Khajuraho's erotic carvings are the most-photographed sex in stone on earth, and the standard readings are "fertility symbolism," "celebration of the senses," or "harem fantasy." White's chapter offers a discriminating test: some of this sculpture depicts actual Tantric rituals, and some is exactly the worldly eroticism it looks like — and you can tell them apart. "No South Asian temple built in the medieval period is without erotic sculpture on its walls, and many of the sexual practices so depicted — because they are condemned in the Indian treatises on erotics (kāmaśāstra) — are likely depictions of Tantric rituals."1 That is the key: the kāmaśāstra (the secular sex-manuals) condemn certain poses. A pose that the worldly sex-experts themselves rule out is unlikely to be a depiction of ordinary or fantasy sex — it is more likely a ritual act, performed for reasons outside the pleasure-economy the kāmaśāstra governs. The "condemned poses" are the tell. Where the carving shows what even the sex-manuals forbid, you are probably looking at a rite.
White assembles the markers that distinguish ritual-sculpture from erotic-sculpture. First, condemned poses: practices the kāmaśāstra rejects — "including rajapāna" (the drinking of female discharge) — are markers of Tantra, not of worldly sex.2 Second, sequence: in Orissa, where maithuna (coupling) motifs appear from the late-sixth-century Śailodbhava period, "explicit depictions of sexual activities multiply significantly near the beginning of the tenth century and are frequently arranged in a sequential manner to imply specific stages of prescribed rites."3 Ordered sequences = rites (see The Aṣṭa-Kāmakalā Rite on Stone); they appear on Śaiva, Vaiṣṇava, and Śākta temples alike. Third, participants: many couples are "clearly aristocrats or members of the royal household, even when… stripped down to their royal headdresses" — and crucially, the presence of ascetics and a preceptor changes everything.4 Fourth, the attendant test: an erotic scene on the Viśvanātha temple at Khajuraho shows orgiastic groups where "the presence of a male attendant indicates that the scene does not represent the harem, where only female and eunuch attendants would have been admitted," and "a preceptor is part of the scene, as if chanting mantras and conducting an orgiastic ceremony."5 A male attendant rules out the harem; a mantra-chanting preceptor rules in the rite. The scene "seems to represent a religious ritual in which kings and Tantric specialists participated."
The page's intellectual integrity is in what White refuses to claim. He does not say all erotic temple sculpture is Tantric ritual. He grants that "many of these erotic carvings may be of harem scenes or representative of flights of fancy in which kings took their kingdom to be 'Kāmadeva's Pleasure Garden.'"6 The claim is discriminating: it is specifically when "condemned poses (including rajapāna) and orgy scenes depicting ascetics and royal figures" appear that "the likelihood is great that these were representations of Tantric rituals."7 So the logic is a set of filters: (1) is the pose kāmaśāstra-condemned (e.g., rajapāna)? (2) is the scene sequential (implying ritual stages)? (3) does it include ascetics and a mantra-chanting preceptor? (4) does the attendant configuration rule out the harem (a male attendant present)? The more filters a scene passes, the more likely it is a rite. On the temples of Modhera, Roda, and Galteśvara, "ascetics are shown making particular hand gestures, as if reciting mantras" within orgiastic scenes — the gesture-marker.8 The discrimination is what makes the claim credible: White is not reading sex into all temple art (the overreach the field rightly fears), but identifying specific markers (condemnation, sequence, ascetics, mantra-gesture, non-harem attendants) that flag ritual.
And then the end of the phenomenon, dated. The depiction of actual Kaula/Tantric rituals on temples "appears to have been a short-lived phenomenon, of no more than two centuries' duration." With the twelfth century, maithunas "begin to be treated as independent motifs… thereby losing their sequential, and ritual, significance."9 The cause White (via Donaldson) names: "the introduction in this period of devadāsīs as temple prostitutes likely contributed to this new glorification of sexual pleasure as an end in itself, as opposed to a ritual means to an end."10 The sexualized Kaula ritual "gave place to hedonistic practices… and most of the kings and feudal chiefs who courted Tantrism were hardly sādhakas."11 So the sculpture's meaning shifted: from ritual sequence (a means to an end) to independent erotic motif (pleasure as the end) — the sublimation visible as an art-historical change, datable to the twelfth century and tied to the devadāsī economy.
This page gives the vault a discrimination method for erotic temple art — the markers (condemned poses, sequence, ascetics, mantra-gesture, non-harem attendants) that separate ritual-depiction from worldly eroticism. It anchors the kāmaśāstra-condemnation test (the sex-manuals' own rejections flag Tantra). It dates the transition (12th-century shift from ritual-sequence to independent erotic motif, tied to the devadāsī economy) — the same sublimation timeline the aṣṭa-kāmakalā page tracks (rajapāna → fellatio → independent motif). And it grounds the royal-participation claim (aristocrats, kings, and Tantric specialists in the scenes).
The sharpest case is the Viśvanātha temple's south-joining-wall scene, because it shows the discrimination method working on a single carving.12 The scene depicts orgiastic groups. The naive reading is "harem fantasy" — a king's pleasure-garden daydream. White applies the filters. Attendant test: there is a male attendant helping the central aristocratic couple — but the harem admitted "only female and eunuch attendants," so a male attendant rules out the harem. Preceptor test: a preceptor is present, "as if chanting mantras and conducting an orgiastic ceremony" — a religious officiant, not a courtier. Participant test: the central couple are aristocratic; attendants are "themselves part of the erotic action." Run the filters and the daydream-reading fails: a male attendant + a mantra-chanting preceptor + participating attendants = "a religious ritual in which kings and Tantric specialists participated," not a harem scene. The case study's force is that it shows the method is falsifiable and discriminating: the same scene that looks like erotic fantasy yields, under the filters, a specific identification (Tantric ceremony) on specific grounds (the male attendant, the preceptor) — and those grounds could have come out the other way (had the attendants been only female/eunuch, the harem reading would stand). White is not reading ritual into all sex; he is reading it out of the evidence, filter by filter. The male attendant is a small detail that decides the whole meaning of the wall.
