Blog Menu

Does the Book of Revelation have a structure?

If you pick up any study of the Book of Revelation, chances are you will quickly come across an analysis of its structure. This is true of any New Testament book or letter, but in the case of Revelation there appear to be an urgency to the quest. As you are about to plunge into the dense undergrowth of the jungle of imagery, the need to have some points of navigation feels more desperate; perhaps, as we wrestle with a text which exercises such mastery over our senses and emotions, we need a way to fight back, and exercise our customary textual mastery over it.

Within academic studies, you will generally find an awareness of the complex challenges of pinning down Revelation’s structure, and then a proposal—but there are a good number of different proposals, and the question then is how can we reconcile them with one another. Within popular studies, you will generally find the claim that ‘my insight gives you THE KEY to unlocking the Book of Revelation!’—and then you find that it just does not fit the text or have any coherence! A common approach here is to propose that Revelation has a chiastic structure, that is, it is composed of a series nested visions (or in a pyramid, depending on how you present it visually) so that the early sections mirror the later ones in reverse order. The only problem with this scheme is that it doesn’t actually match the text, or that the sequencing is out of order, both of which is true in this example. The author must immediately qualify the proposal by claiming that ‘the similarities are not hard and fast rules’, in other words, the evidence isn’t always there! If John had wanted us to see a chiastic structure, couldn’t he have done a slightly better job of making it clear? (I am also intrigued by noting that some of these ‘structuring’ schemes end up being more complex than the text they are supposedly explaining!)


In looking for structure, we need to take into account four features of the text. The first is that Revelation has a very large number of explicit markers of structure, often making use of the number seven. Most obviously, there are seven messages to the Christian assemblies (ekklesiai) in chapters 2 and 3, seven seals that are opened in chapter 6 (concluding after an interlude in 8.1), seven trumpets that blown in chapters 8 and 9, and seven bowls that are poured out in chapter 16. In addition, there are patterns of seven within the text which not explicitly signaled or counted, such as the seven occurrences of the term ‘sickle’ in chapter 14, and the seven unnumbered visions in chapters 19 through to 21 (see introductory comment at 19.11). These features serve to bind together different sections of the text into unified blocks. Complementary to that are the use of particular phrases by John which appear to signal the introduction of a new focus or the arrival or a new character, the most common being ‘And I saw’ (kai eidon) which occurs frequently at the beginning of major sections as well as at the start of subsections within them. John also makes use of noticeable changes of style of writing: the vision of Jesus in chapter 1 takes the form of a vision report, but chapters 2 and 3 have their own distinctive style and structure (see comment introducing this section); the language shifts again at 4:1; then again at 6:1; there is a change of focus at 7:1—and so on. The most noticeable break (observed by all commentators) comes at 12:1, where John uses the language of a ‘sign appeared’ instead of ‘and I saw’; what is less often noticed is that the preceding section in chapter 11 is the only section of the text cast in the future tense.

This leads to the second observation: that within each of these larger units there is a clear microstructure. Some of these structural elements are clearly signaled, such as the shared structure of the seven messages, whilst others are less obviously signaled, but are evident nonetheless, such as the four-fold structure of chapter 12, as it moves from the narrative of the woman and the dragon, to the account of heavenly angelic warfare, to the hymn of praise, and back to the opening narrative of the woman and the dragon. Many of these are indicated in English translations by means of paragraph breaks. There are other structuring features which are only evident on a close reading of the text, such as the six angels in two sets of three that give structure to the harvest visions in chapter 14.

The third feature of Revelation’s structure, which is only evident when looking at the text as a whole, is the striking discontinuity between different units. This is drawn to our attention by the various abrupt changes in language and in the changes of dramatis personae from one unit to another, as is evident when we look at the characters themselves. At first there appears to be little in common between the six main descriptions of Jesus—in the epistolary greeting, the vision in 1:12 that follows, the image of the lamb on the throne, the male child in chapter 12, the one like a son of man in chapter 14 and the rider on the white horse in chapter 19. The angels of the seven assemblies quickly disappear from view, as do most of the angels that we meet however splendid they appear—and the revealing angel mentioned in 1:1 and 22:16 never appears at all! The beast from the land abruptly changes into the false prophet (16:13) and the beast from the land then becomes the scarlet beast ridden by the great prostitute (17:3). Other characters in the drama disappear without trace, most notably the woman clothed with the sun, whom we last see waiting in the desert, protected by God. This discontinuity led an earlier commentator, R H Charles, to argue that Revelation was originally a series of separate units that was edited together by a series of clumsy and ignorant editors, and a similar theory of multiple composition was again proposed by David Aune.


