During the eighth and ninth centuries, when Christian theologians struggled to define the role of icons in their worship, they also re-examined theories of how the human eye perceived and interpreted such paintings. Ancient Greek optics became central to their understanding of sight, touch and imagination, generating a physical as well as philosophical debate about the extromission of rays from the eyes, or the intromission of rays from objects into the eyes. Medical treatises on theories of vision, sensation and perception were investigated to establish the superiority of touch or sight. Arguments moved far beyond the justification for veneration, traditionally based on the statement of Basil of Caesarea that reverence for the imperial image passed on to the individual depicted, to raise philosophical problems of a more complex order.

While iconophile writers like John of Damascus developed arguments that encouraged Christians to use icons to express their devotion without committing idolatry, iconoclasts such as the emperor Constantine v denounced any possibility of painting any image of Christ's divine nature, and therefore elevated the eucharist as the sole legitimate representation that should be venerated in a spiritual manner. These debates were rehearsed and further developed under Leo v, when iconoclasm was officially reinstated at a council held in 815. The emperor had commissioned a much broader search for iconoclast texts by a commission under the direction of a scholar later known as John the Grammarian. After nearly thirty years of this iconoclast domination, the widowed Empress Theodora set about undoing it by removing John from the patriarchate and appointing the iconophile monk Methodios. She also nominated many iconophile supporters to key positions in both church and state administration. She did not, however, summon a council of all bishops to endorse this change, as had happened in 787, but relied on a new liturgical celebration of iconophile belief, the Synodikon of Orthodoxy, composed by Patriarch Methodios. While this condemned iconoclast rulers by name, she insisted on omitting her husband Theophilos from the list, citing his death-bed conversion to iconophile practice – an invention designed to excuse his clear commitment to iconoclasm.

For many years after this reversal of iconoclast theology, iconophile leaders expressed their anxiety about dissident church leaders who did not embrace the change and secretly harboured their traditional beliefs. The generation of iconoclast bishops appointed in 815 and later, reinforced by Theophilos's determination to stamp our icon veneration, may well have clung to what they considered orthodox practice. Whether iconophile supporters of Theodora, her son Michael iii and his co-ruler and later sole emperor Basil i exaggerated fears of a lingering iconoclasm, or used the threat of a possible revival to achieve their own ends, the second half of the ninth century witnessed a continuing preoccupation with the role of icons. Holy images were not immediately restored and were only very gradually reintroduced into major churches.

It is in this tense period of concern over icons that medical and philosophical theories of perception commanded renewed attention. These are the subject of Roland Betancourt's learned study, presented in three parts. The first examines ancient Greek traditions and their later reworking by Byzantine scholars about how the eye sees, which he claims distinguishes vision from touch. The theory of haptic sight, which has become dominant in art historical circles, is roundly denounced. The second focuses on Photios, one of the most erudite of ninth-century scholars, who was actively involved in this debate about the perception of art. And the third extends from the medium of sight into its artistic and linguistic mediation, investigating theories of rhetorical, ritual and visual representations of the divine.

One of the great merits of this dense study, for which all non-specialist readers will be grateful, is that every quotation in Greek is followed by a translation into English. Through a vast range of Christian authors and lay intellectuals – from Aristotle and Plato through the Neo-Platonists, Cappadocian Fathers and later Byzantine philosophers – Betancourt has thus established his own interpretation of terms that were used in a great variety of ways. Many difficult passages in ancient and medieval Greek, including some quite obscure and rarely read, are adduced to support his understanding of the Byzantine theory of perception. Literary experts may well criticise his conclusions about specific genres (progymnasmata, for instance), but the range of texts and their careful translation and interpretation command admiration and respect.

The first part opens with a critical assessment of the theory of haptic vision in the study of Byzantine art: the notion that the eye grasps the object seen as if touching it. Betancourt argues that this neglects the important role of the imagination in the appreciation of objects like icons. He proceeds to analyse ancient Greek theories of sight, often concerned with how a ray of light projected by the mind onto an object (extromission) creates an image. This is contrasted with the rays received from the object seen (intromission), which generated a more passive way of seeing. He analyses these theories and stresses the intermediary medium identified by Aristotle as the diaphanous, which produces a separation between the object and the viewer and activates the transmission of colour to the eye. This emphasis on the medium is common to all sensory awareness, although touch appears to be instantaneous and thus unmediated. But sight is superior because through the diaphanous it brings colour to the vision, which touch alone cannot do. Plato also postulated a medium between the effluxes, two rays from viewer and object, as an intervallic space between them, encapsulated in the mirror (p. 35).

