Fee paying transactions cannot be spam

A consensus-valid transaction that pays the market fee rate cannot meaningfully be called spam.

For (strongest version as argued): Spam implies an externalized cost imposed without payment; block space is a priced, auctioned resource, and anyone who wins the auction has paid the network's asking price by the network's own rules. Every alternative definition requires a subjective judgment about worthy uses, and no one holds authority to make that judgment for a permissionless system. The market is also self-correcting: uses that are merely speculative fade when buyers stop paying (as token-inscription volume did).

Against (strongest version as argued): Fees internalize only part of the cost. A transaction's fee compensates miners, but the permanent storage and validation burden falls on every future node operator forever — a classic unpriced externality, which is exactly what "spam" names. Definitions grounded in purpose are not arbitrary: a payment network can coherently distinguish payments from payload-carrying, and the fact that the protocol cannot enforce the distinction does not make the distinction meaningless. Notably, even participants sympathetic to permissive relay declined to endorse the pure "fee paid = not spam" equation.

Nuance: This is a definitional dispute doing load-bearing work for policy conclusions on both sides. Under a fee-funded security model, high-fee non-monetary demand is revenue; under a money-only model, it is pollution. A debater's position here usually predicts their position on the entire filtering stack.

Common misstatements: "Spam is whatever doesn't pay fees" (nearly all contested data transactions pay competitive fees). "The protocol defines spam" (consensus rules define validity, not spam; spam is a policy/social concept). "Everyone agrees data transactions are/aren't spam" (this is precisely the unresolved center of the debate).