We propose that anchoring is often best interpreted as a scaling effect--that the anchor changes how the response scale is used, not how the focal stimulus is perceived. Of importance, we maintain that this holds true even for so-called objective scales (e.g., pounds, calories, meters, etc.). In support of this theory of scale distortion, we show that prior exposure to a numeric standard changes respondents' use of that specific response scale but does not generalize to conceptually affiliated judgments rendered on similar scales. Our findings highlight the necessity of distinguishing response language effects from representational effects in places where the need for that distinction has often been assumed away.