Research Report #35 (1987)

SYNOPTIC HARMONIZATION:
SOME PRINCIPLES FROM HISTORY
& FROM THE BOOK OF ACTS

Robert C. Newman
Biblical Theological Seminary
Interdisciplinary Biblical
Research Institute

Copyright © 1987 by Robert C. Newman. All rights reserved.
 

EDITOR'S NOTE

Although the author is in agreement with the doctrinal statement of IBRI, it does not follow that all of the viewpoints espoused in this paper represent official positions of IBRI. Since one of the purposes of the IBRI report series is to serve as a preprint forum, it is possible that the author has revised some aspects of this work since it was first written. 

ISBN 0-944788- 35-1



 
 

THE PROBLEM STATED

When discussing the reliability of Scripture, opponents of inerrancy often argue from differences between parallel narratives in the Gospels.[1] Usually such writers pick out some event narrated in two or more of the Gospels and note that one says this and another that, which are apparently inconsistent. A number of such examples are collected and it is argued that probability is decidedly against all of them being explained in a way that avoids error in the Gospels.[2] Such writers often go on to construct elaborate theories regarding the authorship, audience and theology of each Gospel on the basis of such divergences.[3]

In this paper, we want to discuss such a line of argument and how Gospel parallels should be harmonized. Let us first look briefly at five examples of such differences in Gospel accounts to illustrate the problem, without initially making any attempt at harmonization. These examples are neither the most difficult nor the easiest to harmonize. They do show up rather frequently in such discussions and are likely to strike anyone who studies them as problem passages.

1. Jesus Tempted in the Wilderness. The temptations are recorded in some detail in Matthew and Luke, but only very briefly in Mark. The main question which arises concerns the order of the second and third temptations: to worship Satan and to jump from the pinnacle of the temple. Luke gives the order above and Matthew the reverse. Which order, if either, is correct? Isn't this an error in Scripture?

2. The Centurion's Servant Healed. A different sort of problem appears in this incident, narrated in Matt 8:5-13 and Luke 7:1-10. In Matthew it seems that the centurion personally speaks to Jesus, asking him to come heal his servant; when Jesus begins to comply, the centurion decides that he is unworthy to trouble Jesus in this way. In Luke, instead of the centurion coming in person, the Jewish elders intercede on his behalf to ask Jesus to come and heal the servant. Later, on the way, the centurion's friends come with a message from him that he is unworthy to have Jesus come. Did the centurion come directly to Jesus or not?

3. The Demons and Pigs. On the east shore of the Sea of Galilee, Jesus casts out demons into a large herd of pigs which run down the slope and drown in the lake. Matthew mentions two demoniacs; Mark and Luke only one. How many were there?

4. Healing of Blind near Jericho. This account is given in Matthew 20, Mark 10 and Luke 18. Matthew speaks of two blind men being healed; Mark and Luke, only of one. Matthew and Mark narrate this as happening when Jesus leaves Jericho; Luke, as he approaches Jericho. Yet the words that pass between Jesus and the blind are very similar. How are these features to be explained?

5. Peter's Three Denials. Out in the courtyard at Jesus' trial, Peter denies his master three times. Accounts occur in all four Gospels: Matthew 26, Mark 14, Luke 22 and John 18. Mark alone mentions two crowings of the rooster, one after a single denial, the second after all three. The other Gospels mention only one crowing. Details regarding who confronts Peter at each denial are different in the various accounts.

These examples by no means exhaust the number of such variations occurring in the Gospel accounts, but they do give a feel for the type of problem we are considering. Such divergences are most common in the synoptic Gospels Matthew, Mark and Luke, since they have the greatest narrative overlap; they also occur in John on the rarer occasions when it parallels the others.

Rather than immediately discussing how these particular passages might be harmonized, let us next lay some groundwork by considering two related matters. One concerns the nature of history and history writing; the other, the possibility of establishing some sort of control on what the Gospels writers have a real feeling for this complexity.

Complexity of History

To see something of this complexity, consider three features that are important to history: (1) people's words, (2) their actions, and (3) their interactions,

Consider first a person's words. When we read a written account, we are really looking at an abbreviated version of speech, even in the rare case that it is a transcript of what was spoken. It doesn't usually tell us whether the speaker was standing or sitting, speaking loudly or softly, with what emotion or gestures, nor give any indication of changes in volume and intonation. Of course, if the transcript is in our native language, we should be able to make a number of good guesses about such things from shared culture and knowledge of the context, but these are still guesses and many will be wrong. Thus even if we write down a person's every word, we only have a condensed summary of what was actually said.

Imagine a person speaking at a fairly normal rate of speed, say 125 words per minute. It will take about one page, double spaced, to record two minutes' speech. Suppose such a person only speaks for one hour per day. This is probably more than many speak -- though not teachers or pastors! It will take 30 pages per day to record this person, about 900 pages per month, over ten thousand per year! So the simple matter of recording the abbreviated written form of one person's speech can get out of hand very quickly.

In Jesus' ministry, we are concerned with some three years or more of teaching, and we can be sure that on most days he spoke for more than one hour. As we have nothing like thirty thousand pages of Gospel material, it is obvious that the Gospels are strongly abbreviated.

Now consider recording a person's actions. Here the question of level of detail is even more problematic, as there is no standard method for recording actions as there is for words, at least before film and videotape were invented. Even with modern video recording technology, information is lost because the camera has no depth perception and can only be in one place at a time. Such phenomena are familiar to those who watch sports on television, where even multiple cameras and playback techniques don't catch everything.

Let us consider, then, a very restricted level of detail regarding one person's action. Imagine the person keeps a diary and writes one page per day. Though this is less detail than is found in the Gospels for some few days, on the average it is far more than we have recorded there. At one page per day, Jesus' public minstry would still take up over a thousand pages, several times more than we actually have.

Considering just these two points, the words and actions of a single person, it is clear that the Gospels must be condensations of what actually went on. As in all history writing, the actual events are sufficiently complex that any attempt to describe them must be simplified in many ways. The situation becomes even more complex when we consider interactions between people.

