Vern Sheridan Poythress

[Published in the Journal of the Evangelical Theological Society 36 (1993) 41-54.  Used with permission.]


One’s decision about the literary genre of Rev 20:1-6 is one of the most crucial factors in its interpretation.  To what kind of genre does Rev 20:1-6 belong?  Does the passage offer us a straightforward report concerning the future (simply prewritten history), apocalyptic vision, prophetic prediction combined with hortatory implications, or something else?  This question of genre is closely related to how we distinguish between symbolic and literal description.  The literary genre of Revelation guides readers in deciding what in Revelation is intended as symbol and what is intended as a literal or straightforward description of a historical event.


Levels of communication

The nature of symbolic communication can be illustrated in the interpretation of Rev 13:1-8.  In Rev 13:1-8 we can distinguish at least four relevant levels of communication.  First, there is the linguistic level consisting in the textual record in Rev 13:1-8.  Under inspiration John wrote the text of 13:1-8 to the audience consisting of the seven churches.

Second, there is the visionary level, consisting in the visual experience that John had in seeing the beast.  Before writing the text, John received extraordinary visionary experiences “in the Spirit” (1:10; 4:2).  Through the Spirit God gave him visions that formed the basis for what is textually recorded in the written book of Revelation.  One of these visions was the vision of the beast.

Third, there is the referential level, consisting in the historical reference of the beast and of the various particulars in the description.  The beast stands for or symbolizes something that appears in history.  Moreover, some of the details such as the seven heads and ten horns have referents of their own (17:9, 12).  Of course, interpreters have differed among themselves as to just what the beast does stand for.  Nearly all interpreters have agreed that the beast represents some form of antichrist figure, but they differ concerning the time and characteristics of its manifestation in history.  Such differences depend mainly on which school of interpretation one adopts.  Preterists usually think that the beast refers to the Roman Empire or Roman emperor.  Futurists think that it refers to the final Antichrist, who may be associated with a revived Roman Empire.  Protestant historicists find a reference to the papacy.  Idealists find a reference to state persecuting power throughout history.1  My main point is that all interpreters alike find some form of historical reference for the beast.  Such historical reference constitutes the third level of communication.

Fourth, there is a symbolical level consisting in the interpretation of what the symbolic imagery actually connotes about its historical referent.  Rev 13:1-8 does not merely assert that something will occur in history.  The symbolic clothing of the communication conveys something about the characteristics of the historical referent.  Antichrist is powerful, persecuting, demonic, blasphemous, hideous, yet ultimately under God’s control and heading for defeat.  But this information about the beast is conveyed in symbolic, imagistic form rather than through literal photographic depiction or through unadorned prose description.  The description of the beast in Rev 13:1-8 is not intended as a photographic rendering of a literal animal, but a symbolic representation of a human being or a historical institution.  Hence understanding the significance of the imagery involves making a transition from symbols to actual historical significance.

In this respect, Revelation 13 employs symbolism in a way not characteristic of ordinary historical narrative.  In Revelation 13 symbolism is primary, whereas in historical narrative literal description is primary.  For example, the healing of the paralytic in Mark 2:1-12 is part of Mark’s historical narrative.  Primarily, it records a real, literal healing of a literal paralytic.  It is neither a made-up story nor an allegorical, symbolic transposition of an event in which Jesus pronounced forgiveness without healing.  But the event of healing nevertheless also has symbolic, theological connections with the purpose of Jesus to forgive sins (verse 5).  The theological overtones are definitely present, but do not erase Mark’s concern for history.  Similarly, the resurrection of Lazarus in John 11 is both a real historical event and a theological pointer to the significance of Jesus’ coming resurrection (verse 25).  Thus, roughly speaking, historical narratives in the Bible contain both a direct relation to the underlying events and an indirect indication of the theological significance of the events.  In Revelation 13 the relation is reversed.  The symbolism dominates in such a way that the passage expresses directly the theological significance and only indirectly points to the underlying event.  The imagery used in describing the beast is precisely that: imagery.  Imagery captures the symbolical and theological significance “directly,” but we must make a transition to another sphere in order to find the referent.  The challenges associated with this transition are what give rise to the differences among schools of interpretation.

Many biblical interpreters are rightly concerned to retain the direct, factual claims of the Bible and not to dissolve these into mere myths intended for psychological comfort.  Historical narratives must be respected for what they are.  But respect for these factual claims must not be confused with a blind amalgamation of all kinds of praise, prophecy, vision, and apocalyptic, as if they followed exactly the same conventions and laid down the same expectations as do historical narratives.

