Johannes Overbeck was by all accounts a beloved and dedicated pedagogue. Among his contributions to academic life at the university of Leipzig, where he taught classical archaeology for some four decades up to his death in 1895, was the establishment of a clinic for student welfare. A sense of how he pitched lectures is readily gained from his many publications. Overbeck aimed above all to be nützlich, ‘useful’. One of the most useful services he performed for students everywhere was to gather in one volume a magisterial selection of surviving ancient literary testimonies for artists in the classical world.

Published in 1868, Overbeck's Antike Schriftquellen zur Geschichte der bildenden Künste bei den Griechen has indeed proved of durable utility. Since Overbeck did not translate his sources, nor expound them, Anglophones were furnished with derivative textbooks, such as H.S. Jones's Ancient Writers on Greek Sculpture (1895), explicitly tailored for Oxford undergraduates in need of more elementary guidance. Nonetheless, a calfbound copy of Overbeck's collection remains a pleasure to have and to hold. It is sweet and significant that a comprehensive revision and expansion of the work keeps his name.

Five bulky volumes of Der Neue Overbeck (DNO) come at a cost. One loss with respect to the original is of course its portability. Given the option of online consultation, it is a devoted student who will invest in this paperback edition. But a sense of a wonderful thesaurus is undoubtedly heightened by the textual mass of DNO, totalling 4,280 entries plus sundry addenda. Not only does it enlarge upon the original by including a wealth of inscriptions relating to Greek artists (vase-painters, mosaicists and certain other specialists excepted). It gathers sources discovered since Overbeck's time, such as papyrus samples of Posidippus and Herodas; it retrieves incidental references neglected or deemed trivial by Overbeck; and it supplies an apparatus of translation and commentary (in German), along with full bibliography. A number of small-scale monochrome illustrations and line drawings offer a little variety to the lexical load, occasionally complement its information and generally serve to remind us that all this verbiage was once generated by visual experience.

Or was it? At Leipzig it was said of Overbeck that ‘he wrote the best book on Herculaneum and Pompeii, without having seen either spot’ (C. Schurz, The Reminiscences of Carl Schurz [1907], p. 94). Overbeck did eventually visit Italy; but as Freud would argue, we may know an ancient place long before we actually view it for ourselves. A comparable claim could be made about the artworks relating to the texts collected by Overbeck: even if he had favoured travel ‘in the field’ over hours in the library, his ‘material’ was mostly invisible. The margins of DNO tantalise its readers by showing a succession of statue-bases minus the statues; the texts extol in verse and prose a series of must-see paintings that disappeared long ago. To what extent, then, is the corpus of DNO self-sufficient? What if a literary invocation of (say) the pictorial genius of Apelles derives not from direct viewing of pictures by Apelles but from the reception of stories about Apelles and his work?

One of Overbeck's distinguished students was Adolf Furtwängler, who went on to devise a method of gaining virtual knowledge of lost classical sculptures by assembling and comparing presumed copies of such ‘masterpieces’ (‘Meisterwerke’), thereby arriving at some kernel of shared aspect that constituted an idea or essence of the original. Though he never stated as much, Overbeck surely hoped for a similarly positivist use of his anthology. Proceeding in broad chronological phases, commencing with ‘mythical artists’ such as Daedalus and the Telchines, he divided his chosen sources about named individuals into three categories of information: (i) biographical, (ii) attestations of particular works and (iii) technical, stylistic and ‘general’ (‘Allgemeines’). DNO follows both the chronological phases and the tripartite system, dropping ‘Technik’ and ‘Kunstcharakter’, while allowing ‘Allgemeines’ occasional inflexions, for example ‘Allgemeines: Nachleben’. The more prolix groups of entries are completed by a tidy ‘Resümee’. Inserting epigraphic material within this system undoubtedly has the effect of adding historical credibility to artists we might otherwise suspect to be mythologised by the literary tradition. For example, DNO collects 180 ‘Schriftquellen’ for Praxiteles: among them are three statue-bases from mid fourth-century bce Attica naming Praxiteles as sculptor, evidently portraits of females involved in the cult of Demeter and Kore. It is conceivable that Praxiteles and his sons created that honorific stereotype now broadly known as the ‘Herculaneum woman’. The literary tradition ignores such regular worthy commissions, preferring salacious comment about images of the courtesan Phryne at Delphi and Thespiai or gossip concerning the statue of Aphrodite by Praxiteles eventually installed in a temple at Knidos. It is presumably the aim of DNO, in keeping with its eponymous model, to furnish students with the maximum available textual information about the individuals who constructed the entity we call ‘classical art’. Within the complex weave of words there must be threads of truth.

