The Devil, Demons, Judas, and “the Jews”: Opponents of Christ in the Gospels

By Torsten Löfstedt

What place do the four Gospels give Satan, demons, and Jesus' human opponents (including Jewish leaders but also Jesus' disciples) in their accounts of Jesus' life? This study takes a literary-historical approach to the Gospels, examining them as narratives. It shows how the authors were in the process of developing the devil as a character and determining which roles he filled. New interpretations of individual passages in the Gospels are given as well as new understandings of the theological emphases of each author. This study is also a contribution to redaction criticism and the relative chronology of the Gospels. It employs the theory of Matthean posteriority which revolutionizes our understanding of the literary relations between the Gospels and allows for a new understanding of theological development in early Christianity.

• Language: English
• Publisher: Pickwick Publications
• Release date: Nov 9, 2021
• ISBN: 9781666704563

Torsten Löfstedt

Torsten Löfstedt is associate professor in religious studies at Linnaeus University in Växjö, Sweden.

Book preview

Introduction

This study compares the four Gospels to see what place they give Satan, demons, and Jesus’ human opponents in their accounts of Jesus’ life. The four Gospels were written with largely the same function in mind, to provide a theologically informed account of Jesus’ ministry and death. When the Gospels were written, Christian doctrine was not thoroughly systematized. It was up to the individual authors to come up with as coherent a narrative and as cohesive explanations as they could—or wished to—make. They pulled upon the various traditions to which they had access, and might not have been overly concerned if not all parts fit perfectly. This study follows several strands of tradition to show how the characterization of Jesus’ opponents changes over time. It also examines how the Gospels portray Jesus’ disciples (especially Peter and Judas) in their capacity as hindrances or opponents to Jesus. In the process, I show how these early Christian authors were in the process of developing the devil as a character and determining which roles he filled.

The four Gospels present four different stories, and I will examine each Gospel individually. It is apparent that at least three of the Gospels (Matthew, Mark, and Luke) stand in a close literary relationship to each other. In addition to examining the individual Gospels on their own terms, I will try to account for the relationship between them, especially as it applies to how they describe the opponents of Christ, and in so doing show how the narrative about Jesus and his opponents developed over time.

A Literary-Historical Reading of the Gospels

My study is based on close readings of the Gospels, where I treat them as literary works. I will analyze the Gospels one by one rather than harmonizing them. My goal is to approximate the historical authors’ intended meanings by doing close readings of the texts in their literary and historical contexts.¹ I base my interpretation of New Testament passages on the Nestle-Aland reconstruction of the Greek text, the twenty-eighth edition.² When quoting longer passages from the New and Old Testament in English I use reasonably literal translations such as the NRSV, the NASB, and the NIV as specified after each quote.³ Translations of individual words and short phrases, unless otherwise specified, are my own. My translations are unavoidably influenced by the English Bible translation tradition with its roots in the King James Version.

I will do literary-historical readings of the Gospels. The approach I take is often referred to as the historical-critical approach, but that term also has other meanings that do not apply here. I am for instance not interested in reconstructing the historical Jesus nor do I use these texts as source material to learn about the Pharisees in the first century. My aim is to see how the authors characterize Jesus, his disciples, and his opponents and to examine the roles they play in the narrative. I prefer the term literary-historical approach as it puts emphasis on studying the Gospels as literary works of individual authors.⁴ The various parts of each Gospel are understood first and foremost in light of the Gospel as a whole. Where the authors have written other texts besides the Gospel, I look at these to help me better understand their theology and their use of certain key terms. This is the case of Luke, who also wrote Acts, and John, who I hold wrote 1–3 John as well.⁵ In the case of two of the Gospels, Luke and Matthew, we have the advantage of being able to compare how the authors have reworked their source material. With the help of redaction criticism we can identify their agendas more accurately. (I will use redaction criticism to other ends as well, as will be discussed below.)

To understand the intentions of the historical authors it is helpful to know as much as possible about them and their social situation. Unfortunately we know very little about who wrote the Gospels and it is difficult to firmly date the Gospel texts and to establish the historical context in which they were written. I will assume that all four Gospels were written in the late first century by followers of Christ who were colored by Hellenistic Judaism.⁶ It is hard to be more precise than that without becoming exceedingly speculative. We cannot say with any certainty where the Gospels were written (although it appears quite unlikely that Luke was written in Palestine) nor can we say much about the social circumstances of their writing.⁷ This limits my ability to interpret the Gospels in their historical context; my interpretation will therefore have a very strong textual focus. Questions regarding authorship, dating, and the religious background of the authors are discussed in the introduction to the chapters dedicated to each of the four Gospels.

With close readings there is a risk that the interpreter reads more into the text than was intended by the author. I believe that in the case of the Gospels close readings and repeat readings accord with the authors’ intentions. The Gospels are all anonymous texts, in the sense that the authors do not announce their names in the text, unlike what is the case in the Pauline epistles for example.⁸ One explanation for their anonymity is that they were commissioned by a congregation or group of congregations to be used in worship in ways similar to how the Hebrew Scriptures were used in the synagogues and early church.⁹ Judging by the structure and contents of the Gospels we may conclude that the authors set out to write Scripture.¹⁰ They intended for their texts to be read from the beginning to the end, not necessarily in one sitting, and not just once, but many times. Close readings of the texts are therefore warranted. There is reason to believe that the authors have chosen their words with care. This is especially the case with John, where there are numerous examples of foreshadowing and dramatic irony that only become apparent in repeat readings, and with Matthew, who reorganizes his source material thematically into discourses.

Literary characters

Compared to other aspects of New Testament theology, the character of the devil has been neglected. There are various reasons for this. One is that the devil is referred to rather seldom and there is little if any explicit teaching about him, which raises the question whether the devil is more a rhetorical device than a theological reality for the authors.¹¹ Another reason for the lack of research is related to theologians’ aversion to the subject matter itself. Mainstream theologians and exegetes may consider references to the devil and evil spirits as vestiges of superstition that colored early Christian authors but that do not concern Christians today.¹² Whether we believe in the existence of the devil or not, early Christian authors had reason to refer to him, and it is our task as exegetes to try to understand what it is they were trying to say.

