Non-Roman fonts used: SPIonic, SPTiberian

This article is also available in transliteration and text-only formats.

See William L. Petersen's review article in TC 3 (1998).

A Response to William L. Petersen's Review of Hebrew Gospel of Matthew

George Howard
University of Georgia

1. Recently (May 1998) William L. Petersen of Pennsylvania State University published a review article of my 1995 edition of Shem-Tob's Hebrew Matthew (Howard 1995). Petersen's article, "Some Observations on a Recent Edition of and Introduction to Shem-Tob's 'Hebrew Matthew,'" is published in TC: A Journal of Biblical Textual Criticism (Petersen 1998). I appreciate TC allowing me an opportunity to offer this response.

2. I declined to respond to Petersen's review of the first edition of my book (Howard 1987) that appeared in the Journal of Biblical Literature 108 (1989): 722-726, and I take no pleasure in responding to the present review. But Petersen threatens a third something (review? see Petersen 1998: par. 133) if I do not reply, so I have decided to comply with his wishes.

3. Petersen's writing is highly unusual for academia. It is filled with vituperation and invective (virtually every word is with a drop of blood). He scolds me, castigates me, and accuses me of "chicanery" (Petersen 1998: par. 124). Toward the end of his review, he preaches a sermon in which he pours out on me his righteous indignation. To my knowledge, I have never met Petersen, so why he writes with such personal virulence, I cannot fathom. I request the readers of this rebuttal to allow me some indulgence to respond in kind, though I will be as considerate as possible. This is not my usual way of doing things, but the occasion seems to call for it.

4. Overall, Petersen's article is a perfect example of how a review should not be written. I consider it not an exaggeration to write that virtually every page swarms with errors of omission, commission, and lack of understanding. He does not understand my book; he does not understand medieval Jewish literature; he does not understand the literary nature of the Hebrew text of Shem-Tob's Matthew, including its many puns, word connections, and alliteration; he does not understand the evolution this text has undergone or where it belongs in its long chain of transmission; he does not understand the heterodox theological patterns running throughout the document or where these fit into the history of Christianity.

5. Much of Petersen's review is taken up with the Middle Dutch Liège Harmony, but I am not convinced that he understands even this document. I will come back to the Middle Dutch Liège Harmony, but I must first deal with Petersen's failure to understand the purpose of my book and his ingenuousness in dealing with medieval Hebrew language and literature.

I. The Purpose of My Book

6. Petersen never picks up on the real aim of my book, which is stated repeatedly throughout the volume. It is to demonstrate that the Hebrew Matthew contained in Shem-Tob's writing predates the fourteenth century. In the Preface to the second edition (Howard 1995: vii), I write, "The main thrust of this second edition is to demonstrate that the Hebrew Matthew contained in Shem-Tob's Evan Bohan predates the fourteenth century." At the end of the volume (Howard 1995: 234), I sum up my findings: "A conclusion that can be drawn from these comparisons is that Shem-Tob's Hebrew Matthew predates the fouteenth century, being preserved primarily by the Jewish community." I do not believe I could have made this point clearer (see also pp. 153, 173-175, 178, 211).

7. Petersen's review gives the impression that the purpose of my book is to prove that Shem-Tob's Hebrew Matthew dates to circa 100 C.E.. What I have argued is that a Shem-Tob type Matthean text (not Shem-Tob's Hebrew Matthew per se) has roots in an early period of Christian history; I do not presume that Shem-Tob's Hebrew Matthew goes back to early times. This is a text that has been revised repeatedly, has taken on much textual baggage during transmission, and is probably no more that a dim reflection of a prior tradition. I am unsure what this prior tradition amounts to, whether a complete gospel or simply an undefined source. But, whatever the case, the point of this book is simply to demonstrate that the tradition lying behind Shem-Tob's Hebrew Matthew predates the fourteenth century, perhaps by several centuries. Nothing more!

8. I set this agenda for my book because traditionally Shem-Tob's Hebrew Matthew has been held to be a translation of the Latin Vulgate made by Shem-Tob Ibn Shaprut himself in the fourteenth century. Throughout the last century, it was thought that Shem-Tob was the author of this text. As late as 1967, Matthew Black wrote in regard to the Hebrew version of du Tillet, "The author of the Hebrew Matthew was probably a certain Shem-Tob ben Shaprut, a famous Jewish polemical writer who flourished in Spain in the fourteenth century" (Black 1967: 295). Robert Lindsey holds a similar view: he identifies S. Münster's edition of the Hebrew Matthew as a version of "Ibn Shaprut's translation" [Lindsey 1969: 67.]

9. At times Petersen misunderstands the purpose of my discussion and imputes theories to me that I do not formulate, agree with, or believe. I offer here two examples.

Example 1

10. In a segment called, "Linguistic Characteristics of the Hebrew Text," I give a brief description of the linguistic nature of the Hebrew text of Shem-Tob's Matthew. I offer this material (consecutive tenses, non-consecutive tenses, infinitives, pronouns, and vocabulary) merely to acquaint the reader with the type of Hebrew employed in the text. In no way do I use this material to argue for the date of the Hebrew text or for anything else. It is pure description.

