When I was at secondary (high) school, as part of studying English literature we engaged in a project on ‘Utopias, imaginary lands and other worlds’. As part of that we read texts as diverse as Thomas Moore’s Utopia, Plato’s Republic, Jonathan Swift’s Gulliver’s Travels, and George Orwell’s Animal Farm. For any of these texts, if we were not given external, verified, biographical information, could we work out when they were written? The style of prose and the range of intellectual ideas might have enabled us to work out the century. But the decade? Or the year?
Animal Farm is commonly thought to be a satire on Soviet Russia, so many people instinctively assume that it was written in the Cold War. But it actually springs out of Orwell’s observations whilst fighting in the Spanish Civil war, and draws on observations of the events immediately following the 1917 Russian Revolution. But how would we know, if were not aware that Orwell himself died in 1950? Its allegorical nature, combined with its observations about human nature and its tendency towards totalitarianism, make it timeless—so perhaps the question of dating is not as important for reading as we might suppose.
These kinds of questions apply to discussions of dating the New Testament. Scholars are in the habit of pronouncing theories about dating with great confidence, often based on the internal evidence of the text, without attending to three key questions. First, why are we engaging in this discussion, and is it even possible to determine the date of a text from internal evidence? Secondly, what is our methodology, and does it bear scrutiny? Thirdly, what difference does it make to our reading of the text? When people discover that I have written a commentary on the Book of Revelation (here: https://ivpbooks.com/revelation-243 ), one of the first questions they ask is ‘When do you think it was written?’ My immediate response is: ‘Why does it matter? How will it change the way you read it?’
In this response, I will make some observations about these questions by reflecting on Jonathan Bernier’s argument for a date of Revelation ‘between 68 and 70’, by engaging with Ralph Smith’s original paper, and by noting R T France’s comments about dating Matthew.
Bernier’s Dating of Revelation (pp 118–127)
On the basis of the external attestation Justin Martyr and Andrew of Caesarea’s claim that Papias knew Revelation, Bernier sets a terminus ad quem of 160 (for certain) or 120 (likely). He follows Craig Koester in noting that the king list of Rev 17 is not much help, but perhaps offers a terminus a quo of 54. He notes the strong evidence for the gematria of Rev 13.18’s 666 as a reference to Nero, and notes that this could be related to Nero’s reign, but is more likely related to the rumours of Nero redivivus, which circulated from 69 until well into the 90s. Thus far, thus uncontentious.
The central aspect to Bernier’s argument turns on the depiction of the temple and associated events in Rev 11. Interestingly, Bernier notes the potential circularity of arguments that assume the language of the temple here is literal, thus that the temple is standing so the text is pre-70, or symbolic, so the temple is probably no longer standing, so the text is likely post-70, and he criticises John Robinson for assuming the first and not considering the second.
(The most remarkable use of this chapter is by Stephen Smalley, who assumes that this is a literal description of the temple in the process of being literally besieged, and so dates Revelation to a particular month in 70. See his Thunder and Love pp 40–50, summarised in his 2005 commentary p 3.)
Alas, Bernier does not continue this careful consideration when it comes to the later part of chapter 11. Even though the description of the temple might be symbolic, apparently that is not possible for the fall of the ‘great city’, which he takes as a literal description of the fall of Jerusalem. Since the text says ‘a tenth of the city fell’ (Rev 11.13) but we know from Josephus that the destruction was much more extensive, this must have been written prior to 70.
This is all very odd. Not only does Bernier make an odd switch from the possibility of symbolism to literalism, he also fails to notice the possibly symbolic significance of the numbers involved, the ‘tenth’ of the city, and the ‘7000’ who were killed. He sets aside the consistent use of both city and temple as symbolic representations of the people of God, the true Israel of God who are followers of the lamb. We find this in the depiction of the people as 12,000 numbered from each of the 12 tribes in Rev 7, making 144,000 who comprise a square number times a cube, alluding both to the cube shape of the Holy of Holies, and anticipating the New Jerusalem which is itself a giant cube, and thus the Holy of Holies which is now co-terminus with the whole city, and not just a part of it.
(Contra Peter Leithart, I believe that the hearing/seeing structure of Rev 7.4 and Rev 7.9 demonstrated that the 144,000 from the tribes of Israel are the same as the multitude that no-one can count from every nation, tribe, people, and language; see my Tyndale commentary pp 156– 167.)