I will study Satan primarily as a literary character rather than as a historical or theological figure (it is of course not always possible to cleanly distinguish between literary and theological analysis, however, as the Gospels use narrative to express their theology). I follow the approach taken by Neil Forsyth, who writes regarding Satan, the focus of his elegantly written book, The Old Enemy:

Satan is first, and in some sense always remains, a character in a narrative. For Satan is a character about whom one is always tempted to tell stories, and one may best understand him not by examining his character or the beliefs about his nature according to some elaborate and rootless metaphysical system, but rather putting him back into history, into the narrative contexts in which he begins and which he never really leaves. That is, we must see him as an actor, or what Aristotle called an agent, with a role to play in a plot, or mythos.¹³

This approach is helpful in that it allows the scholar to study the character of Satan in the Gospels without having to concern himself with what the different evangelists really believed, which is something we cannot reconstruct with any certainty. The Gospels do not say exactly who the devil is, but they do tell about what he does. His functions define him. Forsyth writes incisively regarding "the Adversary":

His character, indeed his very existence, is a function of his opposition to God, or to man, or to God’s son, the god-man. But he may appear as tempter, tyrant, liar, or rebel, each time taking on the characteristics appropriate to the role. If he appears as the opponent of God, he is (eventually) the rebel, and if God is good, as is often but not invariably the case, then he is evil: if he appears as the opponent of man, then he is the tempter, or the tyrant.¹⁴

When I first embarked on this research project, my intent was to examine the various roles that the devil has in each Gospel, to see how the different authors use this character in their narratives.¹⁵ However, when we read the Gospels we note that some things that the devil is said to do, other characters are also described as doing. The devil tests Jesus, but so too do Peter and the Pharisees, for example. It appeared that there was a connection between what an author said about the devil and what he said about Jesus’ disciples and his opponents. I therefore decided to widen my study to include other characters who opposed Christ. I take the same approach to these figures as I did with the devil; I see them as characters in a story. I am not concerned with knowing what Peter, Judas, the Pharisees or the demons were like in real life.

Conflict in the Gospels

In trying to understand how the evangelists portray the opponents of Christ, I have been guided by the narrative approach taken to the Gospels that was pioneered in the 1980s and early 1990s by Culpepper; Rhoads, Dewey & Michie; Tannehill; Kingsbury; and Powell.¹⁶ Scholars disagree as to whether the canonical Gospels are to be classified as a biography or history or whether they belong to a generic category all their own,¹⁷ but they agree that they are narratives, unlike what is the case with the Gospel of Thomas, which is a collection of sayings without a narrative framework. The evangelists are authors with a story to tell. After some form of prologue, all four Gospels present Jesus’ healing ministry and teaching, and then they describe the events leading up to the crucifixion and the crucifixion itself. This is followed by an epilogue of some kind. Although they differ in details and in emphasis, the story is recognizably the same. The plots differ, however, as the authors have different views of what lay behind the various events.¹⁸

The term plot has been variously defined by students of literature. It may be distinguished from the simple chronological sequence of events in that it also identifies the underlying causes behind those events. E. M. Forster writes in his classic work Aspects of the Novel,

We have defined a story as a narrative of events arranged in their time-sequence. A plot is also a narrative of events, the emphasis falling on causality. The king died, and then the queen died is a story. The king died, and then the queen died of grief is a plot. The time-sequence is preserved, but the sense of causality overshadows it.¹⁹

In defining a narrative’s plot, one must identify causal connections between key events in the narrative. It is not always obvious which events are key events. Identifying the plot of the Gospels is made more complicated by the fact that they include much material that does not relate directly to the plot, such as many examples of Jesus’ teaching. Keefer goes so far as to claim, none of the gospels is primarily plot-driven.²⁰ I do not agree with this view.

The plots of the Gospels are most readily identified by reading the text backwards, from the ending. As Powell explains, the significance of events can only be determined in light of their results.²¹ Not all events described in the Gospels move the plot forward. It is only after we know how the story ends that we can see which of the many events mentioned contributed to that ending. The ending of the Gospels is therefore especially important when we seek to define the plot. As will be discussed in the following chapter, in the case of Mark’s Gospel, it is not certain that the original ending has been preserved, making the identification of the core plot more difficult.

In all four Gospels, conflicts are one of the things that drive the plot forward. When the Gospels speak of evil, they often do so in terms of a battle. On one side are God, Jesus, and God’s people; and on the other are the devil, demons, various personified forces (such as sin and the world), and individual humans. This use of conflict as a structuring device is something that the Gospels have in common with other narratives. Rhoads, Dewey, and Michie write, Conflict is at the heart of most stories. Without conflict, stories would be only a sequence of events strung together without tension or suspense or struggle on the part of the characters.²² Jesus is involved in conflicts in all four Gospels, but can one conflict be said to underlie the others? If so, is the primary conflict in a given Gospel one between Jesus and Israel, or between Jesus and the Jewish leadership, or between Jesus and the devil, or is Jesus’ main opponent something else? As we shall see, on this point, as on so many others, scholars disagree.

The combat myth

Some scholars hold that Christ’s main opponent in the Gospels is the devil.²³ For example, Jeffrey Burton Russell, who wrote a four-volume study of the history of Satan as a literary figure, writes,

The central message of the New Testament is salvation: Christ saves us. What he saves us from is the power of the Devil. If the power of the Devil is dismissed, then Christ’s saving message becomes meaningless.²⁴

Curiously, Russell dedicated only one chapter in his four-volume work to The Devil in the New Testament.²⁵ There he paints a broad picture and does not explain how the Gospels differ in their portrayal of the devil. Russell’s claim about the centrality of the devil is thought provoking, however. If the devil is so central to the Gospels, why do they say so little about him?

