Many arguments about the reliability of the New Testament documents hinge on when they were written. This makes sense: documents written shortly after the events they describe might be reliable, while texts written centuries after everyone described is dead are more likely to be legendary than accurate. Fortunately, biblical scholars are (almost all) very confident about when the books of the New Testament were written, dating some of them within a year or two. But should we believe the dates the experts propose? Some might be inclined to do so based on the greater training of NT scholars, but as a historian, I wish to explore some of the reasons given for the dates, and evaluate for myself how valid those reasons are. My conclusion from doing so is that we know a lot less about when New Testament documents were written than the scholars claim. This claim is neither for nor against the truth of the Christian faith as reported in the New Testament; the documents might as easily be older than scholars claim as newer. It is simply a statement of the lack of evidence, leading to a more humble and open view of the past.
The only date or other attempt by an author of a New Testament document to specify precisely when something happened is included in Luke 3:1, which references the fifteenth year in the reign of Tiberius Caesar. How do we know when he reigned? Ancient chronology is a difficult field, but our notion of when Tiberius reigned is probably based on the ancient calendars used and the dates given by Roman historians such as Tacitus and Suetonius, and by the Jewish historian Josephus. On the other hand, since the date pertains to events decades before Jesus’s public ministry, this date does not provide a very tight bound as to when the gospel of Luke was written: we know it could not have been written very close to that date.
With no other dates to go by, the next most certain method for dating texts is references to events. A text must have been written after any events it describes (with the potential exception of prophecy). The famine under Claudius (Acts 11:28), Roman governors of Judea, and members of the Herod family, are all mentioned by other ancient authors, and give us a sense of when the book of Acts was written. Unfortunately, apart from the book of Acts, this method is less useful, because the events mentioned in the gospels are all in a brief span of time, while the letters (and even more so the book of Revelation) do not clearly refer to events whose date is otherwise known. From this method, we may conclude that Acts could not have been written earlier than around 62 CE, and the other New Testament documents could not have been written earlier than 33 CE or so (because almost all of them refer to Jesus’s death and resurrection). In some cases, we can estimate when Christianity first arrived in certain cities based on relative chronology in the book of Acts, and if Paul wrote a letter to that church, that gives a slightly later date after which the letter must have been written.
But what about prophecy? The bedrock date in NT dating is the destruction of the Temple in 70 CE, referred to by Christ in the gospels. Many scholars assume that such a reference implies the gospels must have been written after 70 CE. But this is to assume that Christ could not have prophesied the destruction, in other words, to assume the gospels are false. (As a side note, even atheist scholars do not need to assume prophesies are always written after their fulfillment.) I argued previously that Christ’s prophecy is of a sort which does not require having been uttered after the events described, even on relatively naturalistic assumptions. So the bedrock date in NT dating turns out not to be as useful for dating the documents at all.
All of the preceding methods are used to establish termini post quem, early bounds for the composition of the text. Any document must have been composed after the dates or events which it describes. But there are also methods for establishing later bounds before which the document must have been composed. The most direct is to know when the author died, as the author could not have written the text later. Unfortunately for the New Testament documents, scholars often (though not always) dispute the traditional authorship, and even where scholarly consensus supports the author identified by tradition, that author’s date of death is unknown. Paul’s letters must have been written before he died, but when did he die? According to Acts, not before around 62 CE, and according to later tradition, under Nero, and therefore not later than 68 CE. If the later tradition is true (I am not aware of scholars who dispute it), then Paul’s letters must have been composed by the mid-60s at the latest, but then, many scholars dispute the authorship of half or more of Paul’s letters.
The most incontrovertible method of establishing a later bound on the date of a text’s composition is to have a manuscript of that text for which the scribe indicated the date he finished copying. But dates in biblical manuscripts only became popular (or survive) from the 400s onward. We have many manuscripts older than 400, approximately dated on the basis of the shape of their letters, but these do not give a clear bound as to dating the composition of a text.
Another method for identifying a later bound for the composition of a text is external attestation: if a later author references an earlier text, the text must have been composed before the dates of the later author. Thus one can look at references to NT books in other early Christian writings. such as 1 Clement, the Didache, and the letters of Ignatius of Antioch. But this faces the difficult of asking how well we know when those were written. The Didache in particular is difficult to date, and 1 Clement is likewise dated mostly on the basis of tradition. Ignatius of Antioch is dated to the reign of Trajan on the basis of his traditional martyrdom. Other early second-century authors are equally uncertain. And even when Clement of Rome or the Didache quotes a portion of a NT text, skeptics have argued that such a quotation does not prove the existence of the whole gospel or letter which now contains that snippet.
Some scholars have argued that useful dating information can be gleaned from comparison with other undated texts, such as the Gnostic gospels. But such a method faces the double difficulty of determining when the comparison text was composed and determining the criteria for discerning which text was composed earlier or later. Similarly, scholars, especially early in the twentieth century, made bold pronouncements about which texts, even within the NT canon, were earlier or later based on their theological or ecclesiastical “development.” But this reasoning is circular, as it assigns dates on the basis of a developmental model, which itself depends upon lining up the evidence in order. How well do we know how Paul’s (or any other author’s) thought developed, apart from reading the texts in the order they were composed? How then can we determine the order in which these texts were composed on the basis of a developmental model? Or, to put it another way, where could we get a developmental model that would tell us when these texts were composed, if not from the order of the texts themselves? Even reasoning on the basis of developments in ecclesiastical hierarchy are precarious, given the potential for multiple different structures to exist in different places at the same time.
It seems to me, as a historian more than as a Christian, that biblical scholars are overly precise in their assertions about when various New Testament texts were written. We know they were all written after about 33 CE (and Acts after about 62 CE), and that they were all written before they are quoted in various second-century texts (many of them quoted by Ignatius of Antioch, during the reign of Trajan in the 110s at the latest). Texts actually written by the traditionally identified author (a category of disputed contents) must have been written by the time that author died (such as the mid-60s in the case of Paul). But beyond those somewhat loose bounds (which are still rather tight compared to various other undated ancient texts!) any more precise dating of New Testament texts is largely speculative. The texts might have been written toward the beginning of the range or toward the end, and which seems more probable is inseparable from each person’s assessment of the truth or falsity of the claims made about Jesus and the origins of what we have come to call Christianity.