Bernier also sidesteps, without any acknowledgement, the debate about the identity of the ‘great city’, which in Rev 16.19 is identified with Babylon. Bernier is right that the description of this city in Rev 11.8 as ‘where their Lord was crucified’ might point to Jerusalem—but of course the ‘place’ where Jesus died was in the Roman Empire, killed by Romans in the Roman way. All the other evidence points to Babylon as being the centre of sea and land trade in the ancient world, dominating the kingdoms of the world, and enriching herself as a result. Only Rome fits this description. (for a full discussion, see https://www.psephizo.com/biblical-studies/is-babylon-rome-or-jerusalem/ )
Bernier’s other considerations of context do not add decisively to his argument one way or another, though they include some refreshing observations about Christology and the context of the ekklesiae, and so does not revise his dating. But we need to note that the specific date he proposes rests entirely on the claim that the ‘great city’ is Jerusalem, and that the account of the city’s fall should be read historically and literally.
What do we learn, then, about the task of dating a text? That there are two major problems with this kind of ‘internal’ argument. First, within the discussion itself, the argument about dating cannot be disentangled from a very specific argument about the nature of a text and its interpretation—and since Bernier is not writing a commentary, he does not offer a defence of his reading strategy. But then, secondly, if he were going to offer commentary, what difference would his view on dating make? If it shaped his reading of chapter 11, then his argument becomes entirely circular, and viciously (rather than virtuously) so. We read Rev 11.13 literally, because of its date—but we only know this date because we are reading this text literally.
At the end of this section (p 127) Bernier notes the external evidence of Irenaeus’ comment, that John had his vision ‘recently’, at the end of Domitian’s reign—but he sets this external evidence aside in favour his own ‘internal’ argument. On what grounds? Because ‘it seems most judicious’ to him!
And Bernier, like many others, appears unaware of other external evidence worth noting which does appear to support a later date of writing. Laodicea was destroyed by an earthquake in AD 60, and the message to the assembly there seems to assume that it is prosperous and well established, which could hardly be the case if John was writing in the late 60s. Polycarp, Bishop of Smyrna, says (Philippians 11) that the church in Smyrna did not exist in the time of Paul which would imply a later date for Revelation. And Epiphanius, writing much later (in his Panarion), notes that it was believed there was no Christian community in Thyatira until late in the first century.
Ralph Smith on Bernier on Matthew 24
Smith offers some important and insightful critique of Bernier’s approach to Matthew, and wider questions of dating.
He is right to highlight the reason for challenges to previous early dating with the rise of post- Enlightenment biblical scholarship. It was not about accounting for the ‘relevant data’, but was driven by an ideological scepticism about the reliability of the NT texts. Richard Bauckham in particular has demonstrated the lack of a proper methodological basis for this approach, in particular the assumptions of form criticism that the stories about Jesus were circulated orally for a generation before being written down.
I do not believe that we need to insist on recovering the theological assumptions about the nature of the biblical text to counter this, as Smith suggests. Rather, we need to be more honest about what the full range of data point to.
In particular, I agree with Smith highlighting the question of what we might call widespread, low- level or practical literacy in the ancient world. We must, of course, resist projecting modern notions of systematic literacy as an anachronism. But Smith is quite right to note the evidence of Robert Gundry, when he cites:
The use of notebooks which were carried on one’s person was very common in the Graeco- Roman world. In ancient schools outline notes . . . were often taken by pupils as the teacher lectured. The notes became the common possession of the schools and circulated without the name of the lecturer. Sometimes an author would take this material as the basis for a book to be published. . . . Shorthand was used possibly as early as the fourth century B.C. and certainly by Jesus’ time. The Oxyrhynchus papyri show that scribes and clerks were often trained in shorthand. Rabbinic tradition was transmitted by the employment of catchwords and phrases which were written down in shorthand notes. (Robert Gundry, The Use of the Old Testament in St. Matthew’s Gospel, 1967, pp 181–182).
