Praxiteles

Ancient writers ranked Praxiteles among the two or three greatest sculptors Greece produced, and Pliny the Elder singled out his marble work as surpassing even his own bronzes: 'though Praxiteles achieved greater fame through marble work, he nonetheless created beautiful pieces in bronze as well.' The reputation was never seriously questioned in antiquity or since.

What is missing is the physical proof a reputation this size would normally rest on. No surviving sculpture bears his signature. Andrew Stewart's catalogue of the testimonia counts eight ancient inscribed statue bases carrying his signature, some of them later renewals, but in every case the statue that stood on the base is lost, so no inscription survives paired with an object anyone can look at today. The corpus now associated with him, whether the dozen or so major types accepted with reasonable confidence or the roughly sixty works Pliny, Pausanias and later testimonia between them attribute to a sculptor of this name, is a modern reconstruction built from Roman marble copies, coins, gems and two ancient authors writing well after his death. An attribution to Praxiteles is a proposition assembled from later evidence, not a record.

The primary ancient source is a small number of sentences in Pliny the Elder's Natural History, written around 77 CE, roughly four centuries after Praxiteles worked. Pliny lists his bronzes in book 34, including 'a youthful Apollo which the Greeks call Sauroctonos, because he is waiting with an arrow for a lizard creeping towards him,' and treats his marble Aphrodite at Knidos in book 36 as the finest statue in the world. Pausanias, writing in the second century CE, adds the Mantinea commission and other temple works seen on his own travels through Greece.

That is close to the entire ancient documentary record. Everything else in circulation, the workshop gossip, the anecdotes about his mistress, the precise dating of individual works, comes from later compilers, from Roman-period rhetorical set pieces such as the dialogue on love called the Erotes, or from modern inference. This page follows the house discipline throughout: a fact from Pliny or Pausanias is treated as primary and dated to them; everything else is dated to whoever actually said it, and graded accordingly.

Pliny records that Praxiteles carved two figures of Aphrodite for sale at the same price, one draped and one nude. The people of Kos, who had first choice, took the draped version 'and for this reason was preferred by the people of Cos, who had an option on the sale, although he offered it at the same price as the other.' Knidos bought the nude figure the Koans had rejected, 'gaining immense fame' from it. It was the first time in Greek art that a full-scale female figure had been carved fully nude, standing rather than in the heroic male tradition, and it set a type that Roman sculptors reproduced for the next five centuries.

The original is lost. What survives is copies: the classicist Kristen Seaman has catalogued 192 surviving ancient versions and variants, which would make the Knidia the single most reproduced sculpture of antiquity if the count holds. The version most scholars treat as closest to the lost original is the so-called Colonna Venus, now in the Vatican's Pio-Clementino Museum, following an argument first made by the Danish archaeologist Christian Blinkenberg that remains widely, though not universally, accepted. A second well-known Vatican version, in the Museo Chiaramonti, gives a different sense of the same pose. Neither is the original. Both are Roman.

Two anecdotes attached to the Knidia are worth separating by how well they are sourced. The first is King Nicomedes of Bithynia's attempt to buy the statue from Knidos by offering to discharge the city's entire public debt: Pliny states plainly that 'the Cnidians, however, preferred to suffer anything but this, and rightly so; for with this statue Praxiteles made Cnidus a famous city.' That is Pliny, writing as a near-contemporary compiler of well-known Roman-period lore, and this page treats it as a reported ancient tradition rather than a verified transaction, since no independent record of Nicomedes' offer survives.

The second anecdote is more vivid and less securely Plinian: a story, told at length in the dialogue on love traditionally but not certainly attributed to Lucian of Samosata, of a young man who broke into the temple at Knidos at night, left a mark on the statue, and afterwards threw himself from a cliff in shame. The temple itself was built open on all sides specifically so the statue could be viewed from every angle, which is itself telling about how the object was designed to be encountered. This page records the stain story as what it is: a widely repeated ancient anecdote about the statue's uncanny effect, not a documented event, and older than any surviving manuscript that reports it.

Ancient sources also disagree, without possibility of adjudication from where this page stands, about who modeled for the figure. Athenaeus and the later rhetorician Choricius of Gaza name the courtesan Phryne, said elsewhere to have been Praxiteles' mistress; Clement of Alexandria instead gives the name Cratina. Both claims are centuries removed from the event they describe, and no source explains how either writer would have known.

On 8 May 1877, excavators working for Ernst Curtius's German expedition to Olympia uncovered a marble group of Hermes carrying the infant Dionysos inside the ruined Temple of Hera. It had been protected by a layer of clay and survived in exceptional condition: the face and torso are highly polished, the back is left rasped and chiselled where it would not have been seen, and traces of cinnabar remained in the hair at the time of discovery, along with gilding on the sandal straps. Pausanias, describing the same temple two thousand years earlier, mentions a marble Hermes carrying Dionysos as a work of Praxiteles, and the identification of the find with that sentence was made almost immediately.

