Why Buddhist Traditions Preserved Themselves, Not “The Buddha’s Words”
Five theses from the Pāli Canon’s own narrative
1. Oral transmission was never about verbatim preservation
The Canon shows scattered sangha teaching in local languages during the Buddha’s life—no memorised suttas, no rigid formulas. DN 16 has him say “be islands unto yourselves” with Dhamma-Vinaya as teacher, not script-recitation as mandate. Yet the same Canon inserts a First Council where top arahants abandon that instruction within days.
Oral transmission specialists competent enough to preserve bulk material would spot wholesale additions—they didn’t miss them, they welcomed them.
2. The First Council reveals itself as institutional change
The Canon’s version breaks its own Vinaya rules: no open sangha invitation, no proper sanghakamma agenda. Needing 500 arahants to “agree” a version proves multiple variants already existed from which they would choose. Ananda gets criticised despite his being an arahant—propping up the council’s credibility whilst trashing his. The story needs Buddha’s inner circle for legitimacy but directly contradicts DN 16’s no-successor teaching.
3. Vinaya was process, not ironclad rulebook
Rules arose reactively for local situations, handled by republican sanghakamma votes fitting regional needs. The passage where the Buddha allowed minor rules to be dropped—clashes with his claim that he was never in charge. DN 16’s “Vinaya” can’t mean frozen Iron Age rules from someone preaching impermanence; it means the ongoing self-governing process already in play. The Parinibbāna passage allowing changes looks like a later insertion designed to curb that autonomy.
4. The competence dilemma nails tradition’s priorities
If memory specialists were sharp enough for near-perfect recall over centuries, they noticed wholesale inserts like council myths. Either they were incompetent (then discard the entire Canon), or competent and approved the changes. No third option. Big shifts weren’t errors that were not noticed—they were endorsed institutional upgrades.
5. No pristine original ever existed to “preserve”
The council needing arahants to hash out “the agreed version” admits that alternative versions existed from the start. Purana rejects their compromise. Scattered sangha across regions and languages guaranteed divergence even before Paribanna. The Canon narrates negotiation between branches of a spreading tree, not the establishment of a master recording.
What This Means
Early Buddhism scholarship searches for, and believes in, oral faultlessness. Instead, the Canon shows purposeful rewrites all along the way, and those scholars cry “corruption” when they don’t find what they believe in. But oral tradition didn’t fail at static fidelity—it never tried. It succeeded at continuing a living sangha.
All surviving traditions prove their worth by endurance: those different communities kept what succeeded for them.
Ehipassiko isn’t “inspect the relic”—it’s “step up and encounter the flux directly”. The Buddha’s anti-reification project (impermanence, no-self, dependent arising) demands live encounter with dependent change, not loyalty to Iron Age utterances. Chasing the origin-artifact reifies exactly what he dismantled.
Below is the conclusion to another article, here, which discusses the size of the “true” teachings of the Buddha that traditions want to own. When one weighs up what’s at stake, it quickly becomes apparent that meaning is more important than the text that contains the meaning. Therefore it’s more important to get started with dhamma you can put into practice.
The Scale of Survival: A Quantitative Reality Check
If we treat the Buddha as a historical teacher operating over approximately forty-five years, even extremely conservative assumptions about how often he taught lead to a striking conclusion: only around four hours of teaching per year have survived in the Pāli Canon.
The four Nikāyas, when read aloud at an ordinary speaking pace (including their extensive repetitions), amount to only a few hundred hours of speech. Set against several decades of active teaching, instruction, dialogue, correction, and debate, this represents a vanishingly small fraction of the Buddha’s spoken output. Even generous adjustments to the estimate do not alter the basic picture: well under one percent of what was spoken can plausibly be said to survive.
This fact alone makes it impossible to think of the Canon as anything like a comprehensive record. The overwhelming majority of teachings were never preserved, regardless of how faithfully later material may have been transmitted. Loss is not an unfortunate by-product of oral transmission; it is the dominant feature of the historical situation. This means that selection pressure, not transmission accuracy, is the primary explanatory variable in the formation of the Canon.
Once this scale is acknowledged, the problem of selection becomes unavoidable. The question is no longer how accurately teachings were transmitted, but why these particular teachings survived at all, while tens of thousands of hours of equally sincere, equally authoritative instruction did not.
The Pāli Canon itself shows that the Buddha anticipated loss, degeneration, and the emergence of new teachings. In DN 16 he outlines how later teachings are to be evaluated against the Dhamma he taught, implicitly acknowledging that variation and distortion would occur. What appears to have survived this process is not a comprehensive body of instruction, but a small set of recurring commitments: impermanence, dependent arising, non-self, suffering, and compassion.
What millennia of winnowing seem to have preserved is not doctrinal completeness, but a persistent resistance to reification—something that can be realised through investigation. Results generate confidence in further inquiry and practice, whereas absolute belief rests on confidence in the revealer, himself.
Full article Pre-Transmission Loss in Early Buddhist Orality