The probability that those who heard the teaching of Jesus might well have taken notes has never been addressed, and we might even have direct evidence of this in Luke’s phrase ‘servants of the word’ ὑπηρέται τοῦ λόγου in Luke 1.1. Whilst the term ὑπηρέτης has a general sense, it is next used in Luke 4.20 of the ‘attendant’ in the synagogue who gives to Jesus and receives from him the scroll of the haftorah from which he reads. Could these ‘servants of the word’ be those who had gathered and guarded notes about Jesus from the eye-witnesses who heard him?
In addition to this, we might also note:
- The evidence of the Vindolanda Tablets at Hadrian’s wall in 1973, demonstrating common literacy in making lists, communicating informally, and even issuing written invitations to a birthday party. The tablets include what was at the time the earliest example of writing in Latin by a woman.
- The evidence of signs written on the walls or Pompeii and Herculaneum, along with informal graffiti, which demonstrate assumptions of multilingualism and widespread use of gematria/ isopsephism (see the discussion here). The prima facie question is: if inscriptions and wall writing was not able to be read fairly extensively, why were they used?
- The evidence within the New Testament of those who were scribes (Tertius, Rom 16.22) letter carriers (Phoebe, Rom 16.1) and lectors within each congregation (‘the one who reads aloud…’ Rev 1.3).
Given the extensive awareness of the importance of letters and writing in the first century world, through work of Harry Gamble (Books and Readers in the Ancient World, 1995), Alan Millard (Reading and Writing in the Time of Jesus, 2001, cited by Smith) and Larry Hurtado (The Earliest Christian Artifacts: Manuscripts and Christian Origins, 2006), it is puzzling that this data is not a major part of discussions about dating. This is all the more surprising given the rise in interest in ‘material culture’ as vital background to reading the New Testament well in its context. (For an excellent and accessible study of material culture, see Bruce Longenecker’s In Stone and Story: Early Christianity in the Roman World, 2020, discussed here).
This makes Smith’s claim, that Matthew might well have taken notes of Jesus’ teaching, certainly plausible.
Smith is correct to challenge Bernier’s grounds for his early dating of Matthew on his agreement with James Crossley that Matt 24 predicts the ‘end of the world’ which did not happen—and thus must have been an early account, since it had not yet proven to be mistaken.
What Bernier is doing here is what he did in relation to Rev 11: basing his argument about dating on a particular interpretation and reading of the text, one that is open to significant challenge. I agree with a central part of Smith’s challenge here: Bernier fails to notice that the language of ‘the coming of the son of man’ is from Daniel 7.13, and relates to Jesus coming from the earth to the throne of God, that is, it is concerned with the exaltation of Jesus, and not with his parousia or return. We find the same language of exaltation not only in the trial narrative in Mark 14.62, but also in the stoning of Stephen in Acts 7.56. (I especially appreciated Smith’s exegesis of Matt 24.30 καὶ τότε φανήσεται τὸ σημεῖον τοῦ υἱοῦ τοῦ ἀνθρώπου ἐν οὐρανῷ; the grammar tells us that it is the Son of Man who is in heaven, and not the sign itself.)
Where I part company from Smith is in his subsequent understanding of the parousia. Jesus in Matthew is not identifying his parousia with his exaltation and the judgement of Jerusalem, but distinguishing the two. Thus the single question from the disciples in Mark 13.4 (‘Tell us, when will these things be…?’) becomes a two-fold question in Matt 24.3:
Tell us, when will these things be,
and what will be the sign of your coming (παρουσία, beUer ‘royal presence’) and of the end of the age?
The first part of Matt 24, up to verse 35, is about ‘these things’; the parousia is not discussed or even mentioned, except in Matt 24.27 where Jesus explains that all these events are not about his παρουσία! But then, from Matt 24.36, the focus of Jesus’ teaching changes, as he now switches from talking about ταῦτα, ‘these things’, to δὲ τῆς ἡμέρας ἐκείνης, about ‘that day’. The first part of this runs parallel to the brief comments at the end of Mark 13—but then follows all the eschatological parables of Matt 25, so that we have a full exposition of the parousia as distinct from the teaching about the fall of the temple, the preaching of the gospel, and the exaltation of Jesus. (For details on this, see my article here: and my Grove study Kingdom, Hope and the End of the World here)
For me, Smith’s argument is weakened because he is doing what he challenges Bernier on, but in reverse—basing an argument for dating on an alternative reading of the text. Moreover, we cannot rely on arguments about ‘when a text needed to be written’ to infer anything about its actual date. Texts are not always written when they are needed, and authors do not have the benefit of surveying past and future history in deciding whether and when to write!
