[This essay has since been published as a peer-reviewed article that can be accessed online here.]
One of the subjects that first got me interested in Classics, even before I began my Ph.D. graduate studies, was the genre of ancient Greek and Latin historiography. I read early on the annalistic histories of Cornelius Tacitus and the historical biographies of Suetonius Tranquillus (I also later completed my Classics M.A. thesis about Suetonius) . Ancient historical prose has a very distinct style, in which the historian often discusses the methodology of his research, the sources he consulted, the differences between multiple traditions about a person or event, and his judgment as an inquirer into past affairs. History, derived from the Greek ἱστορία (“inquiry”), is not merely a narrative about past people, places, and events, but is an investigation that one conducts in the present in order to formulate a hypothesis of what probably took place in the past, based on the available evidence .
The genre of ancient historical prose has key features that are crucial to understanding which works belong to the category and why they are generally more trustworthy than sources that do not. It is not enough for a text to simply talk about things that took place in the past, even when the content deals with real people and locations. A historical text must further investigate and probe these matters, discussing the research process involved, so that it does not merely provide a story, but a plausible interpretation of what took place.
As someone who studies ancient historical writing in the original Greek and Latin languages, it is clear to me that the Gospels are not historical writing. These texts instead read like ancient novelistic literature . In all but Luke, we do not hear anything about the written sources that the authors consulted, and even the author of Luke does not name them, explain their contents, or discuss how they are relevant as sources. The authors of the Gospels do not discuss how they learned their stories or what their personal relations are to these events, and even when John claims to have an eyewitness disciple “whom Jesus loved,” the gospel does not even bother to name or identify this mysterious figure (most likely an invention of the author) . Instead, the Gospels provide story-like narratives, where the authors omnisciently narrate everything that occurs rather than engage in any form of critical analysis. Accordingly, the Gospels all fall short from the criteria that can be used to categorize a piece of historical prose.
So what are these criteria? The ways in which the Gospels diverge from and fall short of the historical writing of their time are perhaps too numerous to exhaustively treat here, but I will discuss ten relevant areas of distinction that are helpful for understanding how historical writing is different.
1. Discussion of Methodology and Sources
Ancient historical works at their beginning (or somewhere else within the body of the narrative) are often prefaced with statements from the author about the period they will be investigating, the methodology they will be using, and the types of sources they will be discussing . None of the Gospels, with the exception of a very brief statement at the beginning of Luke, even come close to following this convention. Furthermore, the opening of Luke is hardly substantial enough to consider it of the same caliber as actual historical prose. As scholar Marion Soards (Oxford Annotated Bible, pg. 1827) notes, “The initial four verses of the book are a single Greek sentence that forms a highly stylized introductory statement typical of ancient historical writings … After this distinctive preface, however, the narrative shifts into a style of Greek reminiscent of the Septuagint.” As such, while Luke mimics some of the conventions of historical writing at the beginning of the gospel, the rest of the narrative reverts into the storytelling typical of the other Gospels.
Consider the very sparse information that the author of Luke (1:1) provides about his written sources (none of whom are identified in any capacity) :
Many have put their hands to compiling a narrative concerning the matters that have been fulfilled among us…
Such a statement is rather vague and does not tell us a great deal about the written sources that the author consulted. We can tell, however, from source analysis that the author of Luke derived a large portion of his material from the Gospel of Mark (another anonymous text even more silent about where it obtained its material). Now, contrast this with the introductory discussion of a historical author–Dionysius of Halicarnassus–in how he lays out the sources that he used for his Roman Antiquities (1.7.1-3):
Thus, having given an explanation for my choice of subject matter, I wish now to discuss the sources that I used when setting out to write my history. For perhaps readers who are already familiar with Hieronymus, Timaeus, Polybius, or any other historian that I mentioned a short while ago as being careless in their works, when they do not find many things in my own writings that are mentioned in theirs, will suspect me of fabricating them, and will want to know where I learned of such things. Lest anyone should hold such an opinion of me, it seems better that I should state in advance what narratives and records I have used as sources. I sailed to Italy at the very time when Augustus Caesar put an end to civil war, in the middle of the one hundred and eighty-seventh Olympiad [30 BCE], and having spent twenty-two years in Rome from that time to the present, I learned the Latin language and familiarized myself with Roman literature, and during all this time I remained devoted to matters bearing upon my subject. Some of my information I learned orally from the most educated men whose company I shared, while the rest I gathered from the histories that were written by esteemed Roman authors–such as Porcius Cato, Fabius Maximus, Valerius Antias, Licinius Macer, the Aelii, Gellii and Calpurnii–as well as other men who are noteworthy. Setting out with these works, which are similar to the Greek annalistic accounts, as my sources, I then put my hands to writing my history.
This is but a snippet of Dionysius’ extensive introduction about methodology, in which he gives an account of how he came upon his sources, how he learned the relevant languages, names his sources, and even explains why some of his readers will not be familiar with the information in his narrative taken from Roman sources that were less known in the Greek world. The Gospel of Luke does not even come close to this level of historical rigor and the other Gospels are even less substantial.
What is even a greater problem with the Gospels’ historical reliability, however, is not their failure to cite the written sources that they consulted about the life of Jesus, but rather their fulfillment of scripture citations derived from the Old Testament. In the Gospel of Matthew, for example, the Old Testament is frequently cited regarding the fulfillment of “prophecies” about Jesus. As NT scholar Bart Ehrman (“Matthew’s Fulfillment of Scripture Citations”) explains:
What is perhaps most striking about Matthew’s account is that it all happens according to divine plan. The Holy Spirit is responsible for Mary’s pregnancy and an angel from heaven allays Joseph’s fears. All this happens to fulfill a prophecy of the Hebrew Scriptures (1:23). Indeed, so does everything else in the narrative: Jesus’ birth in Bethlehem (2:6), the family’s flight to Egypt (2:14) Herod’s slaughter of the innocent children of Bethlehem (2:18), and the family’s decision to relocate in Nazareth (2:23). These are stories that occur only in Matthew, and they are all said to be fulfillments of prophecy.
These citations are not to historical sources, but to Jewish literature that had been written centuries before Jesus. These are literary sources, therefore, rather than historical sources. Because the author of Matthew shaped so much of his narrative around earlier Jewish literature, many NT scholars doubt the historical reliability of these stories that he relates about Jesus, which are based on the Old Testament. M.D. Goulder in Midrash and Lection in Matthew, for example, casts doubt out on the historical reliability of Matthew’s infancy narrative–including Jesus’ descent from David, his birth in Bethlehem, and Joseph’s flight to Egypt–due to the fact that these stories were probably invented by the author of Matthew to draw connections to earlier Jewish scripture, such as Isaiah 11:1, Micah 5:2, and Hosea 11:1 .
2. Internally Addressed and Analyzed Contradictions among Traditions
Contradictions among sources are inevitable when undertaking historical analysis, whether the author be Pagan or Christian. For this reason, as I explain in my essay “Bible Contradictions: Why Are They There? What Do They Entail?,” such contradictions are not ipso facto a major reason why I am distrustful of the Gospels. Rather, it is the way that the Gospels treat contradictions that makes them less credible.
Consider the well-known contradictions between the accounts of Jesus’ birth in the gospels Matthew and Luke . Luke 2 has Jesus’ family travel from their hometown in Nazareth to Bethlehem, due to the Census of Quirinius (6 CE). There is no room at any inn, so Jesus is born in a manger. After waiting the appropriate time required by Jewish law, Jesus’ parents take him to the Jewish Temple, perform the rituals surrounding a male childbirth, and then return to Nazareth.
Matthew’s account (chapters 1-2) is very different. In Matthew, Joseph and Mary are already living in Bethlehem, with no mention of a journey from Nazareth. Jesus is born during the reign of King Herod (who died in 4 BCE), which places the account about a decade before the account in Luke. Rather than travel from Nazareth to Bethlehem, a reverse sequence of events takes place. After Herod orders the slaughter of all the male infants in Bethlehem (an atrocity that no outside historian corroborates), in the hope of killing the future King of the Jews, Joseph is warned by an angel to flee into Egypt in order to escape the slaughter (an excursus of which Luke makes no mention). After Herod’s death, an angel tells Joseph that he can return, but since Archelaus is ruling in Judea, Joseph does not return to his home in Bethlehem, but instead withdraws into Galilee and moves to the town of Nazareth.