History does not consist of just one person speaking and acting, but of the interaction between persons, not to mention various impersonal phenomena. As soon as we consider the words and actions of two people, we must add the interaction between these people. With three people, there are three interactions; with four, six interactions. With n people, the number of interactions is given by 1/2 n times n-1. Thus an account of Jesus and his apostles involves 13 people or 78 relationships. Including the 120 believers in Jerusalem between the ascension and Pentecost, we have 7140 interactions. When we consider the feeding of the 5000 or the 3000 converted at Pentecost, the number of possible relationships goes over a million. Of course, not all these people will actually have any significant interaction, as a sort of saturation effect sets in with large numbers. Yet most of us still have minor contacts with thousands of people in our daily lives. A narrator describing such interaction thus resorts to considerable simplification, concentrating on only a few events of the many that actually happened.

Differing Viewpoints and Emphases

Consider next the differing perspectives and interests that writers bring with them to the narration of history. Of course, at the time a particular event is happening, some people are present and some are not. Of those present, some are located so as to see certain details better; others may see other details better. Some observers are older and more experienced; other younger, with less experience. The personalities and interests of the obervers will also differ.

One of the best ways to get a feel for this sort of thing, if we have grown brothers and sisters, is to get together at a family reunion and reminisce about old times. We quickly find out that although we remember a lot of the same events, we often remember entirely different details about these events. This is because we were at various ages when the events happened, interested in different things, standing in various places, and perhaps some were in on secrets that others were not. If you have no brothers or sisters, try this out at your next high school or college reunion.

Thus, in any group of narratives of the same historical events, we will find differing viewpoints and emphases without even bringing in the question of error at all. If, in addition, the writer is going to restrict himself to fifty or one hundred pages as our Gospel writers do, he cannot write down everything of interest anyway. Unless we assume the Gospel writers had far less connection with the events of Jesus' ministry than historical evidence indicates (and that there was far less oral information around than even most liberal NT scholars assume), the writers must have used much less material than they knew. They must have restricted themselves, among all the things they know about the subject, to those particular matters which would produce the emphases they desired to give.

A writer's choice of materials will thus be influenced by his own personality, interests, and information, plus his specific purpose in writing. If the writer is aware of other narratives on the same subject, he may choose to supplement rather than merely repeat.

Order in a Narrative

A third area to consider is the order in which a narrative is written. When a person sits down to write a narrative, he must write in some sort of linear order. That's the way human language is. One word comes after another after another in a string. This is not to say that the writer didn't go through a number of rough drafts, and that the first word in the published edition was necessarily the first word he originally wrote. Rather, the final form as it circulates to the public is in some specific order. A narrative, of course, could be disorderly in the sense that the writer did not trouble himself to impose much order on what he put down. But that is not our point here. The fact is that an author can say only one thing at a time, but history has many things going on at the same time.

The natural order for a narrative is obviously chronological. Yet it is nearly impossible, and certainly undesirable, to use a strict chronological order everywhere in a narrative, because the writer must deal with the problem of relationships discussed above. If the writer is following a single person as central to his narrative (as the Gospels focus on Jesus), other persons must still come into the narrative as they interact with him. To give the reader some sense of what is happening, the writer usually tells a little about the newly introduced character, digressing from his chronological order to give a flashback, and often carries him forward somewhat as he leaves the narrative. Thus we have persons introduced in the Gospels such as a fellow who was blind from birth or a woman who had suffered for years from doctors trying to cure her hemorrhage. Similarly, when a person is dismissed from the narrative, the author may say that he went out into the cities to tell of Jesus. Thus the narrator will go backward or forward chronologically before coming back to pick up the main thread of his narrative. This sort of departure from chronological order is also quite natural.

Another technique frequently used in a narrative is to deal with a particular subject all in one place rather than leaving references to it scattered in several places. A writer narrating what Jesus had to say about the kingdom of God might leave his remarks in separate places where they occurred or he might bring them together into one place, perhaps a sermon which emphasized this subject. Alternatively, he might select a sermon on this subject for detailed reporting and omit Jesus' references to the matter elsewhere, without actually moving any material chronologically. It would probably be impossible to tell which of these two schemes was followed in a particular case, but it is certainly not uncommon for an extended narrative to handle some subjects topically.

For purposes of emphasis, an incident might be moved within a short narrative to put it in a prominent place, usually first or last. It is a psychological fact that people remember the beginning and end of a sequence better than its middle.

Thus any historical narrative can normally be expected to be chronological in order, but not exclusively so. Some reasons for departure from chronological order include: picking up the background of a person being introduced; following up a person being dropped; dealing with a particular topic at one place in a narrative; and emphasizing a particular incident. Doubtless other reasons also exist.

Compression in a Narrative

Lastly we should mention the matter of compression. We have already suggested that, in view of the complexity of history, all narrative is simplified. But when we begin to compare narratives of the same incident, we see that some are shorter than others.

One obvious reason for this is the differing total length of a narrative. A one-volume work on church history will typically have shorter narratives on particular events than a seven-volume work. Another reason for differing length is specialization. A college text of the history of western civilization might have only a chapter on the Renaissance, whereas a specialist might write a whole volume on the subject. In either case, the shorter account, having less material on a given subject, is likely to be more simplified than the longer account. Of course an author may choose to shorten by deleting events rather than by giving less detail in the particular events he does report. This is one way of preserving vividness in a brief account. In any case the shorter account cannot deal with as many events in as much detail as the longer account. The shortening must involve either simplification of events or deleting of events.

Now the Gospel accounts are short compared to other ancient histories. Josephus' Jewish War contains seven books and his Antiquities twenty, each book being roughly the length of a Gospel. The pagan historians Tacitus, Suetonius, Herodotus and Thucydides wrote comparably long works. The reason for this difference in length is not certain; probably the secular historians were writing for relatively wealthy readers who could easily afford multivolume works, whereas the Gospel writers wanted their works to be available more widely, to readers of modest means. In any case, the need for brevity in the Gospels put more pressure on their authors to compress the narrative.

With a given amount of space an author may compress a particular narrative because he wishes to save room for something another writer has not included or because he wishes to emphasize something the other writer is giving little space to. A writer pressed for space will give more space to things he wishes to emphasize, less space to matters he needs to mention but not emphasize, and no space to things he can get along without.