The identification has never stopped being argued. In 1927 the German archaeologist Karl Blumel published a technical study of ancient carving methods and concluded, on the evidence of the tool marks, that the Hermes was not a fourth-century original at all but a Roman-period copy. A few years later the Swedish scholar Oscar Antonsson took the opposite position, that the statue is an authentic work of Praxiteles but has been reworked, and went further, arguing the figure is not Hermes at all but Pan carrying the infant Dionysos. By 1948 Blumel had revised his own view again, proposing that the statue is an original, but by a different, later sculptor who happened to share the name Praxiteles, active in the second century BCE. Brunilde Sismondo Ridgway's survey of fourth-century sculpture treats the piece within the wider question of what an original of this date should look like at all, without resolving it. The most commonly repeated modern position, that no ancient replica of the group has been identified anywhere else in the corpus, is itself used on both sides: to argue that it must be an original, since nothing else copies it, and to argue that it cannot be one of Praxiteles' famous, endlessly reproduced works, since none of his famous works went uncopied.

No consensus exists. This page states that plainly rather than picking a side its sources do not support.

In 2004 the Cleveland Museum of Art bought a life-size bronze Apollo from Phoenix Ancient Art, the Geneva and New York gallery of the brothers Ali and Hicham Aboutaam, for a price it has never published. The type, an Apollo leaning against a tree trunk and aiming at a small creature, is the Sauroktonos Pliny describes; roughly forty marble versions of it are recorded across collections from the Louvre to the World Museum Liverpool. The Cleveland bronze is the only life-size bronze version known. Everything else that survives of this famous invention is stone.

That single fact is why the object matters far beyond its own gallery label. A marble copy preserves a silhouette. A bronze preserves an alloy, a casting method, a corrosion history and, through its lead, a possible mine of origin, none of which any marble copy can offer. In 2013 the museum's own curator of Greek and Roman art, Michael Bennett, went further than attributing the bronze to Praxiteles: he argued the small creature at the god's feet is not Pliny's lizard at all but the Python of Delphi, and renamed the object accordingly, so that its current catalogue title carries both readings joined by 'or.' On that reading the object is not a copy of anything: it is the statue Pliny described, and every marble Sauroktonos in the Louvre or the Vatican is a copy of it rather than the reverse. Other scholars, including Jennifer Neils, have disputed the Python reading directly, on the grounds that the surviving creature is too small to read as the serpent of Delphi, and Martin Robertson had floated the Apollo-Python association as a possibility rather than a finding well before Cleveland adopted it as fact. Writing in 2017 on the Sauroktonos type as a whole, not on the Cleveland bronze specifically, Neils herself concluded that 'the documentation does not permit us to recognize this work in detail and the context of its creation completely eludes us' - a judgment aimed at the type's whole documentary record, not only at Cleveland's copy of it, and worth holding against any confident attribution anywhere in this page.

A full metallurgical study by Colleen Snyder, Ernst Pernicka and Peter Northover, published by the Getty in 2017 with sample tables included, is the best evidence either side has. It found that a thin bronze display plate under the statue's feet was soldered on in the eighteenth or nineteenth century, meaning the object was already mounted for display well before any of the discovery stories that surface from the 1990s onward. It found the figure had been through a fire early in its history and had never been in salt water, directly rebutting a Greek allegation that the bronze had been fished from the sea between Italy and Greece. And it found that the lead in the bronze is not Aegean, which rules out the largest ancient Greek lead source, Laurium, without telling anyone where the statue actually stood.

None of that dates the sculpture to the fourth century BCE, and none of it proves Praxiteles' own hand rather than a later follower working in his style, a Hellenistic reduction of a lost bronze, or a Roman-period version. The museum's own catalogue range, c. 350 to 200 BCE, spans two and a half centuries and effectively concedes that the object could postdate Praxiteles' working life by more than a century. That range, printed by the institution that owns the piece and argues for the strongest version of its attribution, is the clearest single admission in this whole dispute that the case is not closed.