Apologists can twist themselves in pretzels trying to reconcile these contradictions, but what is important to note is that we have strong differences between these two versions of Jesus’ birth. This is not unprecedented for other historical figures. The historical biographer Suetonius notes in his Life of Caligula (8.1-5) that there were different versions of the emperor’s birthplace:
Gaius Caesar, surnamed Caligula, was born the day before the Kalends of September [August 31st], during the consulship of his father Germanicus and Gaius Fonteius Capito [12 CE]. Contradicting sources have made the place of his birth uncertain. Gnaeus Lentulus Gaetulicus writes that he was born at Tibur, whereas Pliny [the Elder] says that he was born among the Treveri, in a village named Ambitarvium, above the confluence of the Rhine and the Moselle. Pliny further adds as evidence alters located there, bearing the inscription: ‘For the delivery of Agrippina.’ Verses that were in circulation shortly after he became emperor state that he was born in the winter-quarters of the legions: ‘Born in camp and weaned amidst the arms of his country, he was already a sign of future imperium.’ I myself have found it published in the acta diurna that he was born at Antium. Pliny has discredited Gaetulicus, on the grounds that he falsified his account through flattery, so that he could exalt the praises of a young and vainglorious prince who was even from a city sacred to Hercules; and that he made this lie with greater confidence, since there had actually been a son born to Germanicus at Tiber, nearly a year earlier, who was also named Gaius Caesar (concerning whose lovable innocence and premature death I have already spoken about above). Accurate chronology disproves Pliny. For the historians who committed the reign of Augustus to memory agree that Germanicus was not sent to the region of Gaul and Germany until the close of his consulship, when Caligula was already born. Nor does the inscription on the alter in any way confirm the judgement of Pliny, since Agrippina twice gave birth in that region, and any childbirth, regardless of sex, is called ‘puerperium.’ Indeed, from ancient times people called girls ‘puerae,’ just as they called boys ‘puelli.’ There is also a letter of Augustus that survives, addressed to his granddaughter Agrippina just a few months before he died, which is written about the Gaius in question (since there was not any child of that name who was still alive by that time), in which he states: ‘I arranged yesterday for Telarius and Asillius to bring your boy Gaius, should the gods be willing, on the fifteenth day before the Kalends of June [May 18th]. I am sending with him a physician from my staff of slaves, and I wrote to Germanicus that he can keep him, if he wishes. Farewell, my dear Agrippina, and take care that you come to your Germanicus in good health.’ I think that it is abundantly clear that Caligula could not have been born in a place, to which he was first taken from Rome when he was almost two years old. This letter also diminishes the evidence of the verses, which in any case are anonymous. The sole source that remains, therefore, is the public record stating that he was born at Antium, which we must accept as the only remaining testimony. Furthermore, Caligula always loved Antium above all other places and preferred it as none other than his own native soil, and he is even said to have considered transferring the seat and capital of the empire to Antium, when he had grown weary of Rome.
Here, Suetonius acknowledges that there is a contradiction, but as a historical author he instead engages in a rigorous analysis of the various forms of evidence, ranging from the works of previous historians, to inscriptions, to personal letters, to public records, in order to get to the bottom of the discrepancy. He discusses his sources and methods to give context to the conclusions that he has reached.
Notice that this problem is addressed consciously within a narrative, rather than between narratives by two authors who give their own versions of events without any discussion of sources or method . Suetonius, as a historical author, is interpreting events based on evidence, rather than telling a story as religious propaganda. Furthermore, Suetonius’ account is far more plausible, as there would be little reason to invent Caligula’s birthplace in Antium, whereas both Matthew and Luke almost certainly invented Jesus’ birth in Bethlehem, although he was known to be from Nazareth, to fulfill the expectations that the Messiah would be born in Bethlehem.
The ways in which historical sources treat contradictions between traditions is very different from that of the Gospels, and we can see clear methodology in the former category and uncritical, omniscient forms of narration in the latter .
3. Authorial Presence in the Narrative
Notice in the two examples above how both Dionysius and Suetonius have active roles in the narrative, as historians who are interjecting their own voice, in order to discuss their sources and relation to events. We learn details of how Dionysius traveled to Rome and learned Latin, and how Suetonius was acquainted with Augustus’ own letters. The Gospel authors are silent about their identities and give context about their relation neither to their sources nor to the events they contain. The Gospel narratives instead read like novelistic literature, told from a camera-like perspective, which omnisciently follows around the characters with minimal methodological analysis. This third person style of narration further casts doubt on whether the Gospels’ authors are relating eyewitness experiences. As Irene de Jong (Narratology & Classics: A Practical Guide, pg. 17) explains:
It is an important principle of narratology that the narrator cannot automatically be equated with the author; rather, it is a creation of the author, like the characters.
A major characteristic of historical prose, however, is that the role of author and narrator are (generally speaking) the same, which means that ancient historians frequently interject their own voice into the narrative using the first person . Even among ancient historical works in which the author does not specifically give his name within the narrative, historians very frequently discuss their relation to the events that they are analyzing. For example, even though he does not name himself in his Histories, the historian Tacitus (1.1) describes his career and relationship to the persons and events he is documenting:
I myself was not acquainted with Galba, Otho, or Vitellius, either by profit or injury. I would not deny that my rank was first elevated by Vespasian, then raised by Titus, and still further increased by Domitian; but to those who profess unaltered truth, it is requisite to speak neither with partisanship nor prejudice.
Here, Tacitus discusses, out of the individuals that he is writing about, which ones he knew and which one he was more distanced from. He clearly explains his role during the time period and his relation to the events within it, so that the narrator’s identity and background are clearly understood with regard to the events that he is investigating. This is a hallmark of history as a genre, which is an investigation in the present of past events, rather than a mere story set in the past.
4. Education Level of the Audience
While a high school-level education in history is universally taught to inhabitants of modern Western nations (still not as well as I would like), historical writing was very exclusive in antiquity. In order to fully evaluate and appreciate historical prose, one had to be educated, literate, trained in oratory, and skilled at critical thinking. Authors writing to such an audience had to demonstrate their research ability, credentials, and methodology. As scholar Pheme Perkins (Oxford Annotated Bible, pg. 1743) explains, “Greco-Roman biographies were addressed to a social and literary elite, which may explain why the Gospels, addressed to a much broader audience, do not match them very closely.”
The Gospels, in contrast, are written for a far less educated and critical audience. Far from the refined prose of Greek historical writing, the Gospels are written in a low language register in the Koine dialect. For anyone who reads ancient Greek, the difference in quality between a historian like Thucydides versus the authors of the Gospels is on par with comparing an elevated English work like Shakespeare’s Hamlet to a far simpler text like J.R.R. Tolkien’s The Hobbit . Historical writing was simply far more complex and rhetorically polished (as well as much more critical and analytical), whereas the Gospels read as basic stories, which were taught to encourage the faith of people who probably already believed and trusted in Christianity.
5. Hagiography versus Biography
Rather than read as the unmitigated praise of a saint who can do no wrong, ancient historical works and historical biographies were far more critical of their subjects, whom they analyzed less one-dimensionally and more as complete persons. Even for a popular and well-liked emperor like Augustus, his biographer Suetonius in his Life of Augustus still did not hold back from describing Augustus’ acts of adultery (69) and lavish behavior (70). Good historians are concerned with telling the past as it really is rather than just heaping praise upon individuals as propaganda.
The Gospels, in contrast, are not historical biographies but can be more aptly described as “hagiographies,” written in unquestioning praise of their messianic subject. Although the genre of Christian Lives of Saints developed after the Gospels, they can still be regarded as hagiographical in that they function as laudatory biographies, praising the subject, rather than as critical biographies. As a good representation of the scholarly consensus about the rhetorical aims of the Gospels, the Oxford Annotated Bible (pg. 1744) explains, “Neither the evangelists nor their first readers engaged in historical analysis. Their aim was to confirm Christian faith.” Such works, written for an audience of converts, are not chiefly concerned with being critical or investigative, but rather serve the religious agendas and ideologies of the communities that produced them.
6. Signposts about Authorial Speculation
Even when they dutifully followed the sources available, ancient historians frequently did not know the exact words spoken by individuals in famous speeches or the exact order in which things had taken place in past events. In order to provide elegant rhetorical prose, however, creative liberties had to be taken on the part of the author to retell these dialogues as they plausibly could have taken place. This does not entail direct lying on the part of the author, since the speeches were written to represent plausible versions of the original and historians would often signal that the words were approximate. The historian Thucydides, for examples, prefaces in his History of the Peloponnesian War (1.22):
Now, as much as particular persons gave speeches, either entering the war, or when it was already taking place, it has been difficult for me to remember precisely the exact words that were spoken, either from those that I heard myself, or from those that I was informed of by others. And so, my practice has been to make each speaker say what I regard as the most suitable words that the occasion demanded, while adhering as close as possible to the general sense of what was actually spoken. As for the events of the war, I did not think it fitting to write what I learned from the first source at hand, nor what I speculated to be true, but only those things for which I myself was present, or for which I had inquired each detail to the best of my ability from others. And even these matters have been difficult to learn, on account of the fact that those present did not always say the same thing about the events they witnessed, either because they had imperfect memory, or because they were partial to one side or the other.
