The traditional authors of the canonical Gospels—Matthew the tax collector, Mark the attendant of Peter, Luke the attendant of Paul, and John the son of Zebedee—are doubted among the majority of mainstream New Testament scholars. The public is often not familiar, however, with the complex reasons and methodology that scholars use to reach well-supported conclusions about critical issues, such as assessing the authorial traditions for ancient texts. To provide a good overview of the majority opinion about the Gospels, the Oxford Annotated Bible (a compilation of multiple scholars summarizing dominant scholarly trends for the last 150 years) states (p. 1744):
Neither the evangelists nor their first readers engaged in historical analysis. Their aim was to confirm Christian faith (Lk. 1.4; Jn. 20.31). Scholars generally agree that the Gospels were written forty to sixty years after the death of Jesus. They thus do not present eyewitness or contemporary accounts of Jesus’ life and teachings.
Unfortunately, much of the general public is not familiar with scholarly resources like the one quoted above; instead, Christian apologists often put out a lot of material, such as The Case For Christ, targeted toward lay audiences, who are not familiar with scholarly methods, in order to argue that the Gospels are the eyewitness testimonies of either Jesus’ disciples or their attendants. The mainstream scholarly view is that the Gospels are anonymous works, written in a different language than that of Jesus, in distant lands, after a substantial gap of time, by unknown persons, compiling, redacting, and inventing various traditions, in order to provide a narrative of Christianity’s central figure—Jesus Christ—to confirm the faith of their communities.
As scholarly sources like the Oxford Annotated Bible note, the Gospels are not historical works (even if they contain some historical kernels). I have discussed elsewhere some of the reasons why scholars recognize that the Gospels are not historical in their genre, purpose, or character in my article “Ancient Historical Writing Compared to the Gospels of the New Testament.” However, I will now also lay out a resource here explaining why many scholars likewise doubt the traditional authorial attributions of the Gospels.
Coming from my academic background in Classics, I have the advantage of critically studying not only the Gospels of the New Testament, but also other Greek and Latin works from the same period. In assessing the evidence for the Gospels versus other ancient texts, it is clear to me that the majority opinion in the scholarly community is correct in its assessment that the traditional authorial attributions are spurious. To illustrate this, I will compare the evidence for the Gospels’ authors with that of a secular work, namely Tacitus’ Histories. Through looking at some of the same criteria that we can use to evaluate the authorial attributions of ancient texts, I will show why scholars have many good reasons to doubt the authors of the Gospels, while being confident in the authorship of a more solid tradition, such as what we have for a historical author like Tacitus.
How do we determine the authors of ancient texts? There is no single “one-size-fits-all” methodology that can be used for every single ancient text. We literally have thousands of different texts that have come down to us from antiquity, and each has its own unique textual-critical situation. There are some general guidelines that can be applied broadly across all traditions, however, from which more specific guidelines can further be derived when assessing a particular tradition.
Scholars generally look for both internal and external evidence when determining the author of an ancient text. The internal evidence consists of whatever evidence we have within a given text. This can include the author identifying himself, mentioning persons and events that he witnessed, or using a particular writing style that we know to be used by a specific person, etc. The external evidence consists of whatever evidence we have outside a given text. This can include another author quoting the work, a later critic proposing a possible authorial attribution, or what we know about the biography of the person to whom the work is attributed, etc.
For the canonical Gospels there are a number of both internal and external reasons why scholars doubt their traditional authors—Matthew, Mark, Luke, and John. I shall begin by summarizing the problems with the internal evidence.
To begin with, the Gospels are all internally anonymous in that none of their authors names himself within the text. This is unlike many other ancient literary works in which the author’s name is included within the body of the text (most often in the prologue), such as Thucydides’ History of the Peloponnesian War (1:1), which states at the beginning: “Thucydides, an Athenian, wrote the history of the war between the Peloponnesians and the Athenians, as they fought against each other.” The historians Herodotus (1:1), Dionysius of Halicarnassus (1.8.4), and Josephus (BJ 1.3) all likewise include their names in prologues. Sometimes an author’s name can also appear later in the text. In his Life of Otho (10.1), for example, the biographer Suetonius Tranquillus refers to “my father, Suetonius Laetus,” which thus identifies his own family name.
It should be noted that the Gospels’ internal anonymity also stands in contrast with most of the other books in the New Testament, which provide the names of their authors (or, at least, their putative authors) within the text itself. As Armin Baum (“The Anonymity of the New Testament History Books,” p. 121) explains:
While most New Testament letters bear the names of their (purported) authors (James, Jude, Paul, Peter, or at least “the Elder”) the authors of the historical books [the Gospels and Acts] do not reveal their names. The superscriptions that include personal names (“Gospel according to Matthew” etc.) are clearly secondary.
Two exceptions are the Book of Hebrews and 1 John, which are anonymous texts, later attributed to the apostle Paul and John the son of Zebedee, respectively. Modern scholars, however, also doubt both of these later attributions. As the Oxford Annotated Bible (p. 2103) explains about the authorship of Hebrews:
Despite the traditional attribution to Paul … [t]here is not sufficient evidence to identify any person named in the New Testament as the author; thus it is held to be anonymous.
And about the authorship of 1 John (p. 2137):
The anonymous voice of 1 John was identified with the author of the Fourth Gospel by the end of the second century CE … Since the Gospel was attributed to the apostle John, the son of Zebedee, early Christians concluded that he had composed 1 John near the end of his long life … Modern scholars have a more complex view of the development of the Johannine community and its writings. The opening verses of 1 John employ a first person plural “we” … That “we” probably refers to a circle of teachers faithful to the apostolic testimony of the Beloved Disciple and evangelist. A prominent member of that group composed this introduction.
As such, it is not unusual for scholars to doubt the traditional authorship of the Gospels, considering that the authorial attributions of the other anonymous books in the New Testament are also in considerable dispute.
The internal anonymity of the Gospels is even acknowledged by many apologists and conservative scholars, such as Craig Blomberg, who states in The Case for Christ (p. 22): “It’s important to acknowledge that strictly speaking, the gospels are anonymous.” So, immediately one type of evidence that we lack for the Gospels is their authors identifying themselves within the body of the text. This need not be an immediate death blow, however, since ancient authors did not always name themselves within the bodies of their texts. I have specifically chosen to compare the Gospels’ authorial traditions with that of Tacitus’ Histories, since Tacitus likewise does not name himself within his historical works. If the author does not name himself within the text, there are other types of evidence that can be looked at.
First, even if the body of a text does not name its author, there is often still a name and title affixed to a text in our surviving manuscript traditions. These titles normally identify the traditional author. The standard naming convention for ancient literary works was to place the author’s name in the genitive case (indicating personal possession), followed by the title of the work. Classical scholar Clarence Mendell in Tacitus: The Man And His Work (pp. 295-296) notes that our earliest manuscript copies of both Tacitus’ Annals and Histories identify Tacitus as the author by placing his name in the genitive (Corneli Taciti), followed by the manuscript titles. For the Histories (as well as books 11-16 of the Annals), in particular, Mendell (p. 345) also notes that many of the later manuscripts have the title Cor. Taciti Libri (“The Books of Cornelius Tacitus”). This naming convention is important, since it specifically identifies Tacitus as the author of the work. An attribution may still be doubted for any number of reasons, but it is important that there at least be a clear attribution.
Here, we already have a problem with the traditional authors of the Gospels. The titles that come down in our manuscripts of the Gospels do not even explicitly claim Matthew, Mark, Luke, or John as their authors. Instead, the Gospels have an abnormal title convention, where they instead use the Greek preposition κατα, meaning “according to” or “handed down from,” followed by the traditional names. For example, the Gospel of Matthew is titled ευαγγελιον κατα Μαθθαιον (“The Gospel according to Matthew”). This is problematic, from the beginning, in that the earliest title traditions already use a grammatical construction to distance themselves from an explicit claim to authorship. Instead, the titles operate more as placeholder names, where the Gospels have been “handed down” by church traditions affixed to names of figures in the early church, rather than the author being clearly identified. In the case of Tacitus, none of our surviving titles or references says that the Annals or Histories were written “according to Tacitus” or “handed down from Tacitus.” Instead, we have a clear attribution to Tacitus in one case, and only ambivalent attributions in the titles of the Gospels.
Furthermore, it is not even clear that the Gospels’ abnormal titles were originally placed in the first manuscript copies. We do not have the autograph manuscript (i.e., the first manuscript written) of any literary work from antiquity, but for the Gospels, the earliest manuscripts that we possess have grammatical variations in their title conventions. This divergence in form suggests that, unlike the body of the text (which mostly remains consistent in transmission), the Gospels’ manuscript titles were not a fixed or original feature of the text itself. As textual criticism expert Bart Ehrman (Jesus: Apocalyptic Prophet of the New Millennium, pp. 249-250) points out:
Because our surviving Greek manuscripts provide such a wide variety of (different) titles for the Gospels, textual scholars have long realized that their familiar names do not go back to a single ‘original’ title, but were added by later scribes.
The specific wording of the Gospel titles also suggests that the portion bearing their names was a later addition. The κατα (“according to”) preposition supplements the word ευαγγελιον (“gospel”). This word for “gospel” was implicitly connected with Jesus, meaning that the full title was το ευαγγελιον Ιησου Χριστου (“The Gospel of Jesus Christ”), with the additional preposition κατα (“according to”) used to distinguish specific gospels by their individual names. Before there were multiple gospels written, however, this addition would have been unnecessary. In fact, many scholars argue that the opening line of the Gospel of Mark (1:1) probably functioned as the original title of the text:
The beginning of the gospel of Jesus Christ…
This original title of Mark can be compared with those of other ancient texts in which the opening lines served as titles. Herodotus’ Histories (1.1), for example, begins with the following line which probably served as the title of the text:
This is the exposition of the history of Herodotus…
A major difference between the Gospel of Mark and Herodotus’ Histories, however, is that opening line of Mark does not name the text’s author, but instead attributes the gospel to Jesus Christ. This title became insufficient, however, when there were multiple “gospels of Jesus” in circulation, and so, the additional κατα (“according to”) formula was used to distinguish specific gospels by their individual names. This circumstance, however, suggests that the names themselves were a later addition, as there would have been no need for such a distinction before multiple gospels were in circulation.
So, in addition to the problem that the Gospel titles do not even explicitly claim authors, we likewise have strong reason to suspect that these named titles were not even affixed to the first manuscript copies. This absence is important, since (as will be discussed under the “External Evidence” section below) the first church fathers who alluded to or quoted passages from the Gospels, for nearly a century after their composition, did so anonymously. Since these sources do not refer to the Gospels by their traditional names, this adds further evidence that the titles bearing those names were not added until a later period (probably in the latter half of the 2nd century CE), after these church fathers were writing. And, if the manuscript titles were added later, and the Gospels themselves were quoted without names, this means that there is no evidence that the Gospels were referred to by their traditional names during the earliest period of their circulation. Instead, the Gospels would have more likely circulated anonymously.
As discussed above, Tacitus’ name is not affixed to his Histories using an “according to” formula (which in Latin would have been secundum Tacitum). Instead, Tacitus’ name was attached to the title in the genitive (“The Histories of Tacitus”). This kind of construction is not likely as a secondary addition, since the name Tacitus is not being used to distinguish multiple versions of a text, but is rather being used to indicate Tacitus’ personal possession of the work itself. That being said, there are substantial variations between the titles of Tacitus’ earliest manuscript copies. As Mendell (p. 345) explains, “the manuscript tradition of the Major Works [the Annals and Histories] is not consistent in the matter of title.” These variations report different names for the historical works that are attached to Tacitus’ name. The manuscripts of the Histories, for example, can also include the terms res gestae, historia Augustae, and acta diurna within the titles, in addition to historiae.
These title variations appear much later than those of the Gospels (which appear only a couple centuries after their composition), however, since we do not possess manuscripts of Tacitus’ historical works until several centuries after he composed, during the medieval period. (For an explanation of why the survival of fewer and later manuscript copies has no bearing upon the historical value of Tacitus versus the Gospels, see here.) Due to the negligence among medieval scribes in preserving manuscript copies of old Pagan literary works, both Tacitus’ Annals and Histories also contain large portions of missing material. Since these are far later copies, with large lacunas in the manuscripts, the title variations may have crept in later in the tradition.
Nevertheless, Mendell (p. 345) notes that we have strong contemporary evidence to suggest that the title “Historiae” was originally associated with Tacitus’ Histories:
Pliny clearly referred to the work in which Tacitus was engaged as Historiae: Auguror nec me fallit augurium Historias tuas immortales futuras [“I predict, and my prediction does not deceive me, that your Histories will be immortal”] (Ep. 7.33.1). It is not clear whether the term was a specific one or simply referred to the general category of historical writing. The material to which Pliny refers, the eruption of Vesuvius, would have been in the Histories. Tertullian (Apologeticus Adversus Gentes 16, and Ad nationes 1.11 cites the Histories, using the term as a title: in quinta Historiarum [“Tacitus in the fifth book of his Histories“]. It should be noted that this reference is to the ‘separate’ tradition, not to the thirty-book tradition, so that Historiae are the Histories as we name them now.
The evidence for the original title of the Histories is not fully conclusive, but what is noteworthy is that Pliny the Younger (a contemporary) writes directly to Tacitus and says that he is writing a “Historiae,” and Tertullian, the next author to explicitly cite passages in the Histories, refers to the work by that title.
For the purposes of authorship, however, the name of the work itself need not fully concern us. The evidence is certain in the case of Tacitus that the earliest manuscript tradition of his Histories clearly identifies him as the personal author. This manuscript tradition, though late in the process of textual transition, is corroborated by Pliny (a contemporary of Tacitus), who states that Tacitus himself was authoring a historical work about the same period and events covered in the Histories. This evidence is important, because it shows that Tacitus was known as the author of this historical work from the beginning of its transmission. And, although Pliny was writing while the work was still being composed (and thus does not cite passages from the text), the first source to cite passages from the Histories after it was published, Tertullian, clearly refers to Tacitus as the known author of the text. In Tacitus’ case, therefore, we have a clear claim to authorship, which dates back to the beginning of the tradition.
In the case of the Gospels, the first church fathers who allude to or quote the texts for nearly a century after their composition do so anonymously. Since the Gospels’ manuscript titles were likewise probably later additions (most likely after the mid-2nd century CE), this means that there is no evidence that the Gospels were referred to by their traditional names from the beginning of the tradition. Instead, these names only appear later in the tradition, which is the evidence to be expected if the Gospels first circulated anonymously, and were only given their authorial attributions in a subsequent period. Likewise, even when the later titles were added, the attributions were listed only as “according to” the names affixed to each text, which still entails considerable ambiguity about their authors.
Beyond the titles, we can look within the body of a text to see if the author himself reveals any clues either directly or indirectly about his identity. For Tacitus, while the author does not explicitly name himself, he does discuss his relation to the events that he is describing in the Histories (1:1):
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, while he does not name himself, the author of the Histories reveals himself to be a Roman politician during the Flavian Dynasty, which he specifies to be the period that he will write about. This matches the biographical information that we have of Tacitus outside of the Histories. For example, we know outside the text that Tacitus was writing a historical work about the Flavian period, since we have letters from Pliny the Younger (6.16; 6.20) written to Tacitus, where he responds to Tacitus’ request for information about the eruption of Mount Vesuvius (which Tacitus also alludes to Histories 1.2). Pliny’s letters also refer to Tacitus’ career as a statesman, such as when he gave the funeral oration for the Roman general Verginius Rufus (2.1). So we know from outside the Histories that Tacitus was a Roman politician writing a history about the Flavian era. This outside information is corroborated exactly by the evidence within the text. Thus, we have good reason to suspect that the author of the Histories is Tacitus, as the internal evidence strongly coincides with this tradition.
This kind of first person interjection from the author, described above, where Tacitus mentions his own relation to events within the narrative, stands in stark contrast with the anonymous style of narration in the Gospels. Although Tacitus does not overtly name himself in his historical works, he still uses the first person to discuss biographical details about himself. The gospels Matthew and Mark, in contrast, do not even use the first person, spoken by the author, anywhere in the text! Instead, both narratives are told in the third person, from an external narrator. This style of narration casts doubt on whether either author is relating personal experiences. As Irene de Jong (Narratology & Classics: A Practical Guide, p. 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.
The narrators of both Matthew and Mark describe the events in their texts from an outside point of view. This is a subtle aspect of both texts, but it is a very important consideration for why scholars describe them as “anonymous.” Neither narrative is an overt recollection of personal experiences, but rather focuses solely on the subject—Jesus Christ—with the author fading into the background, making it unclear whether the author has any personal relation to events set within the narrative at all.
The author of Luke-Acts only uses the first person singular in the prologues of his works (Luke 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 (discussed further by William Campbell under the “External Evidence” section below). John is thus the only gospel to include any kind of eyewitness construction, through the mention of an anonymous “beloved disciple” (John 21:24); however, modern scholars doubt that this “beloved disciple” was the actual author of John (see endnotes 31 and 32 below), and the character’s complete anonymity fails to explicitly connect it with the experiences of any known figure within the narrative.
Another piece of internal evidence that scholars look at is the linguistic rigor and complexity of a text. Based on the writing itself, we can tell that a certain level of education was required to author it. On this point, it is worth noting that Tacitus, as an educated Roman politician, would have had all of the literary, rhetorical, and compositional training needed to author a complex work of prose, such as his Histories. That is to say, from what we know of the Tacitus’ background, he belonged to the demographic of people whom we would expect to write complex Latin Histories.
As we will see for the Gospels’ authors, we have little reason to suspect, at least in the case of Matthew and John, that their traditional authors would have even been able to write a complex narrative in Greek prose. According the estimates of William Harris in his classic study Ancient Literacy (p. 22), “The likely overall illiteracy of the Roman Empire under the principate is almost certain to have been above 90%.” Of the remaining tenth, only a few could read and write well, and even a smaller fraction could author complex prose works like the Gospels.
Immediately, the language and style of the Gospel of John contradicts the traditional attribution of the text to John the son of Zebedee. We know from internal evidence, based on its complex Greek composition, that the author of this gospel had advanced literacy and training in the Greek language. Yet, from what we know of the biography of John the son of Zebedee, it would rather improbable that he could author such a text. John was a poor, rural peasant from Galilee, who spoke Aramaic. In an ancient world where literary training was largely restricted to a small fraction of rich, educated elite, we have little reason to suspect that an Aramaic-speaking Galilean peasant could author a complex Greek gospel. Furthermore, in Acts 4:13, John is even explicitly identified as being αγραμματος (“illiterate”), which shows that even evidence within the New Testament itself would not identify such a figure as an author.
Likewise, the internal evidence of the Gospel of Matthew contradicts the traditional attribution to Matthew (or Levi) the tax collector. While tax collectors had basic training in accounting, the Gospel of Matthew is written in a complex narrative of Greek prose that shows extensive familiarity with Jewish scripture and teachings. However, tax collectors were regarded by educated Jews as a sinful, “pro-Roman” class (as noted by J. R. Donahue in “Tax Collectors and Sinners: An Attempt at Identification“), who were alienated from their religious community, as is evidenced by the Pharisees accusations against Jesus in Mark 2:15-17, Matthew 9:10-13, and Luke 5:29-31 for associating “with tax collectors and sinners” (μετα των τελωνων και αμαρτωλων). Regarding the authorship of the Gospel of Matthew, scholar Barbara Reid (The Gospel According to Matthew, pp. 5-6) explains, “The author had extensive knowledge of the Hebrew Scriptures and a keen concern for Jewish observance and the role of the Law … It is doubtful that a tax collector would have the kind of religious and literary education needed to produce this Gospel.” For a further analysis of why Matthew the tax collector would have probably lacked the religious and literary education needed to author the gospel attributed to his name, see my essay “Matthew the τελωνης (“Toll Collector”) and the Authorship of the First Gospel.”
We have no such problem, however, in the case of Tacitus. As an educated Roman senator, who belonged to a small social class of people known to author Latin Histories, Tacitus is the exact sort of person that we would expect to author a work like the Histories, whereas we would have no strong reason to believe that an illiterate peasant, like John, or a mere tax collector, like Matthew, would have been able to author the Greek gospels that are attributed to them.
Furthermore, the sources used within a text can often betray clues about its author. In the case of the Gospels, we know that they are all interdependent upon each other for their information. Matthew borrows from as much as 80% of the verses in the Gospel of Mark, and Luke borrows from 65%. And while John does not follow the ipsissima verba of the earlier gospels, its author was still probably aware of the earlier narratives (as shown by scholar Louis Ruprecht in This Tragic Gospel).
Once more for the Gospel of Matthew, the internal evidence contradicts the traditional authorial attribution. The disciple Matthew was allegedly an eyewitness of Jesus. John Mark, on the other hand, who is the traditional author of the Gospel of Mark, was neither an eyewitness of Jesus nor a disciple, but merely a later attendant of Peter. And yet the author of Matthew copies from 80% of the verses in Mark. Why would Matthew, an alleged eyewitness, need to borrow from as much as 80% of the material of Mark, a non-eyewitness? As the Oxford Annotated Bible (p. 1746) concludes, “[T]he fact that the evangelist was so reliant upon Mark and a collection of Jesus’ sayings (“Q”) seems to point to a later, unknown, author.”
Apologists will often posit dubious assumptions to explain away this problem with the disciple Matthew, an alleged eyewitness, borrowing the bulk of his text from a non-eyewitness. For example, Blomberg in The Case for Christ (p. 28) speculates:
It only makes sense if Mark was indeed basing his account on the recollections of the eyewitness Peter … it would make sense for Matthew, even though he was an eyewitness, to rely on Peter’s version of events as transmitted through Mark.
To begin with, nowhere in the Gospel of Mark does the author ever claim that he based his account on the recollections of Peter (Blomberg is splicing this detail with a later dubious claim by the church father Papias, to be discussed below). The author of Mark never names any eyewitness from whom he gathered information.
But what is further problematic for Blomberg’s assumption is that his description of how the author of Matthew used Mark is way off. The author of Matthew does not “rely” on Mark rather than redact Mark to change important details from the earlier gospel. As scholar J. C. Fenton (The Gospel of St. Matthew, p. 12) explains, “the changes which he makes in Mark’s way of telling the story are not those corrections which an eyewitness might make in the account of one who was not an eyewitness.” Instead, many of the changes that Matthew makes to Mark are to correct misunderstandings of the Jewish scriptures. For example, in Mark 1:2-3 the author misquotes the Book of Isaiah by including a verse from Malachi 3:1 in addition to Isaiah 40:3. As scholar Pheme Perkins (Introduction to the Synoptic Gospels, p. 177) points out, “Matthew corrects the citation” in Matthew 3:3 by removing the verse from Malachi and only including Isaiah 40:3.
