According to Featural Relativized Minimality, the local relation between an extracted element and its trace is disrupted when it crosses an intervening element whose morphosyntactic featural specification matches the specification of the elements it separates. This approach naturally leads to a system able to capture degrees of deviance: the relative acceptability of an intervention configuration will vary as a function of the total, partial or zero featural overlap between the intervener and the target. In a nutshell, configurations involving a lesser degree of featural overlap should be more acceptable than sentences involving a higher degree of overlap. Three acceptability judgment experiments systematically investigated predictions ensuing from Featural Relativized Minimality in extraction from weak islands. Four configurations of feature overlap were systematically tested with different methods of data collection and on a large set of linguistically naïve participants. Results from the three experiments are highly consistent in returning that predictions from Featural Relativized Minimality are globally borne out, except for the configuration involving two lexically restricted wh-elements, for which tentative explanations in terms of grammar or processing are sketched out.