I have written quite a bit at this site about the argument from undesigned coincidences for the veracity of the gospel accounts. Undesigned coincidences can be defined as the phenomenon where sources reporting an event fit together in a way that points to truth. If you are not familiar with the basic argument and relevant examples, I strongly suggest you read my previous articles on the subject before continuing to read what I have to say here. At this site, I have responded at length to two critics of the argument from undesigned coincidences, namely, Dr. Richard Carrier and John Nelson. In this article, I turn my attention to address the criticisms of a third individual, Dr. Kurt Jaros, who recently delivered a presentation at the National Conference on Christian Apologetics titled “The Synoptic Problem for Undesigned Coincidences.” Jaros contends that, although there are good examples of undesigned coincidences between the synoptics and John, there are no good examples of inter-synoptic undesigned coincidences, given the well established literary relationship that exists between them. In this article, I will address Jaros’ concerns regarding the four examples of undesigned coincidences that he singles out in his presentation.
First, however, a note of clarification about the nature of undesigned coincidences. It is not our contention that the only possible explanation for undesigned coincidences is the historicity of the accounts to which they pertain. Rather, it is our contention that undesigned coincidences constitute evidence for the hypothesis of historicity by virtue of the fact that they are more probable on the hypothesis of historicity than on its falsehood. Thus, a demonstration of the existence of alternative possible scenarios is not by itself a refutation of the argument, provided that those scenarios are not more probable than the hypothesis that the undesigned coincidence arises as a result of the two accounts being grounded in truth. In this article, then, I attempt to show that the hypothesis of historicity is still the best, most probable, explanation of the inter-synoptic undesigned coincidences that Jaros addresses.
An important feature of undesigned coincidences, which I think is all-too-often overlooked by critics, including Jaros, is the failure to understand the evidential significance of an appearance of casualness. This is what drives many to assume that the evangelists had to have no knowledge of each other’s work before we can argue for an undesigned coincidence. Jaros claims that Lydia McGrew asserted that the two human witnesses in her story at the beginning of Hidden in Plain View had to be “crime-scene separated” before we could have an undesigned coincidence between their statements. Of course, being crime-scene separated would be fantastic, but it is not necessary. We can still have the appearance of casualness in the way that two accounts fit together. The whole point is the casual nature of their statements and the fact that they do not appear to be deliberately trying to corroborate each other, which provides evidence for relevant independence.
The Paired Disciples
One undesigned coincidence in the synoptic gospels relates to the fact that Matthew’s listing of the twelve disciples groups the disciples into pairs (Matthew 10:1-4) but nothing in Matthew explains why he groups them into pairs. Mark’s gospel, however, fills in the missing detail, by telling us that Jesus sent out the disciples two by two (Mark 6:7), thereby illuminating why Matthew pairs up the disciples in his list. While Mark does give the list of the disciples, he does not group them into pairs (Mark 3:13-19), and while Matthew writes about the sending out of the disciples, he does not mention that they were sent out two by two (Matthew 10:5). Thus, a natural question is answered in one account that is answered incidentally by another, thereby pointing to the truth of both accounts.
Jaros offers a few alternative scenarios which he thinks can also explain this data. His first proposed scenario is this:
Mark writes his Gospel first, then Matthew’s listing of the disciples at the commissioning could quite easily be explained by a thematic Designed Corroboration to enlighten readers of those pairings given the absence of the particular pairings in Mark’s gospel.
However, this makes little sense, since the pairing of the twelve disciples in Matthew’s list does not explain anything in Mark, since Mark explicitly indicates that the disciples were sent out in pairs. What is there, then, for Matthew to explain? If Matthew was trying to enlighten his readers about whose partner was which among the twelve disciples, then why does he fail to mention that the disciples were sent out in pairs?
Jaros’ second suggestions is that “Matthew writes his gospel first, then Mark intentionally summarizes Matthew’s list.” However, indicating that the disciples were sent out two by two does nothing to “summarize” a list of the twelve disciples that groups them into pairs. Furthermore, there is no need for Mark to summarize Matthew’s listing of the twelve disciples, since Mark has his own list of the twelve elsewhere (Mark 3:13-19).
Jaros’ third suggestion is that Matthew’s and Mark’s accounts are literarily independent but used a common source. However, this theory doesn’t account for the fact that Mark’s list of the disciples does not group them into pairs. Indeed, the fact that the list of the disciples and the sending out of the twelve are found in two different parts of Mark’s gospel suggests that, if the accounts, as Jaros suggests, are based on oral tradition, that they were found in two different circulating oral traditions. In that case, the undesigned coincidence would still remain.