Beyond overt methodological statements this, there are other constructions that historians could use to signpost authorial speculation, using phrases such as δοκεῖ μοι (“it seems to me”), as well as constructions that employ paraphrase, such as ἔλεγε τοιάδε (“he spoke words like these”). We have no such honesty and signposts in the Gospels. The Gospels are not even written in the same Aramaic that Jesus spoke. The authors of Matthew and Luke may have had the diligence to copy certain saying from an earlier Q source–a hypothetical collection of sayings reconstructed by modern scholars –but even then they do not signal that they are obtaining this material from an earlier source nor do they specify how this source would be trustworthy.
Furthermore, the Gospels predominantly employ direct speech, where they script dialogue verbatim, rather than indirect speech, which was more frequently used by ancient historians to signal when they did not know the exact words of their subjects. As Richard Pervo (“A Nihilist Fabula: Introducing the Life of Aesop,” pg. 81) explains, the frequent use of dialogue in narrative is “a mark of novelists” rather than historians . John is the least reliable of the Gospels, in which Jesus gives whole speeches in a prose style that is very different from the short, formulaic sayings and parables of Jesus in the Synoptic Gospels. In short, the author of John probably made up a lot of Jesus’ sayings and yet did not signpost his speculation in the same way that a historian like Thucydides did.
7. A Greater Degree of Authorial License
When I discuss how ancient historians and historical biographers are far more prone to cite their sources, note contradictions between traditions, interject authorial judgements, and signpost speculation, one impression that I do not wish to create is that they still did not take a great amount of creative liberties in fashioning their narratives. The standards for what constituted historical writing in antiquity were very different from those that are used today in professional historiography. Ancient historians like Tacitus, for example, frequently imagined the speeches given at key sections of the narrative, and likewise characterized their historical subjects in ways that are highly dramatic and conjectural. In fact, Classicist Holly Haynes in the The History of Make-Believe: Tacitus on Imperial Rome has even described many of Tacitus’ literary techniques as “make-believe.” As Haynes (pg. 28) explains:
The inescapable and regrettable fact about ancient historians, according to much of the scholarship, is that they made things up. By “making things up,” we commonly mean “falsifying” or even “lying,” both of which are antithetical to the scholar of ancient history and his or her project of arriving at the truest possible account of the past. Speaking generally, this view is prevalent throughout the scholarship of ancient history and historiography alike, as the latter does little to combat the dyarchic structure of words and deeds embedded in the discourse of the study of antiquity. Rather, the emerging discipline of historiography has emphasized the rhetorical richness of ancient history [and has] urged us to view the “make-believe” in its own right…
While this proviso can be made about historians like Tacitus, however, it should be noted that it applies to a far greater extent in the case of the Gospels. As previously discussed, the authors of the Gospels do not even signpost speculation, nor do they cite or analyze their historical sources. Likewise, the Gospels include many more instances of direct speech and dramatic dialogues in their narratives, which their authors must have frequently imagined and invented. This is especially true in John, where Jesus engages in long discourses, distinct from the short, formulaic sayings in the Synoptic Gospels, which critical scholars have long recognized are probably not authentic words spoken by Jesus. As such, there is a far greater degree of authorial license in the Gospels, even if ancient historians and historical biographer engaged in creative liberties that would exceed the boundaries of modern historiographical techniques.
8. Independence versus Interdependence
One thing that amazes me as a Classicist is just how interdependent the Gospels are upon each other. Matthew borrows from much as 80% of Mark’s material, and Luke borrows from 65% of the material in the earliest gospel. While John does not follow the ipsissima verba of the Synoptics, the author is still aware of the same basic skeleton and is probably familiar with one or more of the earlier gospels (as shown by scholar Louis Ruprecht in This Tragic Gospel). In fact, I know of almost no other texts from antiquity that share as much material as the canonical Gospels . While it is true that the Gospels are not entirely derivative of each other, in that they do have some independent sources not copied from one another (such as the M-Source for Matthew, or the L-Source for Luke), the interdependence that is seen between the canonical Gospels is still far greater than what is typical of ancient historians. This is very bad for historical reliability, since independent attestation can be very helpful for verifying historical claims, and yet the Gospels all fail this criterion miserably.
The same is not true for ancient historical works. Consider just the four most extensive sources that we have for the life of the emperor Tiberius: Paterculus, Tacitus, Suetonius, and Cassius Dio. All four authors obtain their material from a much broader range of sources, rather than simply copy from each other, they write in a far more diverse range of styles, and yet they independently corroborate each other’s claims. Paterculus is an eyewitness historian writing a brief universal history of the known world, which concludes with Tiberius’ military campaigns (which he himself took part in). Tacitus is writing a year-by-year annalistic history of the Julio-Claudian period, but rather than just copy Paterculus for Tiberius, he instead draws from a whole array of authors who wrote during the Julio-Claudian period, as well as public records and other sources. Suetonius, who is writing almost at the same time as Tacitus, does not produce a carbon copy of his Annals, and very likely does not rely on Tacitus as a source at all (as argued by Tristan Power in “Suetonius’ Tacitus”), but instead writes a historical biography, not in chronological order, that is very different from the earlier sources in its style. And yet, Suetonius independently corroborates the claims of these earlier authors. Dio, who is writing a full history of Rome from its foundation in Greek prose, a different language than the earlier Latin sources, has only one part of his massive history dealing with Tiberius. Dio probably used Tacitus, but also many other earlier sources, and writes his own unique narrative that is still consistent with the other independent sources.
For the life of Tiberius we have a wide array of independent sources corroborating each other, whereas for Jesus we have sources that are all copying and redacting one another, not providing as much independent information or research, but instead repeating and adding to growing legends.
One proviso that should be noted is that ancient historians do not always corroborate each other on every claim, and there are likewise occasionally contradictions between their narratives. The historians Polybius and Livy, for example, sometimes contradict each other on the details of Hannibal crossing the Alps, even when they were drawing upon a common historical source. This criterion does not imply, therefore, that ancient historians and historical biographies always drew upon independent sources, or never contradict each other about the details of an event. Nevertheless, there is a far greater degree of independent corroboration seen in ancient historical writing than what can be found in the Gospels, which diminishes the latter’s historical reliability.
9. Miracles at the Fringe versus the Core of the Narrative
Simply because ancient historical authors conducted more rigorous research does not entail that they were skeptical of the supernatural. Unbelievable stories still crop up in the writings of Greek and Latin historians, ranging from Herodotus (8.36-39) claiming that, when the Persians attacked the Temple of Apollo at Delphi, its armaments came alive of their own accord and defended the temple, with lightning bolts suddenly raining down upon the Persian army, to Josephus (BJ 6.5.3) claiming that a cow gave birth to a lamb as an impending sign of Jerusalem’s destruction, to Suetonius claiming (Gal. 1.1) that, prior to Nero’s death, lightning struck the Temple of the Caesars and simultaneously decapitated all of the emperors’ statues, even dashing the scepter from the hand of Augustus’ statue (quite a remarkable feat for lightning alone). Of course, I do not believe such stories and their placement in these narratives does make me less trustful of their authors. But fortunately, for ancient historical authors, these ridiculous tall tales are usually at the fringe rather than the core of the narrative .
The Gospels, in contrast, simply narrate unbelievable claim after unbelievable claim about a man who can feed whole crowds with one tuna sandwich, cause dead saints to rise from their graves, himself resurrect from the dead, and then fly into space in broad daylight. These unbelievable tall tales make up the bulk of the narrative. As philosopher Stephen Law points out, following the principle of contamination, the frequency of these unbelievable stories casts doubt on even the mundane details in the narrative. It is not as if their genre is relatively historical, but merely peppered with a few miracles here and there. Rather, the Gospels are entirely fantastical and legendary. The Gospels are so contaminated by unbelievable claims that they should be treated as generally untrustworthy until there is good reason for believing specific details.