So much for complexity, viewpoint, order and compression as they relate to history and history writing. These observations do not solve the problem passages we began with, but they do suggest some factors to be kept in mind when thinking about possible solutions.

Let us turn now to consider the phenomenon of parallel passages in a single author.

PARALLEL PASSAGES FROM A SINGLE AUTHOR

In attempting to deal with parallel passages in the Gospels, it is helpful to study parallels in a controlled situation where questions of multiple authors and differing viewpoints do not arise. This will help us to evaluate claims that the differences in such accounts indicate that one author was unaware of information he did not include, that he had a differing theological viewpoint, was biased, or wished (implicitly) to correct the statement of an earlier author.

Parallel passages are much rarer elsewhere in the New Testament than in the Gospels, especially so in narrative materials. Yet there are several in the book of Acts, and most scholars agree that this was written by a single author who also wrote the Gospel of Luke.[4] Of course, some complain that the author had axes to grind, but we will probably not get agreement on this question until the last judgment. However, Luke was certainly a competent author, and one cannot easily say Luke did not know what he doesn't mention in one account if he mentions it in a parallel. Thus we here have an opportunity to examine what one writer, narrating a particular event, does when he mentions it on more than one occasion -- what things he puts in or leaves out, how he represents various details. Let us see if this does not give us some help for the phenomena we find in parallel narratives in the Gospels.

Let us consider six examples of parallel passages, all found in Acts -- though one version of one parallel occurs in the Gospel of Luke, generally conceded to be written by the same author.

A. The Ascension (Luke 24:50-53; Acts 1:1-12)

The first example we consider is the one with parallels in both Luke and Acts. The account is very brief in Luke and much more extensive in Acts. In fact, following the textual variant which omits "was carried up into heaven," it is not even obvious that Luke gives the ascension. It merely says Jesus parted from his disciples, which he did on several occasions. It is only by Luke's remarks at the beginning of Acts that we are assured that the end of Luke is the ascension:

The first account I composed, Theophilus, about all that Jesus began to do and
teach until the day when He was taken up, after He had by the Holy Spirit given
orders to the apostles whom He had chosen.

Consider, then, the details in each passage:

LUKE 24  ACTS 1
Bethany (50) Mount of Olives (12)
Blessed them* (50-51) Several details (4-8), no
blessing mentioned
Parted from them (51) Lifted up, clouds received (9-11)
Men not mentioned Two men (angels) speak (10-11)
Returned to Jerusalem (52) Returned to Jerusalem (12)
With joy* (52) Mentions prayer, choice of
apostle (13ff)
Continually in temple* (52) Upper room (13); temple not
mentioned until 2:46

Obviously the Acts passage, being much longer, will have more detail. Yet even so, the Luke account has several details (marked *) not found in Acts.

Consider the differences. Luke mentions going to Bethany; Acts, to the Mount of Olives. One who knows the geography of Palestine will recall that Bethany is on the east slope of the Mount of Olives. One who doesn't might think we have two different locations. In fact the traditional shrines of the ascension are on top of the Mount rather than in Bethany, but this is another matter. Where did the ascension take place? On the Mount of Olives? At Bethany? A reasonable answer: both!

The blessing mentioned in Luke is not found in Acts. This is rather interesting. Though Acts devotes several verses to what Jesus said to his disciples before his ascension, it doesn't happen to mention that he blessed them. Didn't Luke know it? Of course he did. But for some reason we can only guess at (already in Luke? not his concern here in Acts?) he doesn't use this information here.

When we come to the actual ascension itself, the account is sketchy in Luke but detailed in Acts. In Luke, "He parted from them," with a variant addition "and He was carried up into heaven." Acts adds that "a cloud received Him out of their sight," and goes on to narrate the words of the angels. Notice that the two angels are not mentioned in Luke. Could someone narrate Jesus' ascension and not mention the angels? Yes, Luke did in his Gospel. Why? We don't know. Some have suggested that Luke did not know these additional details when he wrote his Gospel, but we need to remember that this is just a guess. It is equally possible that Luke is already planning to write Acts and to begin with the ascension, so that he cuts off the Gospel at this point with a very brief summary pointing forward. Josephus does the same thing in ending book 3 of his Jewish War.

Both accounts mention that the disciples return to Jerusalem, but Luke adds "with joy." The Luke account has the disciples "continually in the temple" but Acts mentions only the upper room and doesn't explicitly mention the temple until nearly the end of chapter two. Since the Gospel account is heavily condensed, it is possible that the Luke is summarizing the disciples' activity for some months after the ascension. Whatever the case, the Gospel emphasizes worship: mentioning joy, praise and the temple. Acts puts the stress on proclamation, though at this point in the narrative we are still awaiting the coming of the Spirit before the proclamation begins.

The chronology of these two accounts is most striking. Reading through the Luke account and making the simplest assumptions regarding time, one would suppose the ascension occurred sometime during the night following the resurrection. In fact, this is not the case, but all the evidence for a longer span comes from the Acts account. Thus we have great chronological compression here in Luke, a possibility we should keep in mind when we consider differences between parallel accounts in the Gospels.

Apparently these two quite divergent accounts are descriptions of the same events by the same author! This in itself undercuts a great deal that has been written about even smaller divergences between Gospels. To avoid this conclusion some liberals have assumed that Luke is inconsistent or is correcting his earlier ascension account in the Gospel by his later one in Acts.[5] Yet we have no other reason to assume this is so, and the use of compression and supplement provide more likely explanations in view of what we know about history writing.

B. Paul's Conversion (Acts 9:1-9; 22:4-11; 26:9-18)

Consider next the accounts of Paul's conversion. Here we have three parallels, as Luke reports the events not only at the proper place in his narrative, but also when Paul recounts them in the mob scene at his arrest and again at a hearing before the rulers Festus and Herod Agrippa II. These three accounts written by a single author in one book may profitably be compared to the triplets found frequently in the synoptic Gospels.

The accounts are of similar length and content, as may be seen in the chart below.