The main problem with these theories is that they do not adequately take into account the fourth feature of Revelation’s structure: the widespread occurrence of links and connections between different sections. The most obvious of these occur in the messages to the seven assemblies in chapters 2 and 3, in each of which the opening greeting links back to the first vision of the exalted Jesus, and the closing comments include an anticipation of the New Jerusalem in chapter 21. Connections can also be seen in the hymns at the end of chapters 7 and 11 which also anticipate the final vision of the new creation. Some sections do not obviously belong to one larger section or another, but function to link one to another. For example, 1:9–11 looks like a continuation of John’s epistolary greeting from verse 4, but it also looks back to the opening verses in reiterating John’s commission, and then leads into the vision of the exalted Jesus. In 11:1, John is ‘told’ as he was in 10:11, so these verses look like a continuation of the encounter with the mighty angel of 10:1—but they lead directly into the account of the two witnesses which continues to verse 13. And the ‘woes’ announced by the eagle in 8:13 are not completed until 12:12, and have some connection with the ‘woes’ of 18:10, 16 and 19.

But there are other, more subtle connections throughout the text. Although in other ways chapters 11 and 12 are quite distinct, not only in the characters featured but also in the style of John’s writing, they are linked together by the use of the threefold time period described as ‘time, times and half a time’, ’42 months’ and ‘1,260 days’ (11:2, 3, 9 and 11 and 12:6 and 14) which are only used in these two chapters.

Overall, then, we have stronger marks of structure and continuity in Revelation than in any other biblical book. But alongside that, we also have stronger marks of difference and discontinuity than in other biblical books. I think the reason for so many conflicting theories is simply that each approach gives different weight to these different aspects of the text, or prioritise some markers or continuity over others, which then gives different answers to the question of the structure of the text.


This is nicely illustrated when reading the messages to the assemblies in the seven cities (not, please note, letters to the seven churches). When writing my commentary, I found it helpful to lay out the text visually:

There is much that can be said about this—but in relation to structure, three things are striking. First, there is a very clear, explicit shape to the letters, and at time this is so strong that whole sentences are repeated (mostly at the beginning and end) verbatim from one message to the next. But, secondly and equally clearly, there is a good deal of variation within this structure. So the messages are all of different lengths, and there is no obvious pattern in this. Thirdly, the overall shape has some marks of a pattern in it—there is some symmetry, not least in the fact that the second (to those in Smyrna) and the second to last (to those in Philadelphia) both lack any rebuke. But, fourthly, there are also very clear structural disruptions to the structure. The most striking is the reversal of the order of ‘Whoever has ears…’ and the promise of victory from the first three to the last four messages (based on other series that come later, if there was a change we would expect it to be 4/3 and not 3/4). Also important, and often missed, is the addition of affirmation following rebuke to Ephesus and Sardis. This disrupts any attempt to offer a comprehensive explanation of an overall, controlling structure. Perhaps there were things that John, passing on the words of Jesus, just needed to say?


All this raises some interesting questions about what we are doing when we a looking for the structure of Revelation. First: are we looking for a structure that is there in the text, and seeking to highlight what John was doing? Or are we in fact imposing our own structure in order to help us (and other readers) understand what is going on in this complex text? There is nothing wrong with the latter—so long as we admit that is what we are doing, and don’t then claim that this is the structure of the book. This leads to a second question: why do we seek a structure? Some readers seem to treat Revelation’s structure as another elements of its secret code, and finding it gives them power not only over the text, but over other readers who are looking to them for guidance. If, instead, we are wanting to help people engage with the text, then our proposals should be easy to see and easy to defend, pointing people back to the text rather than our structural interpretive scheme.

Perhaps the most interesting question of all (and the one almost never asked) is: why did John structure his text in this way at all? My sense is that, in offering a vision which included images of chaos and destruction, he is presenting it within a carefully crafted vessel, reminding his readers that the chaotic events of the world they live in are held in the ordered and ordering sovereignty of God.


Follow me on Twitter @psephizoLike my page on Facebook.


Much of my work is done on a freelance basis. If you have valued this post, would you consider donating £1.20 a month to support the production of this blog?

, , , , ,

7 Responses to Does the Book of Revelation have a structure?

  1. Justin October 12, 2017 at 8:59 am #

    Great article–I’ve seen proposals from different writers like Bauckham and others (Beale?) that significant verbal markers would have been necessary for auditors who wanted to interpret the visions together–I’m drafting a paper now with the working title “Signs and Numbers: Numerical Intratextual Cues and Narrative Structure in John’s Apocalypse”

    The chart with the seven churches is very good in terms of visualizing structure–is it okay to share this with my Revelation class session? We are discussing the seven churches.

  2. Christopher Shell October 12, 2017 at 1:42 pm #

    Chiasmus is ABA at its simplest. So it is ubiquitous, but not always intentionally.
    Inclusio is ABA at its simplest. So it is ubiquitous, but not always intentionally.
    One therefore has to be cautious about appeals to either of these. But they were both genuinely popular in the time and culture in question.

    Revelation of all books wears its heart on its sleeve as a structured book. One will not find those who class it as an unstructured book tout simple. Some class it (mistakenly, an increasing number think) as disarranged or incomplete – but not as showing no signs of deliberated structure. It will always therefore attract attempts to crack the structure. The more such attempts are made, the further down the road we should get.