Finally, Jaros suggests that there are not really intentional pairings in Matthew’s list at all since there is an additional και (“and”) in Matthew’s list. McGrew actually addresses this very point in her book. She notes that,
There is one extra “and” in some texts of Matthew that is not represented in this translation, though the NASB shows it. This extra “and” comes between Andrew and James. Other than that, Matthew’s list in the Greek consistently and precisely sorts the disciples into pairs using the word “and.” That is to say, with that one exception, there are “ands” between the members of each pair but no “ands” between the pairs.
Indeed, I do not think the additional “and” really undermines the consistent pairing in Matthew’s list of the twelve.
Herod Speaks to His Servants
Jaros then turns his attention to a second undesigned coincidence, namely, that involving Herod’s conversation with his servants about Jesus (Matthew 14:1-2). The account naturally raises two questions: (1) Why is Herod speaking with his servants about Jesus? and (2) How does Matthew know what Herod said to his servants behind closed doors, presumably in the privacy of his palace? The answer is provided in Luke 8:3, where we learn that one of the women who followed Jesus from Galilee was Joanna, who was married to Chuza, Herod’s household manager. This then illuminates why Herod would be talking to his servants about Jesus and how Matthew could come to know what Herod was saying to his servants.
Jaros offers three options to account for this coincidence. The first option is that “Luke intentionally mentions Joanna to confirm Matthew’s description.” Jaros concedes that this, though a logical possibility, is highly unlikely, especially since the two relevant passages are completely unrelated to each other and Luke does not mention Herod’s conversation with his servants. Jaros’ second suggestion is oral tradition. He says,
Suppose the info originally came from Joanna. If she had a written account (notes) or simply were the source of an oral tradition, then this calls into question the coincidental nature of the material. In this case, Matthew provides that detail whereas Mark (6:14-29) and Luke (9:7-9) redact it.
It is unclear to me exactly what Jaros is suggesting here. If Jaros’ idea here is that Joanna’s name was included in oral tradition as the source of the story about Herod and his servants — meaning simply that Matthew spoke to Joanna but omitted to include her name in his account of the death of John the Baptist — then I fail to see how this alternative scenario serves as an alternative to the undesigned coincidence. Furthermore, on the ur-source hypothesis — that is, that there is a source, whether literary or oral, that lies behind the accounts in the gospels — the ur-source would presumably have to include both stories (i.e. Herod’s conversation with his servants and the listing of the women who followed Jesus from Galilee). However, oral stories would typically be stories of a given pericope. If there were multiple circulating traditions containing between them both of the relevant stories, then the undesigned coincidence would still exist between the multiple circulating traditions, and the problem the hypothesis sought to address still remains.
Another scenario suggested by Jaros is that Matthew made a plausible conjecture concerning the people to whom Herod would presumably have been speaking when he said that Jesus was John the Baptist raised from the dead, recorded in Mark 6:14. However, Mark 6:14 probably originally read “Some were saying, “John the baptizer has been raised from the dead, and because of this, miraculous powers are at work in him.” Not only is this the majority reading, but the reading that “He said…” is much easier to account for as resulting from from a scribal copyist, whether consciously or subconsciously, attempting to conform the text to Matthew. In Mark, then, it is “some” who are saying that Jesus is John the Baptist raised from the dead. They are not the ones to whom it is being said. It seems like a stretch, therefore, to think that the reference to “some” in Mark 6:14 would have caused Matthew to conjecture that Herod was talking at all, much less that he was having a conversation with his servants. Jaros’ proposal also fails to account for the reference in Luke 8:3 to one of Jesus’ female disciples, Joanna, being married to the manager of Herod’s household. Jaros’ speculation thus does not explain why the details fit together.
Jaros’ third and final speculation on this coincidence is paraphrasing or compression on the part of Matthew. However, it is not at all clear to me how Jaros thinks this would explain the undesigned coincidence. Matthew 14:1-2 does not present a compressed version of Luke 8:3, nor vice versa.
Woes to Bethsaida
The third coincidence that Jaros turns his hand to addressing is the coincidence relating to the woes that Jesus pronounces upon the unrepentant cities of Chorazin and Bethsaida because they didn’t repent in spite of such mighty works being performed there (Matthew 11:21). Matthew doesn’t explain what mighty works were done in Bethsaida. The gospel of Luke, however, does inform us that the event of the feeding of the five thousand took place in Bethsaida (Luke 9:10). Although Matthew does narrate the feeding of the five thousand, he doesn’t mention where the event took place and it is in fact found a few chapters after Jesus has pronounced the woes on Bethsaida, since Matthew is reporting a-chronologically (Matthew 14:13-21).
How does Jaros account for this coincidence? He again offers a few conjectures. His first is that, “If Matthew drew from Luke’s Gospel, there would be strong literary dependence because the woe passage between Matthew and Luke (10:13-15) is nearly identical.”