ACTS 9 ACTS 22 ACTS 26
persecution (1-2) same (4) same (9-11)
authority from high priest (1-2) same (5) same (12)
on way to Damascus (2-3) same (6) same (12)
approaching city (3) same (6) not mentioned
time not mentioned about noon (6) at midday (13)
light from heaven (3) very bright light from heaven light from heaven br than sun
Paul falls to ground (4) same (7) all fall (14)
men stood speechless (7) not mentioned not mentioned
voice (4) same (7) Hebrew voice (14)
Saul, Saul (4) same (7) same (14)
not mentioned not mentioned hard to kick (14)
Who are you, Lord? (5) same (8) same (15)
Jesus (5) Jesus the Naz (8) Jesus (15)
men saw no one (7) men saw light (9) not mentioned
men heard voice (7) didn't hear (9) not mentioned
not mentioned what to do? (10) not mentioned
Arise & enter the city (6) same (10) Arise & go to Gentiles (16-18)
Paul blind (8) same (11) not mentioned

These accounts overlap quite substantially, as may be seen from the frequent occurrence of the entry "same" in the chart above. All note Paul's activity in persecuting Christians, his authority from the high priest, his trip to Damascus, the light from heaven, Paul's falling down, the voice "Saul, Saul, why are you persecuting me?", Paul's question, "Who are you, Lord?", the voice identifying itself as Jesus, and the command for Paul to arise.

Yet in spite of such agreements there are a number of differences, some merely adding detail, others apparently inconsistent. Only the recountings give the time of day and the fact that the light was very bright, though the latter is probably implied in the fact that Paul is now blind. Yet Gospel commentators have sometimes made much of differences this small: "In source A, Paul is supernaturally struck blind; in B, it is a natural result of the bright light. A is therefore the later, more developed account"!

One apparent inconsistency involves what happened to Paul's companions. Did they "fall to the ground" as in 26:14 or "stand speechless" as in 9:7? These are easily reconciled on the assumption that Luke knew what he was talking about and didn't have a lapse of memory. One possibility is that the men fell down but then got up and stood speechless listening to the voice. But the verb histemi can mean both "stand up" and "stand still," that is, "come to a stop."[6] So another possibility is that Paul's traveling companions stopped when they heard the voice, falling to the ground. Since we were not there, we do not know whether one of these alternatives is correct, or something even more complex is. We only know what the accounts say. We can assume Luke knew what he was writing about or that he didn't, and build hypotheses from each. There is no way of using statistics to convert our ignorance to knowledge in such matters. But if the doctrine of inerrancy is biblical, it assures us that the accounts are accurate and so reconcilable, even though it does not tell us how to reconcile them.

The third account adds the details that the voice from heaven spoke in Hebrew and that it mentioned how hard is was for Paul to kick against the goads. The second account notes that the voice identified itself as "Jesus the Nazarene," and that Paul responds by asking "What shall I do?" These may indicate features which Paul mentioned at these particular retellings, but in any case Luke would have access to all this material in composing his original narration in chapter nine. Obviously, Luke has chosen not to use all he knows about Paul's conversion in his first telling, thereby providing more interest for the retellings and incidentally provoking pages of scholarly argument in the centuries to follow! That a writer may intentionally or incidentally omit material he knows about is a sufficiently common phenomenon to make arguments from silence in these matters quite precarious.

Another apparent contradiction occurs where 9:7 says the men "heard the voice" from heaven and 22:9 says they "didn't hear the voice." The NASB has softened this by translating the second "hear" as "understand," which is probably the right solution. The verb akouo can mean either "hear" or "understand," and the noun phone either "voice" or "sound."[7] Probably the writer intends something parallel to John 12:28-29, where a voice from heaven speaks to Jesus; some in the crowd think it is an angel voice, while others think it is thunder. The men on the road to Damascus apparently heard the sound but could not decipher the voice, a phenomenon familiar to us in conversation overheard from another room. Here again an apparent contradiction has a rather straightforward explanation. Both accounts occur in one book which in its final form was the responsibility of one competent writer who had all the material before him.

What about the divergence between "arise and enter the city" in Acts 9 and 22 and "arise and go to the Gentiles" in Acts 26? Not being at the original event, we cannot say that Jesus did not give Paul such instructions within moments of his conversion. But it is also possible that Luke or Paul is here summarizing: that Paul gave a longer speech which Luke is compressing, or that Paul himself condensed the account here to move on to his remark about not being disobedient to the heavenly vision. The phenomenon of compression in history writing is too common for us to be sure of our reconstructions without much more data than we have. Apparently neither Luke nor the other Gospel writers intended to provide data for modern historians to build their own reconstructions -- certainly not for reconstructions which snatch at apparent inconsistencies, pit one author against another, or deny the miraculous.

C. Paul and Ananias (Acts 9:10-19; 22:12-16)

Let us move on to the immediately following material in Acts 9 and 22, the account of Paul's contact with the Damascus Christian Ananias. The first version is about twice as long as the second, which is Paul's retelling of the event to the Jerusalem mob. Though obviously the same event, much of the detail is different.

ACTS 9 ACTS 22
Ananias, disciple (10) Ananias, devout re/ law (12)
vision to A (10-12) not mentioned
A's response (13-14) not mentioned
God's answer (15-16) not mentioned
comes to Paul (17) same (13)
lays hands on (17) stands near (13)
Bro Paul, receive sight (17) slightly shorter (13)
be filled with HS (17) not mentioned
scales fell off (18) not mentioned
regained sight (18) same (13)
not mentioned A's speech (14-16)
baptized (18) implied not explicit (16)
ate (19) not mentioned

One interesting difference between these two accounts may be due to the different audience of each -- the Christian reader of Acts versus the Jewish listeners to Paul's speech. Acts 9 calls Ananias a disciple, i.e., a believer in Jesus' messiahship. This is passed over when Paul speaks to the Jewish mob, where Ananias is described as "devout with respect to the law." This probably reflects Paul's choice of words so as not to antagonize his audience further.

The whole section about the vision to Ananias (9:13-16) is omitted in the recounting. This might be either Paul's or Luke's compression of the narrative, which moves directly to Ananias' coming to see Paul. Note that the longer speech of Ananias given in Acts 22:14-16 does imply that Ananias has somehow learned what God's intention is, presumably through some sort of revelation like the vision in Acts 9.