Another thing that should be noted is that, while ancient historians occasionally report miracles, they often use specific grammatical constructions that distance themselves from affirming the stories and make clear that they are only reporting claims. The historian Titus Livy, for example, in reporting some of the miracle stories of regal Rome, frequently uses terms like ut dicitur (“as it is said”) or ferunt (“they claim”) to specify that he is not endorsing these claims, but only recording that they were made. One such example is when Livy (1.39) discusses the tale of how the king Servius Tullius’ head, when he was a child, caught on fire while he was sleeping, but did not harm him, as it was a sign that he would be a future king. Livy’s careful use of the verb ferunt (“they claim”) indicates that he is distancing himself from gullibly believing in this fable. The Gospels, in contrast, just throw out miracle after miracle, asking us to believe every single one of them, in a manner that presumes far less critical thinking on the part of the reader.
10. Important Characters and Events Do Not Disappear from the Narrative
Prior to entry in my Classics M.A. program, I wrote as a writing sample a paper about the Roman prefect Sejanus and his alleged conspiracy against the emperor Tiberius in 31 CE. Both Tacitus and Dio invest extensive portions of their narratives introducing Sejanus and explaining the steps he took in gaining power under Tiberius. Whatever Sejanus was planning, it did not come to fruition, as he was executed by Tiberius in 31 CE. Part of the accusations levied against Sejanus was that he had many allies in the Roman Senate who were helping him in the conspiracy.
Now, imagine if, after Sejanus’ death, there was no aftermath or follow-up and the narrative merely moved on to another subject. This sequence of events would not at all be logical and would leave many questions unanswered. Instead, both Tacitus (book 6) and Dio (book 58) spend a considerable amount of narrative space discussing the senators who were accused and condemned for being co-conspirators with Sejanus. This makes logical sense, as the event and its instigator were were both of a very important nature and we would not expect that they would suddenly disappear from a narrative in which they played crucial roles.
And yet in the Gospels, earth-shaking events take place that then receive no follow-up and strangely disappear once they have played their symbolic role in the narrative. Take the Gospel of Matthew, for example. While Jesus is being crucified, the sky grows dark for three hours at midday (27:45). Next, Jesus’ death (27:51-53) causes an earthquake that rips the curtain in the Jewish Temple in twain. The earthquake likewise opens the tombs of the saints, from which dead people resurrect and then appear throughout Jerusalem. This is an extraordinary event, indeed, and yet there is no follow-up in the Gospels or Acts of how the city was affected by this. Then, the Jewish authorities are so worried that Jesus’ tomb will be found empty, lest people believe that a miracle has occurred (as if the midday darkness and the ripping of the Temple curtain weren’t already convincing enough), that they convince Pilate to station guards at the tomb. When the guards are foiled, however, and Jesus’ body is found missing, the Jewish authorities claim (28:11-15) that the disciples stole the body. Grave robbery was a capital offense in ancient Judea, and yet there is no follow-up prosecution of the disciples for this charge, even when they are brought to court on other issues. Furthermore, what happened to Joseph of Arimathea? His tomb was the one that was supposed to remain occupied, and yet, when it is found empty, he is not even questioned on the matter. The Jewish authorities had gone to great lengths to ensure that Jesus’ body did not go missing, and yet, when Jesus is claimed to have risen, they do not even undergo an investigation into the circumstances.
This sequence of events does not logically make sense, if the Gospels were narrating actual historical events. Instead, the Gospels are reporting fantastical legends, where people act in bizarrely symbolic ways and do not rationally respond to what has taken place. For this reason, as I explain in my essay “Let’s Presuppose That Miracles Happen: The Gospel Resurrection Stories Are Still Unworthy Of Belief,” the Gospels are not believable accounts, even in a universe where miracles actually happen. Actual historical writing is not so abrupt, and reasonable consequences occur after events that are important to the sequence of the narrative .
11. Even Good Historical Texts Should Not Always Be Trusted
An additional final point, which is not so much a criterion of distinction, but rather a reason why even the lack of these differences would still not save the Gospels, is that not even the real historical works that we have from antiquity should be taken at face value. Their authors still have their biases, they still speculate over past events, they sometimes make errors about dates and places, they had limited evidence afforded to them, and they still report a number of unbelievable claims.
I certainly do not trust miracle claims, simply because a historical text records them. Many ancient historians report miracles that are far better attested and independently corroborated than those in the Gospels. The historians Tacitus (Ann. 6.20), Suetonius (Gal. 4), and Cassius Dio (64.1) all independently corroborate that the emperor Tiberius used his knowledge of astrology to predict the future emperor Galba’s reign. These same historians likewise independently corroborate that Vespasian could miraculously cure the blind and crippled (Tacitus, Ann. 4.81; Suetonius, Vesp. 7.2; Dio 65.8). As I explained above, Tacitus, Suetonius, and Dio are not simply copying each other, whereas the Gospels are heavily dependent upon each other for information. This does not entail that the Pagan miracles are true, but it does show that they were not invented by these historians and most likely derive from an earlier common source (I think that most of these stories go back to roughly contemporary claims about miracles when Galba and Vespasian became emperors, which I elaborate on further in my paper “The Propaganda of Accession of the Roman Emperor Galba”). In contrast, since the Gospels copy from each other, many of their miracles can very easily have no earlier source, and when one earlier gospel author invented a miracle, a later gospel could merely pass it along in a game of telephone.
Likewise, an additional point is that virtually no serious Classical scholar that I am aware of argues that historians can use Pagan texts to prove miracles. As I explain in my essay, “History and the Paranormal,” professional historians normally bracket such claims as philosophical and theological questions that extend beyond the scope of ordinary historical inquiry. It should also be noted that Classics and New Testament Studies deal with the same historical period, working with the same languages, and using the same historical methodology. If Classicists are not in the business of seeking to prove miracles using ancient texts, then this provides a good outside model for the limitations of New Testament Studies. As such, I think it is completely fair to bracket the supernatural claims in the Gospels as matters that extend beyond the scope of history.
The main point to take away from my analysis of the criteria above is that the Gospels certainly do not measure up to the high historiography and historical biographies of antiquity. Many of my Classics professors who specialize in such texts, when they read the Gospels, comment on how much more rudimentary and story-like their narratives are compared to the researched and analytical characteristics of historical writing. Even Luke only has a few brief lines at the beginning that mimic historical prose, before jumping into pure hagiography like the other Gospels.
Ancient historical texts are some of my favorite works from antiquity for their sophisticated writing style, elaborate research, and intellectual rigor in investigating past events. I cannot say the same for the Gospels, although I do think they provide interesting symbolism and allegories as novelistic texts, and are also complex works of religious scripture. After analyzing the Gospels under the historiographical criteria that I discuss above, however, they must be placed in a different literary genre than the actual historical works of antiquity.
A final note about modern historical criticism is that such authors of ancient historical prose, who demonstrate their research, have independent corroboration, discuss their methodology, and reach conclusions through critical investigation, should generally be trusted, until proven otherwise. In contrast, ancient novelistic and religious texts, such as the Gospels, that are packed full of legends and religious propaganda, should not be given the benefit of the doubt, until there is good reason for overcoming their overall unreliability in order to trust a specific detail. I do think that there are some precious kernels of truth in the Gospels, but given their overall lack of critical analysis and the creative liberties taken by their authors, I do not think that we can take many of their stories at face value .
 For the purposes of this article, the term “historical writing” can refer to both ancient historiography and historical biography. To be sure, historiography and biography were not the same genre in antiquity, as the former was based on the history of a broader period or event, while the latter was based on the life of an individual. Nevertheless, the two can both be sufficiently described as “historical writing,” especially since many of the narrative conventions between the two are similar. Plutarch in his Parallel Lives, for example, compares his source material and makes historical judgements in a manner very similar to Dionysius of Halicarnassus in his Roman Antiquities, even if Plutarch wrote historical biographies while Dionysius wrote a Roman history. It should be noted, however, that not all Greco-Roman biographies in antiquity were historical biographies–of the sort of Plutarch and Suetonius–since there were also many less critical biographical texts–such as the Alexander Romance and the Life of Aesop–which would include far more novelistic and mythical elements. On this point, see my essay “Greek Popular Biography: Romance, Contest, Gospel,” in addition to my essay “Are the Gospels Ancient Biographies?: The Spectrum of Ancient Βίοι,” as well as the analysis in footnote 3 below.
 The Greek word ἱστορία (“inquiry”) was first used by Herodotus (1.1) in the 5th century BCE to refer to a genre of historical writing. Previously the word had been used in Greek philosophical literature to refer to scientific inquiry. As François Hartog (“The Invention of History,” pg. 394) argues, Herodotus innovated the use of the term ἱστορία in writing about past events as substitute for the Muses in epic poetry. Rather than derive his knowledge of the past from divine inspiration, Herodotus instead conducts an “inquiry” into past events through the critical use of sources. For further discussion on this point, see my essay “History and the Divine Sphere: Homer, Herodotus, and Thucydides.”