R T France on dating Matthew
In his NICNT commentary (2007) his discussion of dating is remarkably brief (at less than one and a half pages) but judicious.
He notes in turn that:
- The main ‘majority’ arguments offered for dating Matthew rely on an overall scheme which relates the gospels to one another.
- A key datum for all dating arguments is the lack of reference in Acts to the Neronic persecution of 64/65, which provides a terminus ad quem for the dating of the gospels (something with which Bernier agrees).
- The patristic tradition, as offering external data, that Mark was based on the eye-witness account of Peter, so must have been written in his lifetime.
- The strong arguments for a pre-70 date for Matthew set out by Gundry have not been addressed in their own terms, but are dismissed on the basis of wider overall schemes of dating—precisely those which Bernier is challenging.
For me, this leaves the Olivet Discourse still raising some questions about the relationship between Matthew, Mark, and Luke:
- Are the changes ‘from’ Mark 13 ‘to’ Matthew 24–25 (the disciples’ question divided in two, the clearer distinction between ‘these things’ and ‘that day’, and the expanded teaching on the parousia) significant for our discussion of dating? These would make sense if Mark was written before the fall of Jerusalem, and Matthew after, since the focus has now shifted to the longer eschatological horizon. But does that necessarily imply anything?
- Are the further changes in Like 19 of any significance? The details of the siege (‘they will set up a barricade, surround you, tear you to the ground’, Luke 19.43–44) seem graphic—but are they a reflection of knowledge or general language about destruction?
And yet the larger questions remain: Can the dates of texts be determined with any certainty from internal evidence alone? What is our motivation for asking questions about dating? Is our methodology rooted in the cultural context of the texts, and can it avoid circular arguments? And to what ends do we put our conclusions?
Ian Paul is a theologian, author, speaker, and academic consultant. He is an adjunct professor at Fuller Theological Seminary, Managing Editor at Grove Books, and tweets @Psephizo.
When I was at secondary (high) school, as part of studying English literature we engaged in a project on ‘Utopias, imaginary lands and other worlds’. As part of that we read texts as diverse as Thomas Moore’s Utopia, Plato’s Republic, Jonathan Swift’s Gulliver’s Travels, and George Orwell’s Animal Farm. For any of these texts, if we were not given external, verified, biographical information, could we work out when they were written? The style of prose and the range of intellectual ideas might have enabled us to work out the century. But the decade? Or the year?
Animal Farm is commonly thought to be a satire on Soviet Russia, so many people instinctively assume that it was written in the Cold War. But it actually springs out of Orwell’s observations whilst fighting in the Spanish Civil war, and draws on observations of the events immediately following the 1917 Russian Revolution. But how would we know, if were not aware that Orwell himself died in 1950? Its allegorical nature, combined with its observations about human nature and its tendency towards totalitarianism, make it timeless—so perhaps the question of dating is not as important for reading as we might suppose.
These kinds of questions apply to discussions of dating the New Testament. Scholars are in the habit of pronouncing theories about dating with great confidence, often based on the internal evidence of the text, without attending to three key questions. First, why are we engaging in this discussion, and is it even possible to determine the date of a text from internal evidence? Secondly, what is our methodology, and does it bear scrutiny? Thirdly, what difference does it make to our reading of the text? When people discover that I have written a commentary on the Book of Revelation (here: https://ivpbooks.com/revelation-243 ), one of the first questions they ask is ‘When do you think it was written?’ My immediate response is: ‘Why does it matter? How will it change the way you read it?’
In this response, I will make some observations about these questions by reflecting on Jonathan Bernier’s argument for a date of Revelation ‘between 68 and 70’, by engaging with Ralph Smith’s original paper, and by noting R T France’s comments about dating Matthew.
Bernier’s Dating of Revelation (pp 118–127)
On the basis of the external attestation Justin Martyr and Andrew of Caesarea’s claim that Papias knew Revelation, Bernier sets a terminus ad quem of 160 (for certain) or 120 (likely). He follows Craig Koester in noting that the king list of Rev 17 is not much help, but perhaps offers a terminus a quo of 54. He notes the strong evidence for the gematria of Rev 13.18’s 666 as a reference to Nero, and notes that this could be related to Nero’s reign, but is more likely related to the rumours of Nero redivivus, which circulated from 69 until well into the 90s. Thus far, thus uncontentious.