11. Petersen, apparently not understanding this material as profile, accuses me of arguing for an early date of Shem-Tob's Hebrew Matthew by my discussion of the Hebrew waw consecutive. He writes:

Howard elaborates no further, so it is not entirely clear what his point is. One surmises, however, that what he wishes to argue is this:
  1. Shem-Tob's Hebrew Matthew contains a feature which, according to Howard, is distinctive of "Biblical Hebrew"; and
  2. this feature fell from use, according to Howard, after the decline of Biblical Hebrew; therefore,
  3. Shem-Tob's Hebrew Matthew must have been composed in the period when Biblical Hebrew was still in use--in other words, in early Christian antiquity. [Petersen 1998: par. 18].

12. Once Petersen has concocted this theory and attributed it to me, he ridicules me for believing it. "Howard never considers the possibility that the waw-consecutivum might be an artificial, 'antique' feature used by the translator to give his medieval Hebrew translation of Matthew an 'antique' flavor" (Petersen 1998: par. 20). "Howard seems unaware that there is empirical, textual evidence that the waw-consecutivum was used long after the decline of Biblical Hebrew" (Petersen 1998: par. 21). He then cites some little known codex, Vaticanus 32 of the late sixteenth or early seventeenth-century, made by a certain Dominicus of Jerusalem, as empirical proof that the waw consecutive survived the biblical period. (I will return to this codex momentarily.)

13. I must confess to some bewilderment. Petersen creates a theory ex nihilo--one that I do not formulate myself and with which I disagree--then takes delight in criticizing me for it. If he had just read the opposite page (same opening), he would have realized that his theory does not reflect my beliefs at all. I write,

If it were a matter of an original Jewish composition in the late Middle Ages, one would expect BH [Biblical Hebrew] or even archaic BH to play a dominate role, as is the case with most texts written during this time" [Howard 1995: 178].
I am fully aware that the waw consecutive was used in the late medieval period and clearly state this. If Petersen had read the context carefully, he would have known this.

Example 2

14. A second example is Petersen's treatment of my section on Shem-Tob and the Gospel of Thomas. I conclude the section by writing, "It is highly unlikely that Shem-Tob had direct contact with the Gospel of Thomas. The agreements of his Matthew with Thomas, therefore, must be traced to the early centuries of the Christian era" (Howard 1995: 205). With these words, I attempt to demonstrate that the readings under consideration predate the fourteenth century and therefore were not created by Shem-Tob. I think Petersen is in basic agreement with this statement; he just wishes to place some intermediary gospel texts between Shem-Tob and Thomas. And this is fine with me. It fits quite well with my understanding of the revisionary nature of Shem-Tob's Hebrew Matthew.

15. The problem that I have with Petersen is that he thinks I believe that the agreements between Shem-Tob and Thomas prove that Shem-Tob's Hebrew Matthew is ancient. He writes, "Howard's most important evidence for the antiquity of Shem-Tob's Hebrew Matthew consists of lists of agreements between it and texts known in ancient Christianity" (Petersen 1998: par. 36). He then invents another theory ex nihilo, imputes it to me, and accuses me of it. His theory, in abbreviated form, is the following:

(1) Parallels exist between the Hebrew Matthew in Shem-Tob's Even Bohan . . . and these early Christian texts whose circulation was in Eastern Christendom. (2) None of these early Christian texts was known in or circulated in the West during the medieval period when Shem-Tob composed the Even Bohan. . . . (3) . . . One is--according to Howard--driven to conclude that these common readings must be the result of this Hebrew Matthew having been composed in a time and in a place where works such as the Gospel of Thomas, the Vetus Syra, and Codex Sinaiticus were in circulation--in other words, in ancient Christianity [Petersen 1998: par. 36].

16. Again, I must confess to bewilderment. I do not make this argument in my book and believe none of it. I do not know when Shem-Tob's Hebrew Matthew dates or where it was created. I clearly state my position: "It is highly unlikely that Shem-Tob had direct contact with the Gospel of Thomas. The agreements of his Matthew with Thomas, therefore, must be traced to the early centuries of the Christian era" (Howard 1995: 205). You will note that I write, "the agreements of his Matthew with Thomas . . . must be traced to the early centuries of the Christian era." I do not write "Shem-Tob's Hebrew Matthew must be traced to the early centuries of the Christian era."

17. But Petersen does not recognize this nuance in my writing, and lashes out against me, apparently thinking that I do not state the issue correctly. "Howard appears confused when he states that 'it is highly unlikely that Shem-Tob had direct contact with the Gospel of Thomas,' for it is clear to everyone--and Howard himself has argued this point . . . that Shem-Tob did not translate this Hebrew Matthew, but incorporated an already existing (and, therefore, older) Hebrew Matthew into his Even Bohan" (Petersen 1998: par. 50). I wish to assure Petersen that this is not clear to everyone. It has never been clear to everyone. In fact, the reason I wrote this book is to prove that Shem-Tob did not create this Hebrew Matthew.

