Over the past decade, many have come to consider magnitude estimation the 'gold standard' methodology among acceptability judgment tasks, often arguing that it provides either more accurate or more precise data than other tasks (e.g. n-point scale tasks). This paper presents four experiments designed to test this claim by investigating the core of the magnitude estimation task itself: the role of the modulus in the estimation of acceptability. There are two logically possible measurement roles for the modulus in linguistic magnitude estimation: (i) as a unit of measure, and (ii) as a reference point along a linear scale of acceptability; however, the results from these experiments suggest that the modulus plays neither role, and in fact, plays no measurement role whatsoever in linguistic magnitude estimation. These results suggest that, to the extent that the claims of more accurate or more precise data are predicated upon the measurement role of the modulus, linguistic magnitude estimation may not yield more accurate or more precise data.