Jaros’ second scenario is that “If Luke drew from Matthew, then he (Luke) could have copied the woe passage and thought it worthwhile to mention the city of one of the ‘mighty works done’ in that city.” On that scenario, Luke may have been expected to also think it worthwhile to mention what mighty works were done in Chorazin as well, though he does not. This provides evidence against the idea that this is what Luke is doing. The allusion to mighty works done specifically in Chorazin is also an unexplained allusion, itself a feature of the gospels that points to truth. Another issue with this suggestion is that the two relevant pieces are found in completely different passages. That sort of separation itself is evidence of casualness, and one may even take this example to be an undesigned coincidence that is internal to Luke’s gospel alone if one chooses.
Jaros’ third suggestion is that both Matthew and Luke are dependent upon a common source. However, as just discussed, the two parts of this coincidence — namely, the pronouncement of the woes and the detail that the feeding event took place near Bethsaida — are found quite widely separated from each other. They are not found in the same account even in Luke, and, as already stated, the statements are sufficiently casual that one may even view this as an example of an undesigned coincidence that is internal to Luke.
Jaros concludes his discussion of this coincidence by suggesting that Galilee being Jesus’ base of operation accounts for the statement that the mighty works were done in Bethsaida and Chorazin. He observes that Matthew elsewhere indicates that great crowds in Galilee followed Jesus and performed many healings (Matthew 4:23-25). However, the pronouncement of the woes took place in Capernaum in Galilee, and thus it is not unreasonable to gather that Jesus had in mind mighty works that were done specifically in Chorazin and Bethsaida. Neither Chorazin nor Bethsaida is immediately neighboring Capernaum, though all are in the general region of Galilee. The question thus naturally arises what mighty wonders Jesus is referring to that occurred specifically in those towns of Galilee. This is answered by the statement in Luke 9:10 that the feeding of the five thousand miracle took place in Bethsaida.
Joseph of Arimathea’s New Tomb
The final example that Jaros touches on is the coincidence concerning the statement in Luke 23:53 that “Then he [Joseph of Arimathea] took it down and wrapped it in a linen shroud and laid him in a tomb cut in stone, where no one had ever yet been laid.” The question arising from this statement in Luke is “How did Joseph happen to have access to a new tomb in which nobody had previously been laid?” The answer is given in Matthew 27:59-60, in which we read that it was “his own new tomb, which he had cut in the rock.” To explain this coincidence, Jaros offers two conjectures. First, he suggests that the information that Jesus was buried in Joseph of Arimathea’s new tomb was found in the oral tradition. I do not see any reason, however, why the oral tradition must be placed between the gospels and the historical events they report. On the hypothesis of traditional authorship, which I think is quite well supported, Matthew was himself an apostle. He himself would have thus been close up to the facts. The way that Jaros and many other modern scholars treat the oral tradition is as a source that contains all of the facts found in the canonical gospels. This is in violation of Occam’s razor, since it multiplies entities beyond necessity. Instead of speaking of “oral tradition” as a source, why not instead talk about the evangelists who were close up to the facts and are reporting either what they had seen and heard (Matthew and John), or what they had received from those who had (Mark and Luke)? For a fuller discussion of the objection to undesigned coincidences relating to oral tradition, I refer you to my previous article addressing it here.
The other scenario offered by Jaros is that Matthew “asked around” for more information concerning the burial and found out that the tomb was owned by Joseph of Arimathea. However, even if true, this scenario would not undermine the undesigned coincidence, since it is not a problem for an author to go and find out an additional fact — that simply means that the evangelist has access to additional reliable sources of information. Furthermore, most scholars argue that Luke was written after Matthew, and so Matthew would not have had access to Luke, since it would not have been written at the time of his writing. And Mark, which predates Matthew (at least on the consensus view of Markian priority) does not mention the facts, provided by Luke and John, that the tomb was new and that nobody had previously been laid in it. Matthew also does not tell us that the tomb was new and that nobody had previously been laid in it. He thus gives us no indication that he is trying to explain anything.
To conclude, while I appreciate Kurt Jaros’ attempt to offer a critique of the argument from undesigned coincidences — and indeed I welcome healthy in-house critique between Christian scholars — his attempt to address the argument, in my judgment, falls short of convincing. Nonetheless, I hope that my interaction with his points here will serve to clarify the evidential significance and value of undesigned coincidences between the gospels. It is important to bear in mind throughout this discussion that it is not my contention that undesigned coincidences prove the historicity of the events to which they pertain, or that no possible alternative explanations exist. Rather, it is my contention that they provide evidence for historicity, meaning, that their presence is more probable on the hypothesis of historicity than on its falsehood. Undesigned coincidences are also not all of equal weight, and some may be of greater evidential value than others. Since undesigned coincidences are also more probable on the hypothesis that the gospels represent substantially reliable sources than on its falsehood, they may cumulatively be taken as evidence for that proposition as well.