 The comparison of the Gospels and Acts (along with other early Christian and Jewish literature) to the ancient novel has been made by several NT scholars, including Ronald Hock (ed.) in Ancient Fiction and Early Christian Narrative, Jo-Ann Brant (ed.) in Ancient Fiction: The Matrix of Early Christian and Jewish Narrative, Marília Pinheiro (ed.) in The Ancient Novel and Early Christian and Jewish Narrative: Fictional Intersections, and Richard Pervo in Profit With Delight: The Literary Genre of the Acts of the Apostles. Michael Vines has also compared the Gospel of Mark specifically with the genre of the Jewish novel in The Problem of Markan Genre: The Gospel of Mark and the Jewish Novel.
The comparison of the Gospels to the ancient novel is often contrasted with the comparison to Greco-Roman biographies. The Gospels have been compared to Greco-Roman biographies by Richard Burridge in What Are the Gospels?: A Comparison with Graeco-Roman Biography and Dirk Frickenschmidt in Evangelium als Biographie: die vier Evangelien im Rahmen antiker Erzählkunst. It should be noted, however, that there was a great diversity of biographical literature in antiquity, which means that not all Greco-Roman biographies are similar in terms of their style and methodology. As Tomas Hägg (The Art of Biography in Antiquity, pg. 155) argues, regarding Burridge’s study (which compares the Gospels to a canon of ten ancient biographical texts):
There is a great diversity within each of the two groups, the four gospels and the ten ancient biographies; and it is this very diversity, we should note, that makes it possible always to find a parallel in one or several of the ten Lives for each feature occurring in one or more of the gospels. What is proven is that the investigated features of the gospels are not unique in ancient biographical literature; but no control group is established to show which features may be regarded as significantly typical of this literature.
This article does not dispute the comparison of the Gospels to Greco-Roman biographies, due to the fact that there were many novelistic biographies in antiquity–such as the Alexander Romance and the Life of Aesop—which overlapped with both the ancient novel and the Greco-Roman bios. Because of this, the Gospels can still be categorized as novelistic writing while still having biographical elements. In fact, the genre of the Gospels has been compared to the novelistic Life of Aesop by Lawrence Wills in “The Life of Aesop and the Hero Cult Paradigm in the Gospel Tradition” (which I review here), as well as Whitney Shiner in “Creating Plot in Episodic Narratives: The Life of Aesop and the Gospel of Mark” (which I review here). I have also made a similar comparison to another ancient novelistic biography–The Certamen of Homer and Hesiod–in my essay “The Certamen of Homer and Hesiod and the Gospels: Some Comparanda.”
For the purposes of this article, therefore, I classify the Gospels as novelistic biographies distinct from historical biographies, which allows for the Gospels to make use of many biographical conventions, while still lacking critical and historical inquiry.
 For problems relating to the authorship of John, including discussion of the anonymous “beloved disciple” in Jn. 21:24-25, see my article “Why Scholars Doubt the Traditional Authors of the Gospels,” especially footnotes 30, 31, 32, 33, and 34.
 When ancient historical authors do not cite or discuss their sources at the beginning of their works, they frequently cite and discuss them elsewhere in the text. For example, Plutarch’s Life of Alexander does not a list of sources at the beginning of the biography; however, as J. Powell in “The Sources of Plutarch’s Alexander” (pg. 229) explains, “Plutarch cites by name no fewer than twenty-four authorities” elsewhere in the text. Furthermore, Plutarch also cites named authorities in his other biographies. As Maria Schettino (“The Use of Historical Sources,” pg. 417) explains, “In the Lives, Plutarch cites a large number of historians, around 135, of which one hundred write in Greek.” Nevertheless, none of the NT Gospels cite any of their written sources by name, besides books of the Old Testament, which are literary and not historical sources.
Likewise, even when ancient historians do not identify many of their written sources by name, they often discuss their sources anonymously. For example, the historian Tacitus identifies few of his written sources by name in his Annals, with the notable exception of Pliny the Elder in Ann. 1.67; however, Tacitus still engages many of his written sources anonymously in ways that are atypical of the canonical Gospels. For example, Tacitus uses formulas like quidam tradidere (“some have related,” Ann. 1.13), diversa apud auctores (“conflicting accounts among historians,” Ann. 1.81), and secutus plurimos auctorum (“having followed the accounts of most historians,” Ann. 4.57). In Tacitus’ Histories, as well, the author cites several contradicting opinions, about events which had taken place only 30 years prior to the text’s composition (c. 109 CE). As Classicist Clarence Mendell (Tacitus: The Man And His Work, pg. 201) explains:
In the Histories there are sixty-eight instances in which Tacitus indicates either a recorded statement or a belief on someone’s part with regard to something which he himself is unwilling to assert as a fact; in other words, he cites divergent authority for some fact or motive.
Such methodological statements are virtually absent from the canonical Gospels, with the exception of the first few lines of Luke (1:1-4). Likewise, the Oxford Annotated Bible (pg. 1827) notes that these few lines are atypical of the style elsewhere in Luke’s gospel: “The initial four verses of the book are a single Greek sentence that forms a highly stylized introductory statement typical of ancient historical writings … After this distinctive preface, however, the narrative shifts into a style of Greek reminiscent of the Septuagint.”
The source criterion should furthermore not be interpreted as meaning that ancient historians always cited their sources in every instance. Very frequently they do not, and there were no footnotes in antiquity. Nevertheless, the critical analysis of sources was a key feature of historical prose that goes all the way back to Herodotus, who was the first to introduce the genre. As Irene de Jong (Narratology & Classics, pg. 172) explains about the difference between Herodotus and the epic poetry that preceded him:
[T]he Herodotean narrator is clearly indebted to the Homeric narrator. But unlike that narrator, the Herodotean narrator has no Muses to help him, and at times he admits that he does not know something, gives more than one motive for a character’s actions, and reaches the ‘borders’ of his story and is unable to tell what lies outside them … [H]e is present in his own text as a person who travels and talks with informants and, as a histor, compares and weighs sources.
As such, discussion of sources and methodology is considerably more present in ancient historical works, who owe their influence to Herodotus, compared to the canonical Gospels in which it is virtually absent, reflecting a difference in style and genre.
 For every Greek and Latin passage quoted in this article, I have provided my own translation.
 Isaiah 11:1 predicts that the Messiah would descend from King David, but this also served as a good motive for the author of Matthew to invent that Jesus was descended from David. For a discussion of how Matthew’s genealogy of Jesus (Mt. 1:1-17) was likely fabricated by the author using names from the Old Testament, see Paul Davidson’s essay “What’s the Deal with Matthew’s Genealogy?.” It should also be noted that there were several other false genealogies that were invented in antiquity, such as the Roman emperor Galba claiming to be descended from Pasiphaë, the wife of the legendary King Minos of Crete (Suetonius, Life of Galba 2.1), which historians likewise doubt is historical. Micah 5:2 predicts that the Messiah would be born in Bethlehem, but it is far more probable that the historical Jesus was actually born in Nazareth, as explained by Bart Ehrman in “Was Jesus from Nazareth?” and “Bethlehem and Nazareth in Matthew.” Hosea 11:1 is not even a prophecy about the Messiah but simply states, “When Israel was a child, I loved him, and out of Egypt I called my son.” In context, this passage refers to the Jewish Exodus in Egypt, but the author of Matthew later invented an interlude in which Jesus’ father Joseph brought him to Egypt (something none of the other Gospels mention), in order to draw an allusion to this passage. The allusion is further drawn by the legendary patriarch Joseph’s journey to Egypt.
The influence of the Old Testament on the Gospels can even be seen in their source materials, which must have been produced prior to their composition. In fact, through the literary convention known as Midrash, in which NT characters and episodes are designed to mimic OT characters and episodes, we can tell that whole sections of the Gospels’ narratives are derived from imitation of earlier literature.
For example, there are two sets of miracle collections used in the Gospel of Mark, both of which are designed to model Jesus after Moses. As R.C. Symes (“Jesus’ Miracles and Religious Myth”) explains:
Gospel stories about Jesus’ miracles are a type of Midrash (i.e., contemporizing and reinterpreting) of Old Testament events in order to illustrate theological themes. Among the many miracles in Mark’s original narrative, there are two sets of five miracles each. Each set begins with a sea-crossing miracle and ends with a miraculous feeding. He uses this literary construct so his readers will recall the role of Moses leading his people through water towards the promised land, and feeding them with manna from heaven. Jesus does something similar. And with each water and feeding miracle, there is one exorcism and two healing miracles that are to remind readers of the works of the prophets Elijah and Elisha and how Jesus surpasses them. The parallels between events in Jesus’ life to those in the lives of Moses, Elijah and Elisha and others are too close for a coincidence. This points more to constructing religious myths in the gospel for theological reasons, than to reporting historical facts.