If multiple levels exist at one point in Revelation (namely, the beast), might they also exist at other points?  The vision of Christ in Rev 5:6-8 constitutes another example.  For this passage, the linguistic level consists in the textual description sent from John to the seven churches (the actual linguistic material in verses 6-8).  The visionary level consists in the visionary experience that John had of seeing Christ represented in the form of a lamb.  The referential level is the reference to the living Christ, enthroned at God’s right hand.  The symbolical level consists in the symbolic significance of the imagery used.  What is connoted by the imagery of a lamb, the seven horns, the seven eyes, and the taking of the scroll?

Similarly, there are four distinguishable levels in the marriage supper of the Lamb in Rev 19:7-8.  The linguistic level consists in the textual description of 19:7-8.  The visionary level consists in a vision of a bride and fine linen clothing.2  The referential level involves the glorified saints enjoying communion with Christ after his Second Coming.  The symbolical level involves the significance of communion, joy, and beauty attached to the wedding imagery.


Symbolic significance in Rev 11:1-13

This scheme of four levels might seem to involve only fairly easy distinctions.  But the difficulties become apparent when we try to apply the scheme to a more controversial passage like Rev 11:1-13.  Is the temple mentioned in vv 1-2 the literal temple of stone in Jerusalem,3 or is it a symbolic representation of the church4  or of the Jewish people?5  On the linguistic level we have the linguistic description, including the word ναός and related terms.  On the visionary level we have John’s vision, which presumably included seeing a temple of stone.  But this level need not be simply identified with the referential and symbolical levels.  Perhaps this visionary temple, belonging to the visionary level, does refer directly to the temple at Jerusalem on the referential level.  But perhaps not.  In view of the presence of multiple levels elsewhere, it is certainly possible that the historical referent is not a temple of stone, but the true church (cf. 3:12).  In that case, on the symbolical level, the church is depicted as a temple to show that Christians are true heirs of the Old Testament and of Jewish claims (cf. 2:9; 3:9), and that they have the glory of God and his protection, analogous to what was associated with the Old Testament temple.

Likewise, we must ask questions about the significance of the two witnesses of 11:3-13.  Are verses 3-13 a straightforward linguistic prediction, unaccompanied by any vision?  Then we should assume that the two witnesses are two individual human beings who will appear at a particular time in the future.6  But now suppose that verses 3-13 belong to a visionary or quasi-visionary genre.7  Since the beast in verse 7 is a symbolic, visionary figure, the witnesses may be also.  In this case, we need once again to interpret the passage in terms of four levels.  The linguistic level consists as usual in the textual description.  The visionary level consists in a series of visionary representations of events.  John sees a beast, two witnesses, a temple, plagues, and so on.  The referential level consists of the historical referents.  The beast stands for a human antichrist or an antichrist institution (cf. 2 Thess 2:3-12).  The two witnesses, in view of the mention of lampstands in 11:4 and 1:20, probably stand for the witness of the churches, and especially of her martyrs.8  The plagues stand for judgments of God that fall on evildoers when the churches pray and testify against evil.  Such judgments may of course include literal plagues, but they are not confined to the literal.

There are two witnesses, rather than seven witnesses corresponding to the seven lampstands in 1:12,20, in order to alert readers to the correspondence between the present calling of the churches and the prophetic calling of Zerubbabel and Joshua in Zech 4:1-14.  There are further allusions to the prophetic ministry of Moses and Elijah, the great miracle-working prophets of the Old Testament.  These symbolic connotations constitute the fourth, symbolical level.

If this line of interpretation is correct, the passage is less interested in reportorial precision concerning some one-time future event than in revealing the theological significance of the testimony of the churches.  For this purpose, it combines a series of allusions in order to produce a multiple symbolic effect.

What do we say about the death and resurrection of the two witnesses in 11:7-13?  On the visionary level, John sees a death and resurrection just as described.  But the vividness and detail of the description do not result in canceling the symbolic character of the whole sequence of events.  The referential level and the symbolical level still exist as distinguishable levels.  On the referential level, the church undergoes martyrdom and other forms of persecuting pressure from antichrist.  The church is then vindicated by God, and given a more exalted position than before.9  On the symbolical level, the symbolism of death and resurrection is used, both because literal martyr death and bodily resurrection are included in the total plan of God for his churches, and because the pattern of death and resurrection is rooted in Christ’s death and resurrection and in the promises of the Old Testament (Ezek 37; Ps 23:4-6; 16:9-11; Dan 12:2).  The church is being taught to see its own experiences as taking place in union with Christ (Phil 3:10-11; 1:21; 2 Cor 1:9; 4:10-11).