There are also other instances where Matthew adds Jewish elements that Mark overlooks. For example:
- Mark 9:4 names Elijah before Moses. Instead, Matthew 17:3 puts Moses before Elijah, since Moses is a more important figure to Jews than Elijah.
- Mark 11:10 refers to the kingdom of “our father” David. Ancient Jews would not have referred to “our father” David, however, since the father of the nation was Abraham, or possibly Jacob, who was renamed Israel. As such, not all Jews were sons of David. Instead, Matthew 21:9 does not refer to “our father” David.
These are subtle differences, but what they demonstrate is that the author of Matthew was not “relying” on Peter via Mark, but was redacting the earlier gospel to make it more consistent with Jewish scripture and teachings! This makes no sense at all for Blomberg’s hypothesis. Matthew is described as a tax collector (a profession that made one a social outcast from the Jewish religious community). Peter, in contrast, is described as a Galilean Jew who was Jesus’ chief disciple. Why would Matthew redact the recollections of Peter via the writings of his attendant in order to make them more consistent with Jewish scripture and teachings?
Instead, many scholars argue that the anonymous author of Mark was more likely an unknown Gentile living in the Jewish Diaspora outside of Palestine. This is strengthened by the fact that Mark uses Greek translations to quote from the Old Testament. Likewise, the author is unaware of many features of Palestinian geography. Just for one brief example: in Mark 7:31 Jesus is described as having traveled out of Tyre through Sidon (north of Tyre) to the Sea of Galilee (south of Tyre). In the words of scholar Hugh Anderson in The Gospel of Mark (p. 192), this would be like “travelling from Cornwall to London by way of Manchester.” These discrepancies make little sense if the author of Mark was a traveling attendant of Peter, an Aramaic-speaking native of Galilee.
Instead, scholars recognize that the author of Matthew was actually an ethnic Jew (probably a Greek-speaking and educated Jew, who was living in Antioch). As someone more familiar with Jewish teachings, he redacted Mark to correct many of the non-Jewish elements in the earlier gospel. This again makes little sense if the author of Matthew was actually Matthew the tax collector, whose profession would have ostracized him from the Jewish community. Instead, scholars recognize that the later authorial attributions of both of these works are most likely wrong. In fact, even conservative New Testament scholars like Bruce Metzger (The New Testament, p. 97) have agreed:
In the case of the first Gospel, the apostle Matthew can scarcely be the final author; for why should one who presumably had been an eyewitness of much that he records depend … upon the account given by Mark, who had not been an eyewitness?
And Christian scholar Raymond Brown (An Introduction to the New Testament, pp. 159-160) likewise acknowledges:
That the author of the Greek Gospel was John Mark, a (presumably Aramaic-speaking) Jew of Jerusalem who had early become a Christian, is hard to reconcile with the impression that it does not seem to be a translation from Aramaic, that it seems to depend on oral traditions (and perhaps already shaped sources) received in Greek, and that it seems confused about Palestinian geography.
The way that the Gospel of Luke uses Mark as a source likewise casts doubt on the tradition that John Mark, the attendant of Peter, was the original author of the text. As discussed above, the author of Luke borrows from as much as 65% of the verses in Mark. This is all very interesting, since the author of Luke is likewise the author of Acts, and John Mark, the attendant of Peter, has an appearance in Acts (12:12). This means that the author of Luke-Acts includes within his later narrative the alleged author of an earlier gospel, from which he has even borrowed a substantial amount of his material. Yet, never once does the author of Luke-Acts identify this man as one of his major sources! As Randel Helms points out in Who Wrote the Gospels? (p. 2):
So the author of Luke-Acts not only knew about a John Mark of Jerusalem, the personal associate of Peter and Paul, but also possessed a copy of what we call the Gospel of Mark, copying some three hundred of its verses into the Gospel of Luke, and never once thought to link the two—John Mark and the Gospel of Mark—together! The reason is simple: the connecting of the anonymous Gospel of Mark with John Mark of Jerusalem is a second-century guess, on that had not been made in Luke’s time.
Apologists here will merely try to dismiss this point as being an argument from silence. But again, as in the case of Matthew, the way that the author of Luke uses Mark strongly suggests that he was not “relying” on the recollections of Peter via his attendant, as Blomberg suggests, but was redacting an earlier anonymous narrative. For example, Bart Ehrman in Jesus Interrupted (pp. 64-70) discusses how the author of Luke makes changes to many of the details of the passion scene in Mark. In the Markan Passion, Jesus is depicted in despair and agony, whereas in the Lukan Passion, key details are changed to instead depict Jesus as calm and tranquil during his crucifixion. For example, Jesus’ last words are altered from a despairing statement in Mark 15:33-37 to a more tranquil one in Luke 23:44-46. But why would Luke—the mere Gentile attendant of Paul—redact and change the recollections of Peter—the chief disciple of Jesus—about the passion, crucifixion, and death of Jesus? The reason why is that the author of Luke most likely did not believe that Mark was based on the teachings of Peter. Instead, the anonymous author of Luke redacted and changed Mark, which was written by another anonymous author, to suite his own theological and narrative purposes.
A final note about the Gospels borrowing material from each other is that such works, which are not entirely original in their composition, but are largely redactions of earlier traditions, generally lack authorial personality. We saw above that Tacitus (Histories 1:1) discusses his relations with the Roman emperors during the Roman civil war of 69 CE and the Flavian Dynasty, which is the period that his Histories is written about. The Gospels, in contrast, are not written to tell the recollections of any one person, let alone an eyewitness. Instead, the Gospels are highly anonymous, not only in not naming their authors, but in writing in a collective, revisionist manner. New Testament expert Bart Ehrman (Forged, p. 224) explains that the general anonymity of the Gospels, in part, derives from the anonymous narrative structures of Old Testament texts, which served as their model of inspiration:
In all four Gospels, the story of Jesus is presented as a continuation of the history of the people of God as narrated in the Jewish Bible. The portions of the Old Testament that relate to the history of Israel after the death of Moses are found in the books of Joshua, Judges, 1 and 2 Samuel, and 1 and 2 Kings. All of these books are written anonymously … [T]he message of the Gospels … is portrayed … as continuous with the anonymously written history of Israel as laid out in the Old Testament Scriptures.
The authors of the Gospels were thus more concerned with gathering a collection of their communities’ teachings and organizing them into a cohesive narrative, similar to the anonymous, third person narratives found in the Old Testament. As Armin Baum (“The Anonymity of the New Testament History Books,” p. 142) explains, “The anonymity of the Gospels is thus rooted in a deep conviction concerning the ultimate priority of their subject matter.” This is not at all the case for Tacitus. We might be suspicious of the authorial attribution (at least to the extent that the Histories can be considered Tacitus’ own version of events), if Tacitus had merely copied from 80% of the material of an earlier author (as the Gospel of Matthew did) in order to write a highly anonymous narrative. Instead, Tacitus wrote in a unique Latin style that distinguished him as an individual, personal author, and he likewise comments on the events within his narrative from his own personal point of view.
We have seen above that the internal evidence does not support Matthew, Mark, or John as the authors of the gospels attributed to them. What about Luke? The Gospel of Luke and Acts are attributed to Luke, the traveling attendant of Paul. This is all very interesting, since we possess 7 undisputed epistles of the apostle Paul in the New Testament (6 of the traditional letters of Paul—Ephesians, Colossians, 2 Thessalonians, 1 and 2 Timothy, and Titus—are of disputed authorship and are possibly forgeries, as explained by Ehrman in Forged: Writing in the Name of God.) If Luke was Paul’s attendant, then corroborating details between Acts and the Pauline epistles may support the claim that Luke authored Acts; however, scholars often find the opposite to be the case. To name a few discrepancies:
- In Acts 9:26-28, Paul travels from Damascus to Jerusalem only “days” (Acts 9:19; 9:23) after his conversion in Acts 9:3-8, where Barnabas introduces him to the other apostles. However, in his own writings (Galatians 1:16-19), Paul states that he “did not consult any human of flesh and blood” after his conversion (despite consulting Ananias and preaching in the synagogues of Damascus after his conversion in Acts 9:17-22), but instead traveled into Arabia (which Acts makes no mention of), and did not travel to Jerusalem until “three years” after the event, where he only met Peter and James.
- In Acts 16:1-3, Paul has a disciple named Timothy, who was born from a Greek father, be circumcised “due to the Jews who lived in that area.” However, this goes against Paul’s own deceleration (Galatians 2:7) “of ministering the gospel to the uncircumcised.” Likewise, in Galatians 2:1-3, Paul brings another Gentile disciple, Titus, to the Jewish community in Jerusalem, but particularly insists that Titus not be circumcised. Likewise, in 1 Corinthians 7:20, Paul states regarding circumcision, “Each should remain in the condition in which they were called to God.”
- In Galatians 2:6, Paul makes it clear that his authority is equal to the original apostles, stating, “Of whatever sort they were makes no difference to me; God does not show partiality—their opinions added nothing to my message.” However, in Acts 13:31, Paul grants higher authority to those who originally “witnessed” Jesus. Likewise, Acts 1:21 restricts the status of “apostle” to those who had originally been with Jesus during his ministry, despite Paul’s repeated insistence that he was an apostle within his own letters (1 Corinthians 9:1-2).
In light of these and other discrepancies between Paul’s own recollections and how he is depicted in Acts, many scholars agree that the author of Luke-Acts was probably not an attendant of Paul (the speculation that he was is based largely on the ambiguous use of the first person plural in a few sections of Acts, to be addressed below). Nevertheless, the author of Luke-Acts clearly had a strong interest in Paul. However, the Oxford Annotated Bible (p. 1919) points out that the author “was probably someone from the Pauline mission area who, a generation or so after Paul, addressed issues facing Christians who found themselves in circumstances different from those addressed by Paul himself.” Hence, we once more have an anonymous author who was distanced from the various traditions and stories that he later compiled as a non-eyewitness.
The same problem of discrepancies between a text and outside epistolary evidence does not exist in the case of Tacitus. For example, we have Pliny the Younger’s letters (6.16; 6.20) written to Tacitus about the eruption of Mount Vesuvius in Campania. This outside evidence is corroborated within the text, when Tacitus mentions the burial of cities in Campania in the praefatio of his Histories (1.2), as well as when Tacitus mentions the volcanic eruption itself in his Annals (4.6). That Tacitus alludes to the volcanic eruption in the introduction to his Histories shows that he used Pliny’s account when describing the disaster more fully later in his narrative (although these later books did not survive the bottleneck of texts lost during the Middle Ages). Thus, in the case of Tacitus, we have harmony between outside epistolary evidence and the internal evidence of the text, whereas in the case of the Gospel of Luke, we have discrepancies between Paul’s letters, showing that the author was probably not a companion of Paul.
So far I have addressed the internal evidence for the authorship of Tacitus’ Histories and the Gospels. As has been shown, Tacitus has passed the criteria with flying colors, while all of the Gospels have had multiple internal problems. However, there are likewise external reasons to doubt the traditional authors of the Gospels.
In terms of external evidence for the authorship of Tacitus’ Histories, we have Pliny the Younger (a contemporary) writing directly to Tacitus while he was authoring a work that Pliny calls a “Historiae.” This historical work that Pliny describes was further identified as the Histories that we possess today by Tertullian (c. 200 CE), who was the next author to directly refer to it. Tertullian names Tacitus as the author in Apologeticus Adversus Gentes 16, and refers to the “fifth book of his Histories” (quinta Historiarum). Regarding subsequent citations of Tacitus’ historical works, Mendell (Tacitus: The Man And His Work, p. 225) explains:
Tacitus is mentioned or quoted in each century down to and including the sixth.
Thus, Tacitus was identified as the author of his Histories from the beginning of the tradition, rather than being speculated to be the author later in the tradition. This is very strong external evidence. We have precisely the opposite situation in the case of the Gospels. As New Testament expert Bart Ehrman (Forged, p. 225) explains:
The anonymity of the Gospel writers was respected for decades. When the Gospels of the New Testament are alluded to and quoted by authors of the early second century, they are never entitled, never named. Even Justin Martyr, writing around 150-60 CE, quotes verses from the Gospels, but does not indicate what the Gospels were named. For Justin, these books are simply known, collectively, as the “Memoirs of the Apostles.”
Below are the first references and quotations of the Gospels among external source, which treat them anonymously for some decades after their composition:
Ignatius (c. 105-115 CE) appears to quote phrases from Matthew (see here), and to allude to the star over Bethlehem (Matthew 2:1-12) in his Letter to the Ephesians (19:2); however, Ignatius does not attribute any of this material to the disciple Matthew nor does he refer to a “Gospel according to Matthew.” Polycarp (c. 110-140 CE) likewise appears to quote multiple phrases and verses from Matthew, Mark, and Luke (see here), and yet he neither attributes any of this material to their traditional authors nor refers to their traditional titles. There is scholarly dispute, however, as to whether Ignatius and Polycarp are quoting written texts, or instead interacting with oral traditions. As such, it is uncertain whether these two authors are directly referencing the Gospels that we possess today.
A stronger case can be made that the Epistle of Barnabas (80-120 CE) quotes Matthew (22:14), particularly because the epistle says “it is written” (4:14), when referring to the verse “many are invited, but few are chosen”; and yet, the Epistle of Barnabas does not attribute this verse to a text written by the disciple Matthew. What is further worth noting is that the Epistle of Barnabas (4:3) also refers to the Book of Enoch, and states “as Enoch saith,” showing that the epistle refers to traditional authorship elsewhere, when it was known.
Even more important, however, is when the Didache (c. 50-120 CE) directly quotes the Lord’s prayer (8:3-11), which is written in Matthew 6:9-13. This quotation is important, because the Didache attributes these verses to “His (Jesus’) Gospel” (ο κυριος εν τω ευαγγελιω αυτου) without referring to a “Gospel according to Matthew.” What the Didache is probably referring to, therefore, is the original title of the Gospels, before they were attributed to their traditional names. As discussed under the “Internal Evidence” section above, the Gospels were most likely originally referred to under the title το ευαγγελιον Ιησου Χριστου (“The Gospel of Jesus Christ”); however, when later there were multiple gospels in circulation, the construction κατα (“according to”) was added, in order to distinguish individual gospels by their designated names. The Didache likely preserves, therefore, a trace of their original titles, which were anonymous.
Justin Martyr (c. 150-160 CE) later makes explicit references and quotations of the Gospels (see here), but ascribes them under the collective title of “Memoirs of the Apostles,” without making any explicit mention of their traditional names. Finally, Irenaeus (c. 175-185 CE) refers to the Gospels by their traditional names in the late-2nd century (see here). As such, there is a clear development in which the Gospels were first referred to anonymously by external sources, and only later associated with their traditional attributions. For this reason, Ehrman (Forged, p. 225) concludes:
It was about a century after the Gospels had been originally put in circulation that they were definitively named Matthew, Mark, Luke, and John. This comes, for the first time, in the writings of the church father and heresiologist Irenaeus [Against Heresies 3.1.1], around 180-85 CE.
So, it is not until some century after the Gospels’ original composition, around the time of the church father Irenaeus, that they were even given their traditional authorial attributions. Incidentally, Irenaeus wanted there to be specifically “four gospels” because there are “four winds” and “four corners” of the Earth (Against Heresies 3.11.8). This was the kind of logic by which the Gospels were later attributed…
Ehrman (Forged, p. 226) goes on to explain:
Why were these names chosen by the end of the second century? For some decades there had been rumors floating around that two important figures of the early church had written accounts of Jesus’ teachings and activities. We find these rumors already in the writings of the church father Papias [now lost, but still partially preserved by Eusebius in Historia Ecclesiastica 3.39.14-17], around 120-30 CE, nearly half a century before Irenaeus. Papias claimed, on the basis of good authority, that the disciple Matthew had written down the saying of Jesus in the Hebrew language and the others had provided translations of them, presumably into Greek. He also said that Peter’s companion Mark had rearranged the preaching of Peter about Jesus … and created a book out of it.
So, we do have references to works written by Matthew and Mark, which date prior to Irenaeus, in the writings of the church father Papias. Unlike the sources mentioned above, however, Papias does make allusions to or quote any passages from these texts, so that it is unclear whether he is referring to the texts that we know today as the Gospel of Matthew and the Gospel of Mark. Papias’ own writings are likewise no longer extant, and so his reference to these works is only preserved in the later writings of the 4th century church father Eusebius. Incidentally, Eusebius (Historia Ecclesiastica 3.39.13) elsewhere describes Papias as a man who “seems to have been of very small intelligence, to judge from his writings.” Likewise, another fragment of Papias tells a story about how Judas, after betraying Jesus, became wider than a chariot and so fat that he exploded…
Here is what Eusebius (Historia Ecclesiastica 3.39.15) preserves regarding Papias’ claim that Mark, an attendant of Peter, had written an account about Jesus:
Now, the presbyter would say this: “Mark, having become the interpreter of Peter, accurately wrote down as much as he could remember, though not in order, about the things either said or done by the Lord. For he had neither heard nor followed the Lord, but only Peter after him who, as I said previously, would fashion his teachings according to the occasion, but not by making a rhetorical arrangement [ου μεντοι ταξει] of the Lord’s reports, so that Mark did not error by thus writing down certain things as he recalled. For he had one intention: neither to omit any of the things which he heard nor to falsify them.”
Regarding the account written by Matthew, Eusebius (Historia Ecclesiastica 3.39.16) records Papias as stating:
These things are recorded in Papias about Mark, but concerning Matthew this is said: “Matthew organized the reports in the Hebrew language, and interpreted each of them as much as he was able.”
Since Irenaeus clearly knew Papias’ works (Against Heresies 5.33.4), he probably drew the connection between these texts and the gospels Matthew and Mark from his testimony. However, a major problem with this tradition, noted above, is that Papias never quotes from the works that he attributes to these authors, and he could very well not be referring to the texts that were later called Matthew and Mark. This is especially true for Matthew, which Papias claims was written in Hebrew/Aramaic, even though the Gospel of Matthew that we possess today is a Greek text. But for Mark as well, Papias’ statement that the gospel “lacked rhetorical arrangement” (ου μεντοι ταξει) does not mesh very well with the internal evidence the text itself, which is actually pretty sophisticated in its plot and rhetorical devices.
Papias himself had never met any of the apostles (Historia Ecclesiastica 3.3.2), and he was relying on a tradition reported by an unknown figure named John the Presbyter, or “elder John.” It could be the case, therefore, that this oral tradition was referring to other, unknown texts that were later conflated with Matthew and Mark. Even if Papias is correctly referring to our Gospel of Mark, however, there are still problems with this attribution. As New Testament scholar Michael Kok, who argues that Papias is referring to the text that is known as Mark today, explains about Papias’ source (The Gospel on the Margins, p. 105):
His main source was the elder John, a figure who remains as elusive as ever. It is unlikely that he was a personal disciple of Jesus; he was probably a second-generation charismatic leader in Asia Minor. We have no clue about the elder’s connections outside Asia Minor or his general reliability. We know that Papias naively gave credence to local traditions about an original Hebrew or Aramaic edition of Matthew and other marvels (Hist. Eccl. 3.39.9, 16). If the foundation laid by Papias is rotten, how can we trust what subsequent writers build on it?
Some have tried to connect this mysterious John the Presbyter, or “elder John,” with John the son of Zebedee, who was a disciple of Jesus. Papias (Historia Ecclesiastica 3.39.4) describes the elder John in a list where he first names “the disciples of the Lord” Andrew, Peter, Philip, Thomas, James, John, and Matthew, and then names in addition two other figures, Aristion and the elder John. A problem with this list is that Papias names John the disciple, who was John the son of Zebedee, already in the same passage, before mentioning the “elder John.” As Michael Kok (p. 59) points out:
A reason for differentiating the latter two figures from the previous list of seven disciples is that it is redundant to name John twice [Historia Ecclesiastica 3.39.4] … Another reason for discerning two distinct figures is that the second John is prefaced with the title πρεσβυτερος (elder)…
Another problem with this identification is that Papias Fragment 10.17 records a tradition in which John the son of Zebedee was martyred alongside his brother James (44 CE), whose death is mentioned in the Book of Acts 12:2 (this fragment is discussed further in endnote 33 below). If this fragment is authentic, Papias could hardly have been reporting that John the Presbyter was John the son of Zebedee, and regardless of this fragment, what Eusebius preserves of Papias’ testimony elsewhere is too ambiguous to claim that Papias had received his information from one of the original twelve disciples. These circumstances make the Presbyter, who is the crucial source behind Papias’ authorial attributions to Matthew and Mark, a completely unknown figure, undermining the reliability of the authorial attribution itself.
Kok (pp. 159-160) also points out that a clear trail can be established for how the figure of John Mark was spuriously assigned to the Gospel of Mark:
1. In the earliest reliable evidence (Phlm 23; Col 4:10), Mark is casually named among a group of Jewish missionary partners of Paul. Mark may have been involved alongside his cousin Barnabas in a short-lived spat with Paul over mixed table fellowship in Antioch (Gal 2:13; cf. Acts 15:36-38), but it is unlikely that this led to Mark’s association with [Peter] as much later literature continues to remember Mark firmly in the Pauline camp (2 Tim 4:11; cf. Acts 12:25; 13:5).
2. In the next stage, the pseudonymous author of First Peter plucks a few of Paul’s co-workers at random, Mark and Silvanus, from the Pauline sphere to suit a centrist vision of the Christian community under the leadership of Peter. First Peter popularly circulated throughout Asia Minor sometime between 70 and 93 CE, leaving its mark on the general milieu of the elder John…
3. The third and most important step was taken by the elders of Asia Minor who assigned an anonymous gospel to Peter through his intermediary Mark and passed the report on to Papias … The rest of the patristic writers are derivative on Papias, though each develops the tradition in distinctive ways.
As such, there is little mystery behind how the authorship of John Mark could have been invented. There is a clear trail for illustrating the process by which 2nd century speculation selected the figure, in order to connect the text with the apostolic authority of Peter and Paul.