When Ananias arrives, according to Acts 9 he lays his hands on Paul, whereas Acts 22 only mentions that he "stands near." Here again we have a pair of expressions which can be read as contradictory, but need not be. One can stand near without laying on hands, but if one is laying on hands, it is natural to be standing and near.

Ananias' speech in 9:17 precedes Paul's healing, whereas in 22:14-16 it follows. This could be a case where Luke has rearranged the order in one case or the other. But since the speeches are rather different, it may be that Luke has merely compressed the account differently in the two cases. It is doubtful that Ananias spoke only a couple of sentences while he was with Paul, particularly as Ananias is probably the one who subsequently baptized Paul.

What we see here are two accounts, one somewhat longer, each containing details not found in the other. Some of the compression in each could be viewed as intended to avoid duplication: e.g., Acts 9 gives the material on God's call in the vision to Ananias, whereas in Acts 22 Luke (or Paul) has not chosen to narrate this vision, so God's call is mentioned in Ananias' speech to Paul. Whether this was done by selection from the complex of original details or by chronological shuffling, we weren't there and thus don't know enough to be sure.

D. Paul in Jerusalem (Acts 9:26-30; 22:17-21)

Continuing on in Paul's career, the next set of parallels involves two accounts of very similar length with almost no overlap of detail.

ACTS 9 ACTS 22
Paul to Jerusalem (26) same (17)
disciples afraid of him (26) not mentioned
Barnabas intercedes (27) not mentioned
Paul admitted (28) not mentioned
speaks out, argues with Hellenistic Jews (28-29) not mentioned
they try to kill Paul (29) not mentioned
not mentioned praying in temple (17)
not mentioned falls into trance (17)
not mentioned warning from Lord (18)
not mentioned Paul's answer (19-20)
not mentioned God: Go to Gentiles (21)
disciples send to Tarsus (30) not mentioned

Clearly, both narrate Paul's first visit to Jerusalem after his conversion and both note that Paul leaves the city because of opposition from the Jews there. Yet the emphasis of each narrative is quite different, the first concentrating on Paul's interaction with the disciples, the second on his interaction with God. IN general, a narrative can easily be compressed by concentrating on only one such interaction, as here. In seeking to justify himself to Jewish religious opponents, Paul would naturally not concentrate on his connections with Christians nor on the fact that Jews were also trying to kill him earlier, whereas he might well emphasize that his mission to the Gentiles was divinely ordained. Luke, writing up his account, might then use the other material about Paul's relations with the disciples as a supplement.

How are we to handle the divergence that Acts 9 pictures the disciples sending Paul away, whereas Acts 22 has God sending him off? Given that Luke is competent, that he wrote both accounts and knows what he has given in each, Luke is telling us that both God's warning and the disciples' help were instrumental in getting Paul out of Jerusalem. Not being there ourselves, our reconstructions are necessarily uncertain. But it is not unreasonable to suppose that God's warning convinced Paul to leave when otherwise he felt his opponents would surely listen to him; and that the believers aided him physically and perhaps financially in making good his escape.

E. Peter's Vision (Acts 10:9-20; 11:5-12)

The events surrounding Peter's evangelistic work in Caesarea are narrated by Luke in Acts 10 and then recounted by Peter in Acts 11 to justify his behavior to the believers in Jerusalem. The result is a pair of accounts both of Peter's vision at Joppa and his speech at Cornelius' house. Let us look at the first of these in this section and the second in the section following.

ACTS 10 ACTS 11
Peter in Joppa (9:43) same (5)
on housetop, hungry (9-10) not mentioned
vision (10-11) same (5)
object lowered (11) same (5)
animals in it (12) adds "wild beasts" (6)
Arise, P, kill & eat (13) same (7)
Peter's answer (14) same (8)
God's response (15) same (9)
happened 3 times (16) same (10)
withdrawn to sky (16) same (10)
Peter perplexed (17) not mentioned
men arrive (17-18) same (11)
HS's instructions (19-20) summarized briefly (12)

The Acts 10 account is somewhat longer than the recap in Acts 11, twelve verses rather than eight. The parallels are rather what we would expect, not at all unusual like several of the cases we have discussed so far. Acts 10 being longer, it is usually more detailed, having several items like the time of day, Peter's location and hunger, and his perplexity after the vision, that are not in the shorter account. Acts 11 does have one detail, the reference to wild animals in verse 6, that is not in Acts 10. Otherwise Acts 11 is pretty much a shortened repetition of Acts 10. Thus even a single writer might basically repeat an account when giving a parallel on one occasion, whereas elsewhere he might provide parallels that are largely supplementary.

F. Peter at Cornelius' House (Acts 10:23-48; 11:12-17)

Moving on in the narrative, we come to the incidents which occur when Peter arrives at Cornelius' house. In this case, the recap which Peter later gives to the Christians at Jerusalem is much condensed. Also we have a triplet section for Cornelius' vision, since Luke narrates the vision at 10:3-8, then has Cornelius recount this to Peter in 10:30-33, and Peter to the Jerusalem believers in 11:13-14.
ACTS 10A ACTS 10B ACTS 11
went next day (23) not mentioned
some brethren also go (23) six also go (12)
Cornelius' reception (24-25) not mentioned
Peter's reaction (26) not mentioned
Peter's speech (28-29) not mentioned
Cornelius recaps vision (30-33) same (13-14)
9th hour (3) same (30) not mentioned
angel (3) man in shining garments (30) angel (13)
prayer & alms ascended (4) prayer heard, alms remembered (31) not mentioned
same (5-6) send to Joppa for Simon P (32) same (13)
not mentioned not mentioned words of salvation (14)
details on sending (7-8) not mentioend not mentioned
Peter's speech detailed (34-43) as I began to speak (15)
HS falls on all listening (44) HS fell (15)
amazement of Jewish Xns (45) not mentioned
speaking in tongues (46) just as on us at beginning (15)
not mentioned P remembers Jesus' wds (16)
Peter orders baptism (47-48) who I to stand in God's way? 

