The aim of this learned and enterprising book is to elucidate the structure and intention of Clement's Stromateis by comparing it with pagan texts from the first and second centuries of our era which belong, as we might now say, to the same genre. This term, which is chaperoned by quotation marks on p. 15, has proved itself heuristically indispensable, but has no closer equivalent in ancient Greek than genos, which is as likely to denote the style or metre of a work as its place in a critical taxonomy. Strict conventions governed versification and the composition of speeches for given occasions, but it is we who have all but invented the epyllion and coined our own names for the novel, the autobiography and the didactic poem. While Heath proposes on p. 138 to render Stromateis as ‘layout’, ‘miscellany’ is the term that is now most commonly applied to this and other ancient texts whose amorphous character seems to resist taxonomy. As Heath observes, however (p. 24), there are all too many specimens of Greek and Latin writing which are in some sense miscellaneous: she might have quoted the thesis of her namesake, Malcolm Heath, that abrupt transitions, divagations and surprises were not aberrations from the classical norm, but calculated devices to heighten the pleasure or whet the interest of the reader, both in poetry and in prose.Footnote ¹ The culture of ubiquitous imitation was also a culture of unceasing improvisation, and both practices are amply illustrated in Heath's comparison of the Stromateis with four books from the second century to which it bears an obvious resemblance: the Natural history of Pliny the Elder, the Convivial questions of Plutarch, the Attic nights of Aulus Gellius and the Deipnosophistae of Athenaeus.

None of these titles signifies a miscellany, and one of Heath's most fruitful contributions to scholarship in the present monograph is her unfolding of the Christian symbolism which is latent in the four images – ‘meadows’, ‘honeycombs’, ‘Helicon’, peplos or ‘robe’ – which Clement juxtaposes with his own title at Stromateis 6.1.2.1 (pp. 174–95). All four are prefigured in a list of rejected titles for the Attic nights (p. 171), while the meadow and the honeycomb perform the same office in Pliny (p. 168). The plucking of flowers from a meadow is the etymological root of our terms ‘florilegium’ and ‘anthology’, but Heath finds a hint of paradise in Clement's appropriation of the conceit (p. 176; cf. 179–80). Helicon was the haunt of the Muses and a source for inspired potations, but mountains and the moving of them gave their own significance in the Gospels (pp. 186–92). The peplos of Athena had a distasteful association with pagan cult, but could also symbolise the poikilia, or variety, which was widely perceived as a virtue in literary composition (p. 194; cf. 355–74). Since it is one of Heath's recurrent claims that in the Stromateis the Logos supplants the Muses who aided the memories of previous miscellanists, we are not surprised that the bee, which metonymically stands for the Muse (and indeed for chastity and priesthood) in classical verse, should furnish Clement with a symbol for the divine nourishment of souls (pp. 175–92). In this interpretation she differs from many who identify the Sicilian bee of Stromateis 5 with Clement's teacher Pantaenus (pp. 182–3). She is equally reluctant to believe that this self-effacing author thinks of himself as our pedagogue to Christ the Didaskalos: the former appellation, she argues, is clearly reserved for Christ at Paedagogus 1.7.55.2 (p. 146), while Clement's promised Didaskalos is not simply to be identified with the Stromateis but should rather be understood as another non-title which underscores the function of the Logos as preceptor to both barbarians and Greeks (p. 202). There is much to debate here, but Heath is clearly right to insist that Christ is the central figure of the Stromateis, against the older critics who maintained that Clement had found a place in his library for everything but the Gospel. A work must be judged by its audience, and if the Stromateis is in any sense addressed to pagans, we cannot require it to be as explicit as the Hypotyposes, which, as Heath conjectures, may have been the fourth and culminating element in Clement's design (pp. 156, 197).

The most original and ambitious chapters are those which undertake to show that the esotericism of Clement – or to use his own term, the intimation of something hidden – is not a peculiarity which sets him apart from classical miscellanists but a characteristic feature of this tradition. Rightly dismissing the common charge that Clement is merely setting up a wall between his own Gnostic circle and the unlettered Christian, she discovers five modes of hiddenness in the Stromateis: the theological ignorance which arises from the inscrutability of God himself; the anthropological mirroring of this ignorance in the invisibility of our most god-like elements, soul and intellect; the exegetic riddles posed by the parabolic and symbolic idiom of the Scriptures; the withdrawal from the social order of those who prefer the company of God to the acclaim of their fellow-mortals; and the cosmic economy of the unseen Creator who is known only by his works (pp. 248–9). The theological mode of hiddenness underlies the other four, while our anthropological likeness to our Maker is in each case the key to knowledge, on the philosophical principle that like is known by like (p. 255). Although Heath does not say so expressly, it would seem that each of the four subsidiary modes is exemplified in one of the classical miscellanists: the anthropological in Plutarch's co-option of the Muses for the ‘personal formation of the reader’ (p. 229); the exegetic in Athenaeus’ hints that true erudition is the cultivation of mousikê (p. 232); the social in the vigils of Aulus Gellius which surreptitiously draw his readers into the cult of the Muses (p. 223); and the cosmic in Pliny's gradual unveiling of the numinous figure of Nature (p. 218). All these authors anticipate Clement in borrowing from the vocabulary of the mysteries; at the same time, it is not they but the Middle Platonists who provide Clement with his model of enlightenment by distinguishing three successive phases in the pursuit of wisdom – ethics, physics and epoptics – of which the last takes its name from the climactic revelation which was experienced by the initiate at Eleusis (pp. 276, 285 etc.). Clement of course is not merely their disciple, any more than he is merely an imitator of the miscellanists: as a Christian he sets the incarnation of the Word against the delusive mysteries of the fallen world in the Protrepticus, while in the Stromateis it is the leavening of the intellect by the same Word that enables faith to see where even philosophy is blind.