Two qualifications need to be made about the above analysis.  First, John need not have had specific, distinct visions concerning all the things recorded in Revelation.  For example, no word εἶδον (“I saw”) occurs in 11:1-13.10  John does not explicitly claim to have had a distinct vision in which he saw the things described in verses 1-13.  Perhaps the whole content of verses 3-13 was given to him audibly rather than in visionary form.  But contrariwise, verses 1-13 never explicitly deny that there was a vision behind them.  We cannot know for sure.  In view of the similarity in style between verses 1-13 and many other parts of Revelation containing εἶδον, it is wisest to consider them part of the same genre.  Even if some of the parts did not arise from separate visions, the material as a whole invites interpretation along common lines.

Second, the use of future tenses in Revelation, including 11:3-13, may easily promote the idea that we are dealing with straightforward prose prediction.  But Revelation regularly uses loose oscillations between future, present, and aorist tenses in the visionary passages (note, for example, 13:1-8, 11-18).  The oscillation is natural because the historical referent is future, the visionary experience is past, and the present tense can encompass both past and future through the categories of historical present, present for the future, and possibly gnomic present.11  Hence the use of a variety of tenses never moves us clearly beyond the over-arching visionary framework provided in 1:12-22:5.


Symbolic significance in Rev 20:1-6

How do our observations apply to Rev 20:1-6?  Once again, we may expect the presence of four distinct levels.  The linguistic level consists in the text of 20:1-6.  The visionary level consists in John’s actual visions of a descending angel, a dragon, a pit, the seizing of the dragon, the sealing of the pit, the thrones, and so on.  The referential level consists in the historical referents of the dragon, the pit, the thousand years, and the first resurrection.  The symbolical level consists in the symbolical significances of the various figures and events depicted.

Many premillennialists, I suggest, neglect the possibility of the presence of a visionary and a symbolical level.12  Instead, they move almost immediately from the linguistic level to the referential level.  The language of “living” and “first resurrection” is understood in a literal sense.  ἀνάστασις ”resurrection” elsewhere in the New Testament is always used of bodily resurrection.  And, it is claimed, the context of Revelation 20 does not point away from this normal understanding.  Hence “resurrection” must here mean bodily resurrection.  Hence the first resurrection refers to the bodily resurrection of believers at the Second Coming.

Over against this understanding, one must introduce the distinction between a visionary level and a referential level.  What took place on the visionary level?  John saw saints come to life and reign (verse 4).13  In the context of a vision, one could hardly imagine that John’s experience was anything other than seeing a bodily resurrection and its results.  John had to see bodies in order for any information concerning people to be conveyed in a visionary format.  The visionary level thus includes bodily resurrection and its results.  On the symbolic level the text pictures new life and vindication.  And what takes place on the referential level?  The referent is some kind of new life, but the exact form remains to be determined.  The mere fact that the visionary level involves concrete physical representation does not by itself determine the nature of the referential level.

One may illustrate this point by returning to the interpretive situation in Revelation 13 and its vision of the beast.  Some parts of the description translate into historical referents fairly easily.  For example, the language concerning the beast’s blasphemies (verses 5-6) has a literal embodiment in history.  Whether the beast primarily represents a single individual or an institution or a repeated pattern, much of Satan’s persecuting opposition includes literal blasphemy, so it is natural to expect a literal embodiment of this feature.  Moreover, the man of lawlessness of 2 Thess 2:4 involves a similar if not identical pattern.

But not everything is equally as easy.  In Rev 13:3 John saw an actual head and a wound that had been healed.  The visionary material is bodily in nature–concrete, physical, and vivid.  Moreover, it refers to something in history on the referential level.  The actual historical Antichrist will suffer a cataclysmic defeat and yet will return stronger than ever.  What form will his cataclysmic defeat take?  Will it involve an actual physical wound to a human individual?  Or will it involve structural damage to an institution?  Or simply a temporary eclipse of power and influence?  Or all of these?  We may have opinions on these matters.  But we cannot directly read off the answer from 13:3.  Everyone agrees that the head that is wounded is symbolic, not the literal head of a literal animal.  If so, it is quite awkward as well as unnecessary to interpret the wound on the head in a totally nonsymbolic manner.  Rather, the wound is primarily symbolic of defeat or damage.  The vivid character of visual imagery invites this sort of physical representation of things, whether the actual referent involves direct physical damage or not.  Since this example involves counterfeit new life, it suggests that similar flexibility is in order when we come to images of genuine new life elsewhere in Revelation.