Scholar William Telford discusses these miracle collections further in Interpretation of Mark (pg. 18), as does Richard Horsley in Hearing the Whole Story (pg. 106). I also discuss the role of Midrash in Gospel myth-making further, in my essay “Patterns of Myth-Making Between the Lives of Alexander the Great and Jesus Christ.”
This actually means that Mark’s narrative is being built around earlier outlines of Jesus’ miracles (suggesting that even the mundane narrative details may have been invented to narrativize the miracles). But we can tell further that these miracles were themselves based on parallels with the Old Testament, such as the alleged miracles of Moses. That speaks strongly in favor of the hypothesis of legendary development, since we can tell that stories about Jesus were being made up to parallel him with OT figures. It should also be noted that these are pre-Gospel traditions, meaning that we can detect legendary development surrounding Jesus before the Gospels were even written.
Likewise, NT scholar Dennis MacDonald has argued, through mimesis criticism, that a number of the episodes in the Gospels may be based around earlier Greek mythology, particularly episodes in the Odyssey. It should be noted that, while mimesis of OT literature is widely accepted among NT scholars, mimesis of Greek epic is far more controversial. If MacDonald is correct that a number of characters and episodes within the Gospels are based on earlier Greek literature, however, then this would also cast doubt on whether such content is derived from actual historical events.
 For an in-depth analysis of how the Matthean and Lukan accounts of Jesus’ birth contain irreconcilable chronological contradictions, in addition to historical implausibilities, see ancient historian Richard Carrier’s “The Date of the Nativity in Luke.”
 It should also be noted that many of the contradictions in the Gospels are not accidental discrepancies, but rather intentional alterations and differences between the authors. NT scholars have long known through redaction criticism that many of the changes that the later Gospels make to Mark, for example, not only borrow Mark’s material, but also change the order of events. As NT scholar L. Michael White (Scripting Jesus, pp. ix-xi) explains:
[E]ach of the Gospel authors has woven such episodes into the story in distinctive ways, changing not only the running order of the narrative, but also certain cause-and-effect relationships within each story. For example, in the Synoptics–especially the Gospel of Mark–it is the cleansing of the Temple that serves as the immediate cause of Jesus’ arrest and execution. In the Gospel of John there is no connection between these events, as the cleansing is two full years earlier. In contrast, for the Gospel of John the immediate cause of Jesus’ execution is the raising of Lazarus (11:38-44), an event never discussed in the Synoptics. Thus, the story works differently in each of these versions because of basic changes in narrative…
[T]he Gospel writers … reshape and recombine both old and new episodes, teachings, and characters that circulate[d] about the central figure, Jesus. Storytelling was essentially an oral performance medium in the ancient world, even when those stories were eventually written down. Thus, any particular performance might highlight different elements in the light of the circumstances of the author and the audience … Different actors, different settings, different periods of history–all of them create a different climate. Even when a script gets written down, the performances and emphases can change or be reinterpreted…
In this sense, the authors were playing to an audience. They are ‘faithful’ in that they were trying to instill and reaffirm the faith of those audiences, albeit sometimes in new and different ways. Even so, the stories are just that–stories–and not ‘histories’ in any modern sense.
Because of this rearrangement in material, therefore, which often includes re-ordering of events, we cannot assume that any individual Gospel gives an accurate chronological narrative. There are simply too many discrepancies between the texts.
Likewise, redaction criticism can reveal legendary development and other forms of embellishment between earlier and later texts. For example, even Christian scholars, such as James McGrath, have acknowledged that Jesus’ burial is embellished in the later Gospels–Matthew, Luke, and John. As McGrath (The Burial of Jesus: History and Fiction, pg. 70) explains:
Our earliest account of Jesus’ burial, the Gospel of Mark, records a fundamental truth that later Christian authors tried desperately to ignore: Jesus’ disciples were not in a position to provide Jesus with an honorable burial. Mark tells us that a pious Jewish leader named Joseph of Arimathea made sure that Jewish law was observed, and, learning that Jesus had died, got the permission to take the body and bury it.
However, McGrath points out that the later Gospels make a number of changes to Mark’s version of the story, in order to exalt and embellish Jesus’ burial. Luke (23:53) adds the detail that the tomb had never been used before, making the burial more honorable, and Matthew (27:59-60) adds both the detail that the tomb was unused and that it was even Joseph’s own tomb. John (19:39-41) even further adds the detail that Jesus was anointed with 75 pounds of myrrh and aloes before his burial, even when this contradicts Mark 16:1, which implies that Jesus was not anointed before his burial. John also adds the detail that Jesus was buried in a garden. These kinds of embellishments suggest that the tradition of Jesus receiving a private burial in a tomb, which had never before been used, is probably a later embellishment (McGrath, for the record, supports the view that Jesus was buried in a common, criminal cemetery). Because of this, historians can thus doubt the accounts of Jesus’ burial in Matthew, Luke, and John.
It should also be noted that, because the later Gospel authors derived so much material from Mark (which itself is based on earlier Greek pericopes and oral traditions), it casts strong doubt on whether any of their narratives are based on “memories.” Instead, the borrowing and redaction of materials suggests that the Gospels were stitched together based on material that had been circulating for some decades, which was likewise redacted at multiple stages of composition. Because of this, we have to treat the Gospels as received material, rather than first-hand accounts.
For a discussion of why Markan priority is the dominant view among mainstream New Testament scholars, see Michael Kok’s “Markan Priority or Posterity?.”
 Some apologists have sought to excuse the fact that the Gospels do not mention contradictions and varying reports between their sources, by arguing that the Gospels were composed too close to the life of Jesus for divergent accounts to have emerged. Craig Keener (The Historical Jesus of the Gospels, pg. 128), for example, claims, “The Gospel writers do not identify specific sources … probably at least partly because they discuss events of a recent generation of which sources have not yet greatly diverged.” This argument is poorly supported, however, by trends in historical biographical literature written only a generation or two after the subject’s death.
As the Oxford Annotated Bible (pg. 1744) explains about the dating of the NT Gospels, “Scholars generally agree that the Gospels were written forty to sixty years after the death of Jesus.” There is an abundance of historical biographical literature written within this same timespan, however, that nevertheless spends considerably more space citing contradictions and varying reports between sources. Suetonius’ Twelve Caesars, for example, was written c. 121 CE, which placed it about fifty years after the Roman civil war of 69 CE. During that war, three successive emperors died violently, and Suetonius cites contradictions and divergent accounts about all of their lives. For the the emperor Galba, Suetonius cites contradictions about the manner of his assassination (Gal. 20.1). For Otho (the subsequent emperor), Suetonius cites contradictions about the manner in which he overthrew Galba (Oth. 6.2). For the next emperor, Vitellius, Suetonius cites contradictions about his ancestry (Vit. 1.1). The historical biographer Plutarch also wrote about the civil war of 69 CE, and his biographies of the emperors Galba and Otho were probably published, at the latest, during the reign of the emperor Nerva (96-98 CE), only about thirty years after their deaths. And yet, Plutarch cites contradictions about the events of Galba’s assassination (Gal. 14.4), and likewise cites contradictions about Otho’s civil war with Vitellius (Oth. 9.2-3). Suetonius also wrote about the Flavian dynasty (69-96 CE), fifty to twenty-five years after the emperors who reigned. And yet, Suetonius cites contradictions about the occupation of the emperor Vespasian’s father (Vesp. 1.2), and different interpretations of the meaning behind the emperor Titus’ last words (Tit. 10.2), as well as various rumors about the emperor Domitian’s youth (Dom. 1.1).