II. Petersen and Medieval Hebrew Language and Literature

18. One would expect that a person who chooses to write a review of a book containing a Hebrew text from a medieval Jewish polemical treatise would be well acquainted with medieval Hebrew language and literature. Otherwise, how could he/she write an intelligible review? Unless I am badly mistaken, Petersen does not approach medieval Hebrew language and literature with a great deal of sophistication. Please observe the following points.


19. Petersen tells us that Hebrew Matthews have been known and used for centuries. He mentions those published by Sebastian Münster and Jean du Tillet. He continues, "However, in addition to these two, at least five other Hebrew Matthews (mostly fragmentary) are known: (1) the Book of Nestor, (2) the Milhamot HaShem, (3) the Sepher Joseph Hamekane, (4) the Nizzahon Vetus, and, now, (5) the work edited by Howard" (Petersen 1998: par. 3).

20. This statement is revealing. The Book of Nestor, the Milhamot HaShem, the Sepher Joseph Hamekane, and the Nizzahon Vetus are not Hebrew Matthews. They are medieval anti-Christian polemical treatises written in Hebrew. Moreover, Shem-Tob's Hebrew Matthew, the only Hebrew Matthew in the lot, is not fragmentary; it includes the entire Gospel of Matthew. Petersen's knowledge of medieval Jewish writings is clearly faulty on this point.


21. To return now to Petersen's codex Vaticanus 32 (cf. above, par. 12), he writes, "Howard seems unaware that there is empirical, textual evidence that the waw-consecutivum was used long after the decline of Biblical Hebrew." He then cites codex Vaticanus 32 of the late sixteenth or early seventeenth-century, made by Dominicus of Jerusalem, as empirical proof that the waw consecutive survived the biblical period.

22. I agree with Petersen that the document he offers provides the empirical evidence he is looking for. But why refer to this document? It is like going into a library, checking out a book inaccessible to most readers, finding a picture of an automobile on one of the streets of America, and then announcing to the world that this is empirical proof that automobiles have been driven in America. Granted you have proven your point. But would it not be easier simply to walk outside and observe the traffic? Petersen is apparently unaware that there is a whole corpus of medieval Hebrew literature out there that is written primarily in biblical Hebrew!


23. Petersen's assessment of my translation tells us a whole lot more about Petersen than about my translation. He describes my English translation as "not entirely accurate." What disturbs him, among other things, is the way I render the waw consecutive. (He seems to have difficulty getting beyond the Hebrew conjunction.) He writes, "It is very odd, then, that Howard's translation often fails to render this feature, which he considers to be a key piece of evidence for the antiquity of this Hebrew Matthew." He then cites Matt 2:11 and writes, "The waw-consecutivum actually occurs six times in this sentence, but five of them are omitted by Howard's translation." Next, he argues that my translation sometimes "does not accurately reflect the word order of the Hebrew." He cites Matt 3:9 where the Hebrew reads literally, "Father-our Abraham," and writes, "Howard translates this as 'Abraham is our father,' instead of 'Our father is Abraham." He then speaks of these "subtle moves away from a literal translation" (Petersen 1998: pars. 13-15).

24. I am aware that some people feel the need for an interlinear translation of the Bible, but these are usually people to whom biblical languages are basically a mystery. Without a word-for-word translation of the original language, they are lost in that language. But, once one gets beyond the initial stages of the alphabet and basic grammar, one usually no longer feels the need for a word-for-word, woodenly literal rendition of the text. While there may be disagreement on how the finished product should read, there is little or no disagreement that a translator is permitted to have some freedom in making a readable version. Petersen is apparently one of those who needs an Aquila-type translation, thinking that anything less is "not entirely accurate."


25. For some reason, Petersen ignores the section in my book that deals with medieval Hebrew documents, e.g., pp. 155-173. In fact, unless I am overlooking the obvious, he avoids almost everything that requires a knowledge of medieval Hebrew language and literature that is beyond the most elementary level. As demonstrated above, his one skirmish into Hebrew grammar, that regarding the waw consecutive, is pedestrian. Perhaps Petersen should have recused himself from writing a review of my book on the grounds that he is unacquainted with the area.

III. Petersen and the Middle Dutch Liège Harmony

26. I will begin by congratulating Petersen for pointing out that some of Shem-Tob's readings agree with the Middle Dutch Liège Harmony. His discovery confirms what I have argued all along about Shem-Tob's Hebrew gospel text, namely, that it underwent a significant period of transmission and development during which it accumulated considerable textual baggage. I had already pointed out numerous agreements between Shem-Tob's text and other documents and text forms. Now Petersen adds the Middle Dutch Liège Harmony. He reports on a long list of parallels he has found between Shem-Tob and the Middle Dutch Liège Harmony. But, given his way of counting parallels (see below), I suspect there will be some disagreement about the actual length of this list. Petersen is obviously very fond of (obsessed with?) the Middle Dutch Liège Harmony. His application of this document in arguing for the origin and nature of Shem-Tob's Hebrew Matthew, however, leaves a great deal to be desired. Let me explain.