As pointed out above, Papias’ claim that the Gospel of Matthew was written in Hebrew/Aramaic, when the Matthew that we possess in manuscripts is written in Koine Greek (and based heavily on the Greek in the Gospel of Mark), is also a major blow to the authorial attribution of this text. Imagine if our earliest outside author to describe Tacitus’ works claimed that he wrote his Histories in Greek, when the Histories that we possess is in Latin! I can guarantee you that, if that were the case, scholars would have many, many more problems with Tacitus’ authorial attribution.
In fact, even Christian scholars like Raymond Brown (An Introduction to the New Testament, p. 210) acknowledge that this discrepancy is a major problem for connecting Papias’ statement with The Gospel of Matthew:
The vast majority of scholars … contend that the Gospel we know as Matt was composed originally in Greek and is not a translation of a Semitic original … Thus either Papias was wrong/confused in attributing a gospel (sayings) in Hebrew/Aramaic to Matthew, or he was right but the Hebrew/Aramaic composition he described was not the work we know in Greek as canonical.
One possible cause for the discrepancy may be that Papias is actually referring to an earlier literary source, which the (otherwise anonymous) author of Matthew later used when writing the gospel. This possible source has sometimes even been connected with the hypothetical “Q source.” Bruce Metzger (The New Testament, p. 97) discusses this possibility, stating:
As a solution of this difficulty it has often been suggested that what Matthew drew up was an early collection of the sayings of Jesus, perhaps in Aramaic, and that this material, being translated into Greek constitutes what scholars today call the Q source. In that case, the first Gospel was put together by an unknown Christian who utilized the Gospel of Mark, the Matthean collection, and other special sources.
If this is the case, however, then the disciple Matthew only authored a collection of Jesus’ sayings, which hardly entails that he stamped his eyewitness approval upon other material in Matthew, such as its legendary infancy narrative or miracles. Furthermore, even if Matthew did author such a source, then this would still entail that the Gospel of Matthew was misattributed when it was connected with his name (or at least that the attribution was oversimplified), by conflating a source for the text with its author.
These are the kinds of issues that make the authorial attributions of the canonical Gospels more problematic than the attributions of many other Classical texts from antiquity. The later Christian sources claiming that the Gospels were written by the apostles or their attendants simply have far more discrepancies, show greater speculation, and involve more implausibilities (or at least oversimplifications) than other, more solid authorial traditions that we possess from the same period, such as for works like Tacitus’ Histories.
Irenaeus’ notion that the author of Luke-Acts was an attendant of Paul likewise comes from speculation over a few passages in Acts where the author ambiguously uses the first person plural (16:10-17; 20:5-15; 21:1-18; 27:1-28:16). However, scholars studying these passages in Acts, such as William Campbell in The “We” Passages in the Acts of the Apostles (p. 13), have pointed out:
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.
Thus, the attribution to Luke the attendant of Paul is likewise unsound, being based on speculation over vague narrative constructions in the text. Luke was probably chosen, in particular, because the pseudonymous letter 2 Timothy (4:11) associates him with Paul’s company (presumably while he was staying at Rome).
Likewise, Ehrman (Forged, p. 227) explains how Irenaeus’ notion that John the son of Zebedee authored the fourth gospel is based on speculation:
The Fourth Gospel was thought to belong to a mysterious figure referred to in the book as ‘the Beloved Disciple’ (see, e.g., John 21:20-24), who would have been one of Jesus’ closest followers. The three closest to Jesus, in our early traditions, were Peter, James, and John. Peter was already explicitly named in the Fourth Gospel, so he could not be the Beloved Disciple; James was known to have been martyred early in the history of the church and so would not have been the author. That left John, the son of Zebedee. So he [Irenaeus] assigned the authorship to the Fourth Gospel.
As can be seen, Irenaeus’ attribution comes from little more than speculation over the identity of an unnamed character in the text. (As will be shown below, the actual internal evidence within John suggests that the anonymous “disciple whom Jesus loved” was probably the fictional invention of an anonymous author.)
Thus, we have a fairly clear trail for how all of the Gospels’ authors were probably derived from spurious 2nd century guesses: Matthew and Mark were based on an oral tradition reported by Papias that originated from an unknown John the Presbyter. Luke was speculated to be an author based on little more than vague narrative constructions using the first person plural in the text of Acts, and John was based on speculation over an unnamed “disciple whom Jesus loved.” Thus, not only is the external evidence weak, but all of it can be completely explained as later, spurious misattributions.
That the attributions were based on speculation is even reflected in the later titles. As I discussed above, the use of the construction κατα (“according to” or “handed down from”) in the titles already signifies that the attributions were redactional and largely speculative. As Ehrman (Jesus: Apocalyptic Prophet of the New Millennium, p. 42) points out, “Suppose a disciple named Matthew actually did write a book about Jesus’ words and deeds. Would he have called it ‘The Gospel According to Matthew’? Of course not… if someone calls it the Gospel according to Matthew, then it’s obviously someone else try to explain, at the outset, whose version of the story this is.” Thus, the traditional attributions come from later Christians in the 2nd century CE speculating over the different versions of the Gospels to assign apostolic traditions and names to the texts.
Apologists like Blomberg, however, will still attempt another escape hatch. In
The Case for Christ (p. 27), he argues:
These are unlikely characters … Mark and Luke weren’t even among the twelve disciples. Matthew was, but as a former hated tax collector, he would have been the most infamous character next to Judas Iscariot, who betrayed Jesus! … So to answer your question, there would not have been any reason to attribute authorship to these three less respected people if it weren’t true.
It is not clear to me why Matthew, as a reformed tax collector, would be hated next only to Judas, the man who betrayed Jesus. But Blomberg’s claim about these names being “unlikely” attributions is already refuted above, where a clear trail is demonstrated for how the traditional authors were speculated and assigned.
Additionally, Michael Kok has recently argued in The Gospel on the Margins that the attribution to John Mark might not be unlikely at all. The Gospel of Mark was the least favorite gospel among the 2nd century church fathers, and receives the least quotations, compared to the other canonical Gospels, on matters of doctrine and theology. Nevertheless, Kok argues that the Gospel of Mark was too early and too foundational to be removed from the canon. Likewise, the church fathers needed to protect the text from being used by heretical sects, such as those led by Valentinus, Basilides, and Carpocrates. In order to preserve the text’s canonical status, while downgrading its importance to doctrine and theology, therefore, the church fathers attributed the text to the lesser figure of John Mark, who had only imperfectly recorded the teachings of Peter. I discuss Kok’s theory in greater detail in this book review.
Furthermore, Ehrman (Forged, pp. 227-228) explains how the reasoning that apologists use to make this argument is specious at best:
Some scholars have argued that it would not make sense to assign the Second and Third Gospels to Mark and Luke unless the books were actually written by people named Mark and Luke, since they were not earthly disciples of Jesus and were rather obscure figures in the early church. I’ve never found these arguments very persuasive. For one thing, just because figures may seem relatively obscure to us today doesn’t mean that they were obscure in Christian circles in the early centuries. Moreover, it should never be forgotten that there are lots and lots of books assigned to people about whom we know very little, to Phillip, for example, Thomas, and Nicodemus.
Ehrman’s last point about other misattributions is likewise noteworthy. One thing that cannot be forgotten is that, in the context surrounding the Gospels, there were tons of misattributions and forgeries circulating in the early church. As Ehrman (Forged, p. 19) explains, “At present we know of over a hundred writings from the first four centuries that were claimed by one Christian author or another to have been forged by fellow Christians.” In such a context, there were canonical disputes over which texts were authoritative, which led later authors, such as Irenaeus in works like Against Heresies, to speculate and spuriously attribute texts to early figures in the church. As Ehrman (Forged, pp. 220-221) summarizes:
When church fathers were deciding which books to include in Scripture … it was necessary to ‘know’ who wrote these books, since only writings with clear apostolic connections could be considered authoritative Scripture. So, for example, the early Gospels that were all anonymous began to be circulated under the names Matthew, Mark, Luke, and John about a century after they were written … None of these books claims to be the written by the author to whom they are assigned … They are simply false attributions.
Such is the case for why scholars doubt the traditional attributions of the Gospels. To return to Tacitus, however, there was no prevailing context of doctrinal and canonical disputes that would have encouraged a later author to assign the Histories to the Roman senator. Furthermore, while forgery and misattribution could happen with secular texts, scholars have found no evidence of any in the case of Tacitus (here is an article explaining why). The later external references that mention the work simply quote Tacitus as the known author of the text, whereas the Gospels are a clear case of later speculations and misattribution.
Why do apologists attempt to go against the majority scholarly consensus to defend the traditional authors, anyway? The fact is that scholars over the last 150 years have recognized, after thorough study of the New Testament, that we do not possess the writings of a single eyewitness of Jesus. The closest thing we have are the 7 undisputed letters of Paul, who was not an eyewitness, but was writing decades later, and who provides few biographical details about Jesus’ life (I discuss what historical details Paul does provide here, and Ehrman likewise discusses this topic in his series “Why Doesn’t Paul Say More About Jesus?“). Because of this, most of our knowledge of Jesus, outside of a few vague references in Paul, comes from little more than garbled oral traditions, legendary development, and finally, after half a century, anonymous hagiographies, like the Gospels, that are not even written in the same language that Jesus spoke. Our sources for Jesus are thus very problematic and unreliable. None of this entails that Jesus did not exist, but we can only scarcely reconstruct a general biography of his life, let alone prove any of his miracles. For a summary of the minimal historical details that I do think can be said about the life of Jesus, see my essay “When Do Contemporary or Early Sources Matter in Ancient History?“
An apologist may still argue that, even if the Gospels’ authorial attributions are wrong, their real authors may have still had personal access to original eyewitnesses. However, many scholars likewise find this scenario to be unlikely. For Mark, the earliest gospel, Ehrman (Forged, p. 227) explains:
There is nothing to suggest that Mark was based on the teachings of any one person at all, let alone Peter. Instead, it derives from the oral traditions about Jesus that “Mark” had heard after they had been in circulation for some decades.
The situation only gets worse from there, since the anonymous author of Matthew then borrows from as much as 80% of the material in this earlier anonymous source, which itself was based on oral traditions. Likewise, the anonymous author of Luke copies from 65% of the material of the anonymous author of Mark. Furthermore, the author of Luke even suggests that he did not have personal access to eyewitnesses, since he specifies in the prologue of his gospel (1:1-2) that he was making use of previous written accounts (none of which he identifies by name, but we can tell that he copied material from Mark), which themselves were based on traditions that were “handed down” over a span of time (allegedly from distant, original eyewitnesses, although the author of Luke names none).
John is the only gospel to claim an eyewitness source, and yet the author does not even name this mysterious figure, but simply refers to him as “the disciple whom Jesus loved.” This is hardly eyewitness testimony, and it is probably the case that the author(s) of John invented this figure. One possibility is that the anonymous beloved disciple is a character already identified within the text. Verbal parallels suggest that the anonymous disciple may be Lazarus from John 11 (verses 1; 3; 5; 11; 36), whom Jesus raises from the dead in the passage. This Lazarus is likely based on the retelling of a story about an allegorical Lazarus in Luke 16:20-31. In the parable, Lazarus is a beggar who was fed by a wealthy man who dies and goes to Heaven, but the rich man dies and goes to Hell. The rich man begs Abraham in Heaven to send Lazarus to warn his family, since, if someone from the dead goes to them, they will repent. In Luke, Abraham refuses to send Lazarus from the dead, arguing that people should study the Torah and the Prophets to believe and will not be convinced even if someone from the dead visits them. In the Gospel of John, however, in which Jesus is more prone to demonstrate his powers through signs and miracles, rather than by appeals to Old Testament verses like in the Synoptic Gospels, the author instead has Jesus raise Lazarus from the dead, so that people might believe in him. The author of John thus very likely is redacting a previous story based on an allegorical character.
Regardless, even if the anonymous beloved disciple is not based on Lazarus, the Gospel of John is still extremely ambiguous about this character’s identity. The text even refuses to name him at key moments, such as the discovery of the empty tomb (20:1-9), where other characters such as Mary Magdalene and Peter are named, and yet this character is deliberately kept anonymous. The traditional identification of the disciple with John the son of Zebedee is undermined, among many other reasons, by the internal evidence of this beloved disciple’s connection with the high priest of Jerusalem (18:15-16), which could hardly be expected of an illiterate fisherman from backwater Galilee. The Gospel of John likewise shows signs of originally ending at John 20:30-31, and chapter 21, which claims the anonymous disciple as a witness, is very likely an addition from a later author. The chapter (21:24) distinguishes between the disciple who is testifying and the authors (plural) who know that it is true, suggesting that (even in this secondary material) the anonymous disciple is not to be understood as the author of the final version of the text. Furthermore, the final composition of John is dated to approximately 90-120 CE, which is largely beyond the lifetimes of an adult eyewitnesses of Jesus. In order to compensate for this problematic chronology, the author even had to invent the detail that this supposed eyewitness would live an abnormally long life (21:23) to account for the time gap. This detail is further explained if the anonymous disciple is based on Lazarus, who was already raised from the dead and has conquered death. Ultimately, all of these factors suggest that the unidentified “witness” is most likely an authorial invention (probably of a second author) used to gain proximal credibility for the otherwise latest of the four canonical Gospels.
Given all of the problems with the traditional authorship of John, even Christian scholar Raymond Brown (An Introduction to the New Testament, pp. 368-369) explains: “As with the other Gospels it is doubted by most scholars that this Gospel was written by an eyewitness of the public ministry of Jesus.”
To repeat the majority scholarly opinion, which I discussed at the beginning of this article, from the Oxford Annotated Bible (p. 1744):
Neither the evangelists nor their first readers engaged in historical analysis. Their aim was to confirm Christian faith (Lk. 1.4; Jn. 20.31). Scholars generally agree that the Gospels were written forty to sixty years after the death of Jesus. They thus do not present eyewitness or contemporary accounts of Jesus’ life and teachings.
I discuss in my article “Ancient Historical Writing Compared to the Gospels of the New Testament” why modern scholars doubt that the authors of the Gospels engaged in historical analysis. In this (rather lengthy) additional article I have also worked to explain why scholars doubt the traditional authorship and eyewitness status of the Gospels. Furthermore, I have shown how the same scholarly methods for determining authorship can be used to doubt the Gospels, while confirming the authors of texts for which we have more reliable traditions, such as Tacitus’ Histories.
To summarize some of the same questions that we can ask about Tacitus’ authorship versus the Gospels, here are a few:
Does the titular attribution clearly identify the author, rather than use a grammatical construction that only ambivalently reports a placeholder name as the attribution?
Did the attributed author likely have sufficient literary training to author the work in question?
Does what we know of the author’s biography align with the internal evidence within the text?
Do the earliest external sources who quote passages within the text either treat the work anonymously or refer to it by a different title?
Do later external sources who attribute the work show signs of speculating over the author?
Was there a prevailing context of misattribution, forgery, and canonical disputes surrounding the text that would increase the likelihood of its misattribution?
As has been shown, the same criteria for determining authorship can be applied to the Gospels as for any secular work, like Tacitus’ Histories. When scholars apply these criteria they find the authorial tradition for Tacitus to be reliable and the authorial traditions for the Gospels to be highly problematic. I have provided just one example here in the case of Tacitus, but textual experts likewise have undergone rigorous analysis of other ancient authors, such as Livy, Plutarch, etc., and found the evidence to confirm their authorship. My main advice for determining the author of any ancient text is to start by looking at what previous scholars have found. You will find that mainstream scholars for the last 150 years have found the authorial traditions for authors like Tacitus and Plutarch to be reliable, whereas the vast majority of scholars have doubted the authors of the Gospels.
A final note is that the criteria I have used above provide qualitative, rather than just quantitative, reasons for doubting the Gospels’ traditional authors. That is, the criteria that I employ are independent of each other (e.g., internal vs. external evidence). This means that the many reasons we have to doubt the authors are not just based on degree, but also vary by category. Sometimes apologists will make quantitative distinctions to argue for the reliability of the New Testament. For example, apologists often claim that Irenaeus’ late-2nd century attributions, despite decades of anonymous allusions and quotations (from the likes of early church fathers such as Ignatius, Polycarp, and Justin Martyr), are not really that far from the original composition of the Gospels, or will tally later 3rd-4th century church fathers (e.g., Tertullian, Origen, and Eusebius) who use these attributions, even though they are simply repeating late-2nd century speculation. Such arguments are one-dimensional and superficial, however, since the amount of time elapsed or the number of later church fathers who repeat these 2nd century attributions is only an argument by degree. And yet, I have shown that categorically there are many sound reasons to doubt the Gospels’ authorial attributions (based on a variety of issues, such as manuscript titles, literacy and education, conflicts between internal and external evidence, the context of 2nd century canonical disputes, etc.), so that the mere degree of any one criterion is insignificant, when multiple other criteria go against the traditional authors. Likewise, it is noteworthy that Tacitus has passed multiple independent criteria for identifying the author. The best explanation for how Tacitus could satisfy multiple categories of inquiry is because he is genuinely the author of the text. In contrast, the best explanation for why the Gospels’ traditional authors fail multiple categories of evaluation is that the later attributions genuinely do not fit the data.
 For Tacitus we possess two major manuscripts of his Annals and Histories. Books 1-6 of the Annals are preserved in a 9th century CE manuscript called the first Medicean manuscript. As Mendell (Tacitus: The Man And His Work, p. 295) notes, the title of this manuscript is P. Cornelii Taciti Ab Excessu Divi Aug. Books 11-16 of the Annals and books 1-5 of the Histories are preserved in an 11th century CE manuscript called the second Madicean manuscript. The titles included in the second manuscript are not complete. Mendell (p. 296) notes, however, that this manuscript still includes Tacitus’ name as the author: “The MS has a covering leaf at the beginning, added for protection. On the recto of this is the present catalogue number: Pl. 68, No. 2, and on the reverse: Cornelius Tacitus … Subsequent books have large decorative capitals and, with the exception of Books 16 and 21, a subscription, reading Corneli Taciti Liber.” The titles of both of these manuscripts use the genitive case to identify Tacitus as the author (i.e., “The Books of Cornelius Tacitus”), indicated personal possession. This standard construction is thus different from the unusual κατα (“according to”) construction found in the manuscripts of the Gospels.
 In fact, even Martin Hengel, who defends an early fixation of the Gospels’ manuscript titles, still acknowledges that the κατα (“according to”) formula downplays the biographical role of the author. As Hengel (The Four Gospels and the One Gospel of Jesus Christ, p. 49) explains:
The unusual titles of the Gospels already indicate that the evangelists are not meant to appear as ‘biographical’ authors like others, but to bear witness in their works to the one saving message of Jesus Christ. As is already shown by the beginning of the oldest Gospel, Mark 1:1, ‘beginning of the Gospel of Jesus Christ,’ the title of this Gospel could not be ‘Gospel of Mark’ because the content of his book was the ‘Gospel of Jesus Christ,’ the saving message of Christ—as subjective genitive (or genitivus auctoris), ‘coming from Jesus Christ’ and objective genitive, ‘about Jesus Christ’—but only ‘the Gospel (of Jesus Christ) according to Mark.’ The real ‘author’ of the one Gospel was Jesus Christ himself.
Further noteworthy is that the κατα (“according to”) preposition does not even have to refer to named individuals. Gospel of the Hebrews, for example, is titled το καθ’ Εβραιους ευαγγελιον (“the Gospel according to the Hebrews”). This construction hardly entails that the Hebrews themselves are the authors of the work, but rather a placeholder name, which refers to a tradition or group that the gospel was associated with. Hengel (p. 49) also notes that the closest parallel to this title convention, in examples beyond the New Testament and apocryphal gospels, is how the patristic church fathers refer to the Greek Septuagint. This translation of the Old Testament scriptures is referred to as κατα τους Εβδομηκοντα (“according to the Seventy”). It should be noted that this title is referring to a translation, and not to the authors of the Old Testament books. As such, when the κατα (“according to”) formula is used in these examples outside the New Testament Gospels, it does not specifically refer to authorship.
Another argument that apologists will raise is that the titles had to use an unusual construction, because the title το ευαγγελιον (του) Ιησου Χριστου (“The Gospel of Jesus Christ”) had already used an objective genitive to indicate the topic of the text, and thus could not also use a subjective genitive to indicate the author. There are Greek constructions that can avoid this problem, however, and still have the author’s name in the genitive. For example, το ευαγγελιον (του) Ιησου Χριστου το (του) Μαρκου (“The Gospel of Jesus Christ, the one of Mark”). Nevertheless, the ancient scribes who added the Gospel titles made use of no such construction that would have more clearly identified an author.
 A number of scholars have also argued that, even if the Gospels originally bore the names “Matthew, Mark, Luke, and John,” these titles may have actually referred to other, unknown individuals, who were later conflated with figures like John Mark and John the son of Zebedee. Maurice Casey (Jesus: Evidence and Argument or Mythicist Myths?, pp. 83-84), for example, argues that an unknown individual named “Marcus,” who had never encountered Jesus, authored the second gospel, and that this figure was later conflated with the John Mark described in Acts 12:12. Likewise, even conservative defenders of the traditional authorial traditions, such as Martin Hengel and Richard Bauckham, argue that the Gospel of John was actually authored by John the Presbyter and later conflated with John the son of Zebedee.
Furthermore, Raymond Brown, who is cited above for noting geographical errors in the Gospel of Mark, which he argues “are hard to reconcile” with the authorship of a “Jew of Jerusalem who had early become a Christian,” such as John Mark, still argues that there may have been a different, unknown figure named “Mark” who was the original author of the gospel. As Brown (Introduction to the New Testament, pp. 160-161) argues:
If those observations do not fit the NT John Mark and one wants to give some credibility to the Papias tradition, one might speculate that earlier tradition attributed the Gospel to an otherwise unknown Christian named Mark, who subsequently was amalgamated with John Mark.
Brown (p. 161) also notes that there may be “elements of truth in garbled form” to the Papian tradition (discussed under the “External Evidence” section above), though he still does not argue that the Gospel of Mark was directly based on the teachings of the historical Peter, but rather only a tradition attached to his name:
Did the relationship of (John) Mark to Peter in Acts and 1 Peter give rise to the Papias tradition that Mark the evangelist drew on Peter? … Again if one wants to grant at least limited credibility to Papias, one might regard “Peter” as an archetypal figure identified with the Jerusalem apostolic tradition and with a preaching that combined Jesus’ teaching, deeds, and passion … Papias could, then, be reporting a dramatized and simplified way that in his writing about Jesus, Mark organized and rephrased content derived from a standard type of preaching that was considered apostolic … Many would dismiss the entirely the Papias tradition; but the possibilities just raised could do some justice to the fact that ancient traditions often have elements of truth in a garbled form.