It should be noted that all of these contradictory and varying reports are recorded about Roman emperors, for whom there would have been considerably more documentation and knowledge about their lives. In the case of Jesus, he was an obscure Galilean peasant, whose life was first related by oral tradition, spanning various regions, and often in Greek, a different language than what Jesus spoke. Keener’s notion that divergent accounts about Jesus would not substantially emerge under such circumstances, when divergence emerged among accounts about Roman emperors, for whom it would have been considerably easier to ascertain the facts of their lives, is thus poorly supported. To his credit, Keener (pg. 128) also offers a better explanation for why the Gospels do not cite contradictions and varying reports between their sources: “The more popular audience anticipated may be a more important factor; popular works of various genres were less likely to cite sources, even when they clearly depend on them.” In this regard, however, when the Gospels are compared to the genre of Greco-Roman biography, they more closely resemble the style of popular-novelistic biographies, rather than historical biographies. These kinds of biographies–such as the Alexander Romance and the Life of Aesop—neither cite their sources nor contradictions between sources, and they are likewise written in a low language register for popular audiences, like the NT Gospels. Such popular-novelistic biographies also do not engage in historical or critical analysis, and make several embellishments about their subjects’ lives, which is reflected in the Gospels, as well.
Beyond the examples given above, it should be noted that each of the major biographers of the Roman Empire–Cornelius Nepos (e.g., Themistocles 9.1), Plutarch (e.g., Otho 9.2-3), Suetonius (e.g., Caligula 8.1-5), Diogenes Laertius (e.g., 1.23), and the author(s) of the Historia Augusta (e.g., Hadrian 4.8-10)–note contradictions between their sources at various points in their biographies. The same is also true of Greco-Roman historians–such as Herodotus (e.g., 1.1-5), Thucydides (e.g., 2.5), Xenophon (e.g., Hellenica 6.4), Polybius (e.g., 12.5), Diodorus Siculus (e.g., 1.64), Sallust (e.g., Catiline War 19.3-5), Dionysius of Halicarnassus (e.g., Roman Antiquities 1.22), Livy (e.g., 1.55), Paterculus (e.g, 1.7), Curtius Rufus (e.g., 6.4), Josephus (e.g., Judean Antiquities 18.2), Tacitus (e.g., Annals 1.81), Appian (e.g., 11.9), Arrian (e.g., Anabasis 2.3), and Cassius Dio (e.g., 55.23). As such, noting contradictions between sources was a normative feature of both ancient historiography and historical biography. Nevertheless, as with the case of citing written sources, discussed in footnote 5 above, historians and historical biographers did not always note contradictions in every instance in which their sources disagreed. This criterion should therefore be interpreted as a distinction of frequency and not universality. Both ancient historians and historical biographers cite contradictions between their sources far more frequently than the canonical Gospels, even if they do not do so in every instance, which reflects a difference in genre.
The closest that any of the Gospels comes to identifying a contradiction between its sources is in Matthew (28:11-15), when the chief priests tell the guards previously stationed at Jesus’ tomb to report that his body was stolen by the disciples during the night. This report obviously contradicts Matthew’s own account, which has Jesus rise from the dead.
There are a number of reasons, however, for why this report was probably invented. To begin with, the very stationing of the guards at the tomb presumes that Jesus had predicted his resurrection to the Jewish authorities (Mt. 27:62-66), which caused them to ask Pilate to guard the tomb. Many historical Jesus scholars, however, doubt that Jesus predicted his resurrection before his death. Furthermore, if the Jewish authorities had actually reported that the disciples stole Jesus’ body, it is extremely bizarre that none of them are later charged with this accusation in the Book of Acts, especially since the disciples have several confrontations with the Jewish authorities. Finally, there is an obvious motive for Matthew to invent this story, in order to create an apologetic for Jesus’ resurrection, by making it clear that Jesus’ body was not stolen.
If the author of Matthew did not invent the report, and actually heard it from the Jews of his day (the author does state, “this story has been widely circulated among the Jews to this very day, Mt. 28:15), then it was probably first circulated by Jews long after Jesus’ death (Matthew wrote his gospel c. 80-90 CE), in response to Christian claims about the resurrection, rather than shortly after the event. For this reason, the report almost certainly does not go back to the time of Pilate. For further discussion of why Pilate stationing guards at Jesus’ tomb is probably a Christian invention, see Richard Carrier’s “Plausibility of Theft FAQ.”
 The author of Luke-Acts only uses the first person singular in the prologues of his two works (Lk. 1:3; Acts 1:1), without describing any biographical details about himself, and it is doubtful that the use of the first person plural, scattered throughout the “we” passages in Acts (16:10-17; 20:5-15; 21:1-18; 27:1-28:16), reflects the personal experiences of the author. As William Campbell in The “We” Passages in the Acts of the Apostles (pg. 13) explains:
Questions of whether the events described in the “we” sections of Acts are historical and whether Luke or his source/s witnessed them are unanswerable on the basis of the evidence currently available, as even the staunchest defenders of historicity and eyewitnessing acknowledge. More important, the fact that Acts provides no information and, indeed, by writing anonymously and constructing an anonymous observer, actually withholds information about a putative historical eyewitness, suggests that the first person plural in Acts has to do with narrative, not historical, eyewitnessing.
As such, the narration in Luke-Acts, outside of the two prologues, cannot be equated with the experiences of the author. The use of the mysterious “beloved disciple” (Jn. 21:24) likewise does not provide any proof of authorial eyewitnessing in John. To begin with, this character’s identity is kept completely anonymous within the narrative, and his role is narrated through the third person, which is very different from the first person authorial interjections seen above, among historical authors like Dionysius, Tacitus, and Suetonius. But furthermore, it is not even clear that this anonymous “beloved disciple” is intended to be the author of the text. As scholar Robert Kysar (The Anchor Bible Dictionary, v. 3, pp. 919-920) explains:
The supposition that the author was one and the same with the beloved disciple is often advanced as a means of insuring that the evangelist did witness Jesus’ ministry. Two other passages are advanced as evidence of the same–19:35 and 21:24. But both falter under close scrutiny. 19:35 does not claim that the author was the one who witnessed the scene but only that the scene is related on the sound basis of eyewitness. 21:24 is part of the appendix of the gospel and should not be assumed to have come from the same hand as that responsible for the body of the gospel. Neither of these passages, therefore, persuades many Johannine scholars that the author claims eyewitness status.
The authors of Matthew and Mark do not even use the first person singular, spoken by the narrator, within their gospels, making these texts even further anonymous.
This anonymous style of narration, in which the author reveals few or no clues about his personal relation to events, stands in stark contrast with the authorial interjections seen in historical biographies. As I discuss in my essay “Eyewitness Recollections in Greco-Roman Biography versus the Anonymity of the Gospels,” every single author who wrote a historical biography during the early Roman Empire, dealing with subjects dating to within a generation or two of his own lifetime–Cornelius Nepos, Tacitus, Plutarch, Suetonius, and Lucian–uses the first person singular to discuss his own personal relation to both his sources and the biographical subject. These authors also include discussion of events that they had personally witnessed. This kind of information greatly enhances the historical reliability of these texts, since we know that their authors either personally witnessed much of what they relate, or had access to eyewitnesses. In contrast, none of the Gospels include discussion of eyewitness material in this way, making their authors’ relation to the events depicted far more ambiguous, and thus the texts themselves less historically reliable.
 As patristics scholar Seumas Macdonald (“Why Classical (Greek) students are better at Greek than Seminary students”) explains:
Classical students generally train in classical grammar, which on the whole is slightly more complex than Koine grammar … Plato, Homer, Greek plays, Attic oratory. These are all high-level literary texts that demand a lot from their students. It’s functionally equivalent to taking a grammar class in English and then in your second year reading Shakespeare and the like … Seminary students, though, move from a grammar class to reading the newspaper. The NT is not a high-register set of documents.
While the Gospels are very simple in their Koine dialect and vocabulary, however, this does not mean that they lack complexity in their compositional methods. The Gospel of Mark, for example, is actually quite sophisticated in its plot, rhetoric, and literary devices, such as ring composition. Whitney Shiner in “Creating Plot in Episodic Narrative” discusses many of the literary techniques used by the author of Mark, which I summarize in a review here.
 One of the leading proponents to argue that the Q source did not exist is Mark Goodacre, who provides a resource on his website here for the arguments against Q’s existence. It should be noted, however, that the existence of Q is the majority opinion in mainstream biblical scholarship.
 In fact, Richard Pervo in “Direct Speech in Acts and the Question of Genre” has found that Acts of the Apostles contains more direct speech than virtually any piece of historiography or historical biography from the same period. Acts consists of 51% direct speech, which is on par with Jewish novels (e.g., Judith: 50%; Susanna 46%), and even greater than the proportion of direct speech in Hellenistic novels (e.g., Ephesian Tale: 38.9%; Alexander Romance: 34.4%). In contrast, both historiography (e.g., Josephus’ Jewish War I: 8.8%) and historical biography (e.g., Plutarch’s Alexander: 12.1%; Tacitus’ Agricola: 11.5%) have a much lower proportion of direct speech. The only piece of historiography to even come close to the frequency of direct speech in ancient novels is Sallust’s Catiline (28.3%), which is a text that contains a large number of Roman Senate orations, and even this text has only about half the amount of direct speech in Acts. Although Pervo’s study is focused on Acts and not the NT Gospels, the Gospels have a similarly high proportion of direct speech, which aligns their genre more closely with the ancient novel, than with ancient historiography and historical biography.