27. Petersen disputes my claim that the Old Latin and the Old Syriac sometimes agree with Shem-Tob's Hebrew Matthew against all other Matthean witnesses. He argues that many of these readings are also to be found in the Middle Dutch Liège Harmony. He writes, "Howard's claim that these readings [agreements between Shem-Tob's Hebrew Matthew and the Old Latin] are 'against all other Matthean witnesses' is simply not true. Of his twelve readings, six--once again, half of Howard's list--are found in just a single source, namely the Middle Dutch Liège Harmony (if the comparison were expanded to include all the Western gospel harmonies, the number would increase)" (Petersen 1998: par. 57, italics his).

28. I find this to be a very strange statement. Does Petersen actually believe that the Middle Dutch Liège Harmony is a "Matthean witness?" It strikes me as inappropriate to describe a gospel harmony as a "Matthean witness." Usually, we reserve this label for such documents as Codex Vaticanus and Codex Sinaiticus. But Petersen wishes to extend it to include the Middle Dutch Liège Harmony. I reviewed my list of readings and found that almost all of them belong to the double or triple traditions of the synoptic gospels. Just how the Middle Dutch Liège Harmony can be a Matthean witness in the double and triple traditions, where words and phrases of the synoptic gospels are intertwined and overlapped, is unclear.

29. In regard to my list of agreements between Shem-Tob's Hebrew Matthew and the Old Syriac against all other Matthean witnesses, Petersen again writes, "[This] is simply false. Numerous readings are parallel in Western harmonized gospel texts; indeed, singling out just one such text, four of his twelve readings are found in the Liège Harmony, and a fifth approximates its reading" (Petersen 1998: par. 59, italics his).

30. He thus reconfirms his belief that the Middle Dutch Liège Harmony is a Matthean witness. If this had happened only once, one might discard it as a verbal slip. But twice? (Actually he does it a third time in Petersen 1998: par. 79.) Again, I reviewed my list and found that almost all of the readings belong to the double or triple traditions of the synoptic gospels.

31. I offer here six examples (more than half of those Petersen singles out) from my lists for you to judge for yourself if the Middle Dutch Liège Harmony is truly another "Matthean witness." It will be instructive to put Petersen under the gun, so to speak, and see if the examples, which he himself approves, actually support his contention. Please keep in mind that the operative question is whether the Middle Dutch Liège Harmony is another Matthean witness that supports Shem-Tob's reading.

I. Shem-Tob and the Old Latin

  1. 32. Matt 1:12

  2. 33. Matt 16:4

  3. 34. Matt 18:9

II. Shem-Tob and the Old Syriac

  1. 35. Matt 8:9

  2. 36. Matt 22:35

  3. 37. Matt 26:10

38. Needless to say, these examples do not inspire a great deal of confidence. In fact, they are worthless! To argue that the Middle Dutch Liège Harmony is another Matthean witness that supports Shem-Tob's Hebrew text in these passages is just grasping at straws. Some of the other examples are perhaps more convincing, but even here it is often a matter of interpretation. And the fact that they almost all occur in the double or triple traditions of the synoptic gospels generally nullifies the Middle Dutch Liège Harmony from being a "Matthean witness."


39. Petersen understands the agreements between Shem-Tob's Hebrew Matthew and the Middle Dutch Liège Harmony as a key to the origin of the Hebrew Matthew. With confidence, he informs us that he knows the origin of this text. "There is no mystery," he writes, "about the genesis of Shem-Tob's Hebrew Matthew" (Petersen 1998: par. 104). He goes on to explain the document's origin: "The tradition behind the Liège Harmony--which we know to be a Latin gospel harmony--must also be the principal element responsible for the textual complexion of Shem-Tob's Hebrew Matthew" (Petersen 1998: par. 104). Again he writes, "The Hebrew Matthew in Shem-Tob clearly incorporates older traditions, but these are merely derived from its Latin Vorlage, a Vorlage which was closely related to [the] Latin Vorlage of the Middle Dutch Liège Harmony" (Petersen 1998: par. 118). He dates the Latin Vorlage of the Liège Harmony to be approximately 1100, or the mid-point, +/- 200 years, between 900 and 1300, the outside limits for the Vorlage's date (Petersen 1998: par. 109).

40. I have two major problems with Petersen's understanding of the origin of Shem-Tob's Hebrew Matthew.

41. (1) One problem is the weakness of Petersen's evidence that Shem-Tob's Hebrew Matthew as a whole goes back to a Latin Vorlage. He believes that the Middle Dutch Liège Harmony goes back to a Latin Vorlage, and since there are agreements between Shem-Tob's Hebrew Matthew and the Middle Dutch Liège Harmony (some of which apparently suggest a Latin base), he believes that Shem-Tob's Hebrew Matthew as a whole also goes back to a Latin Vorlage.

42. But what Petersen does not tell you is that he is ignoring all the evidence that demonstrates that the base text of much of Shem-Tob's Matthew is Hebrew, not Latin. By making a careful selection of readings that appear to have a Latin base, and being silent about those that do not, he proposes that the whole comes from a Latin base. But until he explains the Hebrew puns, word connections, alliteration, and the other evidence pointing to a Hebrew base, his evidence for a Latin Vorlage is nonprobative.