As such, even if the Gospels were written by individuals named “Matthew, Mark, Luke, and John,” this does not necessarily mean that they were the same authors as the figures with whom they were later identified. Furthermore, these later authorial traditions, such as those reported by Papias, even if possessing kernels of truth, could still have easily included oversimplifications and errors.
 Part of why the titles do not appear to be an original part of the text itself is because they can appear at multiple parts of the Gospel manuscripts. This suggests that the titles were a “floating” addition to the text. As Simon Gathercole (“The Titles of the Gospels in the Earliest New Testament Manuscripts,” p. 34) explains:
There are in fact a number of locations in which titles may appear: (i) on a flyleaf, i.e., on a page of its own; (ii) an opening title above or at the beginning of the text of the particular gospel; (iii) in a list of the contents of a codex, or in the title of a kephalaia or capitula list, or in the title of an argumentum; (iv) as a running title, at the top of a page (or across an opening) more or less consistently through a manuscript of a gospel; (v) as a subscriptio at the end of a gospel.
The titles can also appear in multiple variations, such as ευαγγελιον κατα Μαρκον, κατα Μαρκον ευαγγελιον, or just κατα Μαρκον, which suggests that they do not go back to the wording of any original title that was a fixed feature of the text.
To be fair, it should also be noted that the titles for many Classical texts were also added by a second hand. As Yun Lee Too (The Idea of the Library in the Ancient World, pp. 44-45) explains:
In the pre-Alexandrian period in Greece, ‘titles’ of works, almost certainly added by a later hand than the author’s, were written on the outside of rolls. The normal external indication of a title was the σιλλυβος, or σιλλυβον, a strip of papyrus or vellum that hung outside the roll as it lay on the bookshelf. It contained the author’s name and the title of the work … This way of marking texts persisted into the Roman period. In the context of a boast about the way in which Tyrannion has organized his library, Cicero speaks of sillybae made out of parchment that are attached to his books … in his letters to Atticus (cf. Ep. ad Atticum 4.4a.1; cf. 4.8)). Cicero comments that Atticus’ men have beautified his library by binding his books and affixing syllabae to them (Ep. ad Atticum 4.5), while at Ep. at Atticum 4.8 he observes that his house now has a mens—that is, a mind—now that Tyrannio has arranged his books; the sillybae help much. This statement suggests that the organization of books is significant in that knowing how and where to find one’s texts gives meaning and sense to one’s home. Later, Ovid makes reference to the ‘displayed titles’ (titulos … apertos), no doubt the sillybae, visible on bookshelves at Tristia 1.1.109.
The first manuscript titles for Tacitus’ Histories may have thus also been added by a second hand than Tacitus himself. These earliest titles may have also appeared at multiple parts of the manuscript (as they do in the later medieval manuscripts) and included variations in the wording; although a major difference between Tacitus and the Gospels is that his name was not affixed using the κατα (“according to”) formula. Because this construction is clearly used as an expansion upon a previous title (1. “The Gospel of Jesus Christ” + 2. “according to Mark”), it was most likely a secondary addition to the title tradition. Since this is not the case with Tacitus, whenever titles were added to his historical works (probably by copyists, book dealers, or libraries as soon as they were published, if not by Tacitus’ own hand), they would not have been an expansion upon a previous, anonymous title, but would have bore the name “Tacitus” from the beginning.
Likewise, as discussed in endnote 6 below, Tacitus was clearly known as the author of the Histories before it was even published, as is evidenced in his correspondence with Pliny. The Gospels, in contrast, were quoted anonymously for nearly a century after their composition, as discussed in endnote 20. This anonymous quotation, early in the tradition, suggests that the Gospel titles were added in a later period (probably after the mid-2nd century CE), after an initial period of anonymous circulation. Likewise, as noted in endnote 7, Tacitus’ Histories would have been published in a far more professional context than the Gospels, in which it was being publicly recited in connection with the author, and probably also had public libraries or book dealers identify Tacitus as the author of the text. In contrast, the Gospels would have been published in a less sophisticated literary context that would probably not have emphasized their particular authors when they first circulated.
As Christian scholar Raymond Brown (An Introduction to the New Testament, pp. 158; 208; 267) notes, the Gospels’ titles were probably not added until the latter half of the 2nd century CE. The late-2nd century was a time in the Christian community during which there were many canonical disputes, and connecting particular scriptures with figures in the early church was used as a means of gaining authority and canonical status for a text. A minority of scholars have speculated that they were added earlier, possibly even when the Gospels were first composed, such as Martin Hengel in The Four Gospels and the One Gospel of Jesus Christ. However, even Christian apologists find this view difficult to defend. Christian apologist Craig Blomberg (Making Sense of the New Testament, p. 151), for example, while describing Hengel’s thesis as “suggestive and worth serious consideration,” concludes that this view is “ultimately speculative and not provable.”
New Testament scholar Michael Wolter has critiqued Hengel, and argues that the titles could not have been added until, at the earliest, the first half of the 2nd century CE, primarily because the formula “κατα (according to) + the author” only makes sense if it was added when there were multiple gospels in circulation. The original title of Mark was probably just το ευαγγελιον Ιησου Χριστου (“The Gospel of Jesus Christ”), but when other gospels—such as the those later called Matthew, Luke, and John—were being circulated, the names were added to distinguish them from one another. Bart Ehrman, noting that the Gospels’ traditional names are not referred to by Christian authors writing before the late-2nd century, including Justin Martyr, has suggested that the titles were added sometime after 150-160 CE (after Justin Martyr) and before 175-185 CE (when Irenaeus attributed the traditional names). Ehrman notes that Irenaeus attributes the same names that are found in the Muratorian Canon (c. 170-200 CE), and points out that both were from the western half of the Roman Empire. Ehrman thus theorizes that a special edition of the Gospels was probably published at Rome circa 150-185 CE. This edition probably included the four canonical Gospels and added the titles that they are now associated with to distinguish them from one another, which in turn influenced the attributions of Irenaeus and the Muratorian Canon. Likewise, this edition’s probable place of publication at Rome can explain why the titles were subsequently adopted by churches all throughout the Roman Empire, since the Roman church was particularly influential and authoritative by the time of the 2nd century.
Another theory has also been put forward by David Trobisch in The First Edition of the New Testament, which argues that the Gospels (along with the other New Testament books) were given their traditional titles when the New Testament canon was first assembled, and published as an edition of collected works. This edition would have been highly influential on subsequent manuscript copies of the Gospels, and would thus explain why later manuscripts consistently bear the titles “according to” Matthew, Mark, Luke, and John, which begin to appear in the 3rd century CE. Trobisch (“Who Published the New Testament?,” p. 33) argues that this edition was published c. 156-168 CE, and that it likewise had a strong authoritative influence on both Rome and Asia Minor. Trobisch’s theory is largely compatible with the one given by Ehrman above, although Trobisch adds greater emphasis to the authority of the church in Asia Minor, which would have influenced the attributions at Rome.
It should be noted that Hengel (The Four Gospels and the One Gospel of Jesus Christ, p. 50) argues that a four book edition of the Gospels could not have been the origin of their titles, since the Gospels continued to circulate individually for some time. However, a four book edition could still have influenced the titles of individual gospels, even if they weren’t bound in a four book copy. Furthermore, even if the titles were added earlier than Ehrman’s estimate, as Michael Wolter has pointed out, they still were probably not added until at least the first half of the 2nd century CE, and thus decades after the Gospels’ original composition. In addition, David Trobisch (The First Edition of the New Testament, p. 41) argues, “The uniform structure of the titles points beyond the individual writing to an overall editorial concept and was not imposed by the authors of the individual writings. The titles are redactional … This strongly suggests that the present form of the titles was not created by independently working editors but that they are the result of a single, specific redaction.”
It is also worth noting that Hengel’s argument (The Four Gospels and the One Gospel of Jesus Christ, p. 49) that multiple titles of the Gospels should have appeared, if they were anonymous works given titles by different libraries, actually counts against an early fixation date of the titles. As David Trobisch has pointed out, the uniformity of the titles suggests that they were all added in a single redaction. Since this redaction would have taken place when there was already a four gospel canon, it must have occurred at a point after the Gospels were written, meaning that the Gospel titles are indeed secondary. As Trobisch (The First Edition of the New Testament, p. 43) explains:
[T]he focus of Hengel’s study is clearly on the “precanonical” state of the gospels, whereas my study concentrates on the final form of the collection. Hengel presupposes that the uniform titles are based on tradition. I argue that the awkward elements and structure of the titles are better explained as the result of editorial decisions by the collectors, who were trying to unify dissimilar material. As these collectors selected a certain number of writings, then edited and arranged them, they had a deliberate redactional strategy in mind.
Likewise, Michael Kok argues that the Gospels’ titles cannot be shown to be independent of the Papian tradition (discussed under the “External Evidence” section above). As such, they do not provide independent evidence of the authorial attributions. As Kok (The Gospel on the Margins, pp. 67-68) explains:
The titular usage of ευαγγελιον may have predated Marcion, but … it was not widespread in the first half of the second century. Aside from Papias’ discussion of named evangelists … there is no mention of a “Gospel” by name of its author before Theophilus of Antioch … If the Gospels were given titles early on, they were not necessarily the standard titles by which they are known today. Andrew Gregory [The Reception of Luke and Acts in the Period before Irenaeus, p. 51], discussing Irenaeus’ interest in the beginnings of each Gospel (Adv. Haer. 3.11.7), suggests that their opening verses may have once functioned as the titles. We frankly do not know what they were called by in the early second century. Hengel’s case on the uniformity of the titles is tempered by the fact that the earliest evidence rests on three papyri … If I may hazard a guess about the origins of the standard titles, they presume a theological vision that no longer speaks of plural “Gospels” (ευαγγελια) (cf. Justin, 1 Apol. 66.3), but a unitary “gospel” message proclaimed “according to” different messengers. The titles presuppose more than one text under the heading ευαγγελιον, a counterpoint to somebody like Marcion who privileged a single Gospel. There is no evidence that they predate or are independent of Papias.
Hengel (pp. 36-37) also theorizes that Irenaeus may have retrieved the traditional names and the biographical information about the authors from a Christian library in Rome:
The form of this information is also interesting … [I]t corresponds to the short notes about authors in the catalogues of ancient libraries, of the kind that we know, say, from the Museion in Alexandria. Presumably this information comes from the Roman church archive. As I Clement, around AD 100, shows, the Roman community had a respectable library, even containing ‘apocryphal’ books like Esther, Judith, Wisdom, and Sirach, which are still alien to the New Testament. There in Rome where, after the destruction of Jerusalem in AD 70, all the threads of the communities in the empire came together, in the first decades of the second century they must have had the four Gospels, even if they did not recognize them all equally … The relevant information about each writing, its author and its origin were kept in the Roman church archive…
It is circular to assume, however, that the authorial traditions are reliable, if they were derived from this library. It could be the library itself that created the spurious attributions in the first place, which Irenaeus reports. Furthermore, Hengel notes that this library also would have kept named copies of Old Testament and apocryphal books, but the authorial attributions of those texts are likewise doubted among most scholars. Finally, Hengel’s discussion of this library actually provides an excellent explanation of why the authorial attributions were widespread by the late-2nd century CE, and being reported in regions like Lyons, Carthage, Antioch, and Alexandria. Even if a four book edition of the Gospels was not the origin of their traditional names (such as what Ehrman theorizes was published at Rome c. 150-185 CE), Hengel still points out that Rome was the major center of Christianity in the Mediterranean world following the destruction of the Jerusalem in 70 CE, and that the Christian library in Rome was established around the turn of the 2nd century CE. The authorial attributions could have therefore been made at Rome in the early-2nd century (at the earliest, as Wolter argues), and then spread across the Mediterranean by the late-2nd century, which is when the traditional names begin to be reported. This theory is partially compatible with Trobisch, who also argues that the church in Asia Minor may have antecedently influenced the authorial traditions at Rome. Although Trobisch also argues that the attributions were not made until the latter half of the 2nd century, and that they must have come from a single editorial redaction, not individual editors.
Some Christian apologists have also speculated that, even before the later titles were added, there may have been notes, tags attached to the documents, writing on the sides of codices, and oral reports about authorship that circulated along with the Gospels, which related the traditional authorial traditions. But this is little more than speculation. The only evidence we have that the Gospel manuscripts bore named attributions is the manuscript titles that begin to appear in the 3rd century CE. As is discussed above, these titles were probably not added until at least the early-2nd century, and possibly as late as c. 150-185 CE, decades after the Gospels’ original composition. Likewise, these titles cannot be shown to be independent of authors like Papias, who show signs of speculation. Furthermore, even these manuscript titles they bore include the suspicious grammatical construction κατα (“according to”), which, as discussed above, downplayed the biographical role of the putative author.
Another talking point that apologists raise is that the names which appear in the titles are unanimous across manuscripts. These manuscripts only begin to appear in the 3rd century CE, however, after the date range in which scholars think that the traditional titles were added during the 2nd century. As New Testament scholar Keith Reich (“Gospel Authorship Part II: Formal Anonymity of the Gospels“) explains:
[T]his is not surprising at all to NT scholars since the traditional titles of the gospels are certainly established in the 2nd century, so we would expect no less. Of course manuscripts from the 3rd century would bear the traditional titles. So, barring the discovery of earlier manuscripts, the manuscript evidence will not be able to solve the issue of gospel titles, how early they were attached to the gospels, and whether the gospels were originally anonymous.
Furthermore, if Trobisch is correct that these manuscripts are derivative of the first published edition of the New Testament, this circumstance would thus explain why the titles are both uniform and still dependent upon a later redaction. And, even if the titles were not added along with the rest of the New Testament books, an earlier four book edition of the Gospels by itself could have still easily produced the same uniformity.
It should also be noted that it is not strictly true that there are no Gospel manuscripts which lack the traditional names. One example is P.Oxy. 76.5073, which quotes Mark 1:1-2, but in place of the title ευαγγελιον κατα Μαρκον (“The Gospel according to Mark”), instead has the title αναγνωτι την αρχην του ευαγγελλιου και ιδε (“Read the beginning of the Gospel and see”). As Larry Hurtado notes, this manuscript likely dates to the late-3rd or early-4th centuries CE, providing a fairly early witness of a manuscript that has the beginning of a gospel extant, but lack its traditional title. That said, the manuscript is only a strip that was wound up into a Christian amulet, and so it was never a complete manuscript of the text. This may account for its different title, but it still is not strictly speaking true that all of the manuscripts of the Gospels, which have their opening lines intact, are unanimous in providing the traditional names.
 The earliest manuscript copies that we have for Tacitus’ Annals and Histories respectively date to the 9th and 11th centuries CE, about seven to nine hundred years after the original composition of these texts (Histories c. 109 CE; Annals c. 116 CE). On both of these manuscripts, the titles can appear at various locations, similar to the Gospels. There are also different titles that appear for the texts. The Histories, for example, is referred to by other titles than “Historiae.” These medieval manuscripts are in poor condition, however, and have large sections of missing material. For the Histories, book 5 is missing a large amount of material and all subsequent books are lost. For the Annals, books 7-10 are missing and parts of books 5, 6, 11, and 16 are missing. Furthermore, Tacitus’ works would have originally been kept on papyrus scrolls when they were first published; however, in the process of textual transmission during Late Antiquity and the Middle Ages they were transferred onto parchment codices. It is difficult to know how much the titles were altered during this process, but it may be the case that the title variations crept in later during the trail of transmission.
Mendell (Tacitus: The Man And His Work, p. 345) notes that the first external source to quote Tacitus’ Histories by a different title was the Historia Augusta (c. 395 CE) in its Life of Tacitus 10.3. (This biography is about the 3rd century CE Roman emperor and not the 2nd century historian, but the biography refers to Tacitus the historian as one of the emperor’s ancestors.) This late-4th century source claims that Tacitus wrote a “historia Augusta,” which was probably a false quotation of the title, and may have influenced subsequent manuscript titles (some of which include this later title variation), along with additional factors in medieval textual transmission and other problematic quotations from Late Antiquity. Regardless, Pliny the Younger’s ((7.33.1) testimony shows that Tacitus was clearly known as the author of his Histories from the beginning of its transmission, and both Pliny and Tertullian (the earliest source to explicitly cite passages from the Histories itself) refer to the work by “Historiae.”
Unlike the later manuscripts for Tacitus, we possess titled copies of the Gospels that date much earlier in their trail of transmission. New Testament scholar Simon Gathercole in “The Titles of the Gospels in the Earliest New Testament Manuscripts” discusses how the traditional titles of the Gospels appear on manuscripts that date to the beginning of the 3rd century CE. There are a number of problems with these titles, however, which suggest that they were not originally included in the manuscripts. To begin with, as discussed in endnotes 2 and 5 above, the manuscripts of the Gospels do not use the standard title convention for indicating personal possession in antiquity, which was to identify the author’s name in the genitive, but instead use the unusual construction “κατα (according to) + the author” in the titles. New Testament scholars Michael Wolter (here) and Bart Ehrman (here) both agree that this title convention would not have been added until multiple gospels were in circulation, meaning that it would not have been placed in the original texts at their publication.
Likewise, the textual variations between the titles of the Gospels (discussed in endnote 4 above), found even in the earliest manuscript copies to survive, suggest that these titles do not go back to any original, autograph manuscripts, but were instead added (inconsistently) by subsequent scribes. Titled manuscripts of the Gospels appear around 300 CE, a little over a hundred years after the texts were written in the late 1st century CE. Likewise, the Gospels may have first been written on codices instead of scrolls (or, were at least written on codices from very early in their transmission), so that they did not have to be transitioned as drastically from one writing medium to another, especially during late antique and medieval textual transmission, as was the case with Tacitus. Nevertheless, even on these early manuscripts there are already a large number of textual variations in the titles. Thus, Bart Ehrman (Jesus: Apocalyptic Prophet of the New Millennium, pp. 249-250) concludes:
Because our surviving Greek manuscripts provide such a wide variety of (different) titles for the Gospels, textual scholars have long realized that their familiar names do not go back to a single ‘original’ title, but were added by later scribes.
Furthermore, as discussed in endnotes 20 and 28 below, the church fathers who alluded to and quoted the Gospels for the first several decades of their transmission all treat the texts anonymously (whereas Tacitus is referred to the author of his Histories from the beginning of its transmission). These combined circumstances strongly indicate that the Gospels were originally anonymous at the beginning of their transmission and that their traditional names and titles were only added later.
 Beyond just the issue of manuscript titles, another factor to take into consideration is the context of publication behind the authorial tradition in question. Sophisticated literary works published at Rome, such as Tacitus’ Histories, were publicly recited by the author (or by someone else on the author’s behalf), and were often copied by professional book dealers and kept in public libraries. As such, the author of the text was known through more than just the titles on manuscript copies and internal references within the text itself. The author could further be identified by these public recitals, book dealers, and libraries. Likewise, wealthy politicians like Tacitus belonged to elite literary circles in which authors regularly corresponded and consulted each other. For example, Pliny the Younger (7.33.1) knew that Tacitus was writing his Histories before the work was even published, because Tacitus had previously corresponded with him.
Whereas works like Tacitus’ Histories were published in high literary circles, in which the author was well-known, this does not mean that all texts in the ancient Mediterranean world were published in such a context. There were also many less sophisticated literary works that circulated in antiquity, which were often written and transmitted anonymously. In the genre of ancient biography, in particular—which many scholars argue is the literary genre to which the canonical Gospels belong—there were elite scholarly biographies, such as Tacitus’ Agricola, but also more popular biographies, written about figures such as Homer, Aesop, and Alexander the Great, which were far more often kept anonymous. As Tomas Hägg (The Art of Biography in Antiquity, p. 99) explains:
Simultaneously with the emergence of a bookish form of biography in the late classical and Hellenistic periods, vital biographic traditions were in progress at an oral or subliterary level, concerning in the first place legendary figures of great popular appeal … In contrast to the Lives treated in the previous chapter, which are the works of distinctive authors and largely remain under authorial control, these are anonymous; and they are ‘open texts,’ with regard to origin as well as transmission.
The latter example of popular-level biographies that Hägg discusses shows that there were many contexts of publication in the ancient world in which texts originally circulated anonymously. Sophisticated literary works like Tacitus’ Agricola and Histories did not belong to this category. However, as I argue in my essay “Are the Gospels Ancient Biographies?,” the Gospels of the New Testament more closely resemble this oral and subliterary type of biography that Hägg discusses above. It is very likely, therefore, that the Gospels circulated anonymously when they were first published. There are a number of reasons for thinking that the Gospels belong to this anonymous category of biography:
- To begin with, Hägg points out that a defining feature of this category was that these types of biographies operated more as ‘open texts,’ which were subject to expansion, redaction, and adaptation. This feature describes the Synoptic Gospels perfectly, since Matthew borrows from 80% of the verses in Mark, and Luke borrows from 65% of the verses. Accordingly, the Synoptic Gospels operated very much like ‘open texts,’ which shared a large amount of material, whereas scholarly biographies—such as those written by Tacitus, Plutarch, and Suetonius—did not borrow so much of their material from previous texts, but were instead written in a unique style with far more authorial control.
- Another reason why the original publication context behind the Gospels was probably anonymous is the fact, discussed above, that the Gospels mimic the literary style of the Old Testament, and were thus written in the anonymous, third person narrative structure of the Old Testament scriptures. Whereas sophisticated literary works, such as Tacitus’ Histories, were designed to demonstrate authorial research and talent, the Gospels were instead designed to operate as sacred scripture, which was a type of literature less focused on demonstrating the talents of the author rather than providing a third person, theological narrative.