The central aspect to Bernier’s argument turns on the depiction of the temple and associated events in Rev 11. Interestingly, Bernier notes the potential circularity of arguments that assume the language of the temple here is literal, thus that the temple is standing so the text is pre-70, or symbolic, so the temple is probably no longer standing, so the text is likely post-70, and he criticises John Robinson for assuming the first and not considering the second.
(The most remarkable use of this chapter is by Stephen Smalley, who assumes that this is a literal description of the temple in the process of being literally besieged, and so dates Revelation to a particular month in 70. See his Thunder and Love pp 40–50, summarised in his 2005 commentary p 3.)
Alas, Bernier does not continue this careful consideration when it comes to the later part of chapter 11. Even though the description of the temple might be symbolic, apparently that is not possible for the fall of the ‘great city’, which he takes as a literal description of the fall of Jerusalem. Since the text says ‘a tenth of the city fell’ (Rev 11.13) but we know from Josephus that the destruction was much more extensive, this must have been written prior to 70.
This is all very odd. Not only does Bernier make an odd switch from the possibility of symbolism to literalism, he also fails to notice the possibly symbolic significance of the numbers involved, the ‘tenth’ of the city, and the ‘7000’ who were killed. He sets aside the consistent use of both city and temple as symbolic representations of the people of God, the true Israel of God who are followers of the lamb. We find this in the depiction of the people as 12,000 numbered from each of the 12 tribes in Rev 7, making 144,000 who comprise a square number times a cube, alluding both to the cube shape of the Holy of Holies, and anticipating the New Jerusalem which is itself a giant cube, and thus the Holy of Holies which is now co-terminus with the whole city, and not just a part of it.
(Contra Peter Leithart, I believe that the hearing/seeing structure of Rev 7.4 and Rev 7.9 demonstrated that the 144,000 from the tribes of Israel are the same as the multitude that no-one can count from every nation, tribe, people, and language; see my Tyndale commentary pp 156– 167.)
Bernier also sidesteps, without any acknowledgement, the debate about the identity of the ‘great city’, which in Rev 16.19 is identified with Babylon. Bernier is right that the description of this city in Rev 11.8 as ‘where their Lord was crucified’ might point to Jerusalem—but of course the ‘place’ where Jesus died was in the Roman Empire, killed by Romans in the Roman way. All the other evidence points to Babylon as being the centre of sea and land trade in the ancient world, dominating the kingdoms of the world, and enriching herself as a result. Only Rome fits this description. (for a full discussion, see https://www.psephizo.com/biblical-studies/is-babylon-rome-or-jerusalem/ )
Bernier’s other considerations of context do not add decisively to his argument one way or another, though they include some refreshing observations about Christology and the context of the ekklesiae, and so does not revise his dating. But we need to note that the specific date he proposes rests entirely on the claim that the ‘great city’ is Jerusalem, and that the account of the city’s fall should be read historically and literally.
What do we learn, then, about the task of dating a text? That there are two major problems with this kind of ‘internal’ argument. First, within the discussion itself, the argument about dating cannot be disentangled from a very specific argument about the nature of a text and its interpretation—and since Bernier is not writing a commentary, he does not offer a defence of his reading strategy. But then, secondly, if he were going to offer commentary, what difference would his view on dating make? If it shaped his reading of chapter 11, then his argument becomes entirely circular, and viciously (rather than virtuously) so. We read Rev 11.13 literally, because of its date—but we only know this date because we are reading this text literally.
At the end of this section (p 127) Bernier notes the external evidence of Irenaeus’ comment, that John had his vision ‘recently’, at the end of Domitian’s reign—but he sets this external evidence aside in favour his own ‘internal’ argument. On what grounds? Because ‘it seems most judicious’ to him!