Forsyth continues along similar lines of inquiry as Russell. In The Old Enemy, he studies the development of the so-called combat myth. He offers the following summary of the Christian version of the plot:

A rebel god challenges the power of Yahweh, takes over the whole earth as an extension of his empire, and rules it through the power of sin and death. He is the typical death-dealing villain who causes consternation among his subjects, and his depredations and cruelty make them long for a liberator. This dark tyrant, the god of this world as Paul called him, is eventually thwarted by the son of God (or man) in the most mysterious episode of the Christian story, the crucifixion, which oddly combines both defeat and victory.²⁶

Forsyth uses Proppian narrative analysis to trace the development of this combat myth from ancient Near Eastern mythology, to the Old Testament, intertestamental literature, to the New Testament, especially as it comes to fullest, but still fragmentary, expression in Revelation.²⁷ In Forsyth’s view, versions of the combat myth are implicit in the New Testament, as an assumed rather than a revealed truth.²⁸ In other words, the early Christian authors built on existing versions of the combat myth.

The idea that a pre-Christian combat myth would be central to the Gospels may seem implausible, considering that there are very few references to Satan or equivalent characters in the Old Testament.²⁹ Satan is referred to about twenty times in the Old Testament. It is hard to be more precise as it is not always clear whether śāṭān refers to a supernatural character in a given passage (e.g., Ps 109:6);³⁰ in some passages it clearly refers to a human opponent (e.g., 1 Kgs 11:14). It is likewise debated whether (and if so where) śāṭān is used as a name in the Old Testament passages; some scholars argue that śāṭān designates a role in all the passages where it appears in the Old Testament.³¹ (It is further debated what this role is. While the traditional interpretation holds that śāṭān designates an accuser, Stokes argues that in the Old Testament śāṭān designates an attacker or executioner.³²) For my study the following passages referring to śāṭān are especially important: 1 Chr 21:1; Job 1:6–12, 2:1–7; Zech 3:1–2. Many translations, including KJV, NASB, NIV, and NRSV, render the Hebrew word śāṭān, with or without a definite article,³³ as Satan in these passages.³⁴

The prologue to Job (chapters 1–2) is key to understanding the character of Satan in the Gospels. Like Zechariah (3:1–2), Job (1–2) refer to (the) Satan as a member of the divine council whose role it is to find fault with, accuse, and punish people. Job’s book shows tells about the righteous Job, who loses his health, wealth, and family. In a manner quite unlike other Old Testament texts, the author explains how Satan lay behind these events; he had sought and received permission from God to test Job’s dedication to God. Job himself is not aware of Satan’s role in the matter. This text opens up for the reader to interpret other examples of unjustified suffering (whether in scripture or in real life) and other forms of temptation as being Satan’s work. When the Gospels refer to the one who tempts people as Satan they build on a tradition that goes back to Job.

While there are very few references to supernatural opponents in texts that are included in the Old Testament canon today, Satan and similar characters that oppose God and his people (such as Azazel, Asael, Mastema, Shemihazah) have a larger place in other Jewish texts, such as 1 Enoch and Jubilees, which I will discuss in greater detail shortly. As we will see in the chapters that follow, it is likely that some Gospel texts allude to these books. Sectarian texts found among the Dead Sea Scrolls also refer to supernatural adversaries of God’s people. While these texts are not cited by early Christian authors, they show that dualistic ideas were widespread in the first century.³⁵

As was the case in Old Testament literature, the Gospels sometimes speak of superhuman agents operating through human characters. As we shall see, the different Gospels do not always agree on whether a character acted on his own accord in a certain event or under the influence of another, unseen, character. In some passages, the author probably intends for the reader to understand that there are other agents working behind the scenes, even though he does not make this explicit. The challenge is to determine in which passages an author wishes to imply that God or the devil is pulling the strings behind the scenes, even though they are not explicitly mentioned there. In other words, to what extent can we read the combat myth into the Gospel stories? Following Forsyth, I find it very likely that early Christian authors formed their theology with the help of a combat myth, but the exact contours of that story remain murky and it is far from clear whether they all had the same understanding of that story. It is tempting to see the combat myth as a basic structure that underlies all the conflicts in the Gospels, but to keep my study from becoming unduly speculative, I seek to base my analysis on written evidence alone. Here intertextual clues found in the text help to guide the reader. The Gospels often use quotes and allusions to other texts to hint at how the events described are to be interpreted.

The use of the Hebrew Scriptures

One of the most important sources for the Gospels in describing and interpreting aspects of Jesus’ life, ministry, death, and resurrection, are the Hebrew scriptures. All New Testament authors assume the authority of Scripture, although they did not necessarily all define scripture the same way. They support their arguments with quotes and allusions to books that are included in what Christians today call the Old Testament. The early Christian writers did not know the Hebrew Scriptures as a single volume, but as a collection of books. Certain books seem to have been valued more highly than others; Psalms and Isaiah are quoted very frequently, for example. It has been suggested that New Testament authors did not have access to the complete texts of the Old Testament canon, but only to collections of key quotes or testimonies as certain passages in the books are quoted quite frequently and the rest are never quoted.³⁶ While it is evident that early Christian authors had their favorite Bible quotes, I will assume that the authors of the Gospels wrote on behalf of large congregations and had access to the complete texts from which they quote and to which they allude, and I will appeal to the larger textual context in interpreting these quotes. In the case of books like Isaiah, the larger textual context may even be the whole book.³⁷

The list of texts that the various New Testament authors considered to be authoritative did not correspond exactly to the books included in today’s Protestant Old Testament. How closely delineated the Jewish canon was at this period of time is a matter of debate. It is likely that there were several competing canonical lists.³⁸ In some early Christian authors we find likely allusions to works such as Wisdom of Solomon³⁹ and Sirach⁴⁰ that Protestants today count among the Old Testament apocrypha.