- Finally, as scholar Harry Gamble (Books and Readers in the Early Church, p. 94) notes, Christian scriptures like the canonical Gospels were probably not copied by professional book dealers when they first circulated, and Gamble (p. 196) also points out that they would have been stored in very different kinds of libraries than most elite Greco-Roman literature, consisting primarily of private Christian collections. The Christian scribes who copied the Gospels were likewise considerably less trained than those used by elite authors, as discussed by Barbara Aland in “The Significance of the Chester Beatty in Early Church History,” p. 109). The lack of professional scribes available thus also entailed that there was a considerably higher degree of interpolation and textual redactions during the process of making manuscript copies. In fact, Classicist Peter van Minnen (“Dating the Oldest New Testament Manuscripts“) notes that the earliest manuscript fragments that we possess (dating from the 2nd-4th centuries CE) contain some of the greatest scribal errors out of all surviving Greek manuscripts of the Gospels. Given these circumstances, the Gospels would not have been subjected to the same kind of editorial provisions that works like Tacitus’ Histories, which was professionally published at Rome, were subjected to. Although the Gospels were assigned their traditional titles by at least the late-2nd century CE, the segmentary nature of their composition suggests that their final redactions were not finished until several decades after they were first composed. As argued by David Trobisch in endnote 7 above, these final redactions were likely complete around the mid-2nd century, when the individual Gospels were brought together as a combined edition. The manuscript titles would have been added as part of the final stage of redaction, in order to distinguish the individual gospels from each other by their familiar names (e.g., “according to Matthew”). Prior to this redaction, the Gospels would have likely been recited anonymously by earlier readers.
For all of these reasons, the Gospels were almost certainly published in a literary context that would have been considerably more anonymous than Tacitus’ Histories, which was published under circumstances in which the author would have been more well-known.
 Furthermore, certain regions of the Roman Empire had lower levels of literacy than others, such as rural regions like Galilee, and Greek literacy would have been even further limited in these areas if they were fluent in another language, such as Aramaic. In the case of rural Galilee, scholar Mark Chancey in The Myth of a Gentile Galilee and Greco-Roman Culture and the Galilee of Jesus finds that Greek literacy was largely restricted to two major urban centers, Sepphoris and Tiberias, and that the rural Jews of the region had little interaction with the Greek language or Gentiles. These circumstances would have certainly limited figures like Peter and John, both rural peasants from Galilee, from being able to author complex Greek prose, such as in the New Testament works attributed to them. This is, of course, in addition to the fact that such poor persons would not likely have received necessary literary training even in their own language to author such complex scriptures.
Another apologetic response to the problem of literacy is that illiterate persons could have allegedly used scribes to whom they would dictate their works, rather than writing them. However, this assumption misunderstands both the nature of literacy and how scribes were used in antiquity. It is true that literate persons, such as Paul, would dictate (in Greek) to scribes who would write down their words, as is evidenced in Romans 16:22 and Galatians 6:11. This does not entail, however, that an illiterate person (or someone with only partial literacy) could dictate complex prose in a foreign language. One could, of course, further speculate that such a person could tell a scribe the gist of a story, which the scribe would then interpret, organize, and compose in a different language. In such a case, however, the scribe would be the actual author of the work. The person using the scribe would at best be merely a “source” for the text. This may be the relation described between John Mark and Peter, discussed under the “External Evidence” section above, but it does not entail that the disciples Matthew and John could have actually authored the gospels attributed to them.
Furthermore, Ehrman (Forged, p. 77) points out that attributing authorship through such a practice would be improbable even for the (shorter) epistles of the New Testament:
Where in the ancient world do we have anything at all analogous to this hypothetical situation of someone writing a letter-essay for someone else and putting the other person’s name on it—the name of the person who did not write it—rather than his own name? So far as a I know, there is not a single instance of any such procedure attested from antiquity or any discussion, in any ancient source, of this being a legitimate practice. Or even an illegitimate one. Such a thing is never discussed.
Take note that Ehrman is only describing theological epistles, not multi-chapter works like the Gospels, which compile and redact earlier sources and materials into a complex, extended narrative. The notion that an uneducated person could merely use a scribe to author such a narrative is thus even further improbable.
Regarding William Harris’ literacy estimates in Ancient Literacy, it should be noted that subsequent scholars have critiqued and reevaluated some of his conclusions. Most notably, William Johnson (ed.), among other scholars, in the volume Ancient Literacies has called into question Harris’ binary categorization of “literate” and “illiterate.” Instead, Johnson et al. propose that there were many degrees of functional literacy in the ancient world. While non-elites may have had substantially less education, there is still evidence that they were able to engage with administrative documents, to have sufficient literacy for voting procedures in Athens or Rome, to write graffiti, to communicate through short letters, etc. Contributor Rosalind Thomas still notes, however, that the ability to produce complex literature was a skill largely confined to the upper classes. As Thomas (Ancient Literacies, p. 23) explains about advanced literacy (to compose oratory, rhetoric, and literature):
Gossip, oral communication, heralds, and announcements were all essential; much and was conveyed by these methods, but the ‘slow writer,’ to use the term of Roman Egypt, could hardly be equal to a member of the educated elite in their ability to master every aspect of the political system, especially as the elite could probably manipulate texts with relative ease as well compose eloquent speeches.
In fact, even in regions of the ancient Mediterranean that had substantially higher literacy rates (e.g., Athens), Thomas (p. 16) notes that the gap between the educated elite and the poor, in terms of literary abilities, was still quite substantial:
In ancient Athens, the line at which someone is seriously disadvantaged by poor writing skills can be drawn very low, but that does not mean that he was on an educational and political level with the elite. The educated elite, who overlapped considerably with the political leaders, had advanced literacy and cultural attainments that included mousike, music, literary knowledge, and literary composition. We therefore need to examine evidence for differing literacy skills alongside the surrounding social or political demands for writing.
Bear in mind that Thomas is referring to Athens, a place with a much higher common literacy rate than rural Galilee. (She also notes that this level of common literacy probably declined in Athens in the 4th century BCE.) Even if there was a wider range of literacies in the ancient world, therefore, the point still remains that it would be highly unusual for rural peasants outside of the educated elite to compose complex literature like the Gospels. For further discussion of the criterion of literacy in assessing authorial attributions, see endnote 11 below.
I should also note that, when I refer to the Gospels as “complex literary works,” I am referring to their rhetorical arrangement and composition, and not to the their language and dialect. As I discuss in my other article, “Ancient Historical Writing Compared to the Gospels of the New Testament,” the language in the Gospels is very simple and much less sophisticated than the elite historiographical and biographical literature of antiquity. Nevertheless, even though they are written in a simple vernacular (probably because they target a popular audience), the Gospels are still composed using sophisticated literary devices, which reflect a high level of education from their authors. Whitney Shiner in “Creating Plot in Episodic Narrative” discusses many of the literary techniques, which I summarize in a review here.
 The typical apologetic response to this passage is to claim that αγραμματος (“illiterate”) only means “uneducated” or “lacking formal rabbinic training.” However, it was typically educated Jews with rabbinic training that belonged to the small portion of the Jewish population who could author complex prose scripture. Furthermore, while it is possible that the passage is merely referring to rabbinic training, it is far more probable, given the historical context, that the passage also indicates illiteracy. Judaic Studies scholar Meir Bar-Ilan (“Illiteracy in the Land of Israel in the First Centuries CE“) finds that only about 3% of the population was literate, and most of this portion would have lived in cities or large towns (not typical of where the disciples of rural Galilee were from). Likewise, Catherine Hezser in Jewish Literacy in Roman Palestine (p. 496) argues:
If ‘literacy’ is determined as the ability to read documents, letters, and ‘simple’ literary texts in at least one language and to write more than one’s signature itself, it is quite reasonable to assume that the Jewish literacy rate was well below the 10-15 percent (of the entire population, including women) which Harris [Ancient Literacy] has estimated for Roman society in imperial times.
Furthermore, as Ehrman (Forged, p. 73) explains:
Most people outside of the urban areas would scarcely ever even see a written text. Some smaller towns and villages may have had a literacy level around 1 percent. Moreover, these literate people were almost always the elite of the upper class. Those who learned to read learned how to read Hebrew (not Greek).
Likewise, we have archaeological evidence which suggests that Peter, who is described alongside John as αγραμματος (“illiterate”) in Acts 4:13, was in fact illiterate based on excavations of his hometown in Capernaum. As Ehrman (Forged, pp. 74-75) explains:
In order to evaluate Peter’s linguistic abilities, the place to begin, then, is with Capernaum … The archaeological digs have revealed … there are no inscriptions of any kind on any of the buildings … Reed [Archaeology and the Galilean Jesus, pp. 140-169] concludes that the inhabitants were almost certainly ‘predominantly illiterate’ [even in Aramaic] … In short, Peter’s town was a backwoods Jewish village made up of hand-to-mouth laborers who did not have an education. Everyone spoke Aramaic. Nothing suggests that anyone could speak Greek. Nothing suggests that anyone in the town could write. As a lower-class fisherman, Peter would have started work as a young boy and never attended school. There was, in fact, probably no school there.
Bear in mind that John is described as αγραμματος (“illiterate”) alongside Peter in the passage, for whom we have very strong archaeological evidence that he was probably illiterate. Furthermore, both James and John the sons of Zebedee are likewise described as living around Capernaum. The best interpretation of the passage is thus that Acts 4:13 is describing Peter and John as both lacking Rabbinic training and being illiterate.
 New Testament scholar Jonathan Bernier has critiqued literacy as a criterion for determining the authorship of ancient texts in his blog post “Flipping Coins and Writing Gospels” (notably, Bernier does not defend the traditional authors of the Gospels in the post, but only calls into question whether illiteracy is good argument for doubting their traditional authors). His criticism is based on the statistical fact that, even if a certain group of people, such as Galilean fisherman, is 99% illiterate, this circumstance should not lead to the deduction that a particular Galilean fisherman, such as John, could not have authored a Greek scripture. As Bernier argues:
[T]he nature of statistics is such that even if 99% of all Galilean fishermen were persons who could not have written something like the Gospel of John it does not follow that there is a 99% chance that John, son of Zebedee, would be such a person. It’s an example of the coin flip problem: just because you flip a coin ten times and eight times it comes up heads it does not follow that the eleventh coin has a 80% chance of coming up heads; it in fact has a 50% chance; and even if that were not the case 80% is not 100%.
First off, a coin flip is a false analogy. It would only hold if 50% of fishermen were literary experts. If 99% of people of John’s status could not author a gospel, then the prior probability he could do it is indeed 1%. Bernier would need a die that only comes up “bingo” one out of a hundred times, which then in a sequence of rolls comes up “bingo” a different-than-that number of times. Then he would have a correct analogy.
However, this also misinterprets the logic behind how literacy is used as a criterion in determining authorship. The argument is not that it is 99% unlikely the John could have authored a Greek scripture, simply because 99% of the group to which he belonged would have been incapable of such authorship (assuming the number is not even higher). That would be to mistake prior probability with posterior probability. Rather, the argument is about first assessing the demographics of people to which the attributed author of a text belongs, to see if this is an ordinary or remarkable attribution.
Virtually 100% of Roman senators in the early-2nd century CE were literate in Latin. Likewise, Latin historiography was a genre that was primarily written by Roman senators. As Classicist Ronald Mellor (The Roman Historians, p. 4) explains, “History at Rome was written mostly by senators for senators.” So, when an authorial attribution is made to Tacitus, a Roman senator, for writing Latin historical works, such as his Annals and Histories, this is an ordinary attribution.
In the case of rural Galilee, Judaic Studies scholar Meir Bar-Ilan (“Illiteracy in the Land of Israel in the First Centuries CE“) finds that only about 3% of the population was literate, and most of these would have lived in cities or large towns (not typical of where the disciples of rural Galilee were from). Moreover, these people would primarily be taught to read Aramaic and Hebrew, so that even fewer could read or write in Greek. Once you drop the wealthy, urban population out of the equation, the number of poor, rural people who were literate in Aramaic would be much smaller than 3%. Out of this reduced fraction, even fewer could read and write in Greek. The ability to write complex Greek, such as in prose scripture, would have likewise been even rarer.
So, when a claim is made that a Galilean fisherman, like John, and a toll booth collector, like Matthew, authored complex Greek scriptures, it would have certainly been unusual and rare for someone who belonged to their demographics to have authored such texts. This already means that the authorial traditions of Matthew and John are, at the very least, more unusual than the authorship of Tacitus. It would not be unusual or rare for a Roman senator, like Tacitus, to have authored a Latin history, whereas in the case of Matthew and John, these would have been very rare and exceptional individuals to author complex Greek prose, given the demographics to which they belonged.
This does not mean that it would be impossible for Matthew and John to have been able to author complex Greek prose, but the next step is to see if there is any evidence that they were exceptional. Here, there are direct contradictions in our sources for their lives. John (Acts 4:13) is explicitly stated to be illiterate, and Matthew (Matthew 9:10-13) is explicitly stated to have been ostracized from the Jewish community (despite the fact that the gospel attributed him shows the most knowledge of Jewish Law and includes the most allusions to Jewish scriptures out of the New Testament Gospels). So, all of the data that we have for these individuals suggests that they were not exceptional in terms of their literary abilities and education.
The criterion of literacy can vary by degree. Our degree of confidence that Tacitus could author a Latin history is much higher than that a rural fisherman and toll booth collector could have authored Greek scriptures. So, the authorship of Tacitus is far more secure in this respect, even if it is not inconceivable that Matthew and John may have had remarkable literary abilities for their demographics. The same is true of other Classical authors, such as Livy and Plutarch, who belonged to more literate demographics, which is why it is not a good argument for apologists to equate the authorial attributions of the Gospels with these Pagan texts. Such Pagan texts are ordinary attributions, whereas the attributions of the Gospels are highly unusual.
The fact that the attributions of the Gospels are unusual should next lead to consideration over whether there is a reason for such unusual attributions. It could be because Matthew and John were indeed remarkable in their literary abilities and education, but, as shown above, the sources for their lives state the opposite. However, a very plausible explanation for the unusual attribution is the fact that Christians were having canonical disputes in the second half of the 2nd century CE (when the Gospels were first attributed). During this process, claiming apostolic authorship was used as a way to grant authority and canonical status to a text. It should be noted that the primary motivation driving these attributions was based on finding figures of authority. However, when the Gospels were first written, their authors were most likely chosen on the basis of ability. Educated Christians in the late-1st century CE, after the generation of the apostles had faded, were commissioned to write accounts of Jesus’ life and teachings (Luke 1:1-4). These individuals would have probably been chosen on the basis of their literary abilities as the most skilled and qualified individuals in their community. When there were later canonical disputes in the church, however, figures with authority were preferred for assigning authorship to canonical texts.
What the criterion of literacy achieves, therefore, is driving a wedge between the internal and external evidence. Based on nothing but the internal evidence of authorship, the Gospels would appear to have most likely been written by wealthy, urban-dwelling Greek speakers in the Jewish Diaspora, since that was how such texts were ordinarily composed. However, the external evidence says that they were written by individuals such as rural, Aramaic-speaking fishermen and toll collectors from Galilee. This would certainly be unusual, and not something suggested by the internal evidence when considered alone. The next step is to see if these unusual candidates show signs of having exceptional literary abilities and education for their demographics. When evidence outside the attribution, such as in Acts 4:13 and Matthew 9:10-13, states the opposite, the attribution becomes further anomalous. The final step is to look for whether there is any other reason for such an anomalous attribution. When the external evidence can be explained perfectly by canonical disputes in the late-2nd century CE seeking to ascribe certain texts to apostolic candidates, the cause of the anomaly becomes obvious. So, even if it is not statistically impossible that Matthew and John could have authored Greek scriptures, the internal evidence runs against the grain of the external evidence, and the external evidence itself is problematic and unreliable. In such a case, the criterion of literacy does matter when assessing an authorial attribution, especially if an authorial attribution is doubted on the grounds that it was spuriously assigned to a figure of authority, when the actual process behind a text’s composition would have instead required an individual of ability.
 The textual relations described in this section depend on the assumption of Markan priority, namely that the Gospel of Mark was written first, and that both Matthew and Luke used Mark. Although there is some dispute over this theory, it should be noted that Markan priority is the majority view in the scholarly community by a fairly wide margin. As New Testament scholar Michael Kok (“Markan Priority or Posterity?“) explains:
One sign of the consensus is that of the innumerable academic commentaries on Mark’s Gospel, C. S. Mann’s commentary on Mark for the Anchor Bible series is one of the rare exceptions in working from the basis of the Griesbach hypothesis and it has since been replaced by Joel Marcus’ two-volume commentaries that is firmly in support of Markan priority.
As such, Markan priority is assumed as the correct interrelation between the Synoptic Gospels for the purposes of this article.
 Apologists, of course, have come up with a number of attempts to rationalize this problem in geography. However, as scholar C.S. Mann (Mark, p. 322) explains, “While the text is clear enough at this point, the geography is impossible to reconstruct … The attempts of various commentators past and present to make sense of this awkward journey are often more inglorious than enlightening.” The main problem with this passage is that there is no route “through” Sidon to the Decapolis. You can go south from Sidon back to Tyre, to catch the Roman road to Caesarea Philippi. If Mark said that Jesus went up to Sidon and then back down again toward Tyre, and then on his way to the Decapolis, maybe you could say that he took the Caesarea route. But really, he would just say that Jesus went from Tyre to the Decapolis through Caesarea, not “through” Sidon. Sidon would never factor in. Mann further notes that the author of Matthew, who was probably more familiar with the region, in fact changes the itinerary to resolve the geographical problems. As Mann (p. 322) explains, “Matthew has no reference to Tyre and Sidon, nor yet of the Ten Towns, contenting himself merely with the statement that Jesus ‘departed from there and came by the Sea of Galilee’ (15:29).”
Another point that should be made is that the best supplied and thus safest route is not through the highlands from Tyre to Caesarea, but back south along the main road of the coast, then in to Tiberias and thence to Gadara (the beginning of the Decapolis) through the valley route. It hardly need be explained that fertile valleys and major coastal land trade routes are going to have far more ample water security for a hiker than a dodgy highland route built mainly for armies and caravans. And so, even beyond the fact that the geography in Mark is impossible to reconstruct, it makes far more sense that Jesus would have taken the coastal road, rather than attempting to go east along any highland route.
Likewise, this particular problem is hardly the only problem with Palestinian geography in Mark. Another problematic route is in Mark 11:1, which has Jesus and the disciples, in approaching Jerusalem from Jericho, come first to Bethphage and then to Bethany. As Helms (Who Wrote the Gospels?, p. 6) explains, “Anyone approaching Jerusalem from Jericho would come first to Bethany and then Bethphage, not the reverse. This is one of the several passages showing that Mark knew little about Palestine.” Nineham (The Gospel of St. Mark, pp. 294-295) agrees: “Mark did not know the relative positions of these two villages on the Jericho road.” Another problem concerns the location of Geresa (modern Jerash). As Theissen (The Gospels in Context, p. 242) explains, “According to 5:1ff, the town of Gerasa and its surrounding lands lie near the Lake of Galilee, although in reality Gerasa is about 65 kilometers southeast of the lake.” Likewise, Raymond Brown (An Introduction to the New Testament, p. 160) notes, “No one has been able to locate the Dalmanutha of 8:10, and it may be a corruption of Magdala.”
It should be noted that some scholars have argued that, even if the author of Mark makes geographical errors, it may not undermine the notion that he was a native of Palestine. As Michael Kok (The Gospel on the Margins, pp. 49-50) argues:
[A] native Jewish writer was not exempt from the human propensity to err … A slip-up in an otherwise sensible journey from a coastal area (Tyre) to the middle of Decapolis via the Galilean Sea (7:24-31) is that Mark does not realize how north Sidon was in relation to Tyre and the lake … If not all of Mark’s geographical blunders can be excused, the issue may be moot for Markan authorship … Maps were not a ready commodity so there is no reason to expect a fully accurate knowledge outside a person’s home region.
Likewise, there are a number of scholars who argue that the author of Mark was not a Gentile, but a Palestinian Jew. For example, Talmudic scholar Daniel Boyarin (The Jewish Gospels: The Story of the Jewish Christ, p. 105) argues that there are accurate depictions of Jewish customs in Mark, and that the author was likely Jewish. Michael Kok (p. 56) also argues, “In light of recent research, the data may not contradict the authorship of a Palestinian Jew,” even though Kok also argues that the Gospel of Mark was probably not written by John Mark.
 Even if one could somehow rationalize why Matthew (an alleged eyewitness disciple) would depend so heavily in his material upon the version of events related by Mark (who is described as a non-eyewitness), such a textual relation would still cast doubt on the notion that the Gospel of Matthew represents the personal recollections of an eyewitness. Rather, Matthew would have served more as an editor in compiling and redacting received material.
What is especially odd about this textual relationship, however, is that Matthew would not have even told his own version of how he met Jesus. In Mark 2:14, Jesus is described choosing the tax collector Levi the son of Alphaeus to be one of his disciples. In the Gospel of Matthew (9:9) this same scene is described, but Levi’s name is instead changed to “Matthew.” As discussed in endnote 26 below, this name change is probably one of the reasons why the Gospel of Matthew was attributed to “Matthew.” It remains highly peculiar, however, why the disciple Matthew would have taken this scene (one of the only opportunities in the gospel for which he could have related an overt eyewitness experience) from the writings of another (who wasn’t even an eyewitness!). In fact, even conservative scholar Richard Bauckham (Jesus and the Eyewitnesses, p. 112) acknowledges:
[T]he author of Matthew’s Gospel intended to associate the Gospel with the apostle Matthew but was not himself the apostle Matthew. Matthew himself could have described his own call without having to take over the way Mark described Levi’s call.
Bauckham (p. 302) argues that this association could have been pseudepigraphical, or could have derived from Matthew being connected with source material that was used during the composition of the Gospel of Matthew (a possibility also discussed in endnote 24 below). Bauckham further argues that “Levi” and “Matthew” are probably not the same person, since it would be extremely rare to have the same individual identified by two common Semitic names (as opposed to one Semitic name, and one Greek or Latin name). As such, the figure Levi in Mark is probably not the same person as Matthew in Matthew, even though they are described in the same scene, which raises even more difficulties for who the disciple Matthew was, to begin with. Since even Matthew’s very identity is difficult to ascertain, it remains yet more ambiguous to what extent his own “eyewitness” experiences made it into the Gospels.