Next Betancourt surveys a wide range of later early Christian and Byzantine commentators, enriched by Galen's medical views and Stoic commentaries, who adapted the theory of vision reflected in a mirror to stress the ebb and flow of visual streams, the encounter and fusion of optical rays from both viewer and object mediated by an intervening medium. After a masterly survey of thinkers such as Nemesios of Emesa, John of Damascus and Symeon Seth, Betancourt identifies the hybrid theory of vision that finds its final form in Nikephoros Choumnos's claim to have reconciled the opinions of Plato and Aristotle, with emphasis on the ‘light-borne colours being passed along by the air’ (a reference to the diaphanous, p. 48).

Part ii examines Photios's famous Homily 17 on the mosaic of the Mother of God and Child in the apse of Hagia Sophia, known to all students of Byzantium in Cyril Mango's translation published in 1958. Here Betancourt presents the patriarch's interpretation of the image and his understanding of its power. Against an obvious reaction to Photios's claim that sight is more effective than hearing, he stresses the patriarch's understanding of the process of perception: Photios moves beyond the sensory awareness of the image to its apprehension by the mind, its visualisation in the imagination, and its storage in the memory (p. 113). He describes the image, which is very difficult to see clearly from the floor of the church, as he imagines it and as he wants the congregation to understand it. And because he delivered the sermon on 29 March 867 to the emperors Michael iii and Basil i, as well as a great crowd of officials, skilled artisans and uneducated people, he intended it to be understood. Betancourt thus takes it ‘as the closest one can get to an articulate popular understanding of vision’.

The very considerable leap from a sophisticated text to a more general grasp of the theory of perception by the Constantinopolitan population that filled the Great Church on that Easter Saturday may not convince all readers. But it is a valiant attempt to identify what the patriarch wanted his audience to understand. Photios's interpretation of how to view the image draws on a theory of cognitive perception to be exercised by each member of his audience. ‘Has the mind seen? Has it grasped? Has it visualised? Then it has effortlessly transmitted the forms to the memory.’ Betancourt's critical reading of Mango's translation is directed against the theories of haptic vision that have become influential since 2000 and reduce sight to an aspect of touch. While the significance of touching an object, such as an icon, a manuscript or a coin, regularly confined behind glass in museums, may be critical to its evaluation, in the case of the almost invisible mosaic high in the apse of Hagia Sophia, a theory of perception may be more meaningful.

The third section brings together a wide range of texts, grammatical exercises, formal descriptions of works of art and liturgical performances, which communicated theological discourse to a broader range of the Byzantine population. It studies many theories of representation largely derived from Aristotle, Plato and later commentators (also examined in part i), ‘using synesthetic language to distinguish between different states and stages in the perceptual processs’ (p. 202) and demonstrating a rhetorical aim of ‘transmuting words into images’ (p. 206), with attention to the different meanings attributed by different authors to such concepts as phantasia, enargeia and typos. Betancourt concludes this seriously theoretical part with a short section titled ‘Tempted to touch’, which opposes the predominance of haptic visuality in Byzantine art by stressing the importance of keeping the processes of touch and sight distinct, even when they contribute a mutually fruitful understanding of how the body engages with the icon. An analogical and synesthetic approach structures behaviour, such as kissing an icon, and articulates abstract notions that ‘found the cognitive and spiritual aesthetic experience of imperceptible forms, noetic and divine’. In all this, imagination is a key element which Betancourt argues has been neglected by art historians.

This book is not an easy read, but it demonstrates most clearly that in every century of the empire anonymous teachers, readers, scribes and more famous individuals like Photios, Psellos and Metocheites engaged with the ancient Greek inheritance, commented on what they read, and criticised Aristotle and Plato as well as a host of lesser philosophers and medical experts like Galen. Not only did these Byzantine intellectuals preserve the classics by copying their texts, but they also puzzled over their contradictions, modified their conclusions with their own evidence and enormously enriched them. They also aimed to teach the uneducated how to appreciate their icons. In Betancourt's detailed examination, their medieval theories of Sight, Touch and Imagination are given brilliant exposition.