You are looking at erotic temple sculpture and want to tell ritual from worldly eroticism. The discipline: run the filters; the more it passes, the more likely it's a rite. First, the condemnation filter: is the pose condemned in the kāmaśāstra (e.g., rajapāna)? A pose even the sex-manuals reject is unlikely to be ordinary/fantasy sex — suspect ritual. Second, the sequence filter: are the scenes ordered (implying ritual stages) rather than independent motifs? Sequence = rite. Third, the ascetic filter: are ascetics or a preceptor present, especially making mantra-reciting hand-gestures? A religious officiant marks a ceremony. Fourth, the attendant filter: is there a male attendant (ruling out the harem, which admitted only female/eunuch attendants)? Fifth, the participant filter: are royal/aristocratic figures and ascetics together in an orgy scene? Sixth, the date filter: pre-twelfth-century sequential scenes are more likely ritual; post-twelfth-century independent erotic motifs are more likely hedonic (pleasure as the end). The workflow is to treat each filter as a vote: a condemned, sequential, ascetic-and-preceptor-bearing, non-harem, pre-12th-century scene is very likely a rite; a kāmaśāstra-approved, isolated, purely worldly scene is probably the daydream it looks like.
The failure comes in two opposite forms, and avoiding both is the discipline. Over-reading: declaring all erotic temple sculpture to be Tantric ritual — the overreach the field rightly fears, which White explicitly refuses (much of it is harem scenes and flights of fancy). You've done this when you have no filter that could ever come out "just eroticism." Under-reading: dismissing all erotic sculpture as decoration/fertility-symbolism/sensuality, missing that condemned, sequential, ascetic-bearing, non-harem scenes are specifically ritual-depictions. You've done this when a mantra-chanting preceptor in an orgy scene reads to you as "decorative." A third failure: missing the datable transition — reading a twelfth-century independent erotic motif as if it carried the ritual significance of a tenth-century sequence, and so missing that the meaning changed (ritual means → hedonic end). If your reading is all ritual or all decoration, and ignores the date, you've failed the discrimination the markers make possible.
The evidence is the kāmaśāstra-condemnation criterion; the Orissan sequential-maithuna development (late-6th-c Śailodbhava → 10th-c proliferation in ritual sequences); the cross-sectarian spread (Śaiva/Vaiṣṇava/Śākta); the Viśvanātha male-attendant/preceptor scene; the Modhera/Roda/Galteśvara mantra-gesture ascetics; and the datable 12th-century shift to independent motifs tied to the devadāsī economy (Donaldson).13 The tension is the irreducible interpretive judgment: the filters reduce but do not eliminate the ambiguity — a "condemned pose" might still be a flight of fancy about a transgressive act; a "preceptor" might be a misidentified figure; the line between "sequential rite" and "narrative decoration" requires a call. White's honesty (granting that "many" carvings are harem/fantasy) is the strength, but it means the method yields likelihoods, not certainties. The open question is the base rate: of all the erotic sculpture on medieval temples, what fraction passes enough filters to be probable ritual-depiction — a few flagship scenes, or a large proportion? White's examples are selected (the clearest cases), so the prevalence of ritual-depiction (vs. worldly eroticism) across the whole corpus is undetermined — and it matters for whether "erotic temple sculpture" was primarily ritual or primarily worldly, with ritual the exception.
White builds on Thomas Donaldson's and Devangana Desai's art-historical work, and the convergence-and-division is the same one that runs through the aṣṭa-kāmakalā page. Desai supplies the "architectural punning" reading (love-scene as coded ritual surface) and the careful temple-dating; Donaldson supplies the transition account (the 12th-century shift to hedonic, devadāsī-driven eroticism). White synthesizes them into a discrimination method + a sublimation timeline. The convergence: all three read some erotic sculpture as ritual and recognize a change over time from ritual-sequence to independent motif. The tension is causal weight (as on the aṣṭa-kāmakalā page): Donaldson's social-economic cause (the devadāsī system commercializing temple eroticism) vs. White's theological frame (the sublimation of the Kaula hard core). What the synthesis reveals is that the art-historical record is a sublimation archive: the carvings date the transition from operative ritual (sequential, condemned-pose, ascetic-bearing scenes, pre-12th c.) to hedonic ornament (independent erotic motifs, post-12th c.), and the devadāsī economy and the theological sublimation are two pressures legible in the same stones. The temple wall is a timestamped record of the Kaula hard core being aestheticized — readable, scene by scene, as the rite turns into decoration.
Plainly: the same image (two bodies coupling on a temple wall) could be a ritual instruction or a worldly daydream, and telling them apart requires reading for specific markers — which is a problem of interpretation under ambiguity that reaches well beyond temple art.
The Sharpest Implication If a pose condemned by the sex-manuals is, for that reason, likely a ritual, then the most "forbidden" erotic carvings on the temples are the least erotic in intent — they depict not heightened pleasure but ritual acts (like rajapāna) performed outside the pleasure-economy entirely. Take that seriously and the modern gaze at Khajuraho has it backwards: the scenes that most shock the viewer (the condemned poses, the orgies with ascetics) are precisely the ones least about sex-as-pleasure and most about ritual-as-means, while the "tamer," kāmaśāstra-approved couplings may be the actual eroticism. The discomfort is that the carvings we find most transgressively sexual were, to their makers, the most religious — and our titillation is a category error, mistaking the temple's ritual hard core for its erotic content.
Generative Questions