 Of course, the common apologetic retort to redactions of this kind normally goes something along the lines of “different emotions can exist in the same man,” when Jesus is depicted in one gospel in a different manner than another. But such rationalizations greatly oversimplify the problem and miss the importance of the Synoptic Gospels’ interdependence in their source material. The author of Luke had a copy of Mark in front of him when he wrote about the passion and crucifixion of Jesus (or at least had read the text previously and memorized its material). Yet, at key moments, he made significant alterations in the previous narrative. In the Lukan narrative (23:27), a great number of people follow Jesus during his crucifixion, including a number of women, who are instead stated to have remained at a far in the Markan narrative (15:40). In the Lukan narrative (23:42-43), Jesus is crucified between two criminals, one of whom mocks him, while the other repents. Jesus replies to the repenting criminal and states, “I tell you, on this day, you will be with me in paradise.” However, in the Markan narrative (15:32), both of the criminals crucified next to Jesus mock him, and neither repents. In the Markan narrative (15:34), Jesus’ last words convey despair: “Jesus cried forth in a loud voice, ‘Eloi, Eloi, lama sabachthani?’ (This means: ‘My God, my God, why have you forsaken me?’).” However, in the Lukan narrative (23:46), Jesus’ last words convey resolve and tranquility: “And crying forth with a loud voice, Jesus said, ‘Father, into your hands I commit my spirit.’ Saying this, he expired.”
We can try to brush off these differences by rationalizing that each author merely “told one half of the story,” but we know that the author of Luke had access to Mark. The far more natural explanation for the changes is that the author of Luke simply had a different opinion and wished to depict Jesus’ crucifixion in a different way. As I explain in the essay, “Bible Contradictions: Why Are They There? What Do They Entail?,” this is the type of conclusion that we would reach for any secular text. Apologists who have presuppositions of inerrancy, however, often twist themselves in logical pretzels to avoid obvious contradictions and redactions between the canonical Gospels. For secular interpreters, however, I think the discrepancies and changes between the different authors are quite clear. Since the author of Luke changed the narrative in Mark to suite a different theological agenda, I think it is quite unlikely that this author thought that the account in Mark was directly based on the teachings of Peter.
That being said, New Testament scholar Michael Kok (whom I disagree with on this point) does think that the author of Luke-Acts may have been familiar the Petrine tradition about Mark (possibly from direct dependence on Papias or through common oral traditions in Asia Minor, which Kok discusses on pp. 151-153 of The Gospel on the Margins). Nevertheless, Kok also does not think that John Mark authored the Gospel of Mark, and so, even if the author of Luke-Acts was familiar with this tradition, that does not guarantee its authenticity. Furthermore, Kok’s theory is that Mark was spuriously attributed to John Mark, in order to downgrade its importance to doctrine. The author of Luke, therefore, would have made changes to Mark, since he did not fully agree with the content and structure of the narrative. Craig Blomberg, in contrast, suggests that Matthew and Luke were “relying” on Peter’s version of events, via Mark, implying that these authors would have viewed Mark as authoritative and largely derivative of Peter. This interpretation would largely not align with how Kok understands the relationship between these texts. It should also be noted that, even if the author of Luke-Acts was familiar with the Petrine tradition, he still provides no evidence that the title “Gospel according to Mark” was affixed to Mark when he used the text. Instead, this named appellation does not appear among external sources until the latter half of the 2nd century CE.
 Apologists, of course, have attempted to extract authorial personality from selective readings based on a few tenuous passages and uses of vocabulary. For example, apologists will sometimes claim that the author of Luke-Acts uses vocabulary specialized to physicians (the occupation that Luke, the attendant of Paul, was said to have) and takes extra notice of sick people. However, scholar Henry Cadbury in The Style and Literary Method of Luke: The Diction of Luke and Acts has deflated many of these claims by a closer reading of the relevant passages. Cadbury undertook this research when completing his doctorate, and the joke went round in scholarly circles that Cadbury earned his doctorate by depriving Luke of his.
It should also be noted that the author of Matthew actually makes inaccurate descriptions of tax practices in pre-70 CE Palestine, despite the fact that the text is said to have been authored by a tax collector! In particular, Matthew 22:19 implies that the Roman denarius was used for taxation during the time of Jesus. Based on archaeological evidence, however, scholar Fabian Udoh in To Caesar What Is Caesar’s (p. 236) finds:
[T]he imperial denarii were not required for Roman taxation, and they did not form the basis of the silver currency of the region.
As such, the internal evidence of Matthew actually conflicts with the occupational background of the author to which it was attributed, for it runs against the grain of this attribution that a tax collector would make mistakes about tax practices. I discuss this issue further in my essay “Matthew the τελωνης (“Toll Collector”) and the Authorship of the First Gospel.”
 It should be noted that some scholars, such as Richard Pervo in Dating Acts (chapter 4, pp. 51-147), argue that the author of Luke-Acts was familiar with Paul’s epistles, based on verbal and thematic parallels shared between both authors. These parallels, however, point more toward a textual relationship rather than a personal relationship between the two authors. Pervo, for example, argues that the author of Luke-Acts was writing in the early-2nd century CE, and was merely making use of a published edition of Paul’s letters (roughly half a century after he had died). Demonstrating that the author of Luke-Acts was familiar with the Pauline corpus, considering that it was published and circulated collection, therefore, is far less significant to furnishing evidence that the author was a personal attendant of Paul, than something along the lines of the author being familiar with the background behind when, where, and how Paul wrote his letters. Considering that Luke-Acts makes no mention of such details, its author thus affords no evidence of a personal relationship with Paul along these lines.
 Fitzmyer (Acts of the Apostles, pp. 438-439) elaborates further about this passage: “Luke reports the first visit of Saul to Jerusalem after his flight from Damascus (9:26-29; cf. 22:17; 26:20). It is the first of five, or possibly six, postconversion visits to Jerusalem that will be enumerated (the counting depends on a problematic variant reading). Whether they are all individually historical is problematic. It may be that Luke, dependent on different sources, has historicized and individualized some of the visits, when he should rather have realized that he had inherited more than one record of the same visit … In any case, the first postconversion visit of Saul to Jerusalem in Acts is to be taken as that reported in Galatians 1:18: ‘Then after three years I went up to Jerusalem to consult Cephas, and I stayed with him for fifteen days.’ That means ‘three years’ after his experience on the road to Damascus.” As scholar Christopher Matthews (Oxford Annotated Bible, p. 1937) concludes, “In Gal. 1:18 Paul states that his first visit to Jerusalem was three years after his conversion. Luke associates Paul with Jerusalem from the beginning.”
 A common apologetic rationalization for this contradiction is to claim that, because Timothy was born from a Jewish mother, he was racially considered a Jewish Christian, whereas Titus, who was born of both a Greek father and mother, was regarded as a Gentile. However, this interpretation is anachronistic and, as Fitzmyer (Acts of the Apostles, p. 575) notes, belongs “to a later Mishnaic tradition (m. Kidd. 3:12: ‘the offspring is of her own standing’; cf. Str-B, 2.741).” However, as Cohen (“Was Timothy Jewish?,” p. 268) explains, the “vast majority of ancient and medieval exegetes did not think [that Timothy was Jewish] … There is no evidence that Paul or the Jews of Asia Minor thought so. Ambrosiaster and his medieval followers did think so, but in all likelihood this interpretation is wrong because there is no evidence that any Jew in premishnaic times thought that the child of an intermarriage followed the status of the mother.”
 In terms of the first external references and quotations of the Gospels, as is discussed under the “External Evidence” section above, Ignatius (c. 105-115 CE) appears to allude to passages and quote phrases from the Gospel of Matthew. There are number of problems with his testimony, however, which raise dispute over whether Ignatius is directly interacting with Matthew as we know the text today. Ancient historian Richard Carrier discusses these problems in his essay “Ignatian Vexation.” Polycarp (c. 110-140 CE) likewise appears to quote multiple phrases and verses from Matthew, Mark, and Luke, but there is similar dispute over whether he is interacting with written texts.
The Epistle of Barnabas (80-120 CE) most likely does quote Matthew directly, despite mentioning no author for the text. Even more importantly, however, the Didache (c. 50-120 CE) explicitly quotes Matthew 6:9-13, but refers to the text as “His (Jesus’) Gospel” (ο κυριος εν τω ευαγγελιω αυτου), without adding the appellation “Matthew.” As discussed in endnote 5 above, the Gospels were probably originally known under the title το ευαγγελιον Ιησου Χριστου (“The Gospel of Jesus Christ”), before they were further ascribed κατα (“according to”) their traditional names. The Didache likely preserves, therefore, a trace of the Gospels’ original titles, which were anonymous. Justin Martyr (c. 150-160 CE) later refers to the Gospels collectively as “Memoirs of the Apostles,” without making any explicit reference to their traditional names. It is noteworthy that both the Didache and Justin refer to the Gospels by different titles, which suggests that their traditional titles were not added until after the mid-2nd century CE.
Finally, Irenaeus (c. 175-185 CE) refers to the Gospels by their traditional names in the late-2nd century CE. Another important source, close Irenaeus’ time, which also attests the traditional authorship of Luke and John, is the Muratorian Canon (c. 170-200 CE). Since the text is incomplete at the beginning, the canon also probably mentioned the traditional authorship of Matthew and Mark, which are cut off in the manuscript. As can be gathered from the survey of sources above, there is a clear development in which the Gospels were first referred to anonymously by external sources, and only later associated with their traditional names.
It is worth noting that both Irenaeus and the author of the Muratorian Canon were writing in the western half of the Roman Empire, and, as is noted in endnote 5 above, they both probably received the traditional Gospel names from a common source at Rome. This source could have been either a four book edition of the Gospels that was published at Rome c. 150-185 CE, as Bart Ehrman theorizes, or a catalogue used by the Christian library in Rome, which was founded in the early-2nd century, and could have added the traditional names any time after that date (possibly before the mid-2nd century, but more likely in the latter half of the 2nd century). David Trobisch likewise argues that the first edition of New Testament itself may have added the titles, which he argues was published c. 156-168 CE. Although Trobisch thinks that this edition originated from the authority of the church in Asia Minor, he also argues that it would have been influential at Rome, which would thus explain why Irenaeus and the Muratorian Canon report the traditional names in the western half of the Roman Empire.
There is also evidence that The Valentinian teachers Ptolemy (c. 140-180 CE) and Theodotus (c. 150-180 CE) may have associated the fourth gospel with John (though, this testimony is only preserved in later references among orthodox sources). As Richard Bauckham (Jesus and the Eyewitnesses, p. 466) writes:
Other second-century writers who call the author of John’s Gospel an apostle are the Valentinian teachers Ptolemy (Letter to Flora, apud Epiphanius, Panarion 33.3.6) and Theodotus (apud Clement of Alexandria, Excerpta ex Theodoto 7.3; 35.1; 41.3).
It should be noted, however, that these authors are near contemporaries of Irenaeus, and may be writing within the same time span in which the traditional names were added. Theophilus of Antioch (c. 180-185 CE) likewise refers to John 1:1 as being related by “John” (2.22), but also at around the same period as Irenaeus. Some have argued that Marcion may have associated the fourth gospel with John before Irenaeus. For example, Leon Morris (John, p. 19) argues:
It is also worth noting that according to a very probable reading of the evidence Marcion held the Apostle John to be the author of this Gospel. Tertullian (Contra Marcionem 4.3) speaks of this man as laboring “very hard to destroy the character of those which are published as genuine and under the name of the apostles” by drawing attention to Paul’s rebuking even of apostles (Gal. 2:13-14). It is difficult to catch the drift of Marcion’s argument unless he did in fact think that John wrote his Gospel. His point apparently was not that John did not write it, but that John did write it and was wrong! Since Marcion seems to have come to Rome c. 140, this is quite early testimony.
This line of argument is circular, however, since Tertullian does not tell us which gospels Marcion was attacking. There were several gospels attributed to “disciples,” including Peter and Thomas. Marcion could thus be rejecting any number of gospels attributed to Jerusalem apostles in favor of what he saw as Paul’s Gospel. Likewise, we do not possess Marcion’s own words, but only Tertullian’s polemic against him. Tertullian is also writing during the 3rd century CE, when the fourfold Gospel canon and the apostolic traditions authenticating them was more prominent. Since we lack Marcion’s own testimony of which gospels he was seeking to discredit, we cannot assume that a Gospel of John was among them.
It has also been argued (such as by Darrell Bock in Studying the Historical Jesus, p. 32) that Justin Martyr refers to the traditional authorship of Luke in Dialogue with Trypho (103.8). Justin does not mention Luke the attendant of Paul in this passage, however, but only quotes a (likely interpolated detail) from Luke 22:44, namely that Jesus sweated blood before he was arrested. Justin states that this detail is found in “the memoirs drawn up by Jesus’ apostles and those who followed them.” As discussed in endnote 29 below, however, Justin merely uses this this phrase as a collective title for all of the “memoirs” about Jesus, and so this passage is not singling out a specific author. The fact that the memoirs are associated with 1) apostles and 2) their followers likely reflects a development, prior to when the Gospels’ were given their traditional titles and named attributions, in which the Gospels came to be associated with the first and second generations of Christianity. Originally, the Gospels would have circulated anonymously, and later, by Justin’s time, they would have come to be associated with Jesus’ apostles and their followers, as the first and second generations of Christians. After Justin, by the time of Irenaeus and the Muratorian Canon, the Gospels were then given their named attributions, which specified particular apostles and their attendants. But simply because Justin plucked a detail from the third gospel, which he attributes to the collective “memoirs,” hardly entails that he is reporting a later authorial attribution connected specifically with Luke (whom Justin does not even mention) in this passage.
As can be gathered from the survey above, none of the Gospels are referred to by their traditional names until around the latter half of the 2nd century CE. Some Christian apologists will further appeal to sources who report the traditional names after the 2nd century, in order to inflate the evidence. For example, apologists will list Tertullian (c. 200 CE), Origen (c. 220 CE), and Eusebius (c. 325 CE) as sources who corroborate the traditional authors. All of these sources are writing after the 2nd century, however, and are merely repeating the same tradition. If the 2nd century tradition itself was spurious, they add nothing to the evidence of authorship.
In contrast, the first external reference to Tacitus’ Histories is in Pliny the Younger’s epistles (7.33.1), a contemporary source, written to the author himself. This would be the equivalent of the apostle Paul, for example, writing a letter to Luke, in which he discussed the composition of Luke’s gospel. Needless to say, no such evidence of this kind exists for the authorship of the Gospels. Pliny identifies him as the author of a “Historiae,” but since he is writing to Tacitus before the work was published, he does not explicit quote material within it.
Mendell (Tacitus: The Man And His Work, pp. 225-226) notes that Claudius Ptolemy (c. 150 CE) was probably familiar with material from Tacitus’ Annals (4.72-73) in his Geography, and that Cassius Dio (c. 230 CE) was probably familiar with Tacitus’ Agricola. However, neither author quotes Tacitus’ works explicitly, meaning that they are probably just familiar with information in these texts, which is not the same as citing exact verses. Likewise, neither author identifies all of his sources of information. It is also worth noting that both authors are writing in Greek, rather than in Latin, which was the language in in which Tacitus authored his historical works. This is nothing like the earliest external references to the Gospels. The Didache (8:3-11), for example, directly quotes Greek verses in the Gospel of Matthew (6:9-13) as being found in “His (Jesus’) Gospel,” which is specifically referring to the text by a different appellation than when it was later attributed to the disciple Matthew. Neither Ptolemy nor Dio, in contrast, quote Tacitus’ works by different names.
Moreover, there are also later quotations of Matthew in Justin Martyr, who, despite citing explicit passages, simply refers to the text under the collective title “Memoirs of the Apostles.” It is not like Ptolemy and Dio referred to Tacitus’ works as “Memoirs of a Roman senator,” before they were attributed by a later source to Tacitus. Finally, even when the Gospels were attributed, it was with the abnormal formula κατα “according to,” which, as discussed in endnote 2, is far from a direct claim to authorship. Nobody ever said that Tacitus’ works were written secundum Tacitum (“according to Tacitus”), but instead his works are correctly attributed using the genitive, as discussed in endnote 1.
The first external reference to explicitly cite passages from Tacitus’ historical works, therefore, is Tertullian (c. 200 CE), who clearly identifies Tacitus as the author in Apologeticus Adversus Gentes 16, and refers to the “fifth book of his Histories” (quinta Historiarum). Thus, the external evidence for the authorship of Tacitus may be summed up as follows: Pliny the Younger writes contemporary letters to the historian in which he describes him authoring a “Historiae.” Although Pliny wrote his letters before Tacitus published the work, Tertullian is the next external source to directly cite passages in this “Historiae,” in which he likewise identifies Tacitus as the author and uses “Historiae” as the title of the work in referring to its fifth book. For Tacitus, therefore, the earliest external reference to mention his work (Pliny) and the first external reference to cite passages (Tertullian) both consistently agree that he was the author.
 Irenaeus’ dependence upon Papias’ testimony may also be indirect, since, as discussed in endnote 5 above, he may have derived the authorship of Matthew and Mark from a four book edition of the Gospels that was published at Rome (possibly as part of a larger edition that included other books of the New Testament canon), or a Christian library in Rome. Nevertheless, neither of these sources can be shown to be independent of Papias, meaning that they probably derived the authorship of Matthew and Mark from his testimony as well (or at least from Papias’ source of information). New Testament scholar Michael Kok discusses how none of the 2nd-3rd century sources to report the authorial attributions of Matthew and Mark can be shown to be independent of Papias’ testimony in The Gospel on the Margins, which I summarize in a book review here.
 Sometimes Papias’ statement that Mark was written ου μεντοι ταξει, which I have translated as “without rhetorical arrangement,” is instead interpreted as “not in chronological order.” This is not a very probable reading of the passage, however. As Michael Kok (The Gospel on the Margins, pp. 188-189) explains:
Scholars disagree on whether ταξις refers to a chronological or literary arrangement. Chronology was a desideratum of historians and Papias may have borrowed the platitude on neither subtracting nor adding falsehood from them … The difficulty with this is that historians rarely chose the term ταξις for chronology. Instead, they preferred χρονος or καιρος for sequential time … Papias probably had a rhetorical arrangement in mind. Rhetoric had a prominent role in education and Hierapolis was home to the Stoic philosopher Epictetus.
If Papias meant that Mark lacked rhetorical arrangement, this description does not mesh very well with the internal evidence of the gospel. The Gospel of Mark is actually quite polished in its rhetorical composition and uses a number of sophisticated literary devices. 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 possible cause for this dissimilarity may be that Papias is not actually referring to the completed Gospel of Mark, but rather a source text that was used by the author during its composition. As Michael Kok (The Gospel on the Margins, p. 191) explains:
Other scholars apply Papias’ remarks to an earlier draft of Mark before the final version. Based on his expertise of ancient composition practices and the distinction between private notes (υπομνηματα) and published memoirs (απομνημονευματα), the Classics scholar George Kennedy [“Classical and Christian Source Criticism“] proposes that the evangelist started with preliminary notes before integrating them into an organized account. Notebooks may have come in handy to jot down Jesus’ sayings. Kennedy has to presume that Papias was misunderstood as referring to the finished text of Mark…
If this scenario is correct, one possible reading of the evidence is that the Gospel of Mark was not based directly on the recollections of Peter, but may have incorporated some notes that were based on Peter’s teachings. If Papias is only referring to one of the sources of the text, however, this does not necessarily mean that the entire account goes back to Peter’s eyewitness recollections. Moreover, as discussed by Raymond Brown in endnote 3 above, “Peter” may refer to a standard type of preaching material that was considered apostolic, and thus connected with Peter’s name, even though it was not directly based on Peter’s own eyewitness testimony. It also remains open to dispute whether the same figure who compiled the notes was also the author of the gospel itself. George Kennedy (“Classical and Christian Source Criticism,” p. 148) acknowledges that “[c]onsiderable time could have elapsed between Mark’s note-taking … and the composition of the Gospel as we have it.” Kennedy (p. 152) also points out that such notes, even if unpublished, were often circulated to persons other than the one who wrote them. One reason that John Mark was associated with the Gospel of Mark could be due to the fact that this figure was responsible for compiling a notebook as a source text, which was then used by a different author to compose the Gospel of Mark. When the Gospels were later being attributed to their traditional authors, however, John Mark as a source was then conflated with the author of the text itself.
It should be noted that Michael Kok does not accept this view, but instead argues that Papias was referring to the completed Gospel of Mark in his statements quoted above. Kok does not think, however, that John Mark authored the gospel. Instead, Kok argues that Papias’ source attributed the text to an to otherwise obscure figure, like John Mark, in order to downplay its significance to Christian doctrine. Kok notes that Mark was the least popular gospel in the 2nd century CE, and thus received the fewest quotations among the church fathers. Mark was also a text that was too early and too foundational (which the more popular authors of Matthew and Luke even used as a source) to simply be discarded by the early church. Leaving Mark out of the canon also could have left it vulnerable to the use of heretical sects. Kok argues, therefore, that Mark may have been deliberately attributed to an obscure figure, only loosely connected with Peter, both to protect it from misuse, but also to downplay its importance to the canon. To strengthen this view, Kok even notes that John Mark is described negatively in Acts 15:37-39, thus making him a prime candidate to attribute the text to under this scenario. Since Mark’s name would have been plucked from 1 Peter (5:12-13) in the process of this attribution, which also mentions Silavanus, the negative material about Mark in Acts could also explain why Mark was specifically chosen, and not Silavanus, since Mark’s characterization was better suited for this scenario. Kok’s thesis provides, therefore, another explanation for how the Gospel of Mark could have been misattributed. I discuss Kok’s thesis further here.
 It should be noted that even Richard Bauckham (Jesus and the Eyewitnesses, pp. 452-463) argues that both Papias and Polycarp were hearers of a separate “elder” John, and not John the son of Zebedee. As James McGrath explains in “Which John? The Elder, the Seer, and the Apostle,” by the time of Irenaeus in the late-2nd century CE, different figures in the early church named “John”—such as John the son of Zebedee, John of Patmos, and John the Presbyter (or, “elder” John)—had become conflated in their identities. One of the reasons for differentiating John the Elder from John the son of Zebedee, McGrath explains, is the fact that “Papias mentioned his efforts to find out what a variety of key figures, including John the apostle, said (using the past tense), and also what Aristion and John the Elder say (using the present tense).” This change in tense suggests that Papias is referring to John the son of Zebedee as a past figure, whereas the John the Elder was a subsequent figure, alive during his own day. Due to this ambiguity, and the general unreliability of the available sources, it is untenable that either Papias or Polycarp knew John the son of Zebedee, or any of the twelve disciples.
