Reading any work by Timothy Barnes is an exhilarating experience. His formidable command of both sources and bibliography never clouds his lucid prose or incisive arguments. He seems to inhabit a world of infinite clarity and irrefutable certainty. The book under review is vintage Barnes. His admirers will not be surprised to find him upending Momigliano's famous complaint that Otto Seeck ‘never believed anything to be authentic if he could help it’ (p. x). Barnes likes exactly what Momigliano disliked. He presents Seeck's scepticism as the very model of how to proceed in the analysis of documents and literary texts for early Christianity. He associates, not unreasonably, his teacher (and mine), Ronald Syme, with this procedure. But the tone of Barnes's writing is much less austere than Seeck's and markedly less oblique and ironic than Syme's. He distributes unqualified accolades to scholars whom he judges enlightened, and he hurls thunderbolts down upon those who have failed to perceive what he regards as the truth. Barnes is ever impartial, and on occasion we can find the same scholar basking in an accolade and writhing under a thunderbolt.

Barnes likes both contradiction and certainty. He starts off his new book with some arresting assertions: St Peter was not crucified at Rome, upside down or not, but was dressed in a combustible garment and burned alive. We are told that the emperor Domitian never persecuted the Christians but simply acquired his reputation as a persecutor because he had a reputation as a bad emperor. The revisionist account of Peter's death depends entirely on John xxi. 18–19, which Barnes acknowledges to be a text that all historians of the early Church have known, but he claims with characteristic confidence that ‘they have virtually without exception misunderstood it’ ( p. 7). The risen Christ is made to prophesy that, although Peter had dressed himself when he was young and went wherever he wished, someone else will dress him as an old man and take him where he does not want to go (this is explicitly glossed as indicating death). Barnes insists that this cannot allude to crucifixion because Peter's clothes would have been removed for the purpose. Hence Barnes thinks he was clad specially in a combustible garment to undergo a fiery end among those Christians whom Nero burned alive on crosses during the persecution that Tacitus described. For Barnes this does not count as a crucifixion but a parody of one. Before we ask how generations of serious scholars could have been so imperceptive, we have to wonder why Barnes denies Peter a traditional crucifixion in the nude and yet allows him a crucifixion in an inflammable tunic. For that matter, Peter could even have been burned upside down. As for Domitian's persecution, Barnes declares that it was ‘widely accepted as historical by later authors both ancient and modern’ (p. 37). This tradition has been good enough for most people to credit, but Barnes claims that it ‘is not attested by any reliable evidence at all’. When does evidence become reliable? Barnes may conceivably be right in believing that Melito of Sardis made up the whole story, but there is no reliable evidence for that either.

Barnes's contrarian streak surfaces conspicuously in his account of the Greek and Latin texts for the martyrdom of Perpetua. He resists Louis Robert's compelling argument for the priority of the Greek version, and claims priority for the Latin because at 16.3 the Greek provides an equivalent for a Latin phrase that he believes ‘is surely a corruption of the original text’. Barnes helpfully puts up warning flags before his more startling revisions: ‘surely’ and ‘presumably’ are among them, and sometimes even stronger expressions such as ‘it is an irresistible inference’ (p. 101). In the case of Pass. Perp. 16.3, the imprisoned martyr-in-waiting asks why comfort could not be provided since she and the other prisoners were going into the arena on the emperor's birthday: ‘nobis refrigerare noxiis nobilissimis Caesaris scilicet et natali eiusdem pugnaturis’. Barnes is certain that Perpetua would have equipped Geta with his official title nobilissimus, although it is hard to imagine why she would have bothered at that moment. He insists that the transmitted text is a corruption of ‘noxiis nobilissimi Caesaris’. His postulated Greek translator is thought not to have detected this error and therefore to have written onomastois katadikois Kaisaros. But there is no way that onomastois could be a translation of nobilissimis. Instead it looks as if the Latin writer misunderstood the Greek.

The follies of historians of early Christianity are compounded when Barnes reports unequivocally that ‘Christianity was recognised as a lawful religion by the emperor Gallienus in 260.’ He writes of ‘incalculable damage’ caused by the failure of everyone in the past to see this great truth. Yet, the sharp-eyed G. E. M. de Ste Croix, had written, very judiciously, in his classic article of 1954 on the Great Persecution of Diocletian that ‘Since the edict and rescripts of Gallienus in the early 260s, recorded by Eusebius, the Christian Church (or churches) had received a certain measure of toleration and recognition – how complete, and according to what constitutional principles, there is fortunately no need to discuss here’ (reprinted in Christian persecutions, martyrdom, and orthodoxy, Oxford 2006, 38). Barnes neither mentions de Ste Croix nor considers the issues of completeness or constitutional principles.This is a pity because de Ste Croix himself was not unlike Barnes in his relentless search for clarity in ecclesiastical history.

The foregoing observations represent but a small part of the immensely rich material in Barnes's new book. They serve only to illustrate the more treacherous paths into which his crusade against error can sometimes lead him. History will always be full of unsolved problems, inauthentic texts and, alas, unwarranted dogmas. Timothy Barnes deserves the gratitude of all historians for his vigilance and erudition in trying to set the record straight.