43. (2) A second problem is that Petersen apparently believes that a gospel harmony can explain the origin of a single gospel. But Petersen does not tell us how this can be. A gospel harmony can influence the composition and scribal transmission of a single gospel, but it is very unlikely that it can explain a single gospel's ultimate origin. Evidence that Petersen is ambivalent on this point is how he changes terminology. One time he writes, "common Vorlage," (Petersen 1998: par. 105); then he writes, "harmony-influenced separate gospel text" (Petersen 1998: par. 113). Just what a common Vorlage would be for the Middle Dutch Liège Harmony and Shem-Tob's Hebrew Matthew is beyond my comprehension.

44. Perhaps the analogy of the Diatessaron of Tatian and the Evangelion da-Mepharreshe (separate gospels) will help Petersen sort this out. Two extant manuscripts of the Evangelion da-Mepharreshe, the Curetonian and the Sinaitic Syriac, contain a great many harmonistic readings that appear to come from the Diatessaron. A longstanding debate concerns which one of these text forms came first, the Evangelion da-Mepharreshe or the Diatessaron. Generally it is believed that the Diatessaron was the earliest form of the gospel in Syriac and that the translator(s) of the Evangelion da-Mepharreshe, immersed in the wording of the Diatessaron, employed much of the Diatessaron's phraseology. In addition, the Evangelion da-Mepharreshe acquired new readings from the Diatessaron as it was transmitted. Nevertheless, in spite of this, in the final analysis the Evangelion da-Mepharreshe is a translation of some form of the separated gospels (perhaps Greek), not the Diatessaron. F. C. Burkitt explains as follows:

The Evangelion da-Mepharreshe is a translation from the Greek. Even if the translator knew the Diatessaron in Syriac and was greatly influenced by it, it is nevertheless obvious that a text of the Four Gospels cannot be evolved from the Diatessaron alone. . . . According to the view here suggested, the Evangelion da-Mepharreshe in its original form gives in essentials a faithful representation of the text of the Four Gospels as received at Antioch about 200 AD. The wording of the translation has been often influenced by the renderings found in the Syriac translation of Tatian's Diatessaron, a work familiar to the translator of the Evangelion da-Mepharreshe; to this cause also we may put down the many instances of minor harmonistic readings, so far as they have not been brought into our MSS by a similar tendency on the part of Syriac scribes [Burkitt 1904: 191, 209-210].

45. Following this analogy, if Shem-Tob's Matthew received an input of readings from the Middle Dutch Liège Harmony tradition or its Vorlage sometime around or after 1100 CE (I concede this for the sake of the analogy only), this does not explain the origin of the base text of the separated Gospel of Matthew that was already in existence when the input was received. Petersen's confusion over the nature of a gospel harmony has apparently led him to assign more importance to the Middle Dutch Liège Harmony in regard to the origin of Shem-Tob's Hebrew Matthew than is appropriate.

46. It is this base text of Shem-Tob's Hebrew Matthew, or its substratum, that Petersen's theory does not explain. It is this substratum to which I give much attention in my book. Unfortunately, Petersen pays no attention to this, ignoring large sections of my book that describe it. It is a substratum that belongs to a stage in the process of a textual evolution that began in earlier times and culminated in du Tillet in the sixteenth century, or possibly even later. It is a substratum filled with Hebrew puns, word connections, and alliteration that cannot be explained by the Middle Dutch Liège Harmony or by its Latin Vorlage. It is a substratum with its own theology, different from the theology of canonical Matthew or the Middle Dutch Liège Harmony.

IV. Did Petersen Read the Entire Book?

47. It appears that Petersen read only a small portion of my book. Judging from the part of the book he actually comments on, I conjecture that he read between 10 and 12 pages of the analysis, or approximately 15% of 81 pages. Add to this an occasional reference to the text of Matthew and you get 15 pages at the outside. One might argue that Petersen read all the book, and only commented on a few pages. I think this is not the case. Some of his mistakes are understandable only if he did not read the whole book. For example, he writes, "Virtually all of the arguments and evidence employed in that first edition are absent from this second edition" (Petersen 1998: par. 17). This is untrue. If Petersen had read the whole book, he would have known that almost all the evidence employed in the first edition (Howard 1987) is present in the second edition. Petersen also states that I never return to the issue of Shem-Tob and the Gospel of John and never "present any additional argumentation or evidence" (Petersen 1998: par. 71). This also is untrue. He obviously did not read pages 220-221 of my book. For more on this, see below in the Appendix.

48. In the Appendix, I review the material that Petersen ignores, including the development of the Hebrew text, its literary characteristics, and its theology. I will let you be the judge whether this material has anything to offer concerning the origin and nature of the base text of Shem-Tob's Hebrew Matthew.