    I am sure I am not the only person who eagerly anticipated Aune (after all, we are talking Aune here) and of course it is a possession for all time tho’ primarily for classical background and linguistic analysis. In other ways it is so mighty a conception that it must have proven impossible to end up with a satisfyingly unified final vision; similarly with his 1972 thesis.

    There seem to be seven types of contents, very Johannine. Chuck Missler is one who agrees; I am sure there are others. The two good churches which as you note have only 6 are juxtaposed to the 2 that have 8, so that the average is 7. Why not just have 7 7s here? Maybe it is an example of the good outweighing the bad in the ratio 2:1, as often in Rev.. I have this idea that there are 102 positive numbers in Jn and 51 negative (oudeis – emia-en): total 153….

    Seven sevens in overall structure including one of the sevens which is itself divided into sevens is, also, a structural trait that John exhibits elsewhere.

    As is the division here into 3+4 (or 4+3 sometimes, all in due proportion), attained by a midway reversal of the final 2 elements.

    As is also the fact that those elements are nos 6 and 7; a lot of his structuring is of course (don’t misunderstand here…) ‘sixes and sevens’.

    Incidentally I checked Morgenthaler to see if there are spikes in the graph (e.g. 7) for the number of times certain words appear. There absolutely aren’t. But Bauckham is, as was already clear, correct to see deliberate twinning between certain concepts and the number of times they appear, and also to see lots of deliberate 4s, 7s, 28s etc.. One finding is that John had no special reverence for 8 despite Jesus being 888.

    I had not twigged that the Tom’s Midnight Garden reference is actually a misunderstanding, as of course it is. I remember them digging out the reference at the end of the book, I think. There is also a reference to the Cain & Abel story as I remember. I did not ever cross paths with Philippa Pearce though she was a good friend of my sister’s mentor.

    • Ian Paul October 12, 2017 at 2:18 pm #

      Christopher, I don’t think I understand your comment about 6s being next to 8s. 6 what?

      And what is Morgenthaler? Has someone done an analysis of all the word frequencies in Revelation?

    • Ian Paul October 12, 2017 at 5:12 pm #

      Just found Morgenthaler. Goodness me! have you done some analysis on these word frequencies?

  3. Christopher Shell October 12, 2017 at 7:23 pm #

    The 6s and 8s: if you look in your columns you will see that 3 churches have 7 boxes filled in, 2 have 6 and 2 have 8. But the ones that have 6 are next to the ones that have 8, making the average clearly 7. One caveat though: some might think that Laodicea should have only 6 boxes filled in and no plaudits, since it is not clear that we should separate the fact of lukewarmth from the analysis of lukewarmth. It depends whether John’s category is ‘I know your works’ on the one hand or plaudits on the other hand.

    Michael Green preferred ‘lukewarmness’ – he is probably right.

    Morganthaler is 1958 based on 1958 texts. I copied out all the words that appeared thrice or more but saw no patterns at all. For example: 25 words came 5x, 27 words came 6x, 26 words came 7x, 31 words came 8x.

    • Ian Paul October 12, 2017 at 10:14 pm #

      But I am not sure my boxes are all that rigid; others propose division in a different way, so I am not sure we could deduce anything from that.

      On the numbers, I have just analysed Morganthaler myself. If John wanted key words to come e.g. 7 times, he might well have made sure that other words didn’t. So you might find that the frequency of words at 6 and 8 times was greater—which in fact it is.

      The key question is *which* words come 7 times, and which 6 and 8, and whether there is any pattern there.

  4. Christopher Shell October 13, 2017 at 11:09 am #

    Looking at it, I think there is only one place, or at best 2, where the division can easily be altered: within the central ‘report’ part (and to an extent within its overlap with the later ‘instruction’ part). The report category itself may be divided by John into points-for-praise and rebukes, but these are usually clearly separated from one another, so that different people’s analyses would agree at that point.

    This being the case I think that most analyses would come up with 7 divisions (approx.) as a typical number for these messages. Perhaps a combination of 6s and 7s. It is odd for messages to display a relatively uniform division, and that alone suggests careful, minute structuring. Also it is unusual that they are divided into so many discrete parts, given that they are so short. Also it is typically Johannine to create 7s even where not strictly necessary (cf. mikron in Jn 16).

    Re: word-counts: yes. My impression was that Bauckham carried out his harvest well. The sorts of words for which total-occurrence-quantity might be significant have already been identified by Bauckham for the most part. As for avoiding 7s, yes I think John did so where he wanted to, but that is limited to the most ‘charged’ or potent words and concepts. For words that were not sufficiently charged, he would not mind nor count. He has 7 occurrences for lots of neutral words: result being, neither a spike nor a dip. (The difference between 25, 27, 26, 31 is not statistically significant of course, but as you imply this pattern could come about through actively choosing 6, 7, or 8 on different occasions).

Leave a Reply