The museum's published provenance for the bronze has exactly three entries: a German lawyer named Ernst-Ulrich Walter, Phoenix Ancient Art in 2004 and the museum itself. Walter's own account, given to journalists rather than published as a sworn statement, is that he found the statue broken on his family's reclaimed estate in the Lausitz region of former East Germany after reunification in 1990, believed it to be an eighteenth or nineteenth century garden copy, and sold it for around 1,250 deutsche marks to a buyer whose name he says he cannot recall and for which no receipt survives. Lucia Marinescu, then director of Romania's National History Museum, has stated she saw the fragments on the estate in 1994 and later argued in a conference paper that the sculpture is Roman rather than fourth-century Greek, a position that directly contradicts the museum that eventually bought it. Phoenix Ancient Art has never named the person or persons from whom it acquired the object, and the gallery's own owners have separate, later legal histories: Hicham Aboutaam pleaded guilty in 2004, the same week as the Cleveland announcement, to a federal customs charge over an unrelated Iranian silver vessel, and Ali Aboutaam received an 18-month suspended sentence in Geneva in 2023 for violating Swiss law on the transfer of cultural property using forged provenance documents. Manhattan prosecutors have since seized further Phoenix Ancient Art material from other American museums, including a Roman marble bust returned to Turkiye in 2026, none of which names the Cleveland Apollo specifically but all of which documents continuing law-enforcement scrutiny of the network the bronze passed through between 2001 and 2004.

Greece's public position, first raised in December 2006, is that the statue was more likely recovered from the sea between Italy and Greece in the 1990s than found in any German garden; the Getty metallurgy directly contradicts the marine-recovery claim without confirming the garden story either. No claim has been filed by any government, and no restitution agreement exists.

Andrew Stewart's catalogue of Praxiteles' extant and recorded works, compiled from Pliny, Pausanias and later testimonia, runs to fifty-nine entries once the works Stewart himself flags as disputed or misattributed are set aside, and Stewart states outright that the length of the list, 'far greater than one man could produce in a lifetime,' points either to a sizable workshop and following or to the same inflationary process by which any respectable Italian church eventually claimed a Raphael. Praxiteles had two sons, Cephisodotus the Younger and Timarchus, who had already begun independent careers by around 345-340 BCE, while their father was still working, and who inherited his workshop, its clientele and its wealth on his death, which Andrew Stewart's chronology of the family places shortly before 326 BCE, since by that year Cephisodotus the Younger was already shouldering heavy Athenian naval liturgies as the apparent heir to the family fortune; the brothers specialized in marble divinities and bronze portraits and were themselves working sculptors of reputation, not merely custodians of a name.

Modern scholarship accepts only two works as plausible candidates for Praxiteles' own hand rather than a copy or a workshop product: the Hermes at Olympia, for the reasons already given, and the Mantinea base, a set of three marble relief panels of Apollo, Marsyas and the Muses found reused in the floor of a Byzantine church at ancient Mantinea in 1887, connected by the excavator Gustave Fougeres to a Pausanias passage describing a Praxiteles statue group on a base carved with exactly this scene. Even there, scholarly opinion is divided between Praxiteles' own hand and a trusted assistant working from his design. Beyond those two, a handful of further attributions circulate as informed modern proposals rather than settled fact: the Eubouleus, a head type known in several copies; the Marathon Boy, a bronze ephebe recovered from the sea off Marathon in June 1925 and linked to the Praxitelean workshop chiefly through its soft musculature and exaggerated contrapposto, with no ancient text naming it at all; the so-called Aberdeen head; and the Leconfield Aphrodite, a head at Petworth House. None of these four carries anything like the evidentiary weight of the Hermes or the Mantinea base, and this page treats them accordingly, as argued rather than accepted.

Beyond the Knidia and the Sauroktonos, several further types are attributed to Praxiteles with reasonable confidence from Pliny, Pausanias and the consistency of the copies themselves. The Artemis of Gabii, a draped huntress fastening her cloak, survives in a well-regarded Roman copy in the Louvre, Ma 529. The Eros of Centocelle, a winged youth found near Rome on the Via Labicana, is generally connected to the Eros Pliny says Praxiteles made for Thespiae, the statue Cicero cited against Verres as the reason travelers visited that city at all, and it is now in the Vatican's Museo Pio-Clementino. A separate, nude Eros at Parium on the Propontis is recorded by Pliny as matching the Knidia's fame and, in his words, 'the outrageous treatment which it suffered' - a second version of the same stain story, briefer and less quoted, attached to a different statue entirely. The Resting Satyr, a youth leaning against a tree trunk in a moment of repose, survives in some 115 recorded copies, the best known in the Capitoline Museums in Rome; Nathaniel Hawthorne saw that copy on a visit to Rome in 1858 and built his novel The Marble Faun around it, which makes the Capitoline Faun a rare case in this corpus of an object whose modern literary fame is more precisely documented than the ancient statue behind it.

What unites all of these is the same structural gap as the Knidia and the Sauroktonos: a named type, a Roman-period physical record, and no original.

Nothing by Praxiteles' own hand has ever been offered at auction, and nothing plausibly could be: if the Hermes or the Mantinea base are genuine, both are permanently in Greek national museums, and the Cleveland bronze was a private treaty museum purchase from a dealer rather than a sale at auction, so no hammer price or premium exists to record for it. What does occasionally reach the saleroom is a Roman-period marble descended from a Praxitelean invention, and the market there is thin but not empty. See market.results below for the one substantial, fully documented result found in the sources consulted.