Pages 160-173

49. In this section, I show that Shem-Tob's Hebrew Matthew belongs to a process of textual evolution within a Jewish Hebrew milieu that began in earlier times and culminated in du Tillet in the sixteenth century, or possibly later if other text forms are to be taken into consideration. The works I include in this examination are the Hebrew polemical writings of the Book of Nestor (dating perhaps between the sixth and ninth centuries), the Milhamot HaShem by Jacob ben Reuben (1170), Sepher Joseph Hamekane by Rabbi Joseph ben Nathan Official (thirteenth century), the Nizzahon Vetus (latter part of the thirteenth century), and the Hebrew versions of Matthew published by Sebastian Münster and du Tillet. The similarity of Shem-Tob's text of Matthew with the quotations of Matthew in the earlier Hebrew polemical works shows that Shem-Tob's text preserves an already existing Hebrew Matthean tradition that had been in the process of evolution for an unknown period of time.

50. Petersen does not comment on this part of my book except to mention the Book of Nestor, the Milhamot HaShem, Sepher Joseph Hamekane, and the Nizzahon Vetus, and to call them four Hebrew Matthews (see above, pars. 19-20).

Pages 182-183

51. This section, entitled: "Revision and Modification of the Hebrew Text," shows that the Hebrew text of Shem-Tob's Matthew has undergone extensive revision throughout its transmission history. The revision includes alteration designed to bring the Hebrew into line with the Greek and Latin texts used during the Middle Ages, to improve the style of the Hebrew, and to introduce various extraneous material into the narrative.

52. Petersen does not discuss this part of my book.

Pages 184-190

53. In this section I show that Shem-Tob's Hebrew Matthew is characterized by literary devices such as puns, word connections, and alliteration. These are numerous--some parts of the text are virtually saturated with them--and they belong to the very structure of the Hebrew. Their origin cannot be explained by the Middle Dutch Liège Harmony. I offer here some examples.


54. Matt 7:6: "Do not throw your pearls before swine, lest they trample them under foot and turn to attack you." "Swine" (ryzx) and "turn" (rzxy) are similar in Hebrew, both in sound and appearance, and form a wordplay in the text.

55. Matt 16:18: Instead of the famous Greek pun, Pe/troj/pe/tra|, the Hebrew reads, "You are a stone, and upon you I will build my house of prayer." The pun consists of the word Nb) "stone" and hnb) "I will build."

56. Matt 23:27-28: "For you are like whitewashed tombs, which outwardly appear beautiful, but within they are full of dead men's bones and all uncleanness. So you also outwardly appear righteous to men, but within you are full of hypocrisy and iniquity." The pun involves the root rbq "tomb" and brq "within."

Word Connections

57. Word connections are links made by the repetition of the same or similar words designed to tie separate sayings and pericopes together.

58. Matt 4:21-23: "He turned from there and saw two other brothers, James and John, brothers who were sons of Zebedeel [l)ydbz from ydbz "gifts" and l) "God," perhaps meaning "gifts of God"] . . . . Then Jesus went around the land of Galilee teaching their assemblies and preaching to them the good gift (dbz) . . . of the kingdom of Heaven." The repetition of dbz "gift" ties the pericope of the call of the disciples to the pericope of Jesus' first preaching tour in Galilee.

59. Matt 5:9-10: "Blessed are those who pursue [ypdwr] peace . . . Blessed are those who are persecuted [Mypdrnh] for righteousness" [Pdr = to pursue/to persecute].

60. Matt 8:28, 31: There met him [wb w(gpyw] two demon-possessed men. . . . Then the demons entreated him [wb w(gpyw]" [(gp = to meet/to entreat].

61. Matt 14:35-36: "They brought to him all those who were sick [Mylwxh] with various kinds of diseases. They implored [wlxw] him" [hlx = to be sick/to implore].

62. Matt 15:34-37: "They answered, seven [h(b#] and a few fish. So Jesus commanded the people to sit upon the grass. Then he took the seven [h(b#h] loaves . . . . All of them ate and were satisfied [w(b#yw] and from that which was remaining they filled seven [h(b#h] seahs [h(b# = seven; (b# = to be satisfied].

63. Matt 18:23-35: This parable is held together by the catch-word Ml#, first meaning "to repay," then "perfect."

At that time Jesus said to his disciples: the kingdom of heaven is like a certain king who sat to make a reckoning with his servants and ministers. As he began to reckon, one came who owed about ten thousand pieces of gold. But he had nothing to give and his master commanded to sell him and his children and all that was his to repay [Ml#l] the value. The servant fell before his master and implored him to have pity on him and to be patient with him because he would repay [Ml#y] everything. Then his master had pity on him and forgave him everything. But that servant went out and found one of his comrades who owed him a hundred pieces of money and he grasped him and struck him saying . . . Trust me and be patient with me and I will repay [Ml#)] everything. But he was not willing to listen to him; so they brought him to the prison until he repaid [Ml#] him everything. The servants of the king saw that which he did and were very angry and went and told their master. Then his master called him and said to him: Cursed servant, did I not forgive you all your debt when you placated me? So why did you not forgive your servant when he supplicated you as I forgave you? His master was angry with him and commanded to afflict him until he should repay [Ml#y] him all the debt. Thus will my Father who is in heaven do to you if you do not forgive each man his brother with a perfect [Ml#] heart.

64. Matt 19:9-13: He who takes her who has been divorced [h#wrgh] commits adultery. . . . Then they brought children to him . . . but his disciples were driving [My#rgm] them away [#rg = to divorce/to drive away].