In general, we see here what we would expect: the original narration of the vision to Cornelius in 10:3-8 is shortened in 10:30-33 and still more in 11:13-14. Likewise, the original narration of Peter at Cornelius' house in 10:23-48 is shortened in 11:12-17. Even so, it is notable that the account in chapter 11 contains three items not found in the original narrative of chapter 10: the number of believers who accompany Peter from Joppa to Caesarea (11:12); the angel's description of Peter's message as "words of salvation" (14); and Peter's remark that when the Holy Spirit came upon the Gentiles he remembered Jesus' words contrasting his baptism with John's (16).

Also notable is the fact that 10:3 and 11:13 describe the visitor in Cornelius' vison as an angel, whereas 10:30 calls him "a man ... in shining garments." Very frequently in discussions of the resurrection accounts in the Gospels, much is made of the fact that Mark says a man in a white robe appeared to the women, whereas in Matthew he is an angel, as though Matthew has inflated Mark's account to make it more miraculous. Here in Acts a single writer has chosen to call the same visitant an angel in one place and a man in another. Of course, this has precendent stretching all the way back to Genesis, where Abraham's visitors are called men in 18:2 and angels in 19:1. Apparently angels either naturally look like humans or they choose to appear in human form when delivering messages to humans.

The treatment of Peter's speech is different in chapters 10 and 11. Ten verses are given to it in 10:34-43, whereas the whole is summarized in one phrase "as I began to speak" in 11:15. Some might see a contradiction here also -- not that a speech cannot be summarized in a phrase, but that the phrase is "began to speak" whereas 10:34-43 itself seems to be a summary of a rather long discourse.

Not being there ourselves, we don't really know that chapter 10 gives only a summary, but this seems likely given the common practice of compression and the number of subjects mentioned in these verses. The few sentences of 10:34-43 sketch Jesus' ministry, his crucifixion, resurrection, the commissioning of the apostles as witnesses to the resurrection, the testimony of Old Testament prophecy to Jesus, and forgiveness through belief in him. Of course, it is possible that Peter said this much in a few sentences, that he had just finished his (summarizing) introduction to an intended long speech when he was suddenly interrupted. It seems more likely, though, that Peter did speak for a while but was intending to go on much longer when the Spirit intervened. Again we don't have enough information to be sure which of several alternatives is correct.

At least two reasons for this compression in 11:15 might be suggested. Peter may have compressed it, since he did not need to reproduce his speech for his Christian audience back in Jerusalem. After all, he was defending his action in visiting and eating with Gentiles, not the content of his Gospel. On the other hand, Luke may have compressed it instead (or as well), since he had already given a full summary in the previous chapter. Yet again we see both how little we can be certain of details when we try to go beyond what the text provides and how extensively a writer or speaker may compress a narrative if he feels the need to do so.

Another feature in this account which might be seen as a conflict regards the activity of the Spirit. According to 10:44, "the Holy Spirit fell upon all those who were listening to the message," whereas 11:15 says "as I began to speak the Holy Spirit fell upon them just as he did upon us at the beginning." Taking the simplest interpretation of each of these separately, the first would have the Spirit come upon the entire audience including Peter's Christian companions (10:23; 11:12), whereas the latter only upon the Gentiles present.

Actually either of these interpretations could be correct without forcing a contradiction upon the other parallel passage. It is merely necessary to take a slightly more complex interpretation of the passage which seems in conflict. Suppose first that the Spirit fell upon all present. Then one could understand 11:15 (given Peter's audience in Jerusalem) not to deny that the Spirit fell upon his companions also, but as making the point that God poured out his Spirit on Gentiles here just as he did upon Jews at Pentecost. It would be irrelevant to the Jews in Jerusalem that he also fell on a few more Jewish Christians as well. In fact, the way Peter knew that this event was just like Pentecost might be that he personally had the same experience both times.

Or suppose that the Spirit only fell on Gentiles at Cornelius' house. Then one could understand 10:44 to mean that the Spirit fell upon those listening to the message who had never heard it before, the writer not bothering to qualify his statement by excluding the Jews present. Though I prefer the former alternative, we really do not have enough information to make a certain choice.

This example is a good illustration of the danger of pressing a passage as much as possible and then finding contradictions with parallel passages. Instead, we here have parallels written by a competent author who was in possession of many more facts than have come down to us. The parallels thus serve to show us how far each can be pressed and what range of meaning we should derive from the whole account.

Lastly, 10:47-48 has Peter ordering the baptism of the new Gentile converts, while 11:17 seems more passive: "who was I that I could stand in God's way?" Again, if one presses each passage, it may look like either Peter or Luke was bending the facts. But this ignores several features which suggest a reconciliation. First, Peter was certainly the believer in command at Cornelius' house. If water baptism was to be administered, it would naturally be at his word. Second, given the close connection between Holy Spirit baptism and its symbol, water baptism, Peter might well feel that since God had provided the one, he was responsible to provide the other and would be oppposing God if he didn't. Third, the question in Jersualem is whether Peter had gone against God's commands in eating and drinking with Gentiles. Peter responds that his vision about eating unkosher food points to a change in God's law regarding Jew-Gentile relations, and that God's pouring out of the Holy Spirit on the Gentiles at Cornelius' house ratifies this. Rather than disobeying God in eating and drinking with Gentiles, Peter argues he would have been disobeying God had he refused to do so.

CONCLUSIONS

Precedents in a Single Author

Having now looked at these six sets of passages in which Luke narrates the same events on more than one occasion, what conclusions can we draw? I suggest that we here have a precedent in a single author for many phenomena occurring in parallel passages of the synoptic Gospels, features which liberals regularly take as evidence of error, ignorance, bias or suppression in the Gospel authors and from which they conclude that these accounts cannot be inerrant. Nothing of the sort is necessary.

It is, of course, possible to claim that the author of Luke-Acts has here slavishly followed several inconsistent written sources, so that these phenomena are evidence of error, ignorance, bias or suppression in his sources.[8] Yet we have no evidence that Luke used such sources, and his evident care in handling materials where we can check them by means of archaeology or other ancient writers indicates that he was a competent and careful man, familiar with the best historical methods of his period.

