
(H. W. Bailey, N. Sims-Williams, St. Zimmer)
Originally Published: December 15, 1988
Last Updated: August 22, 2011
This article is available in print.
Vol. III, Fasc. 4, pp. 403-406