‘And verily so excellent he was in this perspective, that a man would say, his even, plaine, and flat picture were embossed and raised work.’ Philemon Holland's endearing 1601 translation of Pliny's Natural History (35.50) is one of the earliest documented references in English to ‘perspective’, understood as ‘the art of delineating solid objects upon a plane surface so that the drawing produces the same impression of apparent relative positions and magnitudes, or of distance, as do the actual objects when viewed from a particular point’ (OED). The passage relates to a Sicyonian painter called Pausias, whose penchant for representing oxen front-on, and yet conveying all their bulk and size, clearly impresses Pliny. But in a sour footnote, the Loeb edition comments: ‘there is no proof that perspective is meant’. It is true that the established text of Pliny does not support Holland's claim that Pausias ‘had a singular gift to work by perspective’. With the benefit of Renaissance expositions, Holland presumably knew what perspective was. But did Pliny know – or indeed did any other ancient writer or artist understand the basic principle of a vanishing point? Modern scholarship has been frustratingly incapable of answering that question, with authorities such as Panofsky and Richter concluding that the geometrical know-how existed, but was not applied, and others (e.g. John White) content to accept that artists understood the principle without needing its formal articulation. Given this aporia, one seizes Rocco Sinisgalli's monograph, Perspective in the Visual Culture of Classical Antiquity, with an eager hope that we may resolve the issue. It is a slim volume, and promises that ‘key concepts…are clarified and enhanced by detailed illustrations’ (cover). Alas, the diagrams rarely clarify or enhance the somewhat staccato text; and the central claim of the book, that ‘ancient theories of perspective were based primarily on the study of objects in mirrors, rather than on the study of optics and the workings of the human eye’ (cover), remains (appropriately, it may be) speculative. If read in a certain way, texts of Euclid and Ptolemy imply the use of mirrors by painters. But one has only to glance at the surviving text of Ptolemy's Optics, as translated from Arabic into Latin by a twelfth-century Sicilian admiral (sic) of Byzantine Greek origin, to see that it is the stuff of a very sadistic Latin Unseen; and that even if Ptolemy's terms meta (‘destination’) or nutus (‘sign’) be sympathetically understood as ‘principal vanishing point’, we could wish that Ptolemy – and indeed the other authors deployed by Sinisgalli, such as Lucretius and Vitruvius – had expatiated upon the utilization of a vanishing point by artists. The trompe-l'oeil trick of linear recession is played often enough in Roman wall-paintings – but that is not quite the same thing, and painters seem not to have been concerned about creating a complete illusion. Sinisgalli imagines Augustus at home on the Palatine, in the Room of the Masks: it is proposed that if the emperor was wearing slippers, with an eye-level at 1.48 metres, and in shadowy light, he would share in a unified ‘perspective’. But I fear we are still unsure.