In fact, we confront an analogous situation with Rev 20:1-6.  The intended historical referent may be either a bodily resurrection, or new birth, or an enthronement of the disembodied souls of martyrs to reign with Christ in heaven.  If any of these were the intended referents, the visionary framework would lead us to expect that on the visionary level the events would be vividly visible in concrete bodily form.  The imagery of resurrection and life is appropriate.  Naturally, the linguistic level, as a terse transcription of the visionary level, uses the usual words ἀνάστασις and ζάω to describe the vision.

To put it another way, the words for resurrection and life in 20:1-6 are no less and no more “literal” than are the words for beast and wound in 13:1-8.  John used the words “beast” and “wound” because on the visionary level he saw a beast and a (healed) wound.  In Rev 20:1-6 he used the words for resurrection and life because on the visionary level he saw resurrection and life.  In neither case do the words, by themselves, provide any clue as to whether the symbolic, visionary depiction enjoys a direct or indirect relation to its historical referent.  The nature of the referent remains unsettled.  The vocabulary is what it is because it describes a vision, not because it literally describes the referent of the vision.14  Some particular elements in a vision, like the mention of blasphemies in Rev 13:5-6, may indeed have more “direct” correspondence to a historical reality.  But the vision as a whole involves a system of correspondences some of which may also be more indirect and symbolically mediated.

On final example will illustrate my point.  Consider again Rev 11:1-13, and the resurrection in verse 11.  If Rev 11:7-13 came as a vision, John would necessarily have seen the bodies of two human beings, those bodies would have been raised to life, and they would then have been caught up to heaven.  Within a visionary framework, such a picture powerfully depicts the idea of a vindication coming through new life.  In 11:7-13, this conclusion holds true whether the referential level consists of the bodily resurrection of two individual human beings, or of the spiritual vindication of the churches in the face of persecution.

Can we generalize from these observations?  Yes and no.  Not all forms of literature employ all these levels in a distinct way.  But the levels are definitely there in a natural way through most of Revelation, because of its visionary character.  Even within Revelation there are still variations.  The distinction of levels belongs most specifically with the descriptions of visions, and not equally with the other material in Revelation, such as in 1:1-11, 2:1-3:22, and 22:6-21.  Yet certain themes from the visions do infiltrate the other material to a certain extent.  So we may raise the question whether a more general kind of openness to symbolism may be proper to some other types of prophetic communication besides apocalyptic visions.


Concerning literalism

Many premillennialists have thus skirted some key issues when appealing to the supposed literalness of the first resurrection.  They have neglected the visionary and symbolical levels of the discourse.  In fact, premillennial interpreters have often applied a similar literalistic interpretive strategy to the rest of Revelation and to much of Old Testament prophecy as well.15

In such a literalistic interpretive strategy, the visionary level and symbolical level are virtually collapsed into the referential level.  Throughout Revelation, the visions are then understood to be direct transcriptions of future history.  Partly for this reason most premillennialists are futurist in their interpretation of Revelation.  If the visions are more or less direct reproductions of the events, they cannot have been fulfilled in the past.  The future just prior to the Second Coming remains the only arena in which the strange visionary forms of Revelation could find direct, literalistic embodiment.

Among premillennialists, dispensationalists are the most consistently literalistic.  Thus, a dispensationalist like John F. Walvoord can provide us with a clear illustration of this tendency.  Walvoord acknowledges in principle the presence of symbolism in Revelation.16  But his more detailed characterization of his approach permits in practice a fairly thorough-going suppression of symbolism:

In contrast to the other approaches to the book of Revelation, the futurist position allows a more literal interpretation of the specific prophecies of the book.  Though recognizing the frequent symbolism in various prophecies, the events foreshadowed by these symbols and their interpretation are regarded as being fulfilled in a normal way.  Hence, the various judgments of God are actually poured out on the earth as contained in the seals, trumpets, and vials.17

What does Walvoord mean with the words “being fulfilled in a normal way”?  Such words could mean merely that each particular symbol or cluster of symbols refers, in the final analysis, to particular events in history.  But the existence of such correspondences and references is not any more characteristic of futurists than of preterists and historicists.  For example, people in all these schools believe that the beast refers to something in history.  All these schools believe that the locust plague in Revelation 9 has a definite historical referent.  How then would “being fulfilled in a normal way” favor futurism?

In fact, Walvoord has in mind a much more specific commitment to literalism.  The phrase “more literal interpretation” in the above quotation already suggests his commitment.  But if so, “in a normal way” probably means “by way of a simple, straightforward (literal) correspondence between text and historical event.”  That is, Walvoord maintains that there is usually a direct, transparent correspondence between the linguistic level and the referential level.  The text gives a straightforward, direct, “literal” description of the future.  In practice, this position calls for the elimination of symbol.  Symbolism is not acknowledged unless it is practically inescapable.  Most of the material in Revelation is construed as nonsymbolical.  Thus, Walvoord’s approach is in serious tension with his admission, “recognizing the frequent symbolism in various prophecies.”