We suggest, then, that we have precedent in Acts (and Luke) of a single author producing parallel accounts with the following features also found in the synoptic Gospels.

1. Some parallel accounts are almost identical (Paul's conversion, Peter's vision). Others are so different as to be barely recongnizable as the same incidents (the ascension, Paul in Jersualem).

2. In some cases the parallel accounts are about the same length (Paul's conversion, Paul in Jerusalem, Peter's vision). In others, one parallel is highly compressed relative to the other (the ascension, Paul and Ananias, Peter at Cornelius' house).

3. Longer accounts usually have more details, but even the shorter accounts often have some detail or details missing in the longer parallel (the ascension, Peter's vision, Peter at Cornelius' house).

4. Apparent discrepancies regularly occur in such parallel passages (Paul's conversion, Paul and Ananias, Paul in Jerusalem, Peter at Cornelius' house). Sometimes these can be resolved by information from the contexts, sometimes not. We can, of course, guess at solutions (assuming either error or inerrrancy), but need to remember that any particular solution is a guess.

5. Occasionally one parallel account is compressed in such a way as to skip over a substantial time-period without any indication that it has done so (the ascension, Peter's speech at Cornelius' house).

6. There may be places where one account departs from chronological order for some reason (Paul's conversion: men stood speechless; Paul and Ananias: speech and healing). We often do not have enough information about what happened to be sure of this.

Since we have only these half-dozen examples of parallel passages with a single author in Acts, whereas there are over a hundred parallel passages in the Gospels, we should not expect to find all the phenomena in Gospel parallels matched here. Yet the assumption of a single, competent author for the book of Acts helps us to see that the harmonization of apparent discrepancies is not special pleading.

Although the synoptic Gospels do not have a single human author, they do have a single divine Author if the biblical teaching regarding inspiration is true. Thus, harmonization of Gospel parallels need be no special pleading either, given an appropriate humility about the certainty of our guesses. We are merely recognizing the presence of a single, competent Author behind the Gospel narratives.

Some Suggestions for Problem Passages

Having looked at some features of history and history writing that may be relevant to synoptic harmonization, and having considered what phenomena occur in parallel passages written by a single author, let us return to the problem passages mentioned at the beginning of this paper to see what can be said about them.

1. Jesus Tempted in the Wilderness. The brief narrative provided by Mark is an excellent example of a highly compressed account which yet has an additional detail, the reference to Jesus being with the wild animals in Mk 1:13.

Concerning the main problem, the order of the last two temptations in Matthew and Luke, it is notable that neither account uses connecting words which require them to be understood as giving a chronological order. Luke uses the vague kai and de, while Matthew has tote and palin. Either of the authors might have departed from chronological order, perhaps to give emphasis to one of the temptations by placing it last. On the other hand, given the length of Jesus' time in the wilderness, the experiential fact that temptation is often repeated, and the possibility that these temptations involved visions rather than actual travel, it may be that these particular temptations were repeated, and that both accounts are compressed but in chronological order!

2. The Centurion's Servant Healed. This apparent discrepancy has no close parallel in our Acts passages. It is not likely to be solved by having both the centurion and his friends come to Jesus, as some have suggested. More likely, the Matthew account is a compressed version of what happened in which the indirect nature of the centurion's appeal was omitted to conserve space. According to Jewish law, a man's messenger was legally the same as himself. Though this idiom seems strange to us, it is really not so different from our speaking of hearing from someone when in fact he or she only wrote or phoned.

3. The Demons and Pigs. Regarding the number of demoniacs, two is not equal to one. If Mark and Luke are understood as forbidding two demoniacs, then there is either an error here or two different incidents are narrated. Yet, as we have seen in Acts, a parallel narrative often contains both compression and details not found in the other account. The great similarity of the exorcisms here suggests that we have the same incident, so perhaps one demoniac was relatively insignificant. Maybe the other was the spokesman or more seriously demonized. In either case, Mark and Luke might well omit one demoniac in compressing their narrative.

4. Healing of Blind near Jericho. The apparent discrepancy in number of blind men may be handled as were the demoniacs above, or as two separate incidents, though again the events seem very similar and occur at about the same time and place in each Gospel.

The apparent discrepancy in chronology seems more serious. Did the event occur as Jesus approached Jericho or as he departed? Besides the possibility that two different events are in view, some have suggested that the request for healing was made as Jesus approached Jerusalem but he healed them as he left, or that Jesus was leaving Jerusalem but turned back to heal them. Both of these seem rather forced, though admittedly such an opinion is itself subjective and we weren't there to see what actually happened.

An apparently even less likely suggestion, that Jesus was leaving one Jericho and approaching another, deserves more consideration than one would at first be inclined to give it. At New Testament times, there were two Jerichos less than a mile apart, the old Jewish city surviving from after the exile, and a new Hellenistic city built up around the palaces of Herod and the Maccabees. Travel toward Jerusalem would have Jesus leaving the old Jericho as he approached the new. Matthew and Mark, with a more Jewish outlook, might well focus on the Jewish city, and mention that Jesus was leaving it, while Luke, with a more Hellenistic outlook, might emphasize the new Jericho, especially as he is about to narrate the incident with Zacchaeus which probably took place in the wealthier town where a prosperous tax collector would be expected to live.

Again, we do not know exactly how the parallel narratives are to be historically harmonized, but we can make several suggestions. Here also is a good example of how one historical fact may make a seemingly preposterous solution reasonable, which should warn us about the uncertain value of our subjective impressions of likelihood.

5. Peter's Three Denials. Here we have a very complex problem involving four accounts, each with somewhat different details.

Matthew has a woman accuse Peter as he sits in the courtyard (26:69-70); at the gateway, another woman accuses him (71-72); after a little while, a whole group says his accent gives him away (73-74); then the rooster crows (74).

In Mark, the first accusation is the same as Matthew's (14:66-68); then a rooster crows; at the gateway, the same woman accuses him again (69-70); then, a little later, a group recognizes him as a Galilean (70-72); then the rooster crows again (72).

Luke also has a woman accuse him as he sits in the light of the fire (22:56-57); later a man accuses him (58); about an hour later, another says he must have been with Jesus for he is a Galilean (59-60); then the rooster crows (60).