And Bernier, like many others, appears unaware of other external evidence worth noting which does appear to support a later date of writing. Laodicea was destroyed by an earthquake in AD 60, and the message to the assembly there seems to assume that it is prosperous and well established, which could hardly be the case if John was writing in the late 60s. Polycarp, Bishop of Smyrna, says (Philippians 11) that the church in Smyrna did not exist in the time of Paul which would imply a later date for Revelation. And Epiphanius, writing much later (in his Panarion), notes that it was believed there was no Christian community in Thyatira until late in the first century.
Ralph Smith on Bernier on Matthew 24
Smith offers some important and insightful critique of Bernier’s approach to Matthew, and wider questions of dating.
He is right to highlight the reason for challenges to previous early dating with the rise of post- Enlightenment biblical scholarship. It was not about accounting for the ‘relevant data’, but was driven by an ideological scepticism about the reliability of the NT texts. Richard Bauckham in particular has demonstrated the lack of a proper methodological basis for this approach, in particular the assumptions of form criticism that the stories about Jesus were circulated orally for a generation before being written down.
I do not believe that we need to insist on recovering the theological assumptions about the nature of the biblical text to counter this, as Smith suggests. Rather, we need to be more honest about what the full range of data point to.
In particular, I agree with Smith highlighting the question of what we might call widespread, low- level or practical literacy in the ancient world. We must, of course, resist projecting modern notions of systematic literacy as an anachronism. But Smith is quite right to note the evidence of Robert Gundry, when he cites:
The use of notebooks which were carried on one’s person was very common in the Graeco- Roman world. In ancient schools outline notes . . . were often taken by pupils as the teacher lectured. The notes became the common possession of the schools and circulated without the name of the lecturer. Sometimes an author would take this material as the basis for a book to be published. . . . Shorthand was used possibly as early as the fourth century B.C. and certainly by Jesus’ time. The Oxyrhynchus papyri show that scribes and clerks were often trained in shorthand. Rabbinic tradition was transmitted by the employment of catchwords and phrases which were written down in shorthand notes. (Robert Gundry, The Use of the Old Testament in St. Matthew’s Gospel, 1967, pp 181–182).
The probability that those who heard the teaching of Jesus might well have taken notes has never been addressed, and we might even have direct evidence of this in Luke’s phrase ‘servants of the word’ ὑπηρέται τοῦ λόγου in Luke 1.1. Whilst the term ὑπηρέτης has a general sense, it is next used in Luke 4.20 of the ‘attendant’ in the synagogue who gives to Jesus and receives from him the scroll of the haftorah from which he reads. Could these ‘servants of the word’ be those who had gathered and guarded notes about Jesus from the eye-witnesses who heard him?
In addition to this, we might also note:
- The evidence of the Vindolanda Tablets at Hadrian’s wall in 1973, demonstrating common literacy in making lists, communicating informally, and even issuing written invitations to a birthday party. The tablets include what was at the time the earliest example of writing in Latin by a woman.
- The evidence of signs written on the walls or Pompeii and Herculaneum, along with informal graffiti, which demonstrate assumptions of multilingualism and widespread use of gematria/ isopsephism (see the discussion here). The prima facie question is: if inscriptions and wall writing was not able to be read fairly extensively, why were they used?
- The evidence within the New Testament of those who were scribes (Tertius, Rom 16.22) letter carriers (Phoebe, Rom 16.1) and lectors within each congregation (‘the one who reads aloud...’ Rev 1.3).
Given the extensive awareness of the importance of letters and writing in the first century world, through work of Harry Gamble (Books and Readers in the Ancient World, 1995), Alan Millard (Reading and Writing in the Time of Jesus, 2001, cited by Smith) and Larry Hurtado (The Earliest Christian Artifacts: Manuscripts and Christian Origins, 2006), it is puzzling that this data is not a major part of discussions about dating. This is all the more surprising given the rise in interest in ‘material culture’ as vital background to reading the New Testament well in its context. (For an excellent and accessible study of material culture, see Bruce Longenecker’s In Stone and Story: Early Christianity in the Roman World, 2020, discussed here).
This makes Smith’s claim, that Matthew might well have taken notes of Jesus’ teaching, certainly plausible.
Smith is correct to challenge Bernier’s grounds for his early dating of Matthew on his agreement with James Crossley that Matt 24 predicts the ‘end of the world’ which did not happen—and thus must have been an early account, since it had not yet proven to be mistaken.