65. Matt 26:28, 34-36: "This is my blood of the new covenant which will be poured out for many for the atonement (trpkl) of sins. . . . . Jesus said, Truly I say to you, this night before the cock-crow you will deny (rwpkt) me three times. Peter said to him, If it is possible for me to die with you, I will not deny (rwpk)) you . . . . Then Jesus came with them to the village (rpkl) of Geshemonim." This extended word connection is made up of the words hrpk, meaning "atonement," and rpk, meaning either "to deny" or "village."


66. There are many passages in Shem-Tob's Hebrew Matthew that are structured around alliteration, including paronomasia and rhymes of various sorts.

67. Petersen does not discuss this part of my book.

Pages 192-194

68. In this section I consider the short ending of Matthew in Shem-Tob's Hebrew Matthew and a similar short ending that F. C. Conybeare observed in some manuscripts of Eusebius. Conybeare suggested that the short ending in Eusebius, lacking the Trinitarian baptismal formula, was reflected in Justin Martyr (Dial. 39, 53) and Hermas (Sim. 9.27.4) (see Conybeare 1901). Others have added new evidence for a short ending of Matthew, and this evidence is discussed as well.

69. Petersen does not discuss this part of my book.

Pages 205-212

70. In these pages I consider several ancient writings that have variant readings in agreement with Shem-Tob's Hebrew Matthew. These include the Pseudo-Clementine Writings, The Tol)doth Yeshu, and the Protevangelium of James. The Pseudo-Clementine Writings, made up primarily of the Recognitions and Homilies (but also the Epitomes), go back to an early third century Grundschrift, which is itself a compilation of various works, the oldest of which is the Kerygmata Petrou, or the Preaching of Peter. The Tol)doth Yeshu is a medieval Jewish antigospel which exists in various forms. It dates somewhere between the beginning of the sixth and the tenth centuries. It draws on early traditions reflected in the Talmud and Origen's Contra Celsum. The Protevangelium of James dates perhaps in the second century.

71. Petersen does not discuss this part of my book.

Pages 212-223

72. This section of the book is entitled, "Theological Motifs in Shem-Tob's Matthew." The motifs are made up of various themes such as divorce, swearing, the Gentiles, the Christ, and John the Baptist. In each instance the Hebrew text of Shem-Tob treats these subjects differently than they are treated in orthodox Christian writings. The Middle Dutch Liège Harmony does not explain the origin of the theological patterns contained in these motifs. Three examples follow.

The Gentiles

73. Shem-Tob's Hebrew Matthew envisions the incorporation of masses of the Gentiles into the Kingdom of God not in this present age but only after this present age ends (Matt 25:31-46). Its theology thus corresponds to the Hebrew Bible and later Jewish thought (including some forms of Jewish Christian thought), in which the entrance of the Gentiles is anticipated in the golden age to come. In the canonical texts of the Church, the Gospel of Matthew ends with the Great Commission (Matt 28:19), mandating the disciples to "Go therefore and make disciples of all nations, baptizing them in the name of the Father and of the Son and of the Holy Spirit." This ending clearly opens the doors to the Gentiles for the present time. But this ending does not occur in the Hebrew text. It reads instead, "Go and teach them to carry out all the things which I have commanded you forever." This is said apparently only in reference to the Jews, and nothing is said about teaching or baptizing the Gentile nations.

The Christ

74. Shem-Tob's Hebrew Matthew has the peculiar characteristic of not referring to Jesus as the Messiah/Christ until ch. 16. Before this, the Hebrew text never calls Jesus the Messiah. In the canonical text of Matthew, Jesus is called the Christ from the very beginning. A comparison of the Greek and Hebrew texts suggests that these two text forms have different theological agendas.

John the Baptist

75. Shem-Tob's Hebrew Matthew exalts John the Baptist above the role he plays in the canonical Matthew; cf. Matt 11:11, 13; 17:11. In summary, these passages assert that none is greater than John the Baptist, the prophets and the law spoke concerning John, and John came to save all the world. Traditional Christianity uses this language only in regard to Jesus. Its appearance in the Hebrew Matthew to describe John elevates the Baptist to a salvific role. This clearly flies in the face of the orthodox interpretation of John.

76. In this section, I return to the subject of the Gospel of John in a segment called "John the Baptist: The Fourth Gospel and Shem-Tob's Hebrew Matthew" (Howard 1995: 220-221). In it I discuss the polemic against John the Baptist in the Gospel of John. I demonstrate that Shem-Tob's Hebrew Matthew depicts the Baptist in the exalted terms the Fourth Gospel polemizes against. "If the Fourth Gospel was directed against the followers of John the Baptist," I write, "one could hardly conceive of a more appropriate document to represent this community than a Shem-Tob type Matthew" (Howard 1995: 221).