Of some 3,000 literary texts in DNO, two-thirds are Greek, predominantly taken from Pausanias. Latin texts are 90% prose, with over 70% extracted from Pliny the Elder. Only Cicero, Lucian and Plutarch stand out as other contributors with more than 50 citations to their name. Such statistics immediately alert us to potential bias in this compendium. Given that volumes I–IV of DNO cover from ‘Daedalus’ to artists of the ‘late Hellenistic’ period, its most ‘readable’ sections – that is to say, those parts that offer some kind of narrative about an ancient Greek artist – derive mainly from authors writing several centuries in retrospect. As is well known, many of the texts are anecdotal; and as scientists remind us, the plural of anecdote is not data. Far from being literally ‘unpublished’ (ἀνέκδοτα), these are stories routinely recycled, repurposed and relocated from one artist to another. Searching for those threads of truth therefore becomes a challenge.

I tried the quest with Protogenes. The Carian site of Caunos, alleged birthplace of Protogenes, has become archaeologically less mysterious than it was, yielding (in the 1990s) traces of a late fourth-century bce family exedra monument with a dedicant named Protogenes. But since this cannot be proven to be the late fourth-century artist, there are no inscriptions among the 39 sources for Protogenes assembled by DNO. Earliest mentions come from Varro and Cicero; however, while Pliny is our prime informant for Protogenes – eulogising in particular an image of the Rhodian hunter-hero Ialysus displayed (we assume between 75 and 191 ce) at the Temple of Peace in Rome –, not one of these sources presents itself as an eyewitness account. Pliny (NH 35.101–6) tells a tale about the prodigal effort and technical application summoned by Protogenes for the Ialysus project. This tale culminates with the artist suffering a sort of skill-failure, being unable to achieve the desired effect in just one part of the scene – depicting the foam-flecked mouth of an exhausted dog. Many of us know or can guess what happens next. In a fit of rage, the artist throws a sponge at his picture. The sponge strikes the area of the dog's mouth – and, as if by magic, it produces exactly the desired effect. Pliny states the lesson, fecit in pictura fortuna naturam, ‘luck rendered nature in the picture’, and tells his readers that another painter, Nealces, had a similar happy result when trying to portray saliva about the jaws of an excited horse. (Whether Nealces was following a ‘method’ of sponge-flinging established by Protogenes or enjoyed the same fortunate consequence of angry frustration, is not clear from Pliny's account.)

Commentary on this passage in the DNO is strikingly reserved. There is cross-reference to a variation on the sponge story as told about Apelles and his depiction of a war-horse (DNO 2931–3), but no allusion to the itinerant tale (‘Wanderanekdote’) when invoked without mention of a particular artist. Nor is there any discussion of what the anecdote signifies. For Pliny, it seems, the incident proves that chance, fortuna, can be part of aesthetic success and part of an artwork's market value: further interpretations are possible (see V. Platt, ‘Of Sponges and Stones: Matter and Ornament in Roman Painting’, in: N. Dietrich and M. Squire [edd.], Ornament and Figure in Graeco-Roman Art. Rethinking Visual Ontologies in Classical Antiquity [2018], pp. 241–78). With other writers, the anecdote serves as a paradigm of Pyrrhonic Scepticism (Sextus Empiricus, DNO 2932); testament to the auxiliary power of Tyche (Dio Chrysostom, DNO 2931); illustrative of Stoic pronoia (Valerius Maximus 8.2.ext.7); or a singular exception to the rule that technical effort is essential to artistic achievement (Plutarch, Mor. 99B). All these variations may be radically connected to Aristotle's citation of a one-liner from Agathon, τέχνη τύχην ἔστερξε καὶ τύχη τέχνην, ‘art loves chance, and chance loves art’ (Eth. Nic. 1140a).

