An environment becomes scientifically and publicly legible through operations that are neither reducible to perception nor innocent of their material supports. The instrument that calls one urban surface ‘green space’, the newspaper that makes one expert recognisable, the archive that preserves a disciplinary conversation, and the map that renders a neighbourhood’s risks visible do not merely reproduce pre-existing objects. They establish a resolution at which objects can be distinguished, compared and acted upon. This is not the assertion that representation creates every material condition it describes: pollution, heat, enclosure, illness and uneven access remain materially consequential whether or not an index registers them. It is the narrower and more demanding proposition that the route from condition to evidence passes through classifications, formats, temporal conventions, institutional authority and possible audiences. Park, Brown and Pearson reveal this at the level of green-space audits: the alleged simplicity of a variable conceals divergent instruments for recording quality and divergent assumptions about what might matter to wellbeing. The absence of a single measurement standard is not merely a nuisance to be resolved by technical harmonisation, since the divergences also expose different environmental theories embedded within the measure. A path, a bench, water, maintenance or an opportunity for social encounter may enter the audit, while fear, exclusion or seasonal use may remain difficult to compare. Hoffmann and colleagues make a corresponding displacement in environmental discussion: describing change at the level of individual conduct or systemic responsibility changes perceived efficacy and relational risk. Here the medium of comparison is not a geographic sensor but an account of where agency sits.