That Walvoord largely denies or overlooks the presence of a symbolical level is confirmed by Walvoord’s detailed treatment of the trumpets and the vials in his commentary.  Regarding the first trumpet he says,

The tendency on the part of the expositors has been to read into this judgment a symbol of divine chastening rather than literal hail and fire.  The obvious parallel, however, is found in the tenth plague [sic; actually, the seventh plague] in Exodus 9:18-26.  Inasmuch as in the account of Exodus there was literal hail and fire, and the result of the judgment here is the burning up of the third part of trees and all the green grass, there is no solid reason for not taking this judgment in its literal sense.18

Walvoord rightly notes the parallel between the first trumpet plague and the seventh plague in Exodus 9:18-26.  But this parallel does not provide unambiguous evidence in favor of literalistic fulfullment.  The parallel suggests that the two plagues are homologous, but we still have to decide how they are homologous.  Are they homologous in terms of their physically concrete effects, or in terms of their theological significance, or both?  If the homology is physical, then indeed the plague in Exodus and the plague in Revelation are two disasters of exactly the same physical kind.  Since the plague in Exodus involved physical hail and fire, so does the plague in Revelation.  However, if the homology is theological, the two disasters are of the same kind theologically: the first trumpet, embodying the final outpouring of the wrath of God on the world, is a universalization and intensification of the wrath of God expressed earlier in the epoch of the exodus.

The presence of a distinct symbolical level in Revelation is of decisive importance.  If such a separate symbolical level exists, this level will naturally exploit the associations, allusions, and symbolism available from previous revelatory words and deeds.  Hence, John sees a vision of hail and fire, as a symbolical representation of the referential level.  The vision contains hail and fire, while the referential level involves divine chastening.  The distinction between symbol and referent makes it inappropriate simply to assume that literal fire and hail are involved in the referent.

The same problem can be restated in another way.  Walvoord makes no mention of the difference in genre between the historical narrative of Exodus and the visionary, apocalyptic communication of Revelation.  The difference in genre alerts readers to the probability of pervasive symbolism in Revelation, while it denies the appropriateness of such alertness in reading Exodus.

Walvoord’s remarks on other passages display the same tendency.  On the second trumpet:

As in the interpretation of the other trumpet, the tendency of expositors is to give a symbolic meaning to this great judgment.  It is not impossible, however, to suggest a reasonable literal interpretation.19

On Revelation 11:

The guiding lines which govern the exposition to follow regard this chapter as a legitimate prophetic utterance in which the terms are taken normally.  Hence, the great city of 11:8 is identified as the literal city of Jerusalem.  The time periods are taken as literal time periods.  The two witnesses are interpreted as two individuals.  The three and a half days are taken literally.  The earthquake is a literal earthquake.  The seven thousand men who are slain by the earthquake are seven thousand individuals who die in the catastrophe.  The death of the witnesses is literal as are their resurrection and ascension.20

When Walvoord says that “the terms are taken normally,” he means simply that there is no real symbolism at all.  Apparently, “normal” implies “nonsymbolical” and eliminates the visionary and symbolical levels of communication.

But difficulties arise in trying to apply the literalistic strategy consistently and uniformly.  For example, the explicit interpretive statements in Rev 17:9-13 exclude the possibility of interpreting the beast literalistically.  We must invoke separate visionary and symbolical levels, because 17:9-13 tells us that they exist.  Likewise, the general New Testament teaching concerning God the Father and Christ the Son, as well as interpretive comments like 5:6b, 4:5b, and 1:20, require us to recognize the visionary and symbolical levels in the interpretation of 4:1-5:14 and 1:12-16.  Moreover, grammatical-historical interpretation of Revelation, against the background of the struggles of the seven churches, urges us to take into account apocalyptic style, and moves us in the direction of a more preterist interpretation, in which Revelation refers to events in the Roman Empire and is obviously symbolical in character.

What do we do when faced with such evidence?  Literalistic interpreters all admit the presence of symbolism when it is obvious and unavoidable.  But they begin to differ in the rigidity of their literalism when they venture out into the parts of Revelation that do not offer such direct guidelines.  For example, Seiss interprets the star of Rev 9:1 as symbolic of Satan, but the locusts of Rev 9:1-11 are regarded as literal.21  Walvoord interprets the locusts as a symbolic representation of hosts of demons, while the five months are still literal.22  Walter Scott and G. E. Ladd allow that the five months as well as the locusts and the star may be symbolical.23  Literalists understandably fear the introduction of uncontrolled subjectivity, if we are no longer certain what items are nonsymbolical.24  But in fact it is just as subjective to impose a pedestrian, nonsymbolical reading on a visionary genre to which such reading is alien.