77. When criticizing my earlier treatment of Shem-Tob's Hebrew Matthew and the Gospel of John, Petersen claims that I do not pursue the matter further. He writes, "Howard does not pursue its implications [Petersen 1998: par. 69] . . . . Howard neither answers it, nor does he present any additional argumentation or evidence; one is simply left with his assertion that these Johannine snippets suggest that 'the author of the Fourth Gospel knew a Shem-Tob type of text of Matthew and used [it] when he wrote his gospel [Petersen 1998: par. 71].'" Petersen obviously did not read pages 220-221 of my book.

78. In a recent article (Howard 1998) I discuss more fully the significance of the theological patterns running throughout the Hebrew text of Shem-Tob's Matthew. I show that these patterns reflect a heterodox form of Jewish Christianity.

79. Petersen does not discuss any part of this section.

Pages 223-229

80. In this portion of the book, I discuss passages in Shem-Tob's Hebrew Matthew that receive a different meaning than they receive in the canonical version.

81. Matt 19:22: The Greek reads, "When the young man heard this he went away sorrowful, for he had great possessions." The Hebrew reads, "It came to pass when the young man heard he went away angry because he did not have much property."

82. Matt 26:13: The Greek reads, "Truly, I say to you, wherever this gospel is preached in the whole world, what she has done will be told in memory of her." The Hebrew reads, "Truly, I say to you, everywhere this gospel . . . is proclaimed in all the world, that which this one has done will be said in reference to my memory."

83. Matt 28:6: The Greek reads, "Come, see the place where he lay." The Hebrew reads, "Come, therefore, and see the place where the Lord arose."

84. In some instances, differences between the Greek and Hebrew texts can be traced to possible variations in the Hebrew tradition.

85. Matt 7:6 in the canonical text reads, "Do not give that which is holy to the dogs." The Hebrew text reads, "Do not give holy flesh to the dogs." The difference in Hebrew is: that which is holy = #dq r#); holy flesh = #dq r#b.

86. Matt 7:29 in the canonical text reads, "For he was teaching them as one having authority and not as their scribes." The Hebrew text reads, "For he was preaching to them with great power, not as the rest of the sages." The difference in Hebrew is: as = r#)k; as the rest = r)#k.

87. Matt 8:26 in the canonical text reads, "Why are you fearful?" The Hebrew text reads, "Why do you look at one another?" The difference in Hebrew is: are you fearful = w)rt; do you look at one another = w)rtt.

88. Matt 13:48 in the canonical text reads, "When it was full, they drew it up on the shore." The Hebrew text reads, "When it is full they draw it out." The difference in Hebrew is: on the shore = Pwxl; out = Cwxl.

89. Matt 23:37 in the canonical text reads, "Jerusalem, Jerusalem, who kills the prophets and stones those sent to her." The Hebrew text reads, "Jerusalem, who kills the prophets and removes those who are sent." The difference in Hebrew is: stones = tlqsmw; removes = tqlsmw.

90. Matt 26:23 in the canonical text reads, "He who has dipped his hand in the dish with me will betray me." The Hebrew text reads, "He who dips his hand with me in the dish will sell me." The difference in Hebrew is: will betray me = ynrsmy; will sell me = ynrkmy.

91. Many other examples are listed. Petersen does not discuss this part of my book.

Pages 229-232

92. In this segment, I discuss the use of the divine name in Shem-Tob's Hebrew Matthew. The Hebrew employs the symbol ~h for the Divine Name, which I suggest is an abbreviation for M#h, "the Name." This symbol occurs 19 times in the text, one time written out in full (28:9). Usually the Divine Name appears where the Greek reads ku/rioj, but twice (21:12 mss, 22:31) the Greek reads qeo/j, and three times the Divine Name has no correspondent in the Greek (22:32; 27:9; 28:9).

93. Petersen does not discuss this section of my book.

© TC: A Journal of Biblical Textual Criticism, 1999.


Black, Matthew 1967. An Aramaic Approach to the Gospels and Acts. 3rd ed. Oxford: Clarendon.

Burkitt, F. Crawford 1904. Evangelion Da-Mepharreshe. Vol. 2: Introduction and Notes. Cambridge: Cambridge University Press.

Conybeare, F. C. 1901. "The Eusebian Form of the Text Matth. 28, 19." Zeitschrift für die neutestamentliche Wissenschaft 2: 275-288.

Howard, George 1987. The Gospel of Matthew according to a Primitive Hebrew Text. Macon, GA: Mercer University Press; Louvain: Peeters.

Howard, George 1995. Hebrew Gospel of Matthew. Macon, GA: Mercer University Press.

Howard, George 1998. "Shem-Tob's Hebrew Matthew and Early Jewish Christianity." Journal for the Study of the New Testament 70: 3-20.

Lindsey, Robert L. 1969. A Hebrew Translation of the Gospel of Mark. Jerusalem: Dugith Publishers.

Petersen, William L. 1998. "Some Observations on a Recent Edition of and Introduction to Shem-Tob's 'Hebrew Matthew.'" TC: A Journal of Biblical Textual Criticism 3.

Plooij, Daniël 1929-1970. The Liège Diatessaron. Verhandelingen der koninklijke Nederlandse Akademie van Wetenschappen, Afdeling Letterkunde, vol. 31. Amsterdam: Koninklijke Akademie Wetenschappen te Amsterdam.