The Bharhut Stūpa was built in the 2nd century BCE in central India. Its stone railing — much of which is now preserved in the Indian Museum, Calcutta — provides the single most substantial body of identified yakṣa-deity sculptural material from early Buddhist context.1 The railing was originally circular, with large carved figures at intervals interspersed with narrative scenes from the Buddha's previous lives (Jātaka stories).
What makes Bharhut distinctively important: the large figures along the railing are identified — most carry inscriptions naming the specific yakṣa, nāga, or deva they represent. Other early Buddhist sites have yakṣa-style figures but rarely with inscriptional identifications. Bharhut lets us see what specific deities were considered appropriate for inclusion at a major Buddhist site.
The railing's preserved figures include:
The Four Great Kings (lokapālas): Vaiśravaṇa (Kubera), Dhṛtarāṣṭra, Virūḍhaka, Virūpākṣa — the four directional protectors who become major Buddhist iconographic figures. Two of the four are clearly preserved (Virūḍhaka of the south, Kubera of the north).
Sirima devatā: a goddess identified by inscription as Sirima (which is Sanskrit Śrī-mā = Lakṣmī-mother). The figure represents Lakṣmī, the goddess of prosperity, at this site. Multiple Sirima figures may have been originally present.
Multiple yakṣas and yakṣīs: individual figures identified by inscription as specific named yakṣas (e.g., Cūlakoka, Mahākoka, Suprabuddha) and yakṣīs.
Multiple nāgas and devas: serpent-deities and other deva-category figures.
Gajalakṣmī figures: the goddess on lotus being asperged by elephants, appearing in multiple places on the railing including as decorative motifs.
Mithuna couples: male-female pairs, appearing in railing decoration.
The figures are positioned along the railing with apparent intention — they face the central stūpa as if attending it, with respectful postures.
The Bharhut iconographic programme implements a specific religious-political claim: the Buddha rules over the higher (lokottara) realm; the yakṣa-deities (Buddhism's lesser figures) rule the laukika (this-worldly) realm; the yakṣas attend respectfully on the Buddha's superior authority.
The programme also implements a political claim by the stūpa's patron-builder. By portraying the major regional yakṣa-deities as respectfully attending the Buddha, the patron claims a corresponding status — as a ruler whose authority parallels the Buddha's universal authority and to whom the regional yakṣa-deities (and the regional populations they represent) accordingly attend.
This is what Samuel identifies as political-religious work: the iconographic programme is not just decorative; it operates as a claim about the relationship between the patron's authority, the Buddha's authority, and the various regional deity-cults' authority.
The Sinhalese religious chronicle Mahāvaṃsa (5th c. CE) describes the construction of the Great Stūpa in Anurādhapura. The construction included images of the Four Great Kings, the Thirty-Three Gods, Thirty-Two Celestial Maidens, and Twenty-Eight Chiefs of the Yakṣas — a substantial body of yakṣa-related iconography arranged around the central stūpa with the deities depicted in attending postures.
The Mahāvaṃsa pattern is structurally similar to the Bharhut pattern. The Sinhalese stūpa-construction tradition incorporates yakṣa-incorporation iconography on the same model. The pattern is not unique to Bharhut; it represents a general approach to Buddhist site-construction that traveled across the subcontinent and into Sri Lanka.
Connects to Yakṣa-Yakṣī Cult — the cult that Bharhut incorporates. Connects to Fertility-Auspiciousness Religion at Early Buddhist Sites — the broader iconographic programme. Connects to Laukika vs Lokottara — the operative two-realms architecture. Connects to Maṇḍala as Galactic Polity — the political-religious incorporation pattern foreshadows mandalic structure.
The Sirima devatā figure at Bharhut is one of the earliest clearly-identified images of what becomes Lakṣmī. The figure shows iconographic features that persist into later Lakṣmī iconography — lotus association, prosperity-coding, female-fertility figure type.
The case is instructive because it shows how a yakṣī-class deity (a local-protective female figure) gets identified with what becomes a major Hindu goddess. The transition from local-protective yakṣī to all-India goddess Lakṣmī is partly traced through these Bharhut-era identifications. Sirima at Bharhut is on her way to becoming Lakṣmī of later Hindu tradition.
When you encounter description of an early Buddhist site's iconography, specific questions follow:
What proportion of identified figures are yakṣa-class deities vs Buddhist-specific figures? At Bharhut the proportion is substantial.
How are the yakṣa figures positioned relative to the central Buddhist content? Attending-respectfully (laukika-subordinate-to-lokottara pattern) or independently?
Are figures inscribed with identifying names? When they are (Bharhut case), we can identify specific yakṣas. When they aren't (most other sites), the figures' identity remains uncertain.
The iconographic-analysis discipline lets you read sites as religious-political programmes rather than as decorative ensembles.
Barua (1934-37) is the foundational systematic study; Kala (1951) and Coomaraswamy (1956) developed it. The convergence on Bharhut's iconographic content is substantial.
The tension is about how much of the original railing has been preserved. Estimates suggest perhaps one-third of the original yakṣa figures are recovered. The full original programme is partly conjectural.
The pattern of major-religious-site incorporating subordinate-deity iconography has parallels across traditions.
The handshake to history identifies several parallel cases. Cathedral iconography incorporating local-saint and pre-Christian-deity figures; Roman temple iconographic programmes incorporating subordinated foreign-deity figures; Buddhist temple iconography across East Asia incorporating local-deity figures. The vault's history corpus on religious-architectural-iconographic material can be analyzed under this frame.
The handshake to behavioral-mechanics sits in how the iconographic programme operates as positioning-strategy for the site's patron. The patron is claiming a specific authority-relationship vis-à-vis the broader regional religious-political situation. The vault's behavioral-mechanics corpus on positioning can analyze. The insight neither domain alone produces: religious-architectural patronage is positioning-strategy as much as devotional act; the patron-claim implicit in the iconographic programme is part of what the site does.
Treating the Bharhut iconography as merely-religious (without political-positioning content) misses what the patron was claiming through the iconographic programme.
Treating it as merely-political (without genuine religious content) reduces what was operationally religious work to instrumental-political maneuver.
Strongest evidence: the surviving railing material with inscriptional identifications.
Open questions: What was the original total programme? Who specifically was the patron? What political-religious claims was the patron making through the specific iconographic choices?
Sharpest implication: When the vault discusses Bharhut or other early Buddhist sites, the iconographic programme should be read as religious-political work, not as decorative ensemble.