Regardless of how far they go in identifying symbolic figures, many interpreters still are captivated by the principle of “literal if possible.”25  Such a principle may seem safe, and indeed it works well as a first approximation for historical narratives and New Testament letters.  But with respect to Revelation and other instances of apocalyptic literature it constantly inhibits interpreters in practice from doing justice to the pervasively visionary character of the discourses.  Instead of accepting the visionary, symbolic medium as a natural form, they constantly fight its own inner integrity by requiring explicit proofs of symbolism for each separate, individual vision.

For example, Walvoord says,

In many instances, where symbols are explained in the book of Revelation, they establish a pattern of interpretation which casts a great deal of light upon the meaning of the book as a whole.  This introduces a presumption that, where expressions are not explained, they can normally be interpreted according to their natural meaning unless the context clearly indicates otherwise.26

The phrase “according to their natural meaning” really means interpreting material nonsymbolically.  (Walvoord’s actual practice in verse-by-verse comments confirms this view.)  Thus, the above quote gives us a recipe for interpreting nonsymbolically except when (a) the symbol is explicitly explained elsewhere in Revelation or (b) when “the context clearly indicates otherwise.”  This recipe rightly appeals to the context of Revelation as the appropriate guide to interpretation.  And it rightly maintains that an explicit explanation of a symbol elsewhere or a clear indication from the immediate context are among the ways in which clues to proper interpretation are obtained.  But the word “clearly” biases interpretation in favor of literalism.  It produces too narrow a basis for interpretation, since in many cases clues may be present but are nevertheless not so obvious.  In some cases Revelation may rely primarily on Old Testament background or loose associations to supply the appropriate meaning for a symbol (as when the four horsemen of Revelation 6 are related to the horsemen of Zech 1:8 and the chariots of Zechariah 6:1; or the bowls of Rev 15-16 are related to Jer 25:15-29 and other OT instances of the cup of God’s wrath).  Whether such cases are “clear” or obvious to an interpreter may depend a good deal on whether the interpreter is expecting them on the basis of an overall assessment of the genre of Revelation.

Thus, Walvoord’s formula is really a variant form of the rule “literal if possible.”  Like that rule, it fights the integrity of any literary genre that uses apocalyptic or symbolico-visionary style in its global structure.

In Revelation, we need to recognize the thorough-going centrality and key character of the visions of 1:12-16 and 4:1-5:14.  These offer us a theological center-point, by proclaiming that God is Lord and judge of history, the Alpha and the Omega, the creator and consummator.  But they also offer us a literary and interpretive center-point for the symbolism of Revelation.  The judgments issuing from the seals, the trumpets, and the bowls originate from God’s throne, his angels, and his temple, thus referring us back to 4:1-5:14.  The evil figures of the dragon, the beast, and the false prophet are counterfeits of the Trinity.  The symbolism throughout Revelation is a kind of overflow from the central symbolism of theophany, the visionary appearing of God.  It is thus a foretaste of the consummation, when God fills all in all with his glory (21:23).  Once one sees the symbolism of theophany as the norm rather than an exception for Revelation, the case for maximizing literalistic interpretation and minimizing the presence of visionary and symbolical levels collapses.


Looking afresh at Rev 20:1-6

When we recognize the visionary form of Rev 20:1-6, all the questions concerning the referents for the passage are reopened.  In theory, such a vision could still refer to a millennial period following the Second Coming.  It could also refer to the regeneration of believers, to a postmillennial triumph of the gospel, or to the vindication of martyrs and other persecuted and suffering Christians in a heavenly reign.  The context as well as the exact language of the vision must provide us with clues.  But mere appeal to an apparent “literalness” and vividness of the resurrection does not help.  Such vividness is characteristic of the visionary form as such.

We would have to travel far beyond the scope of this paper if we were to examine afresh all the data not only from Rev 20:1-10 but from contextual influences in the rest of Revelation and the salient OT and NT connections.  However, for the sake of satisfying readers’ curiosity and for indicating what I myself think are fruitful lines of investigation, I may indicate briefly what my own present position is.