 Michael Kok (The Gospel on the Margins, p. 193) notes that only a minority of scholars accept that Papias is referring to a collection of sayings in his statement about an Aramaic version of Matthew:
A minority of scholars equate the λογια with “Q” in contrast to the words and deeds (λεχθεντα η πραχθεντα) that make up Mark. A better term for a sayings collection is λογοι (sayings) rather than λογια (oracles) and, from Papias’ title and the comparison with a narrative Gospel like Mark, λογια seems interchangeable with sayings and deeds.
Kok (p. 195) further points out, however, that there are other possibilities for a type of Matthean source material that Papias is conflating with the Gospel of Matthew:
Dennis MacDonald [Two Shipwrecked Gospels, p. 15] argues that Papias’ postulation of a lost Aramaic original was a device to explain away differences between two Greek texts—Matthew and a Matthew-like source (Q?)—while Matthew Black [“Rhetorical,” pp. 38-39] and Maurice Casey [Jesus, pp. 80-86] are open to an Aramaic vorlage behind some of the double tradition against most Q specialists. It makes sense that Papias confused Matthew’s sources with the notion of an Aramaic original of canonical Matthew, even if he was mistaken.
It is plausible that such source material, used during the composition of Matthew, may have been authored by the disciple Matthew. It remains highly obscure, however, what the nature of this source was and whether we can identify any specific passages in the Matthew that derive from it. As such, even if the disciple Matthew did author such a source, it would still be highly ambiguous to what extent his personal “eyewitness” experiences made it into the final version of Matthew. But, if such a source was later conflated with the final version of Matthew, it could explain why the disciple Matthew was associated with this gospel, even if he did not write the final version of the text itself.
Another possibility (although I think that it is quite unlikely) is that there had been an Aramaic version of Matthew, and not just a collection of sayings or source material, that had preceded the Greek version of the gospel. If so, this version may been translated into Greek. But, as George Kennedy notes, “translation” in the ancient world could refer to a much more creative adaptation of a work than the literal “word for word” understanding of translation that we have today. As Kennedy (“Classical and Christian Source Criticism,” p. 153) explains:
When a work was translated from one language into another, existing traditions in the second language often exercised influence on the form and style of the work, and considerable freedom of rearrangement or restatement was possible even if not inevitable.
If this is the case, there could have been substantial differences between this hypothetical Aramaic version of Matthew and the Greek text as we have it today. (Kennedy also argues that this translation would have been heavily influenced by Mark, and so, many of the stories in Matthew would have likely not derived from this Aramaic version, but were rather adapted from Mark.) Since it is not easy to identify and sift which passage in our current version of Matthew go back to this hypothetical Aramaic text, I do not think that scholars can reconstruct this Aramaic version of Matthew with any certainty. Kennedy (p. 149) also notes that the titles for both Matthew and John may not refer to the authors of the works themselves, but rather to “schools” or traditions that were associated with their names. As such, even if there was an Aramaic version of Matthew, it could have easily been authored by someone other than the disciple Matthew.
Given these possibilities, and the general uncertainty of the evidence itself, I do not think that we can treat the overall content of Matthew as derivative of the disciple Matthew’s eyewitness experiences. Likewise, as discussed by Richard Bauckham in endnote 14 above, there is considerable ambiguity over the disciple Matthew’s very identity, which adds yet more uncertainty to the origins of the text itself.
 It is possible that the “we” sections may derive from an earlier literary source that was used by the author of Acts when constructing the narrative. This possibility has been argued by Stanley Porter in “The ‘We’ Passages in Acts as a Source regarding Paul” (chapter 2 of The Paul of Acts). This source may have been authored by a companion of Paul, and possibly even Luke. Such a source would not entail, however, that the author of Luke-Acts was an actual eyewitness of Paul, or had witnessed the events in Acts. Rather, the author of Acts may have drawn upon (limited) source material that possibly derived from an eyewitness, which does not entail that the bulk of Acts is based on eyewitness experience. If Luke did author such a literary source, however, it could explain why his name was later conflated with the authorship of Luke-Acts, even if he was not the final author of the text.
 David Trobisch likewise points out in The First Edition of the New Testament that the names attached to each gospel can easily be shown to have derived from internal passages within the New Testament, which a later editor used to assign titular labels to each book. Trobisch (p. 47) argues that the title for Matthew was based on the name change from “Levi” to “Matthew” in Matthew 9:9. As discussed in endnote 14 above, Levi and Matthew are probably not even the same person, and it is also unlikely that the disciple Matthew made this name change, rather than an unknown writer. Trobisch (pp. 49-51) also points out that Mark and Luke are even named together in 2 Timothy 4:11, and that based on cross-references with other passages in the New Testament, these figures were assigned to the second and third gospels. As for the Gospel of John, Trobisch (p. 53) explains:
All the readers … have to do is consult the other three Gospels. The preceding synoptic Gospels inform the readers that Jesus occasionally chose to confide in only three of the twelve disciples; these were Peter and the two sons of Zebedee, John and James … If the readers suspect the beloved disciple to be one of these three preferred disciples, they should be able to conclude with confidence that the beloved disciple is identical with John, son of Zebedee. Peter is eliminated as the possible author of the Fourth Gospel since he is mentioned next to the beloved disciple in the final chapter and in the scene depicting the Last Supper … James can be excluded from further consideration because according to Acts 12:2 the Zebedee James died early … Jn 21, however, presupposes that the beloved disciple outlived all the other disciples.
As such, it is far more likely that the authorial attributions were based on speculation over internal passages within the New Testament, rather than earlier traditions that preserved knowledge of the Gospels’ authorship. Trobisch also argues that the titles are meant to call attention to individual books of first New Testament edition. Accordingly, they probably serve more as an organizing paradigm, rather than as biographical authorial traditions.
 There are also many other later Christian texts that were attributed to obscure figures, despite Blomberg’s assertion that unlikely candidates would not be chosen if an attribution was invented. As Robert Price (The Case Against The Case For Christ, p. 19) elaborates, “In fact apocryphal (which only means ‘not on the official list’ for whatever reason) gospels are attributed to such luminaries as Bartholomew, Judas Iscariot, the prostitute Mary Magdalene, doubting Thomas, the heretical Basilides, the even more heretical Valentinus, Nicodemus, and the replacement Matthias. They didn’t always go for the star names.”
 A common apologetic slogan about the church fathers’ attributions is that they allegedly “universally agreed” upon Matthew, Mark, Luke, and John, and thus were not speculating about the authors. However, this is only true of the later church fathers from the latter half of the 2nd century CE onward into the 3rd and 4th centuries—such as Irenaeus, Tertullian, Origen, and Eusebius—whereas the earliest references to the texts treat them anonymously (likewise discussed in endnote 20 above). Ignatius (c. 105-115 CE), Polycarp (c. 110-140 CE), the Epistle of Barnabas (80-120 CE), and the Didache (c. 50-120 CE), for example, allude to or quote the Gospels anonymously. As New Testament scholar Bart Ehrman explains, “in any of the writings of the Apostolic Fathers—ten proto-orthodox writers, most of them from the first half of the second century”—the Gospels are not identified by their traditional names, but are treated anonymously. Later, Justin Martyr (c. 150-160 CE) refers to the Gospels collectively as the “Memoirs of the Apostles.” Later still, Irenaeus (c. 175-185 CE) finally refers to the four Gospel canon with the names of its traditional authors.
This trail reflects a process in which the Gospels were gradually associated with the apostles, until eventually being attributed to specific names, when there were canonical disputes over the status of Christian texts during the latter half of the 2nd century CE. Sometimes apologists will further claim that, if the attributions were invented, we should expect to see multiple names proposed for the Gospels. However, there is little reason to expect this. If there was only one canonizing movement (or one published canonical edition, such as what David Trobisch argues in The First Edition of the New Testament), then there would only be one set of names attributed to the anonymous works, whereas multiple attributions would only be expected if there were separate, conflicting canonizing movements (or separate canonical editions that had been published).
Sometimes it is also argued that, if there was a such a canonizing movement, it should have left some trace in later Christian writings. As David Trobisch (The First Edition of the New Testament, pp. 41-42) explains:
Hengel argues that even if it could be conceded for a moment that a widely successful effort actually had been made to declare a certain selection of writings as authoritative for all, it would be all the more surprising that not one record of this event has survived, not even in the form of legend.
This need not be the case, however, if the canon was first chosen only as an edition of published works, which is quite different from the church councils that exercised control over dogmas and creeds in later centuries. As Trobisch (pp. 42-43) argues:
[A]lthough it is accurate to state that the observance of dogmas and creeds was later enforced by certain hierarchical and centralized structures of the church, this did not hold true concerning early Christian literature. The Canonical Edition was not the only work of the second century distributed specifically to Christians. There were Polycarp’s edition of the Letters of Ignatius; Marcion’s Bible; the Greek editions of Jewish Scriptures published under the names Theodotion, Aquila, and Symmachus; the Didache; the Letter of Barnabas; the Shepherd of Hermas; the Acts of Paul; and the Gospel of Thomas, to name only a few. They did not need an authoritative endorsement by the church to be produced, sold, bought, or read all over the Christian world. Hengel’s presupposition is that an edition of alleged apostolic writings could be distributed successfully among Christians only if backed by a central church authority. But this is clearly disproved by the evidence of the rich Christian literature of the time. In my view, the fact there is no record of a global church decision indicates that the editio princeps of the Canonical Edition was probably just one more ambitious Christian publication of the second century, one that faced strong competition.
As such, since the titles of the Gospels would have gone back to a published edition, rather than to any hierarchical decision in the church, there would not need to be any evidence of such a global canonizing decision, for this published edition to have later exerted influence over the titles and traditional names, which begin to appear on manuscript copies in the 3rd century CE. This edition would also explain why the titles are uniform in their names.
But furthermore, it is also not true that there were no other names proposed. The early 3rd century Roman presbyter Gaius attributed both the Gospel of John and Revelation to the authorship of the Gnostic Cerinthus (Epiphanius, Panarion 51.3.1-2). Likewise, Marcion had edited an abridged version of the Gospel of Luke, which he did not attribute to any named author. As Tertullian (Adversus Marcion 4.2.3) reports:
Marcion, in contrast, attributes no author to the gospel, that is, his own gospel, as if for the same man to whom it was not a crime to alter the very body of the text, it were not also permitted to affix a title. On this point I might have taken a stand, contending that a work ought not to be recognized, which does not erect its head, which displays no courage, and which offers no proof of credibility from the fullness of its title and the requisite authority of its author.
So, indeed, there are other attributions that survive in the record, and if the writings of “heretical” authors had not been stamped out by the orthodox church in Late Antiquity, there could have possibly been more. The sheer fact that the church fathers wrote polemics against figures like Marcion and Gaius of Rome testifies to the fact there were, indeed, disputes about the authorial attributions. In contrast, no ancient source claims that the Histories was written by a different author than Tacitus.
Finally, there are many other interpretations of the Gospels found in early church tradition that modern scholars dispute. For example, during the formation of the New Testament canon, the early church agreed upon Matthean priority, thus placing the Gospel of Matthew first in the New Testament. However, the majority of modern scholars (and even most apologists) through source analysis and redaction criticism agree upon Markan priority, and that the Gospel of Matthew was written after Mark and used Mark. This is almost as radical a deviation from church tradition as doubting its authorial attributions, showing that scholars are not exercising any excessive skepticism when doubting the traditional authors the Gospels, as there are many other claims found in church tradition that modern scholars dispute. For a discussion of why Markan priority is the majority scholarly view, see Michael Kok’s “Markan Priority or Posterity?“
 Justin Martyr’s Dialogue with Trypho (106.3) is sometimes cited as evidence that Justin believed the Gospel of Mark had been written based on the recollections of Peter. The passage refers to Peter’s name being changed from Simon to “Peter,” as well as James and John, the sons of Zebedee, being given the name “Boanerges,” which means “sons of thunder.” This could be a reference to Mark 3:16-17, which includes mention of both name changes. Justin does not explicitly refer to the Gospel of Mark (which he never calls by that name) in this passage, however, and instead claims that this information was found εν τοις απομνημονευμασιν αυτου (“in the memoirs of him”), adding considerable ambiguity to what text Justin is citing here.
To begin with, Justin usually refers to the Gospels under the collective title εν τοις απομνημονευμασι των αποστολων αυτου (“in the memoirs of the apostles of him”). What I have bolded are two words that are missing from the title in 106.3. When this title is normally used, Justin refers to the Gospels collectively as “memoirs of the apostles,” which are “of Jesus,” i.e., “about Jesus.” This is not a title that singles out the specific author of a text. In 106.3, however, Justin only has εν τοις απομνημονευμασιν αυτου (“in the memoirs of him”), which could indicate a particular author or source for the text, referred to as “of him,” i.e., “his memoirs.” This could be read as “Peter’s memoirs,” which may suggest that Justin is familiar with the Petrine tradition about the authorship of Mark. A difficulty with this reading, however, is that elsewhere Justin does not single out particular authors or sources in this way.
In fact, in 106.1 just above this line, Justin refers to the Gospels collectively as εν τοις απομνημονευμασι των αποστολων (“in the memoirs of the apostles”), and in 106.4 just below this line, Justin also has εν τοις απομνημονευμασι των αποστολων αυτου (“in the memoirs of the apostles of him”). This could suggest that there is a scribal error in the text at 106.3. All that it would take to change the Greek in this passage is to remove “of the apostles” (των αποστολων), and what is left is “the memoirs of him.” Such a change could happen very easily, if a later copyist (who was perhaps familiar with the Petrine tradition about Mark) saw what he considered to be a reference Mark 3:16-17, and simply removed the των αποστολων, so that the passage became “the memoirs of him,” i.e., “Peter’s memoirs.” This would not entail, however, that Justin had originally referred to the text in that way, if his original Greek simply had εν τοις απομνημονευμασι των αποστολων αυτου (“in the memoirs of the apostles of him”). The weak nature of the textual evidence in this passage, therefore, makes it difficult to adduce from it that Justin is referring to a specific gospel’s attribution.
Nevertheless, textual critics of Dialogue with Trypho (106.3), such as Egard Goodspeed (Die ältesten Apologeten, p. 222), list the Greek as εν τοις απομνημονευμασιν αυτου (“in the memoirs of him”), which should be taken into consideration, even if a scribal error is possible in this passage. If Justin originally wrote that Peter’s name change, as well as the sons of Zebedee being called “Boanerges,” was found in “his memoirs,” there are three possibilities that scholars have proposed for the meaning of this passage:
- The passage means “the memoirs of Jesus.” This view has been advocated by New Testament scholar Paul Foster. Under this interpretation, the αυτου (“of him”) is serving as an objective, rather than subjective, genitive, meaning that the memoirs are “about Jesus,” rather than being written or possessed by Jesus. If this interpretation is correct, then Justin is probably referring to the Gospel materials collectively as sources about Jesus, and is not singling out a specific gospel, let alone claiming that the Gospel of Mark was based on the recollections of Peter. To strengthen this view, it should be noted that Justin elsewhere uses the objective genitive when he refers to the Gospels as “memoirs of the apostles of Jesus,” i.e., “about Jesus.” Even though 106.3 is lacking “of the apostles” (των αποστολων), as noted above, the passage may still retain this use of the objective genitive, so that it refers to “the memoirs of Jesus,” and not to the specific author or source for the text.
- A challenge to the above interpretation is the fact that “Peter” is the nearest antecedent to the αυτου (“of him”) in the passage, rather than Jesus, meaning that the αυτου could be referring to Peter as a subjective genitive. In that case, the passage means “the memoirs of Peter.” There is still considerable ambiguity at this point, however, since Justin could be referring to the apocryphal Gospel of Peter rather than the Gospel of Mark. This view has been advocated by Bart Ehrman. If that is the case, Justin is not referring to the Gospel of Mark at all, nor claiming that the text was based on the recollections of Peter; however, a major challenge to this view is the fact that our surviving text of the Gospel of Peter is no longer fully extant, and the opening chapters are lost. The reference to Peter’s name change, as well as the sons of Zebedee being called “Boanerges,” is not included in our surviving Gospel of Peter, but it may have been in an earlier lost portion of the text. Since the opening chapters of the text are lost, however, it is impossible to determine whether this story about the name changes was ever included within the Gospel of Peter.
- The passage means “the memoirs of Peter” and is quoting Mark 3:16-17, meaning that Justin is referring to the Gospel of Mark by this description. This view has been advocated by New Testament scholar Graham Stanton. Under this third option, Justin may have believed that the Gospel of Mark was based on the recollections of Peter. Even if this interpretation is correct, however, Justin probably drew this connection from the writings of Papias (Historia Ecclesiastica 3.39.15), who claimed that John Mark had written a text based on the teachings of Peter. (Justin may have also been dependent on a common source with Papias.) In any event, Papias does not quote from our Gospel of Mark, nor does he even appear to have seen the text that he claims was authored by John Mark, since Papias states (Historia Ecclesiastica 3.39.15) that he had only learned of the text from John the Presbyter. Because of this, the text that Papias is describing cannot be identified with any certainty as the Gospel of Mark. If Justin Martyr is referring to the Gospel of Mark as “the memoirs of Peter,” therefore, he may have been the first to suggest that the Gospel of Mark was the unknown text that Papias was referring to. This would have little significance for the analysis above, however, since it has already been argued that Irenaeus (c. 185 CE) probably drew a connection between the Gospel of Mark and the unknown text that Papias claims was authored by John Mark. If Justin (c. 150-160 CE) made this connection before Irenaeus, then that would only mean that the connection was made a couple decades earlier; it would not necessarily mean that Justin corroborates Papias, especially since Papias does not seem to have personally seen the text that John the Presbyter described to him in Historia Ecclesiastica 3.39.15. Likewise, even if Papias is referring to our text of Mark, and Justin is likewise, as Michael Kok notes above, their common source would have still been John the Presbyter (and those dependent upon his testimony), whose identity and reliability remains completely unknown.
Another possibility worth noting is that Justin Martyr may have believed that the Gospel of Mark was based on the recollections of Peter, but did not think that it was written by John Mark. If Justin is referring to “his (Peter’s) memoirs” in Dialogue with Trypho (106.3), he still does not say that this text was authored by John Mark. This may suggest a development in which the Gospel of Mark was first attributed to Peter, but was later attributed to one of Peter’s attendants. A possible motive for changing the attribution was to change the canonical status of the text. As Michael Kok discusses in The Gospel on the Margins (which I summarize in a book review here), the gospel attributed to John Mark was the least popular among the church fathers of the 2nd-3rd centuries in terms of quotations and theology. To downgrade the importance of the text to doctrine, therefore, the church fathers may have attributed the text to a minor figure only roughly connected with Peter. In this case, the attribution to John Mark would be spurious. It is also worth noting, under this scenario, that spuriously attributing a gospel first to Peter would not be in the least improbable, due to the fact that Peter was a very important figure in the early church. If the Gospel of Mark was first attributed to Peter, and later to John Mark, therefore, it is probably the case that the attribution to Peter was spurious as well.
Regardless, Justin does not refer to the Gospel of Mark by name in Dialogue with Trypho 106.3, nor does he refer to any of the Gospels (both canonical and apocryphal) by their traditional names. Instead, Justin uses the formula απομνημονευματα των αποστολων (“memoirs of the apostles”) to refer to the Gospels, which provides evidence that their traditional titles and named appellations had not been attached to the texts by 150-160 CE. Otherwise, Justin would have called them by their traditional names, as Irenaeus (185 CE) and the Muratorian Canon (c. 170-200 CE) did a couple decades later. But, as is noted in endnote 5 above, the Gospels’ titles were probably not added until c. 150-185 CE, possibly from an edition that was published at Rome around that time.
Even if Justin is describing to the Gospel of Mark when he refers to “the memoirs of Peter,” therefore, that still provides no evidence that the Gospel of Mark had been given its traditional title by Justin’s time. Likewise, Michael Kok (The Gospel on the Margins, p. 115) argues, “There is a good probability that Justin was familiar with the Papian tradition on Mark,” which means that Justin is not independently reporting any information beyond the problematic tradition of Papias, discussed under the “External Evidence” section above. But regardless, the meaning of απομνημονευμασιν αυτου (“the memoirs of him”) in Dialogue with Trypho 106.3 is ambiguous to begin with, and could very well not be referring to the Gospel of Mark, or even Peter’s memoirs, at all.
 The internal evidence for Lazarus as the “beloved disciple” is so strong that even conservative New Testament scholar Ben Witherington supports this interpretation of the figure’s identity. In “Was Lazarus the Beloved Disciple?,” Witherington explains:
It has been common in Johannine commentaries to suggest that the Beloved Disciple as a figure in the narrative does not show up under that title before John 13. While this case has been argued thoroughly, it overlooks something very important. This Gospel was written in an oral culture for use with non-Christians as a sort of teaching tool to lead them to faith. It was not intended to be handed out as a tract to the non-believer but nevertheless its stories were meant to be used orally for evangelism. In an oral document of this sort, the ordering of things is especially important. Figures once introduced into the narrative by name and title or name and identifying phrase may thereafter be only identified by one or the other since economy of words is at a premium when one is writing a document of this size on a piece of papyrus (Jn. 20.30-31). This brings us to John 11.3 and the phrase hon phileis [‘the one whom you love’]. It is perfectly clear from a comparison of 11.1 and 3 that the sick person in question first called Lazarus of Bethany and then called ‘the one whom you love’ is the same person as in the context the mention of sickness in each verse makes this identification certain. This is the first time in this entire Gospel that any particular person is said to have been loved by Jesus. Indeed one could argue that this is the only named person in the whole Gospel about whom this is specifically said directly. This brings us to Jn. 13.23.
At John 13.23 we have the by now very familiar reference to a disciple whom Jesus loved (hon agapa this time) as reclining on the bosom of Jesus, by which is meant he is reclining on the same couch as Jesus. The disciple is not named here, and notice that nowhere in John 13 is it said that this meal transpired in Jerusalem. It could just as well have transpired in the nearby town of Bethany and this need not even be an account of the Passover meal. Jn. 13.1 in fact says it was a meal that transpired before the Passover meal. This brings us to a crucial juncture in this discussion. In Jn. 11 there was a reference to a beloved disciple named Lazarus. In Jn. 12 there was a mention of a meal at the house of Lazarus. If someone was hearing these tales in this order without access to the Synoptic Gospels it would be natural to conclude that the person reclining with Jesus in Jn. 13 was Lazarus. There is another good reason to do so as well. It was the custom in this sort of dining that the host would recline with or next to the chief guest. The story as we have it told in Jn. 13 likely implies that the Beloved Disciple is the host then. But this in turn means he must have a house in the vicinity of Jerusalem. This in turn probably eliminates all the Galilean disciples.