When referring to the Hebrew Scriptures, most New Testament authors cite the Greek translation of the Hebrew Scriptures, which scholars refer to as the Septuagint (abbreviated LXX), although on occasion some appear to provide their own translations of the Hebrew text.⁴¹ The Septuagint sometimes differs significantly from the Hebrew text that lies at the basis of most modern translations. Where relevant I will note these differences. When quoting the Septuagint I use Rahlfs’s edition.⁴² For English translations of the Septuagint, I generally use A New English Translation of the Septuagint.⁴³

1 Enoch

As was mentioned, some early Christian authors considered Enoch inspired scripture. Jude (v. 14) quotes 1 En. 1:9. He claims that these words were uttered by Enoch, in the seventh generation from Adam (NRSV). Similarly, Barnabas (a Christian composition from the first part of the second century) quotes 1 En. 89:56 as scripture (Barn. 16:5). Although 1 Enoch is not included in the Jewish Bible today, texts that are included in 1 Enoch were highly regarded by the Essenes at Qumran as evidenced by fragments that were found in the caves by the Dead Sea. 1 Enoch is included in the canon of the Ethiopian Orthodox Church together with the book of Jubilees, which reflects the high status this book had in the early church.

While it is reasonable to assume that some early Christian writers had access to texts found in 1 Enoch, the question is whether they actually had access to all the texts that constitute 1 Enoch today. 1 Enoch is a combination of several texts that have their own history and transmission. The earliest parts of 1 Enoch are thought to have included material that is included in the Book of the Luminaries (1 En. 72–82), which in Nickelsburg’s view has its roots in the Persian period.⁴⁴ The Book of Luminaries seems to have originally circulated in Aramaic separately from other parts of what are now 1 Enoch.⁴⁵ The Book of the Watchers (chapters 1–36) is important to our study. Several church fathers also refer to this part of the Enochic text.⁴⁶ It was probably completed by the middle of the third century BCE.⁴⁷ The idea that early Christian writers may have referred to the Book of Watchers and the Book of Luminaries is not controversial. The Book of Parables (chapters 37–71) is that part of 1 Enoch which is considered the youngest.⁴⁸ This section is especially important for understanding the development of the Satan figure in the New Testament, but its dating has been subject of considerable debate. Although it was probably composed in Aramaic, no Hebrew or Aramaic fragment of this part of the Enochic text has been found; it is only preserved in the ancient Ethiopian language Ge’ez.⁴⁹ Most scholars today are convinced it was composed prior to the fall of Jerusalem.⁵⁰ In his scholarly commentary on the text, Nickelsburg concludes that it was probably composed in Palestine some time "between the late decades BCE and the early decades CE.⁵¹ It reflects ideas that were certainly prevalent during Jesus’ earthly ministry, but cannot be considered a Christian composition. I am open to the possibility that the Gospels allude to this section of 1 Enoch as well. Other parts of 1 Enoch include the Dream Visions, the Epistle of Enoch, the Birth of Noah, and Another Book of Enoch (1 En. 108).⁵² I have not had reason to refer to these parts of the Enochic text in my study.

1 Enoch shows an interest in angels and other spiritual beings and may have been especially important for how early Christian authors understood powers of evil. I will note possible allusions to this book in chapters that follow. These connections will be tentative because we do not know how widely used 1 Enoch was and because we do not have access to the text in the same language as the early Christian writers. Early Christian writers probably read Enoch in Greek translation or perhaps in Aramaic. Although parts of the text have been preserved in the Aramaic and Greek, the full text of 1 Enoch (conventionally divided into 108 chapters) is preserved only in Ge’ez, and it is that text I will refer to. When I cite 1 Enoch I use Nickelsburg’s translation.

Another text in the Enochic tradition that has material about the devil is 2 Enoch. The book is only preserved in medieval manuscripts written in Church Slavonic. The oldest manuscript is from the fourteenth century.⁵³ The date of its original composition is very uncertain. Since it is doubtful that the book antedates the Gospels, I will not treat it here.

Jubilees

Jubilees retells stories also included in Genesis and Exodus and builds on the Enochic literature. Among other things it tells about the origin and classification of angels, which is of interest to our study. The text was originally composed in Hebrew, but is best preserved in Ge’ez. Some fragments in Hebrew have been found at Qumran, and quotes from Greek translation have also been identified. It was probably composed in the second century BCE.⁵⁴ Like 1 Enoch, Jubilees is also included in the canon of the Ethiopian Orthodox church and was highly regarded by church fathers. It is not unreasonable to imagine that the Gospels may allude to Jubilees. I use VanderKam’s translation of Jubilees.⁵⁵

Other texts

Other texts that are occasionally referred to in order to help understand the characterization of the devil in the Gospels are Martyrdom and Ascension of Isaiah, Testaments of the Twelve Patriarchs, Testament of Job, and The Life of Adam and Eve (also referred to as the Apocalypse of Moses⁵⁶).⁵⁷ All four of these books show Christian influence, and it is doubtful whether they antedate any of the Gospels.⁵⁸ I will not use them to interpret Gospel material. 4 Ezra was a Jewish composition written after the destruction of the temple in 70 CE.⁵⁹ I have no reason to believe the authors of the Gospels had access to it. It is of interest, however, as it gives insight into the worldview of some Jews who were contemporaries with the authors of the Gospels but were not followers of Christ.

Development of traditions

The four Gospels give different accounts of Jesus’ opponents. One of the goals of this study is to show how the characterization of Jesus’ enemies changes over time, and from one Gospel to the next. The literary relationship between the four Gospels is therefore of central concern.