The sources disagree as to how many years Protogenes worked on the Ialysus picture, even while ballistic warfare pounded outside his studio. Only Pliny adds the detail that the artist subsisted throughout on a diet of pickled lupins. DNO refers us to Nikias, who became so absorbed in his painting that he had to ask his household staff whether he had eaten or bathed (DNO 2823). But that is a story about the absorption and unworldliness of a creative genius. (According to Plutarch [Mor. 1093E], Nikias was working on a scene of Odysseus in the Underworld. When the picture was completed, the artist refused to let it go, though Ptolemy offered him the normally irresistible sum of 60 talents.) Pliny may not know that lupin beans – gluten-free, low carbohydrate, high protein, probiotic – are a superfood. Still, he emphasises that Protogenes chose this diet as part of an ascetic regime, in order to give full attention to his work.

One could go on. Obviously, the editors of DNO, like Overbeck before them, were at some point obliged to curtail their task with the cry of manum de tabula!, ‘stop writing’. But behind their reluctance to pursue thematic connections within, between and beyond the sources they have compiled, there lies a fundamental problem. Overbeck, as mentioned, made a magisterial selection: that is, for the sake of his students, he justified the exclusion of certain texts on the grounds that they were ‘nichtsnutzige Notizen’, instances where he considered that ancient literature about art had nothing to teach the archaeologist or art historian. He was not entirely consistent about the application of his principles: Philostratus was banned because the Imagines, in Overbeck's judgement, were exercises in fictional ekphrasis, while Homer's description of the Shield of Achilles (Iliad 18.468–608) was admitted in its entirety (DNO 112 permits just four lines, for the mention of Daedalus). Such discrimination had precedent, however, with K.J. Sillig's Catalogus artificium (1827). It rested upon the categorical difference for which G.E. Lessing so aggressively campaigned in his Laokoon essay of 1766. The making of images and the making of texts were to be kept wholly separate. They were never, as Lessing argued, transferable skills. Accordingly, when (for example) Cicero mentions that Pheidias might finish off a statue roughed-out by another sculptor (Fin. 4.13.34), this is not to be regarded as a comment about ancient art. Pheidias merely serves the author as exemplary of an abstract process whereby humans require to be shaped and polished by education. The passage therefore belongs to the genre of philosophical prose, not marble sculpture.

As if aware that Lessing's case for incompatibility of the arts no longer commands general assent, the editors of DNO have reclaimed certain passages damned as ‘useless’ by Overbeck, such as Quintilian's allusion to Pheidias as not only the ‘master-sculptor’ of the Olympic Zeus, but an artist perfectly capable of executing all the minor decorative accoutrements of that enormous statue (DNO 957). The context is advice about learning the art of rhetoric, not the art of chryselephantine statuary; nonetheless, it delivers at least proverbial wisdom about a classical artist. We also find the inclusion of a critique of the Olympic Zeus attributed to Apollonius of Tyana (DNO 983): once more this is metaphorical, yet it serves to indicate theological (Neopythagorean) disquiet about ‘fixed’ effigies of the divine.