 One exception is the genre of popular-novelistic biography. As I explain in my essay “Greek Popular Biography: Romance, Contest, Gospel,” both the Alexander Romance and the Life of Aesop exist in multiple recensions that directly copy and redact material from earlier texts. This kind of open textuality is characteristic only of novelistic biographies, however, and is atypical of the style of historical biographers, such as Plutarch and Suetonius, who would exercise far more authorial control over their texts.
 Of course, there are exceptions where certain ancient historians include miracles in their narratives far more frequently. One example is the Roman historian Valerius Antias, whose history of regal Rome (although no longer extant, but still partially preserved in fragments) shows far more credulity towards miracles than is seen with Livy. As in the case of source citations and noting contradictions between traditions, the criterion that miracles pop up less frequently in ancient historical writing than the Gospels, and are often analyzed more critically, should be treated as a distinction of frequency, rather than universality. Some ancient historians mention virtually no miracles in their narratives, whereas others (like Antias) may include a large number. But, in general, the miracles in the Gospels are far more central to the core of the narrative, and are treated less critically, than what is usually seen in ancient historiography and historical biography.
 Another aspect of the Gospels that points towards legendary development is the presence of a number of characters, who appear to be solely allegorical in their role. For example, in my essay “WLC Tries to Defend the Myth of Barabbas,” I argue that the character Barabbas (whose name means “son of the father”) is probably a mythical character. Barabbas appears right before the scene of Jesus’ crucifixion, when Pontius Pilate asks the crowd to choose one prisoner for release as part of the Passover festival. Not only is this custom to release a prisoner at the crowd’s bidding unattested among any other ancient sources (and is likewise unknown as a practice in any Roman territory, beyond just Judea), but there is also strong reason to suspect that this scene was invented for allegorical purposes.
During the Yom Kippur sacrifice, there were two identical goats selected each year. One was released into the wild bearing the sins of Israel. The other was sacrificed in blood to atone for those sins. Hebrews 8-9 outside of the Gospels already attests to how the early Christians viewed Jesus as the ultimate Yom Kippur sacrifice where Jesus is the atonement for sins. Thus, in this allegory, the Gospel authors are telling their readers to reject the sins of violence and rebellion represented though Barabbas and instead to embrace Jesus’ ultimate atonement sacrifice.
This can be further demonstrated by the fact that the early church father Origen even recognized the symbolism of the allegory in Homily on Leviticus (10.2.2):
Behold! Here you have the goat, who is sent alive into the wilderness, bearing the sins upon himself of the people shouting and saying: Crucify! Crucify! That man [Barabbas] is accordingly the goat sent alive into the wilderness, and he [Jesus] is the goat, who is offered to God as a sacrifice, in order to atone for sins, and has made a true atonement for the people who believe in him.
Because the character of Barabbas may have been invented for allegorical purposes (or at least been depicted in fictional ways, if historical), it casts doubt on whether this story ever actually took place. Such allegorical characters, therefore, are another historical-critical problem for the Gospels’ reliability. Scholar Jennifer Maclean discusses Barabbas’ role as an allegorical character further in “Barabbas, the Scapegoat Ritual, and the Development of the Passion Narrative.”
 Despite the fact that the Gospels have a number of historical-critical problems, such as those outlined above, I state in this article that there are still “historical kernels” in the Gospels. Why do I think this is the case?
In part because the Gospels were written 40-60 years after Jesus’ death. That is a timeframe in which legendary development could have easily occurred. As Kris Komarnitsky (“Myth Growth Rates and the Gospels: A Close Look at A.N. Sherwin-White’s Two-Generation Rule”) explains, in response to A.N. Sherwin-White’s speculation that 40-60 years was too short a time to displace the historical core of the Gospel narratives:
That Sherwin-White did not fully consider the effects of public interest in a figure on the preservation of the historical core after his or her death is evident by the fact that every example he gives in his myth-growth-rate essay of people whom the historical core was preserved — Pisistratus (tyrant of Athens), Hipparchus (tyrant of Athens after Pisistratus), Gaius Gracchus (politician), Tiberius Caesar (emperor), Cleomenes (king), Themistocles (military commander), and all forty-six people in Plutarch’s Lives (every single one a statesman, general, king, emperor, lawmaker, politician, tyrant, or consul) — all are figures of significant public interest.
But what about the presence and influence of firsthand eyewitnesses on the oral tradition, someone might ask. Although a few of Jesus’ closest followers were probably eyewitnesses to a large part of his ministry (such as the Apostles), in an enthusiastic religious movement driven by belief in Jesus’ resurrection and imminent return (I think these were sincerely held beliefs that were not the result of legendary growth), these followers may by themselves have been unable to contain the growth of legend and displacement of the historical core among those in the growing church who did not know Jesus when he was alive or were not eyewitnesses of the specific events being distorted. The ability of a few of Jesus’ closest followers to contain the growth of legend would have been further hampered if the legends were growing in several different locales, for in this case they would have had the nearly impossible task of being present everywhere, stamping out all of the unhistorical legends…
…In conclusion, the Gospels are an understandable exception to what classical historians normally deal with, because classical historians rarely if ever deal with the written records of a highly revered religious figure who had very little contemporary significance to anyone but his followers when he was alive and to his worshippers after his death and where the entire written record comes only from those who worshipped him.
Nevertheless, despite the fact that legendary development could have easily occurred in the period between Jesus’ death and when the first gospels were written about his life, that does not mean that no accurate historical information could have been preserved within them. Half a century is still short enough of a timeframe for a number of accurate traditions to have been preserved. Because the Gospels contain legendary elements–-such as redaction, Midrash, allegorical characters, and fulfillment of scripture citations–-however, we cannot take their accounts at face value. Instead, stories in the Gospels should only be trusted if their are good historical-critical reasons for doubting that the story was invented.
I will provide two counter examples illustrating when the Gospels probably relate historical kernels of information, and when they do not:
For an example of a story that was probably invented, we have the infancy narratives in Matthew and Luke. Not only do the stories contradict each other (as noted in section 2 above), but there is also a strong theological motive to suggest that the authors invented Jesus’ birth in Bethlehem. As argued above, this was likely invented because of the expectation that the Jewish Messiah would be born in Bethlehem. We also have no outside corroboration for a number of the details in these narratives, such as the slaughter of the infants in Matthew (2:13-18), meaning that is fully possible that the Gospels‘ authors invented such details.
The problems with the infancy narratives are so considerable that even evangelical scholars, such as Robert Gundry have doubted a number of the details. For example, Gundry thinks that the author of Matthew placed Magi in the narrative, in order to draw an allusion to Daniel 2:2. Likewise, there are other signs of literary invention in the narrative, such the slaughter of the infants in Mt. 2:13-18 being derived from mimesis of Exodus 1:22-2:1-8, and the flight to Egypt being used as a reference to Moses, and so on. As such, I see no reason to interpret either as “an event that really happened,” when there is ample literary justification for where this passage came from.
In contrast, why is Jesus’ crucifixion under Pontius Pilate a probable historical kernel in the Gospels? Unlike Jesus’ birth in Bethlehem, it was not expected that the Jewish Messiah would be crucified like a common criminal. As such, this is probably not a detail that the Gospel authors would have invented. Likewise, we have Paul’s epistles, and other outside sources, who corroborate Jesus’ crucifixion. However, we have no such outside corroboration for Jesus’ birth in Bethlehem. Because the Gospel authors would not have likely invented Jesus’ crucifixion, in addition to its early attestation in outside sources, such as Paul, we can therefore trust it as a reliable historical kernel in the Gospels. In contrast, because the infancy narratives had a strong theological motive to have Jesus born in Bethlehem, in addition to the fact that this story only appears in later sources, such as Matthew and Luke, who even contradict each other about the order of events, we can therefore doubt that Jesus’ birth in Bethlehem was a historical event.
Using historical-critical methodology like this, therefore, can help historians sift between fact and fiction in the Gospels. As I explain above, I do think that there are some historically reliable stories in the Gospels. But, nevertheless, because of legendary elements–such as redaction, Midrash, allegorical characters, and fulfillment of scripture citations–-I do not think that we can take their accounts at face value.