The authors of the Gospels develop their theology, including their understanding of Jesus’ mission and the opposition that he faced, on the basis of earlier theological traditions. It is likely that Jesus’ own teaching was the single most important source for the evangelists and other early Christian authors. But as Jesus did not write anything himself our understanding of what he taught is based on what others say he said. I will proceed from the written sources that we do have rather than from reconstructions of what Jesus might have said. I have briefly referred to some pre-Christian traditions that the authors of the Gospels had access to. But which other texts did they have access to? Curiously, there is little evidence that any of the evangelists were aware of other books that are included in the New Testament, aside from earlier Gospels or works that they themselves had written.⁶⁰ While we might have expected Luke to refer to Paul’s letters in writing Acts, for example, he does not do so. Nor do we find clear textual connections between 1 Peter and Mark. It is clear that there is some kind of literary interdependence among the Synoptic Gospels,⁶¹ and as I will discuss it seems likely that John is aware of at least two if not all three of the Synoptics.

The most important textual source for later Gospels are earlier Gospels. I will use redaction criticism to show how some of the Gospels modify material from earlier texts. This gives a better understanding of the authors’ intentions and allows us to see how the traditions about Jesus were modified over time. I am not the first to compare how the Gospels portray Jesus’ opponents. Hultgren took a form critical approach to conflict stories in the Synoptic Gospels. He focuses on accounts of conflicts between Jesus and his human opponents and one of his goals is to discover the form and substance of traditions prior to their being written in their present form in the gospels.⁶² I will not be following his approach, which I consider overly speculative. My approach is closer to that of Pagels, who examined the rhetorical purposes behind the use of the term Satan and its equivalents in various parts of the Bible and in the early church father Ignatius.⁶³ In her view, leading figures both in pre-Christian Judaism and in the Jesus movement demonized their opponents by saying they were servants of the devil. As Pagels presents it, New Testament authors use Satan to represent other Jews. The New Testament gospels almost never identify Satan with the Romans, but they consistently associate him with Jesus’ Jewish enemies, primarily Judas Iscariot and the chief priests and scribes.⁶⁴ Pagels finds an increasing demonization of Christ’s opponents, especially the Jews. Pagels writes well and her interpretations are frequently cited, but her work is beset with serious problems. For example, she neglects those passages that speak of the Enemy being within the individual follower of Christ; Jesus’ addressing Peter as Satan is only treated in passing, for example.⁶⁵ Still, I found her ambition to show how the portrayal of Jesus’ enemies changes over time very helpful. I follow her lead in trying to study the development of theology by examining four Gospels in the order in which I think they were written. Since my initial assumptions, including my understanding of the relative chronology of the Gospels, differ from hers, my conclusions also differ significantly from hers.

In order to carry out this project, the relative chronology of the Gospel texts must be established. Scholars generally agree that Mark wrote his Gospel first and that both Matthew and Luke use Mark. Scholarly consensus holds that Matthew and Luke wrote their gospels independently of each other but they both had access to a second source, which has not been preserved. Their dependence on this source, commonly referred to as Q, would account for the large number of passages that Matthew and Luke have in common but are lacking in Mark. When I first began writing this book, I proceeded from this two-source hypothesis. But the more I studied the Gospels and the temptation narrative in particular, the less plausible the two-source hypothesis seemed. I therefore sought an alternative explanation. The one that seemed to account for the material best was the little-known theory of Matthean posteriority, which argues that when he expanded on Mark’s account, Matthew used Luke as one of his sources.⁶⁶ I find this theory promising. I will briefly review some of the more important attempts at solving the Synoptic problem and explain the theory of Matthean posteriority in greater detail. Then in chapters 2 and 3, which focus on Luke and Matthew respectively, I will put the theory of Matthean posteriority to the test.

The Synoptic problem

One of the goals of this study is to better account for the literary relationship between the Synoptic Gospels. The Synoptic problem has generated a vast amount of literature. Here I will briefly explain the two most common explanations scholars give for the considerable similarities and striking differences between the Synoptic Gospels and then explain the theory of Matthean posteriority that I will apply. Most of my readers are familiar with the traditional solutions to this problem so I will keep this discussion brief.

Most scholars are convinced that the similarities between the Gospels of Matthew, Mark, and Luke are such that they must have be some form of literary relation between them. The idea that they would simply all be based on the same oral tradition seems implausible.⁶⁷ Most scholars are convinced that Mark’s Gospel is earliest of the four Gospels. It is the shortest, lacking both a birth narrative and an account of the resurrection, at least in the form we have today. Compared with Matthew and Luke, the language is simple. While some scholars continue to argue that Matthew’s Gospel was written first and that Mark and Luke are directly or indirectly dependent on it,⁶⁸ that standpoint is very difficult to defend. It is difficult to come up with a convincing explanation for why anyone would want to undo Matthew’s coherent structure. Others have argued that Luke’s Gospel is the oldest, but that too is quite implausible.⁶⁹ I will assume Markan priority; it is sufficiently well established that I need not argue for it.⁷⁰

While Markan priority is not controversial, the relationship between Matthew and Luke remains unresolved. Luke shares a good deal of material with Matthew that is not found in Mark.⁷¹ The question is how to account for this overlap. Has Luke borrowed it from Matthew, has Matthew borrowed from Luke, or are they both dependent on a third source that has not been preserved?