In view of the freedom and fluidity in the use of the visionary form, any interpretive solution must involve some tentativeness.  But I think that a reference to heavenly vindication of martyrs is the most sensible interpretation of Rev 20:1-6.27  The immediate reference to “the souls of those who had been beheaded because of their testimony for Jesus and because of the word of God” (20:4) as well as the fact that the seven churches faced threats of persecution and martyrdom (2:10; 2:13) make such a theme more relevant than any of the others.  I also find weighty evidence in the observations of Meredith G. Kline and R. Fowler White concerning the thematic links with the immediate context and with the rest of Revelation.28  Briefly, Kline argues that the linkage between the first resurrection and the second death (verse 6) suggests the fundamentally spiritual character of both, and that the coming of the second death simultaneously with the new heaven and new earth (20:14-21:2) puts “first” things into the category of preliminary or nonfinal, again qualifying the character of the first resurrection.  Thus the first resurrection refers to vindication of the souls of the martyrs in heavenly reign, which is preliminary in relation to the final resurrection of the body.  The first resurrection involves entrance into spiritual life simultaneous with bodily death, just as the second death involves entrance into spiritual death simultaneous with bodily resurrection.  Both are paradoxical or tensive in character.29  White argues that a number of contextual evidences indicate that 20:1 begins a recapitulation rather than a depiction of events chronologically succeeding 19:11-21.  (1) The nations to be deceived in 20:3 have already been definitively destroyed in 19:11-21.  (2) The battle imagery of 20:7-10 like that of 19:11-21 alludes to Ezek 38, indicating that they are the same battle.  (3) The battle of 16:14 is the great battle: the definite article as well as other elements of the description press us not to multiply battles by splitting apart the visions of 19:11-21 and 20:7-10.  Thus there are significant contextual clues available for pointing us towards the superiority of one interpretation over another.

Because some of these clues are subtle or indirect, and because arguments may still be raised in favor of other alternatives, there will surely be continuing debate on the passage.  My aim in mentioning these clues is not to arrive at an irrefutably correct interpretation, but to underline what everyone already knows, namely that literary context plays a decisive role in determining the referents of particular passages.  At the same time, I want to reassure those accustomed to literalistic interpretation of Revelation that there is a sound alternative.  The recognition of pervasively symbolic, visionary mode of communication in Revelation does not result in abandoning objective meaning or objective referents.  It does make the determination of referents a matter of greater subtlety.  But such is bound to be the case in a genre where the specification of referents is not the exclusive concern.

In any case, one general conclusion definitely follows.  The intrinsic flexibility and relative indirectness of the correspondence between vision and referent in Revelation as a whole should make all interpreters hold their views on Rev 20:1-10 with less dogmatism.



1 On the schools of interpretation see, for example, Leon Morris, The Revelation of St. John (London: Tyndale, 1969) 16-18.  For a more expansive history of the interpretation of Revelation, see Isbon T. Beckwith, The Apocalypse of John (London: Macmillan, 1919; reprinted in Grand Rapids: Baker, 1979) 318-36.

2 But see below on the question of how much John actually saw, and how much may have been revealed to him by speech.

3 So John F. Walvoord, The Revelation of Jesus Christ (Chicago: Moody, 1966) 175-76; J. A. Seiss, The Apocalypse (New York: Charles C. Cook, 1865) 2:161; C. Anderson Scott, The Book of the Revelation (New York: A. C. Armstrong, 1906) 213.

4 G. R. Beasley-Murray, The Book of Revelation (London: Marshall, Morgan & Scott, 1974) 182; Morris, Revelation 146; Robert H. Mounce, The Book of Revelation (Grand Rapids: Eerdmans, 1977) 218-20; Henry B. Swete, The Apocalypse of St John (London: Macmillan, 1911; reprinted in Grand Rapids: Eerdmans, 1968) 132; Alan F. Johnson, Revelation (Grand Rapids: Zondervan, 1983) 108.

5 George E. Ladd, A Commentary on the Revelation of John (Grand Rapids: Eerdmans, 1972) 152-53.

6 So Walvoord, Revelation 179; Seiss, Apocalypse 175.  Ladd (Revelation 154) is more cautious, but prefers to see two individuals as included in the larger symbolic significance.

7 Note Mounce, Revelation 218:

[I] understand the entire section [11:1-13] to be symbolic of the fate of the witnessing church during its final period of opposition and persecution…. That the language of prophecy is highly figurative has nothing to do with the reality of the events predicted.  Symbolism is not a denial of historicity but a matter of literary genre.  Apocalyptic language has as one of its basic characteristics the cryptic and symbolic use of words and phrases.

8 Ibid. 223; Beasley-Murray, Revelation 183-84; Swete, Apocalypse 134; Morris, Revelation 147-48; Johnson, Revelation 110.