In John, the woman that keeps the door asks (18:17); then around the fire, the servants and officers ask, while Peter stands warming himself (25); finally, one of the high priest's servants claims to recognize Peter from the garden (26); then the rooster crows (27).

Harold Lindsell suggests a rather elaborate harmonization here, with eight denials in all, three before the first crowing, and five after. This interprets both versions of Jesus' prediction strictly: "before the rooster crows, you will deny me three times" and "before the rooster crows twice, you will deny me three times." But it does so at the cost of having more than three denials when that seems to be the emphasis of the prediction.

It seems more likely to me that there were three occasions of denial separated by some period of time, and that the second and third of these involved several different persons joining in on the accusations. All the Gospel writers have thus compressed a complex series of events in slightly different ways, perhaps using witnesses who remembered different features. I take Mark's version of the prediction with its Semitic parallelism (twice ... thrice) to be the original, which has been simplified by the other writers since the first crowing was only a warning and apparently occurred much earlier than roosters normally crow.

Inerrancy and Induction

Having suggested some possible harmonizations for the five problem pasages with which we began, let us close with a few observations about relations between history, harmonization and inerrancy.

Once the eyewitnesses to a given series of events have passed away, only the information they have recorded or transmitted survives to later generations. The eyewitnesses may have been honest or liars, careful or sloppy, but once they are dead they cannot be cross-examined again until the last judgment.

We as historians, living centuries later, can work with the records they left, but we must not forget our limitations. We can trust their records as somewhat, substantially or entirely reliable, or we can dismiss them as unreliable. We can seek to harmonize the records if we think them reliable, or we can set them against one another if we distrust them, even setting an author against himself if we think him totally incompetent. Yet our procedure will be no better than the accuracy of our assessment of the witnesses.

We have suggested from our study of parallel passages in Luke-Acts that the same sort of phenomena which occur in Gospel parallels also occur here. Given that the author of Acts is competent and reliable, yet in parallel passages he compresses narratives and omits materials, sometimes prooducing very dissimilar accounts or even apparent discrepancies, then we should not be so quick to allege error or bias when we see similar things in the Gospels.

If we agree that the Bible teaches its own inerrancy, we should not be setting one author against another. Yet given that the Bible writers compress and omit materials which they know about, we should not be too confident that our particular reconstruction is what really happened when a variety of reconstructions are possible. Whenever we go beyond what the authors actually say, we are on much less solid ground than when we have their explicit statements.

By the nature of history and the inductive method we use to investigate history, it is not possible either to prove or disprove the inerrancy of Scripture just by looking at the data of Scripture rather than its explicit statements. Though in principle a single counter-example would disprove inerrancy, in practice the flexibility of language, the use of compression in narrative, and the complexity of history conspire to defeat such hopes. So, too, the number of apparent discrepancies make it impossible to convince opponents of inerrancy merely by suggesting possible harmonizations.

In fact, arguments against inerrancy from the data of Scripture such as are in vogue among liberalizing evangelicals are really quite similar to arguments by unbelievers against the goodness of God or the sinlessness of Christ. In all three cases, problem passages in Scripture are brought forward to explain away the teaching of Scripture on these matters. But all three matters are revealed doctrines derived from the express teaching of Scripture, and thus claim to come from God who knows all data.

Of course, just as our doctrine of God's goodness should be consistent with the proper interpretation of all passges on that subject (including, but not limited to the problem passages), so ought our doctrine of inerrancy to be consistent with the proper interpretation of all its pasages. Therefore, on the basis of this study, we suggest the following proposals regarding biblical inerrancy.

Biblical inerrancy does not mean: (1) a passage must give all the details we would like to have; (2) there can be no compression in the narrative; (3) all order must be chronological; (4) there must be no apparent discrepancies; (5) we must be able to explain what actually happened; (6) the simplest interpretation of a given passage is necessarily correct.

Biblical inerrancy does mean: (1) the details given are correct; (2) order must be chronological if the text so indicates; (3) there may be no real discrepancies; (4) there must be an interpretation of the passage or passages consistent with the grammatical usages and idioms of the original language which fits what actually happened.

REFERENCES

1. See, e.g., Paul J. Achtemeier, The Inspiration of Scripture: Problems and Proposals (Philadelphia: Westminster, 1980), chap III; James Barr, Beyond Fundamentalism (Philadelphia: Westminster, 1984), pp 74-75, 78-81; Dewey M. Beegle, Scripture, Tradition, and Infallibility (Grand Rapids: Eerdmans, 1973), pp 60, 192-93; Harry R. Boer, Above the Battle? The Bible and Its Critics (Grand Rapids: Eerdmans, 1977), chaps 5-6.

2. Boer, Above the Battle, pp 82-85; Beegle, Scripture, pp 195-97.

3. See nearly any recent commentary or study on one of the synoptic Gospels; e.g. F. W. Beare, The Gospel According to Matthew (Oxford: Blackwell, 1981), pp 7-13; or even Robert H. Gundry, Matthew: A Commentary on His Literary and Theological Art (Grand Rapids: Eerdmans, 1982), pp 5-10; Etienne Trocme, The Formation of the Gospel According to Mark (Philadelphia: Westminster, 1975); Charles H. Talbert, Luke and the Gnostics: An Examination of the Lukan Purpose (Nashville: Abingdon, 1966).

4. H. J. Cadbury, "Acts of the Apostles," Interpreter's Dictionary of the Bible (1962) 1:38. For evidence identifying this author with Paul's associate Luke, see Donald Guthrie, New Testament Introduction, 3rd ed (Downers Grove, IL: InterVarsity, 1971), pp 98-109.

5. Charles H. Talbert, Reading in Luke (New York: Crossroad, 1982), pp 226-33.

6. Walter Bauer, William F. Arndt and F. Wilbur Gingrich, A Greek-English Lexicon of the New Testament and Other Early Christian Literature (Chicago: University of Chicago Press, 1957), p 383. 7. Ibid., pp 31-32, 878. 8. See Guthrie, N.T. Introduction, pp 368-77, for a discussion of various source theories.