Notably, Witherington interprets Lazarus as a historical person who authored John (which he argues was later edited by John of Patmos), whereas this article favors the view that the role of Lazarus in John is probably based on a redaction of an allegorical Lazarus in Luke 16:20-31. Given this possibility, it is untenable that the Lazarus described in John was actually a historical person, rather than just an allegorical character or literary invention.
Witherington also notes a number of other problems with the tradition that John the son of Zebedee authored the fourth gospel:
One of the things which is probably fatal to the theory that John son of Zebedee is the Beloved Disciple and also the author of this entire document is that none, and I do mean none, of the special Zebedee stories are included in the Fourth Gospel (e.g., the calling of the Zebedees by Jesus, their presence with Jesus in the house where Jesus raised Jairus’ daughter, the story of the Transfiguration, and also of the special request for special seats in Jesus’ kingdom when it comes, and we could go on). In view of the fact that this Gospel places some stress on the role of eyewitness testimony (see especially Jn. 19-21) it is passing strange that these stories would be omitted if this Gospel was by John of Zebedee, or even if he was its primary source. It is equally strange that the Zebedees are so briefly mentioned in this Gospel as such (see Jn. 21.2) and John is never equated with the Beloved Disciple even in the appendix in John 21 (cf. vs. 2 and 7— the Beloved Disciple could certainly be one of the two unnamed disciples mentioned in vs. 2).
 Other proposed candidates for the author of the fourth gospel include John the Presbyter, John Mark, and Thomas. Regardless, even Christian scholar Raymond Brown (An Introduction to the New Testament, pp. 368-369) explains, “As with the other Gospels it is doubted by most scholars that this Gospel was written by an eyewitness of the public ministry of Jesus.” What is further noteworthy is that even many conservative New Testament scholars doubt the traditional authorship of John. Ben Witherington identifies Lazarus as the intended author of the text (discussed in endnote 30 above), and Martin Hengel and Richard Bauckham have proposed John the Presbyter, in place of John the son of Zebedee.
It should also be noted that Mark Goodacre (“NT Pod 38: Who is the Beloved Disciple in John’s Gospel?“), who does think that the author of the fourth gospel is trying to suggest that the beloved disciple is John the son of Zebedee, still does not think that this John authored the gospel. Goodacre notes how the entire use of the beloved disciple in the Gospel of John is highly ambivalent, and how this likely reflects the fact that even the narrator of the text is not making an overt authorial attribution. Another possible explanation of why the beloved disciple’s identity is kept anonymous is because the text is not referring to a historical figure at all, but instead is using the beloved disciple as a fictive narrative device that makes him the ideal witness for Christian readers to emulate.
 It is not clear that the “beloved disciple” described at the end of John is even intended to be understood as the author of the work. 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.”
 Likewise, while church tradition maintains that John the son of Zebedee lived to a very old age, there is also a body of ancient evidence indicating that he died much earlier, being executed alongside his brother James, whose martyrdom is described in Acts 12:2. This body of evidence indicating that John did not live to old age is laid out by scholar F.P. Badham in “The Martyrdom of John the Apostle.” Likewise, even conservative New Testament scholar Ben Witherington (“Was Lazarus the Beloved Disciple?”) explains:
Papias Fragment 10.17 has now been subjected to detailed analysis by M. Oberweis (NovT 38 1996), and Oberweis, rightly in my judgment draws the conclusion that Papias claimed that John son of Zebedee died early as a martyr like his brother (Acts 12.2). This counts against both the theory that John of Patmos was John of Zebedee and the theory that the latter wrote the Fourth Gospel.
While it is historically uncertain whether John died alongside James, this body of evidence casts doubt on the tradition that John lived to an old age and thus raises further problems for the notion that John authored the fourth and latest gospel. As I explain in my essay “March to Martyrdom,” our evidence for any of these church figures is very, very limited. In light of such problematic evidence, in addition to contradictions among our sources, it is not tenable that John the son of Zebedee ever lived to an old enough age to author the gospel later attributed to him. This is just another problem for the authorial tradition of the fourth gospel, in addition to the numerous other ones listed above.
 Sometimes apologists cite Irenaeus’ letter to Florinus (Eusibius Historia Ecclesiastica 5.20) as evidence that he knew, on the basis of good authority, that John the disciple authored the fourth gospel. However, this argument is based largely on speculation. In the letter, Irenaeus states that he knew Polycarp as a child. He also states that Polycarp was a companion of John the disciple. The logic goes that, since Polycarp knew John, he must have told Irenaeus (when he was a kid) that John authored the gospel attributed to his name. However, as scholar R. Alan Culpepper (John, the Son of Zebedee: The Life of a Legend, p. 126) explains, “In this excerpt from the letter, Irenaeus reminds Florinus of their common experience, sitting at the feet of Polycarp. His point is to remind Florinus that he did not learn his Gnostic views from Polycarp … On the other hand, Irenaeus does not say that Polycarp taught that the apostle John was the author of the Fourth Gospel, the Epistles, or Revelation.”
That being said, it has been argued by David Trobisch (“Who Published the New Testament?,” p. 33) that Polycarp may have been responsible for the publication of the first edition of the New Testament. Trobisch acknowledges, however, that this argument is based on probability and not certainty. Trobisch’s broader thesis is that the church in Asia Minor, during the mid-2nd century CE, had a strong influence over the selection of the canonical books. Since Polycarp was an authoritative figure in this region and during the same time period, Trobisch suggests that he may have been the one who selected the names of the Gospel titles, along with the other books of the New Testament. There is a catch, however: Trobisch also argues that the names of the Gospels were selected on the basis of internal passages within the New Testament, as discussed in endnote 26 above. As such, Polycarp would have more likely chosen the names as part of an editorial process, linking the books together, rather than on the basis of personal knowledge that he had of the disciples.
There are also major problems for assuming that Polycarp personally knew John the son of Zebedee. First, as discussed in endnote 33 above, there is a body of ancient evidence suggesting the John died alongside his brother James in 44 CE. Polycarp likewise did not write anything that we can date with certainty prior to c. 110-140 CE. This creates a rather problematic chronology, if the traditions implying an early death of the disciple John are accurate. But it should likewise be noted that, even if John had not been martyred with James, it is still doubtful that he lived and traveled long enough to know Polycarp, who was active in Asia Minor around the mid-2nd century. The sources claiming that John the son of Zebedee traveled to Ephesus (e.g., the Acts of John), and lived to a very old age, are primarily based on later traditions intended to grant special importance and authority to the church at Ephesus. As I explain in my essay, “March to Martyrdom,” our sources for what happened to any of the apostles, after the Book of Acts, are highly problematic, and full of legendary (and often contradictory) information. This situation likewise applies to John the son of Zebedee. Polycarp also does not state in his own writing that he knew or traveled with John or any of the apostles. Irenaeus mentions this detail, but it is likely to aggrandize Polycarp.
Why then was Polycarp associated with John? A far more likely explanation is that he actually knew John the Presbyter (which is argued in endnote 22 above). As discussed by New Testament scholar James McGrath in “Which John? The Elder, the Seer, and the Apostle,” there were several figures named “John” in the early church, whose identities became conflated in the 2nd century and onward, including during the time of Irenaeus. What is very likely the case, therefore, is that Polycarp knew a leading authority named “John,” who was later conflated with the disciple John the son of Zebedee. This conflation likewise happened with Papias, as Michael Kok discusses under the “External Evidence” section above. Since Papias only knew John the Presbyter, or “elder John,” it is likewise probable that Polycarp only knew this figure, as well. And, if that’s the case, it would also explain where Polycarp got the names “Matthew” and “Mark” for the first and second gospels, since these authorial traditions, as Kok explains, derive from John the Presbyter (who likewise, at least in the case of Mark, appears to have derived the name from internal references within other books of the New Testament). In such a case, Polycarp would have only repeated the dubious Papian tradition for the authorship of Matthew and Mark, discussed above, when he assembled the New Testament canon.
As noted, Trobisch’s argument that Polycarp assembled the canon is purely probabilistic. It should also be noted that Polycarp does not refer to the Gospels by their traditional names in his own writing (see here), which is part of the evidence (in addition to other anonymous quotations among the early church fathers) that the Gospels had not yet been given their traditional titles by the mid-second century CE (discussed further in endnote 20 above). Even if Polycarp was responsible for adding the titles after this date, therefore, he still provides evidence that the titular affixation did not take place until the late-2nd century.
 On this point, it is worth noting that Christian apologist Mike Licona (in his review of Bart Ehrman’s Forged) has compared the authorial traditions for the Gospels with Plutarch’s biographies. Licona argues:
Something else must be considered. There were many biographies written in antiquity. Plutarch was one of the most prolific biographers of that time, writing more than 60 biographies of which we still have. It is of importance to observe that Plutarch’s name is absent from all of his extant biographies, which are therefore anonymous like the four Gospels in the New Testament. Yet, modern historians are quite certain Plutarch wrote them. Most classical authors did not include their name. But the manuscript traditions pertaining to the authorship of Plutarch’s biographies are clear. Moreover, the Lamprias catalogue from the fourth century attributes them to Plutarch. Does this provide us with unimpeachable evidence that Plutarch wrote the biographies attributed to him? No. Is it reasonable to believe that Plutarch wrote them? You bet. The same may be said concerning the four Gospels in the New Testament. The traditions concerning the traditional authorship of the Gospels begin within 30 years of the final of the four to be written and continues without debate for centuries. Thus, Ehrman’s argument from the anonymity of the autographs of the four Gospels carries little if any weight.
First, it should be noted that there were debates over the authorship of the Gospels, since (as discussed in endnote 28 above) Marcion did not corroborate the authorship of Luke and Gaius of Rome instead argued that Cerinthus authored the Gospel of John. But, more importantly, Licona’s argument is riddled with methodological problems.
To begin with, there were other works of Plutarch that were attributed to him by external sources much earlier than his biographies. As Marianne Pade (“The Reception of Plutarch from Antiquity to the Italian Renaissance,” p. 532) explains, Aulus Gellius (c. 130-180 CE) identified Plutarch (c. 45-120 CE) as the author of his Moralia in his Attic Nights (17.11.1-6), only about half a century after he was composing (not multiple centuries later). Since Plutarch wrote in a distinct Greek style, we can compare the Moralia with other works that are attributed to him, such as his biographies. Scholars use similar methods when evaluating New Testament authorship, such as in assessing the authorship of Paul’s epistles. A major reason why scholars think that the 7 undisputed letters are genuinely Pauline is because they are written in a very similar Greek style, suggesting a common author. Scholars could thus use similar methods to compare Plutarch’s Moralia with his biographies, and so, the external evidence for the authorship of Plutarch’s biographies would not hinge solely on a catalogue dating from the 4th century CE, since this source would need to be assessed alongside the external evidence for Plutarch’s other works.
But even more importantly, Licona is making a very shallow quantitative argument, when the issue of authorship is far more qualitative. Whether the first external source to mention a text’s authorship dates to decades or centuries after the text’s composition is not the only consideration that is factored in to why scholars consider authorial attributions to be reliable or unreliable. Sometimes the earliest external quotations of a text can count against its traditional authorship, if sources quote the text anonymously or refer to it by a different name. For example, the Didache quotes Matthew, but refers to it as “His (Jesus’) Gospel,” and not the Gospel of Matthew. This preserves a trace of an original, anonymous title, suggesting that the attribution to Matthew was added later. Even if the first external source to mention the authorship of Plutarch’s biographies dates to centuries later, therefore, if there are no earlier sources calling his works by a different name, then the external evidence for this later attribution would still be stronger than for an earlier attribution that was preceded by quotations calling the text by a different name. Since the Gospels are all quoted anonymously or referred to by different titles until the latter half of the 2nd century CE, therefore, before they receive their named attributions from sources like Irenaeus and the Muratorian Canon, this strongly suggests that their traditional names were added later. In contrast, if a text is simply not mentioned by external sources for a couple centuries, and then the author is mentioned by the first external source to discuss the work, this attribution would still be stronger (even if dating later), since there would be no trace of earlier sources calling the text by a different name.
But there are also several more considerations that would need to be factored in to Licona’s comparison. For example, were Plutarch’s biographies attributed within a context in which multiple forgeries and false attributions were being made? If not, there would be greater reason to take his attribution at face value. In contrast, if you were to take all of the works that were attributed to Jesus’ disciples and their followers from the 1st-4th centuries CE (including works like the Gospel of Peter, the Gospel of Thomas, etc.), even apologists would agree that the vast majority were falsely attributed. And so, the canonical Gospels were attributed under circumstances that would have made false attributions far more likely. This, at the very least, means that we need to treat the Gospels with greater scrutiny than works that were attributed under circumstances in which forgery and false attribution were less present.
Another consideration is the wording of the title. Plutarch’s works are not attributed “according to” (κατα) Plutarch, and so, his identification of authorship is far more common. In contrast, as is discussed in endnote 5 above, the wording of the Gospels’ titles strongly suggests that they were a secondary addition. The formula “Gospel of Jesus” with names added “according to” individual authors suggests that, when the first gospel was written, it was simply called “the Gospel of Jesus.” When multiple gospels were in circulation, however, the formula “κατα (according to) + the author” was needed to specify individual works of a multiple gospel canon. This suggests that the named titles were a secondary addition to the Gospels. Since Plutarch’s titles do not suggest such a development, there would be less reason to suspect that his name was not attached to his works when they were first published.
Likewise, as discussed in endnotes 22, 24, and 25 above, the “according to” (κατα) formula may not even be referring to the final author of the text, but rather to a source or tradition that was connected with the affixed name. This is especially true in the case of Matthew, in which the author of the text makes no authorial interjections in the first person, but which has the name “Levi” changed to “Matthew” in Matthew 9:9. What Richard Bauckham suggests in endnote 14 above is that the disciple Matthew may have had some special connection with the text, which caused a later author to make the name change. This also is what probably led to the text being titled “according to Matthew,” as discussed in endnote 26 above. Perhaps Matthew had authored an earlier source material used during the composition of the text, or perhaps the connection is even pseudonymous. Regardless, the relationship that is being designated by the “according to” (κατα) formula is hardly a clear claim to the final author of the text. This is not at all the case for the manuscripts and authorial traditions pertaining to Plutarch. Rather, Plutarch is clearly identified as the final author of his biographies, and so, his authorial tradition is a far more standard case of authorship than is the case for a text like Matthew.
Then there is the issue of literacy. Since Plutarch belonged to social elite demographics, it is far more likely that he would have had the literary training needed to author his biographies. John the son of Zebedee, in contrast, was only a Galilean fisherman, and so it is far less likely that he would have been capable of authoring a text like the Gospel of John. This is one reason why the authorship of most elite works from antiquity is more secure than the authorial attributions of (both canonical and apocryphal) Christian works that were attributed to figures like John, Peter, etc., who (despite being appealing authorial candidates for granting authority to texts) would probably have lacked the education needed to author them.
Likewise, Plutarch’s biographies are not “anonymous” in the same sense as the Gospels. Anonymity can mean that an author does not provide his name within the body of the text, but it can also refer to whether a text is written in the author’s own voice. As discussed above, authors like Tacitus make authorial interjections in the first person (even if they do not provide their name within the text), indicating that they are relating their own personal perspective. The Gospels, in contrast, mostly lack these authorial interjections, and instead are written in a collective, third person manner of narration. Plutarch uses the first person when discussing battle monuments that were located near the town of Chaeronea, which he states could be seen during his own time, in his Life of Alexander (9.3). This passage needs to be considered alongside the fact that Plutarch was said to be a native of Chaeronea. And so, this biography is not fully anonymous, since Plutarch appears to allude to his own eyewitness experience in discussing details about his home town, which we can use to corroborate external evidence claiming that he was an author from that town.
The means of publication also need to be considered, as discussed in endnote 7 above. Many elite works were professionally published by book dealers and kept in public libraries, under the author’s name. The author would also frequently recite his own works, or have someone recite them in his name. Because of this, the author of the text would be associated with it from the beginning of the text’s transmission. In contrast, there were also less sophisticated literary works in antiquity, which circulated anonymously. Since the Gospels are more typical of this latter category, they were probably first published in a very different context than Plutarch’s biographies.
Considerations like these mean that scholars cannot merely crunch numbers when assessing an authorial attribution. The nature of authorship is complex and qualitative. Even if the Gospels were attributed earlier than our first (surviving) external source that discusses the authorship of Plutarch’s biographies, therefore, the circumstances behind the evidence are still vastly different. And so, Licona’s response to Ehrman carries little if any weight as a comparison.
I have chosen to compare the authorship of the Gospels specifically to Tacitus, partly because of an article on the Christian apologetics website Tektonics (“Dates and Authorship of the Gospels“) that makes a similar argument comparing the authorship of Tacitus, which I strongly disagree with, and also because Pliny’s letters provide an interesting parallel with Paul’s letters and Luke-Acts, where we can use outside epistolary evidence to evaluate an authorial attribution. It should be noted, however, that Pliny’s letters provide very early (i.e., contemporary) testimony for Tacitus authoring his Histories, whereas for many Classical texts, such external evidence often does not appear so early. For this reason, I have used my same criteria in this article to also evaluate the authorship of Plutarch’s biographies in this endnote, which do not have external evidence of authorship that appears as early as Tacitus. Nevertheless, through the arguments listed above, I also think that we have a much stronger case that Plutarch authored his biographies than the problematic authorial attributions for the Gospels, and so, even if the external evidence for Plutarch’s authorship appears later, it can still be more reliable. This observation about Plutarch applies equally to the authorship of most elite Classical literature from Greco-Roman antiquity.
 A minority of mainstream scholars have defended the 2nd century attributions and “eyewitness” status of the canonical Gospels, perhaps most notably Richard Bauckham in Jesus and the Eyewitnesses. (Although, even Bauckham does not think that the disciple Matthew or John the son of Zebedee authored the final versions of the gospels attributed to them.) Bauckham’s study deals less with defending the authorship of the titular names affixed to the Gospels, as much as presenting arguments that eyewitnesses lie behind the traditions and sources in the Gospels. For example, in chapter 7 Bauckham argues that Peter plays a greater role in Mark than the other Gospels, and that the text thus reflects a Petrine perspective. This evidence is too vague, however, to suggest a direct connection with the disciple Peter, and it is more likely that Peter is simply a prominent character in the text, rather than a direct source whom the author consulted. As New Testament scholar Michael Kok (The Gospel on the Margins, p. 80) argues:
It is not wrong to stress Peter’s significance on the literary level, but neither the inclusio nor the plural-to-singular-narrative device proves that he was one of Mark’s informants.
Bauckham’s arguments have received considerable criticism by subsequent scholars, including critical reviews from New Testament scholars David Catchpole, Stephen J. Patterson, and Theodore J. Weeden Sr. in Journal for the Study of the Historical Jesus. A good summary of a number of the problems that scholars have raised with Bauckham’s arguments can be found here.
In contrast to the speculative traces of eyewitness informants that Bauckham seeks to identify behind the Gospels, historical Greco-Roman authors were far more explicit in identifying their eyewitness sources of information. For example, in my essay “Eyewitness Recollections in Greco-Roman Biography versus the Anonymity of the Gospels,” I discuss how every single author who wrote a historical biography during the early Roman Empire, specifically those writing about historical figures 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.
Finally, it should be noted that even if the Gospels had been written by eyewitnesses, or the companions of eyewitnesses, or if their authors had access to eyewitness informants, this would still not entail that many of the miracles reported about Jesus are either reliable or literal descriptions of real events. There are several other “eyewitness” accounts throughout history about equally extraordinary miracles that few apologists would defend. For example, Bart Ehrman has recently discussed the eyewitness testimony that exists for the Jewish miracle worker Baal Shem Tov (c. 1700-1760 CE), also called “Besht,” who is considered to be the founder of Hasidic Judaism. As Ehrman (“Another Jewish Miracle Worker“) explains:
Our principal source of information about the Besht comes in a series of anecdotes about his life written 54 years after his death, entitled In Praise of the Baal Shem Tov (in the Hebrew: Shivhe ha-Besht). The book was published in 1814 in Poland. Its author was Rabbi Dov Ber, who, as it turns out, was the son-in-law of a man who had been the personal scribe and secretary for the Besht, a rabbi called Alexander the Shohet. The book contains 251 short tales about the Besht. Fifteen of these are said to have come directly from Alexander; the rest come from other sources, including the rabbi of the author’s own community who had heard them from his own teacher.
Throughout the tales the Besh heals the sick, exorcises dybbuks (restless souls of dead people), and helps barren women conceive. He can ascend to heaven and miraculously shorten a journey. He is often shown to be superior to others he encounters: other rabbinic scholars, medical doctors, and sorcerers. While those outside the Hasidic tradition might consider these stories simply to be pious fictions, legendary accounts based on hearsay, started by gullible devotees, the author Dov Ber himself claims that they are rooted in reliable sources and relate historical realities. As he himself reflects, “I was careful to write down all the awesome things that I heard from truthful people. In each case I wrote down from whom I heard it. Thank God, who endowed me with memory, I neither added nor omitted anything. Every word is true and I did not change a word” …
…There are many, many tales such as these throughout the account. And what is my point? Do I think the Besht actually had supernatural powers and was able to do these things, to be transformed into a divine, glowing presence, to cast out and imprison demons, to ignite trees with his finger, to raise the dead, and all the rest? No, personally, I don’t believe it. But are the reports based ultimately on eyewitness reports? Well, writing some 55 years after the events the author claims they were indeed based on eyewitness testimony. Does that make them reliable?
Likewise, New Testament scholar Richard Miller in Resurrection and Reception in Early Christianity has recently argued that the account of Jesus’ resurrection in the Gospels are not literal descriptions of real events, but “translation fables” that provide narrative symbolism. The thesis of the books is:
Richard Miller contends that the earliest Christians would not have considered the New Testament accounts of Jesus’ resurrection to be literal or historical, but instead would have recognized this narrative as an instance of the trope of divine translation, common within the Hellenistic and Roman mythic traditions.
So, even if the Gospels were written by eyewitnesses, the companions of eyewitnesses, or had access to eyewitness informants, that would still hardly entail that they are literally describing historical miracles.
Copyright ©2017 Matthew Wade Ferguson. The electronic version is copyright ©2017 by Internet Infidels, Inc. with the written permission of Matthew Wade Ferguson. All rights reserved.