9 Morris, Revelation 151; Swete, Apocalypse 140.

10 Some interpreters view the use of εἶδον as mainly a stylistic device, and feel free to deny that John actually experienced the visions that he describes.  But this extreme view is undoubtedly wrong.  It owes its attractiveness to the reductionistic idea that matters of genre are primarily significant as expressions of aesthetic and self-referential expression, and are thereby disconnected from factual claims.  On the relation of factuality to literariness, see Tremper Longman, III, Literary Approaches to Biblical Interpretation (Grand Rapids: Zondervan, 1987) 54-58.

11 Ernest D.W.  Burton, Syntax of the Moods and Tenses in New Testament Greek (3d ed.; Edinburgh: T. & T. Clark, 1898) sections 12, 14, 15.

12 Sophisticated, sensitive premillennialist interpreters like Ladd and Mounce may perhaps have considered such issues in their own minds, but there is still too little sign of it in the actual wording of their commentaries.  See Ladd, Revelation 265-67; Mounce, Revelation 356.

13 Or, as some think, he saw saints who had come to life and had reigned.  See, for example, R. Fowler White, Victory and House Building in Revelation 20:1-21:8: A Thematic Study (Ann Arbor, MI: University Microfilms, 1987) 135-44.  The variations in interpreting the aorists do not affect the major point that I am making.

14 Just this issue is raised by G. C. Berkouwer, The Return of Christ (Grand Rapids: Eerdmans, 1972) 307-315:

We may not tamper with the real, graphic nature of the vision of Revelation 20, nor may we spiritualize the first resurrection.  But one question is still decisive: does this vision intend to sketch for us a particular phase of history?  …this vision is not a narrative account of a future earthly reign of peace at all, but is the apocalyptic unveiling of the reality of salvation in Christ as a backdrop to the reality of the suffering and martyrdom that still continue as long as the dominion of Christ remains hidden.  (P. 307)<

Berkouwer’s words first alerted me to the main issues raised in this article.  Unfortunately, Berkouwer did not become specific enough about the questions of genre and levels of communication.  In the absence of such specificity, it becomes relatively easy for others to dismiss too quickly the questions that Berkouwer raised.

For example, in replying to Berkouwer Alan Johnson pleads for the importance of finding a referent on earth and within history (Revelation 188).  But he does not realize that within Berkouwer’s framework this referent does exist in the form of the vindication of the martyr church (Rev 11:11-13), and martyrs’ prayers that affect earth and history (Rev 6:10-11; 8:4-5).  A distinction between visionary level (where there is a symbolism involving a vision of human bodies) and referential level (where there are events on earth, but not necessarily bodily resurrection) would have helped to clarify Berkouwer’s claim and to show the inconclusiveness of Johnson’s reply.

15 On the slipperiness of the word “literal” and various related terms like “normal” and “plain,” see Vern S. Poythress, Understanding Dispensationalists (Grand Rapids: Zondervan, 1987) 78-96.  In this article I have consistently used the word “literal” to mean prosaic, nonmetaphorical, nonfigurative, and nonsymbolical.  “Literalistic” interpretation tends to find only nonfigurative, literal meanings even when the author intends otherwise.  It corresponds to what my book calls “flat” or “plain” interpretation.

16 Walvoord, Revelation 21-30.

17 Ibid. 21.

18 Ibid. 153.

19 Ibid. 154.

20 Ibid. 175.

21 Seiss, Apocalypse 2:77-79, 92.

22 Walvoord, Revelation 160-61.

23 Walter Scott, Exposition of the Revelation of Jesus Christ (4th ed.; London: Pickering & Inglis, n.d.) 204; Ladd, Revelation 132.

24 Thus Charles L. Feinberg, Millennialism: The Two Major Views (Chicago: Moody, 1980) 42-52, pointedly reminds us of the dangers of subjectivity.  On the desire for objectivity and certainty, see Poythress,Understanding Dispensationalists 62-65.

25 See, e.g., the favorable mention of this principle in Feinberg, Millennialism 46, 48.

26 Walvoord, Revelation 30.

27 Thus my view in substance matches that of Morris, Revelation 233-36.

28 Meredith G. Kline, “The First Resurrection,” WTJ 37 (1974-75) 366-75; idem, “The First Resurrection: A Reaffirmation,” WTJ 39 (1976-77) 110-19; and R. Fowler White, “Reexamining the Evidence for Recapitulation in Rev 20:1-10,” WTJ 51 (1989) 319-44; see also idem, Victory.

29 The standard premillennial concern over ἄχρι and the two uses of ἔζησαν in verses 4 and 5 is defused in Kline’s approach.  The two uses of ἔζησαν refer not to identical kinds of resurrection, but express a homologous relation between the first resurrection and the second death.