The two-source hypothesis

To explain the considerable verbal similarities between the passages shared by Matthew and Luke but lacking in Mark, for more than a century New Testament scholars have assumed that Matthew and Luke were dependent on another written source in addition to Mark (hence this theory is often referred to as the Two-Document or Two-Source hypothesis). Had the shared non-Markan material been orally transmitted, the wording and relative order of the pericopes in question would not have been as similar as they are. (On the relative order of the passages, compare the passages in Luke’s Sermon on the Plain (Luke 6:17–49) with their counterparts in Matthew’s Sermon on the Mount (Matt 5:1—7:29).)⁷² Inspired in part by a cryptic reference attributed to Papias concerning a collection of sayings by Matthew, C. H. Weisse postulated in 1838 that Matthew and Luke both had access to a collection of sayings with which they supplemented Mark.⁷³ Encouraged by the publication in 1897 of Papyrus Oxyrhynchus 1, fragments of a collection of sayings attributed to Jesus that we now know to have been included in the Gospel of Thomas, scholars were convinced that their assumption that Matthew and Luke were dependent on a written text containing Jesus quotes was reasonable.⁷⁴ Attempts at more clearly establishing that written text which came to be referred to as Q have continued ever since.⁷⁵

The historical existence of Q is often treated as a historical fact,⁷⁶ and scholars have even tried to establish how Q developed over time and to describe the community that produced Q.⁷⁷ Although they disagree on details, Q scholars generally imagine that although it included some short narrative sequences (such as the story of the temptation in the wilderness), Q was basically a collection of sayings, similar in structure to the Gospel of Thomas, rather than a sustained narrative. Many also claim that Q lacked a passion narrative, which makes it quite unlike the canonical Gospels, but again similar to the Gospel of Thomas.⁷⁸ In their reconstructions of Q, scholars generally hold that Luke preserves the original form and order of the Q material better than Matthew.⁷⁹ Scholars differ as to which Gospel keeps the original setting of these passages better; according to Fitzmyer, Luke preserves the setting better in some cases, Matthew in others.⁸⁰

There are many difficulties with the attempts to reconstruct Q. The lack of generic cohesion in Q is problematic. The presence of the temptation narrative in the reconstructed text that otherwise mainly consists of sayings is especially difficult; I will discuss this below.⁸¹ To defend the unity of Q one must concede that it was not a pure sayings source.⁸² But then it becomes considerably more difficult to say what kind of a text it was.

The variation found between Matthew and Luke in how they word certain passages such as the Lord’s Prayer that were supposedly borrowed from Q is also problematic. Some scholars who defend Q argue that Luke and Matthew shared other sources, including oral traditions, in addition to Q.⁸³ But once one has started to trace some double tradition material to other sources than Q, the unity of Q is called into question. And once the unity of Q is called into question, the question arises whether it is meaningful to refer to Q at all.

Scholars are surprisingly dogmatic about the appearance of this hypothetical document, especially in regard to what it did not include. As was mentioned, some scholars are convinced that Q did not include a passion narrative.⁸⁴ But if Q is reconstructed on the basis of the material that is lacking in Mark but shared by Matthew and Luke, how would we know whether it included a passion narrative? That is, after all, one thing that Mark does include. Indeed, at least one passage that has been assigned to Q appears to allude to Christ’s coming crucifixion; see Luke 14:27: Whoever does not carry his own cross and follow me cannot be my disciple, with a parallel in Matt 10:38.⁸⁵ To claim that Q was written by followers of Christ that had no interest in the crucifixion is not plausible.

A factor that has often been used in arguments against the two-document hypothesis are the hundreds of so-called minor agreements of Matthew and Luke against Mark.⁸⁶ In some cases where all three Synoptic Gospels include the same pericope, Matthew and Luke are closer to one another than either is to Mark. Compare for instance Matt 26:67–68, Mark 14:65, and Luke 22:63–64, where all three Gospels tell of men beating and mocking Jesus. Both Matthew and Luke have the men beating Jesus ask him, Who is it that struck you?, a question which is lacking in Mark. Another example is the healing of the paralytic (Matt 9:1–8 // Mark 2:1–12 // Luke 5:17–26). While Mark speaks of the paralytic lying on a pallet (or perhaps a mattress; Greek: κράβαττος) according to Matthew and Luke he lay on a bed (κλίνη).⁸⁷ How are these similarities between Matthew and Luke to be explained? Have Matthew and Luke independently of each other modified Mark’s material in the same way? Have later copyists assimilated one Gospel to another? Are these instances where Q and Mark happened to include the same pericope and Matthew and Luke preferred the version recorded in Q to the one found in Mark? Or is it the case that Luke is dependent on Matthew or is Matthew perhaps dependent on Luke?

If there were only a few minor agreements between Matthew and Luke a good case could be made that they both happened to change the text in the same way independently of each other. For example, perhaps Matthew and Luke reasoned that a κράβαττος would not have been sturdy enough to allow someone to be lowered down on it so the paralytic must have been lying on a bed. Similarly, one could argue that Mark’s account of the soldiers mocking Jesus implies that they beat him; that explains why they blindfolded him and then told him to prophesy. Matthew and Luke make what was implicit explicit, and they could have done this independently of each other.⁸⁸ One could also argue that in some cases a scribe at an early stage assimilated one text to the other, as we know scribes continued to do. But there are many minor agreements between Matthew and Luke. There is no agreement regarding exactly how many there are, as scholars disagree as to how to define a minor agreement (if both Matthew and Luke lack a conjunction found in Mark, does that constitute a minor agreement?), but examples of minor agreements between Matthew and Luke may be found in virtually every passage that the three Gospels share.⁸⁹ This presents a significant problem for the two-document hypothesis, which insists that Matthew and Luke wrote independently of each other.⁹⁰

Although the majority of scholars still accept the two-source theory in one form or another, in recent years the notion of a shared Q text antecedent to both Matthew and Luke has been called into question. Outside of scholars’ reconstructions, Q has left no traces. No church father refers to a text that would match modern scholars’ understanding of Q. As was mentioned, it is no longer taken for granted that Q was a single text or that it was necessarily a written text. A few scholars have developed the very reasonable suggestion that in addition to having access to some of the same written material (like Mark’s Gospel and perhaps other texts), Matthew and Luke may have heard some of the same oral traditions.⁹¹ While oral tradition has always been important in the church, it is likely that most teaching in the earliest church was orally transmitted. Jesus taught his disciples not by writing books but by word of mouth, and his disciples were likewise primarily preachers and teachers rather than authors. Many significant differences in passages Matthew and Luke share can be explained if we assume they learned these accounts through hearing them at different occasions and possibly from different people rather than by reading the same text.⁹² Some traditions that they both record may have been part of a larger oral church tradition (such as the Lord’s Prayer) while in other cases they may have talked to some of the same people. It is of course quite likely that Jesus himself said the same thing more than once, using slightly different words; in such cases it would be meaningless to try to decide which evangelist better preserves the original wording.⁹³