Such flexibility is welcome. All the same, libraries should not use DNO as an excuse to jettison their copies of the first major modern syntheses of ancient literature about art, namely the De pictura veterum (1637: translated by the author as The Painting of the Ancients, 1638), and Catalogus architectorum etc. (1694) by Franciscus Junius. The first is a discursive essay, the second a dictionary of names: both have been admirably made available to modern readers (Franciscus Junius, The literature of classical art. Vol. i. The painting of the ancients: De pictura veterum according to the English translation (1638), edited by K. Aldrich, P. Fehl and R. Fehl; Vol. ii. A lexicon of artists and their works: Catalogus architectorum …, translated from the original Latin of 1694, edited and translated by K. Aldrich, P. Fehl and R. Fehl [1991]). Junius served as librarian to one of the first systematic modern collectors of classical art, Thomas Howard, Earl of Arundel. A multilingual European scholar, Junius so fully adapted to English ways that his compilation of stories about ancient artists reads somewhat like Aubrey's Brief Lives, a parade of quirky personalities. Typical of the information that fascinated Junius was the report that another celebrated painting by Protogenes, showing a satyr leaning against a pillar, featured as part of its background detail a partridge (perched above the pillar). The picture went on view at a sanctuary at Rhodes, near to where the image of Ialysus was originally displayed. Spectators were charmed by the partridge; then partridge-breeders began bringing their birds to the temple, and the phenomenon of real partridges calling out to a painted partridge attracted crowds. Crotchety Protogenes, displeased to learn that one of his subsidiary studies or ‘sideworks’ (parerga) was eclipsing the image of Ialysus, insisted the temple-guardians allow him to paint out the partridge (Strabo 14.2.5; DNO 3011).

How Junius extrapolated paradigms from such tales – this one inspired the conclusion to his essay – remains worth following, and not only for its influence upon artists such as Rembrandt and Rubens. What irritated purists such as Lessing and Overbeck, an easy latitude of thematic and technical association across all the arts of classical antiquity, is just what makes a precious if idiosyncratic conspectus. Junius certainly strays afar – citing Fabius Maximus, for instance, to affirm that artists should work without haste. With some reason, however, he deemed that the precepts of Longinus on ‘the Sublime’ should apply as much to the craft of shaping images as to the craft of shaping texts. In DNO, by contrast, the heredity of Overbeck endures to the extent that it admits artists by name alone. General aesthetics therefore fail the entry test.

The irony remains: that knowledge about the makers of classical art comes to us mostly filtered through Roman or Romanised sources. So many Graeculi delirantes, ‘crazy little Greeks’! (Pet. 88, specifically referring to Pheidias and Apelles, but complemented by evocations of Myron and Lysippus – doomed to die destitute, such is their tireless quest for artistic perfection.) And for all their engaging ‘human interest’, what in the end do these stories convey? Is it not essentially a single critical preoccupation, an artistic telos lauded from Daedalus onwards – the insistent praise-song of how lifelike? In Philippe Descola's analysis of figurative art across human cultures, Graeco-Roman obsession with mimêsis (‘imitation’) is merely part of one function potentially fulfilled by art, ‘analogism’ (P. Descola, Les formes du visible: une anthropologie de la figuration [2021]). From a global perspective, the goal of creating virtual reality, art as Nature's analogue, seems a narrow and ultimately somewhat preposterous Western fixation. In that sense, DNO might be regarded as a vanity project. As the pale Roman versions of nobilia opera by Myron and Polykleitos no longer claim prime location in museums around the world, perhaps a textual monument to Roman connoisseurship is ideologically out of date; perhaps the time has come to disown ‘the heritage of Apelles’ (to borrow E.H. Gombrich's phrase). There is no doubt that the literary development of a ‘story’ about classical art privileged individuals over workshops, males over females, competition over cooperation – and Greeks over all other ethnicities (including Roman). There is also no doubt that relics of this literature survived in sufficient quantity to shape both the behaviour and the biographical glorification of artists in early modern Europe. In short, a lot to answer for.

Nonetheless, it is hard to begrudge the legacy. Surely for no other pre-industrial society is the making of images so generously documented. And this is despite the fact that of all the classical literature explicitly devoted to art and artists – we know full well that it existed and conveyed precious expertise about sculpture, painting, architecture and more, from Iktinos, Polykleitos, Euphranor, Xenokrates et al. – not one integral text has survived. In lieu of what has been lost, the substance of DNO is little short of miraculous.