With this observation, Heath dispels the seeming inconsistency between the first and the final element in Clement's trilogy. It may be, however, that she finds more consistency between his practice and that of his pagan forerunners than is warranted by a strict review of the extant literature. The division of philosophy into three branches – ethics, physics and epoptics – appears to have been familiar to Origen, who adopted from Clement the title of his own Stromateis;Footnote ² as Heath concedes (p. 304 etc.), it is not quite present in Clement's Stromateis 1.28.176, where the triad is preceded by a fourth term, the ‘historical’ parsing of Scripture, which might correspond to logic in the four-fold division that Origen eschewed. Bucur,Footnote ³ whose extrapolation of the three terms from other passages in Clement she endorses with little criticism, does not in fact cite one occurrence of all three terms together, or even of two in the requisite sequence. The passage from Plutarch's On Isis and Osiris 382d which Bucur, following Hadot, displays as evidence of the currency of this scheme at the beginning of the fourth century, says only that Plato and Aristotle recognised the epoptic mode as the climax of philosophy. If Origen wrote ‘theoric’, not ‘epoptic’ (as some editors opine), this may prove his acquaintance with the three modes of philosophy – ethical, physical, theological – which Iamblichus purports to find in Nicomachus of Gerasa;Footnote ⁴ if, on the other hand, Origen's word was indeed epoptic (or the all but unknown variant ‘enoptic’), it is possible that he arrived at it by subtracting a term from Stromateis 1.28. By contrast, the tenet that like is known by like is indeed a pagan commonplace, but Clement's application of it assumes an initial decision by God to create human beings in his image and likeness, which is not presupposed in Plato's definition of the goal of life as ‘likeness to God, so far as is possible’ (Plato, Theaetetus 176b). Since Clement is one of the earliest readers of Genesis i.26 to conclude that the likeness is given only by prolepsis, his itinerary of the soul includes not only a fall but a future consummation, to which the advent of Christ was a necessary prelude. The disclosure at the last of all that is hidden is an article of the Christian faith that is not foreshadowed in the private eschatology of the Platonic soul.

We must also grant that, for all the ostentatious promiscuity of its quotations from the historians, philosophers and poets of the ancient world, the Stromateis exhibits much more unity of content and of intellectual purpose than the works of the pagan miscellanists whom Heath cites for comparison. The shaping of the Gnostic has an end beyond this world which is of no concern to the hierophant or his literary adept in the mystagogy of Pliny, Plutarch, Gellius or Athenaeus. While the Stromateis is not a systematic theology, André Méhat's comparison with the City of God Footnote ⁵ will not seem so infelicitous when we reflect that the latter is also not a Summa theologiae but a retaliatory tract which has been purposely allowed to outgrow the conventions of apologetic – including the convention of outgrowing its own conventions – to encompass every question on which Augustine desired the world to know his opinion. The nearest congeners in more modern times are Burton's Anatomy of melancholy and Frazer's Golden bough, the first ostensibly a treatise on medicine and the second a study of magic and religion. In both the stated thesis seems to be merely an excuse for parading all that the author has read; yet Frazer at least was in no want of models (now forgotten) for a more direct and rational presentation of his materials. When we assume that a scientific concentration of interest is the hallmark of expository prose, and that the centrifugal style of an author like Clement presents a riddle, we may be sliding into that same fallacy of retrospection that led us for so long to wonder why the first philosophers chose to write in verse. The prosody of Empedocles and the garrulity of Clement ensure that their works are never at more than one remove from the conversational manner which even now is the usual medium, outside schools and universities, for the communication of knowledge. The information thus conveyed is often received with more pleasure and retained with more precision; can we be sure that our modern way of consulting a book through its index, with the brutal efficiency of a clerk reading a ledger, is the norm by which all other literature should be appraised?