It seems likely to me that certain similarities between Matthew and Luke are the result of dependency on a shared oral tradition, but in other cases lexical similarities are too striking for that explanation to work. I am thinking here of the wilderness narrative again. Either Matthew borrowed the text from Luke, or Luke borrowed it from Matthew, or both borrowed the text from a third written source. While it is possible that Matthew and Luke had access to another shared written source that included this account, perhaps a different source than the so-called Q text with its wisdom sayings, we should also consider the possibilities of Luke being dependent on Matthew or Matthew being dependent on Luke.

Did Luke borrow from Matthew’s Gospel?

The notion that Mark was based on Matthew or on Matthew together with Luke is extremely implausible. But what if we assume that Matthew was based on Mark while positing that Luke was based on both Mark and Matthew? That is what Austin Farrer argues.⁹⁴ Goulder, Goodacre, Gorman, McNicol, and others have further developed Farrer’s thesis.⁹⁵ The idea that Matthew reworked Mark is not problematic. But how do we account for the similarities and differences between Luke and Matthew using Farrer’s assumption? Goodacre argues that before he ever saw Matthew’s Gospel, Luke was well acquainted with Mark’s Gospel. He was inspired by Matthew’s Gospel to also expand on Mark, and he included in this new Gospel material he found in Matthew as well as material he knew from other sources, both written or oral. Mark’s Gospel was Luke’s primary source and Matthew was a source for supplementary material.⁹⁶ The advantage with this view over the two-source hypothesis is that it does not posit the existence of a document that has left no traces in the manuscripts. It also resolves the problem of the minor agreements: triple tradition passages where Luke is closer to Matthew than to Mark are passages that Luke has borrowed from Matthew.

There are difficulties with Farrer’s thesis that make it implausible. Scholars have argued against Lukan dependence on Matthew on the basis of how the two Gospels are structured. Matthew has collected material from various sources including Mark and organized them into five larger discourses. There are no traces of these discourses in Luke’s Gospel, however. If Luke is based on Matthew, he must have taken Matthew’s discourses, such as the Sermon on the Mount (Matt 5–7) or the eschatological discourse (24–25), and divided it up into scattered pericopes. Why would he do that? This is especially curious considering that Luke does not modify the Markan text blocks much. It seems counterintuitive that Luke would so thoroughly reject Matthew’s structure. Why would Luke chop up and redistribute the material found in Matthew’s discourses, as Farrer and his followers must claim?⁹⁷ Goodacre answers that for Luke, Mark was the more authoritative text and that is why he takes fewer freedoms with that text than with the Matthean text.⁹⁸ Goodacre and Gorman also argue that it is misleading to characterize Matthew’s structure as superior to that of Luke. Both Gospels are structured, but they are structured differently. From Jerome’s time forward readers have found Luke stylistically superior to the other Synoptic Gospels.⁹⁹ Gorman argues that one can observe some structural principles in Luke; he places most of Jesus’ deeds toward the beginning of the Gospel, and most of the teaching material comes later.¹⁰⁰ She also notes that when Luke reworks accounts from Mark, he tends to shorten them. He does not want to use more words than necessary. Brevity appears to be especially important in teaching passages for Luke. Goulder argues that Luke prefers teaching pericopes of about twelve to twenty verses, as that is how long a teaching passage can be for the congregation to take it in in one sitting.¹⁰¹ That would explain why Luke chopped up the Sermon on the Mount and scattered the material. While Gorman shows that Luke is well constructed, and while she makes a good case for why Luke would not feel compelled to keep Matthew’s discourses, she does not address the question whether Matthew might equally well have taken over material from Luke to construct those discourses.¹⁰² I agree that Luke’s Gospel is well suited for congregational reading while Matthew’s Gospel can easily become tedious to read, but other arguments may be made against Luke’s dependence on Matthew.

On the level of individual passages, one can ask why Luke would take stylistically well-formed texts and edit them into something less attractive. For example, why would he have significantly shortened the Lord’s Prayer?¹⁰³ Goodacre answers that Luke knew another version of the Lord’s Prayer that he considered to be more authentic.¹⁰⁴ In this case Luke has not actually modified Matthew’s material; rather he has replaced it with similar material from another source. The same question can be made regarding the Beatitudes, and the same responses could be given there as well. Depending on how we define them, Matthew (5:3–12) has eight or nine beatitudes while Luke (6:20–23) only has four; on the other hand in Luke (6:24-26) the Beatitudes are followed by four woes not found in Matthew. Whether we try to argue that Matthew was dependent on Luke or that Luke was dependent on Matthew, we face the same challenge. The challenge can be met in the same way. The author of the later Gospel was acquainted with a different version of the Beatitudes and preferred that version.

