Ritual-first theory: Rappaport, Whitehouse's modes, and the Göbekli Tepe inversion
"Ritual-first" is the family of theories holding that practice precedes belief — that organised collective action (rite, taboo, synchronised performance) is the engine of religion, and doctrine is a later rationalisation of what people already do. It is the deliberate inversion of the older intellectualist picture (Tylor, Frazer) in which religion begins as a belief (in spirits, in an afterlife) that then generates rites. The question matters acutely for origins research because the earliest evidence the vault can touch is overwhelmingly behavioural — burials, deposits, monuments — not propositional; if ritual-first is right, that evidence is religion's actual leading edge, not a pale shadow of lost beliefs. Class 5 (cognitive/theoretical), anchored to one hard archaeological case.
Rappaport (1999): ritual is the "basic social act"; by performing a liturgical order the participant publicly accepts it, and this acceptance — not private belief — is what ritual produces and what makes the sacred. Meaning is enacted into being; the canonical, invariant form of liturgy is logically prior to any doctrine it later acquires. Religion is grounded in performance, not in assent to propositions.
Whitehouse's modes theory: two attractor patterns for how religion is transmitted and remembered — the imagistic mode (rare, high-arousal, traumatic rituals binding small cohesive groups through episodic memory) and the doctrinal mode (frequent, low-arousal, repetitive rituals encoding teachings in semantic memory and enabling large anonymous communities). In both, the ritual frequency-and-arousal structure drives the social form and the belief content, not the reverse.
The Göbekli Tepe inversion (Schmidt 2010): monumental T-pillar enclosures built by pre-pottery, pre-agricultural foragers c. 9600 BCE, with no domestic architecture on site — apparently a place people gathered to do something (feast, congregate, perform) before they farmed or settled. Schmidt's reading — "first the temple, then the city" — is the headline empirical case that organised cult activity can precede, and perhaps drive, the Neolithic transition, rather than following from settled belief.
Three cautions keep this medium-confidence as a general theory of origins. (1) The belief/practice dichotomy is partly a false binary — most scholars now hold the two co-evolve, and "ritual-first" is best read as evidential priority (behaviour fossilises; belief does not) rather than a claim that early humans performed rites while believing nothing. (2) Göbekli Tepe's "temple, then city" narrative has been softened by later excavation: domestic and water-management features and possible residential use have since been reported, blurring Schmidt's pure-sanctuary picture — so the flagship case is itself contested (cross-link schmidt-2010-gobekli-tepe-worlds-oldest-temple). (3) Whitehouse's large-scale empirical extension (the 2019 Nature paper on moralising gods) was retracted in 2021 after a coding-and-imputation reanalysis (Q8), a caution against over-reading the quantitative versions. The durable core: because the deep past speaks to us almost entirely in actions, ritual-first is the methodologically safer default for prehistoric religion — read the practice, infer the belief only with hedges.
ritual-precedes-belief-as-origins-default, paleolithic-mortuary-religion