What Bernier is doing here is what he did in relation to Rev 11: basing his argument about dating on a particular interpretation and reading of the text, one that is open to significant challenge. I agree with a central part of Smith’s challenge here: Bernier fails to notice that the language of ‘the coming of the son of man’ is from Daniel 7.13, and relates to Jesus coming from the earth to the throne of God, that is, it is concerned with the exaltation of Jesus, and not with his parousia or return. We find the same language of exaltation not only in the trial narrative in Mark 14.62, but also in the stoning of Stephen in Acts 7.56. (I especially appreciated Smith’s exegesis of Matt 24.30 καὶ τότε φανήσεται τὸ σημεῖον τοῦ υἱοῦ τοῦ ἀνθρώπου ἐν οὐρανῷ; the grammar tells us that it is the Son of Man who is in heaven, and not the sign itself.)
Where I part company from Smith is in his subsequent understanding of the parousia. Jesus in Matthew is not identifying his parousia with his exaltation and the judgement of Jerusalem, but distinguishing the two. Thus the single question from the disciples in Mark 13.4 (‘Tell us, when will these things be...?’) becomes a two-fold question in Matt 24.3:
Tell us, when will these things be,
and what will be the sign of your coming (παρουσία, beUer ‘royal presence’) and of the end of the age?
The first part of Matt 24, up to verse 35, is about ‘these things’; the parousia is not discussed or even mentioned, except in Matt 24.27 where Jesus explains that all these events are not about his παρουσία! But then, from Matt 24.36, the focus of Jesus’ teaching changes, as he now switches from talking about ταῦτα, ‘these things’, to δὲ τῆς ἡμέρας ἐκείνης, about ‘that day’. The first part of this runs parallel to the brief comments at the end of Mark 13—but then follows all the eschatological parables of Matt 25, so that we have a full exposition of the parousia as distinct from the teaching about the fall of the temple, the preaching of the gospel, and the exaltation of Jesus. (For details on this, see my article here: and my Grove study Kingdom, Hope and the End of the World here)
For me, Smith’s argument is weakened because he is doing what he challenges Bernier on, but in reverse—basing an argument for dating on an alternative reading of the text. Moreover, we cannot rely on arguments about ‘when a text needed to be written’ to infer anything about its actual date. Texts are not always written when they are needed, and authors do not have the benefit of surveying past and future history in deciding whether and when to write!
R T France on dating Matthew
In his NICNT commentary (2007) his discussion of dating is remarkably brief (at less than one and a half pages) but judicious.
He notes in turn that:
- The main ‘majority’ arguments offered for dating Matthew rely on an overall scheme which relates the gospels to one another.
- A key datum for all dating arguments is the lack of reference in Acts to the Neronic persecution of 64/65, which provides a terminus ad quem for the dating of the gospels (something with which Bernier agrees).
- The patristic tradition, as offering external data, that Mark was based on the eye-witness account of Peter, so must have been written in his lifetime.
- The strong arguments for a pre-70 date for Matthew set out by Gundry have not been addressed in their own terms, but are dismissed on the basis of wider overall schemes of dating—precisely those which Bernier is challenging.
For me, this leaves the Olivet Discourse still raising some questions about the relationship between Matthew, Mark, and Luke:
- Are the changes ‘from’ Mark 13 ‘to’ Matthew 24–25 (the disciples’ question divided in two, the clearer distinction between ‘these things’ and ‘that day’, and the expanded teaching on the parousia) significant for our discussion of dating? These would make sense if Mark was written before the fall of Jerusalem, and Matthew after, since the focus has now shifted to the longer eschatological horizon. But does that necessarily imply anything?
- Are the further changes in Like 19 of any significance? The details of the siege (‘they will set up a barricade, surround you, tear you to the ground’, Luke 19.43–44) seem graphic—but are they a reflection of knowledge or general language about destruction?
And yet the larger questions remain: Can the dates of texts be determined with any certainty from internal evidence alone? What is our motivation for asking questions about dating? Is our methodology rooted in the cultural context of the texts, and can it avoid circular arguments? And to what ends do we put our conclusions?
Ian Paul is a theologian, author, speaker, and academic consultant. He is an adjunct professor at Fuller Theological Seminary, Managing Editor at Grove Books, and tweets @Psephizo.
-->