One may also ask why Luke did not preserve Matthew’s birth narrative. Goodacre argues that the very fact that Luke has a birth narrative, while Mark and John do not, suggests that he knew Matthew’s Gospel; the verbal similarities between Matt 1:21 and Luke 1:31 could be cited as supporting evidence. On the whole, verbal similarities in the birth narratives are very few, however.¹⁰⁵ As to why his birth narrative differs so strongly, Goodacre suggests, Luke thought he could improve on Matthew’s account.¹⁰⁶ This argument is not especially strong; the same argument mutatis mutandis has been given by those who advocate Matthean posteriority to explain why Matthew would have replaced Luke’s birth narrative.¹⁰⁷

When we compare passages that are found in all three Synoptic Gospels we can find several examples of how Matthew has modified Mark’s account. In these cases, Luke usually appears to base his account on Mark rather than on Matthew.¹⁰⁸ For example, according to Mark (10:11) Jesus rules, Whoever divorces his wife and marries another commits adultery against her. Matthew (19:9) modifies this ruling by adding an exception for divorce motivated by the partner’s adultery (whoever divorces his wife, except for unchastity, and marries another commits adultery NRSV). Luke (16:18) does not include Matthew’s exception clause. One might ask why; surely it would be of interest to the reader to hear about this exception to the rule. One might argue that Luke in this case favored Mark’s reading because it better fit his understanding of divorce.¹⁰⁹ While occasional difficulties can be explained in this manner, others are harder to account for. For example, Luke generally puts Peter in a more positive light than did Mark (I give examples of this tendency in chapter 2). But Matthew includes some passages that give Peter an even more central position. The best example is Matthew’s expansion of Peter’s confession from You are the Messiah (Mark 8:29) to You are the Messiah, the Son of the Living God (Matt 16:16), which is then followed by Jesus’ benediction of Simon, his giving him the new name Peter, and promising to build the church on him (16:17–19). If Luke had access to Matthew’s Gospel, and if he wanted to put Peter in good light, why did he not include this material?¹¹⁰

Burkett gives several other reasons why the view that Luke used Matthew is implausible. As was mentioned, it is hard to understand why Luke did not preserve Matthew’s structure, and why he would have replaced stylistically eloquent pericopes with less sophisticated versions of the same passages if he had access to Matthew’s Gospel. While individual differences may be accounted for, it is hard to explain why Luke shows no signs of knowing Matthew’s redaction of Markan material.¹¹¹ Burkett notes that themes that are central to Matthew (such as the fulfilment of Scripture) are less developed in Luke, even when these do not run counter to Luke’s theology.¹¹² Another argument against Luke’s dependence on Matthew is that Matthew generally reflects a more developed theology.¹¹³ I will give examples of where this appears to be the case in later chapters.

Scholars are inventive, but ultimately the case for Lukan dependence on Matthew is difficult to maintain. I find that the simplest explanation for Luke’s failure to more closely follow Matthew’s Gospel is that Luke never had access to Matthew’s Gospel.

The theory of Matthean posteriority

A handful of scholars, including Huggins, Hengel, MacEwen, and Garrow, reject the two-source hypothesis and the Farrer hypothesis and have put forward strong arguments that Matthew may have had access to both Mark and Luke.¹¹⁴ They explain most of the similarities between Matthew and Luke by arguing that contrary to received wisdom, Matthew was written later than Luke and the author in fact had access to both Mark and Luke, in addition to other sources.¹¹⁵ According to this theory, Matthew revised Mark’s Gospel but he had access to Luke’s Gospel as well and could compare Mark’s Gospel with Luke’s formulations, which he used when it suited his purposes. Like the Farrer hypothesis, the theory of Matthean posteriority readily accounts for both the passages that Matthew shares with Luke alone and the minor agreements between Matthew and Luke in texts found in all three Gospels.

Those who adhere to the two-source hypothesis in one of its forms tend to assume it is equally difficult to believe that Matthew used Luke as it is to believe that Luke used Matthew, but they do not adequately explain why this should be. Adherents of the Farrer hypothesis similarly dismiss the possibility of Matthew having used Luke, but they do not explain their reasoning.¹¹⁶ It is indeed exceedingly difficult to imagine Matthew’s Gospel as a revision of Luke, but if Matthew’s main source was Mark and he only used Luke as a complementary source these difficulties largely disappear. Luke’s additions to Mark’s text are readily identifiable. While Matthew has placed much of the material he had to work with into different contexts and has imposed a structure of his own to the Gospel material, Luke has largely retained Mark’s structure. He dropped some material (most notably Mark 6:45—8:26 and 9:41—10:12), inserted large blocks of material in a few places in Mark’s text (including birth and infancy narratives, and most of the material in Luke 6:20—8:3 and 9:51—18:14), but where he uses Mark’s material he does not change the order or contents of the pericopes much. This suggests that Luke generally had a conservative approach to his source material. In this light the Farrer hypothesis, which concludes that Luke radically redistributed Matthew’s material while carefully preserving Mark’s material, seems considerably more complicated, and therefore less plausible, than the theory of Matthean posteriority.

This hypothesis of Matthean posteriority is not widely accepted,¹¹⁷ but it is should be taken seriously. In her survey article on the Synoptic problem Walters touches on this explanation only briefly (six lines of text are dedicated to it), noting that it offers answers to major problems. The need for Q or other hypothetical documents is eliminated and the minor agreements and problems with Lukan dependence on Matthew are explained.¹¹⁸ She considers the theory worthy of further attention.

The theory of Matthean posteriority allows us to dispense with the hypothetical Q source. Not all advocates of the theory do so however. In his version of the theory of Matthean posteriority, Garrow does not rule out that Matthew and Luke may also have had access to another source, commonly referred to as Q, but he assigns much less material to Q than scholars have traditionally done. The International Q Project estimates that Q consisted of about 4,500 words; Garrow suggests that the total combined length of any ‘Q’ sources is likely to be somewhat less than 450 words.¹¹⁹ Distinguishing between what Matthew has taken over from Luke and what he has taken over from Q would not be easy. Hengel’s rule of thumb is that where there is close word-for-word agreement in the double tradition material, we can take as starting assumption that Matthew used Luke, and where their wording is different, we can assume that they are dependent on different logia sources.¹²⁰ Considering that Matthew takes considerable liberties in redacting his sources (as we see in the case of material he has taken over from Mark) one should be cautious in trying to determine the source of individual double tradition passages. I will not trace any material to Q in this book. Luke certainly used material from other texts besides Mark in writing his Gospel, but since those texts have not been preserved I will not speculate about how Luke might have modified them nor will I speculate about whether Matthew and Luke might