Resolution Drift

The standard model of linguistic communication treats meaning as a point that survives the channel: the speaker selects a meaning, encodes it in surface form, and the hearer decodes the surface back to the meaning, with miscommunication modelled as channel noise corrupting an otherwise determinate signal. This paper argues that the model omits a transmitted parameter it cannot do without. Speakers do not only transmit meanings; they transmit, and intend to transmit, a resolution: how collapsed the content is meant to be, ranging from a sharp point to a deliberately open distribution over candidate readings. Resolution is part of the message, it is transmitted lossily and asymmetrically, and the gap between the resolution a sender intends and the resolution a receiver reconstructs is a distinct and diagnosable failure mode that the encode-decode model codes, wrongly, as the receiver simply getting the meaning wrong. I name this gap resolution drift, give it a signed typology (positive drift, where the receiver collapses tighter than the sender meant and manufactures definition that was never sent; negative drift, where the receiver holds looser than the sender meant and loses definition that was), predict from Binary Bias (Temte 2026a) that the unmarked receiver default is positive drift, and locate the decision-theoretic cost of the phenomenon in Quantum-Ethical Decision Algebra (Temte 2025c): the receiver who collapses a cloud the sender meant to leave open is committing a premature collapse on the sender's behalf, on borrowed authority. The framework's highest-stakes arena is human-AI interaction, where both parties drift in characteristic directions. The paper closes with the empirical signature of drift, a set of resolution-marking remedies, and the paper's position in the corpus. It is the production-end sibling, anchored on the sender's intended resolution in live intralinguistic communication, to Translation as Probability Cloud (Temte 2026as) at the reception end, anchored on the recovery of a source meaning across linguistic distance.