< type="text/css"


by Doug Ward

The gospels of Matthew, Mark, and Luke contain a great deal of common material, often expressed in very similar-sometimes even identical-language. For this reason, they are called the synoptic gospels. (``Synoptic'' literally means ``seeing together.'')

The issue of how to account for the similarities and the differences among the three synoptic gospels has fascinated biblical scholars for centuries. This issue is known as the ``synoptic problem.'' It involves hotly-debated questions like the following: In what order were the synoptic gospels written? In what language did each one originally appear? What kinds of written sources were available to each of the gospel writers?

Christian tradition has much to say about the synoptic problem. 1 For example, the third-century commentator Origen (c. 182-251 A.D.) gave this explanation of the order in which the gospels were written:

``First to be written was by Matthew, who was once a tax collector but later an apostle of Jesus Christ, who published it in Hebrew for Jewish believers. The second was by Mark, who wrote it following Peter's directives, whom Peter also acknowledged as his son in his epistle: `The church in Babylon greets you...and so does my son Mark' [I Peter 5:13]. The third is Luke, who wrote the Gospel praised by Paul for Gentile believers. 2 After them all came John's'' (quoted in Eusebius, Ecclesiastical History 6.25).

Our Bibles today reflect the order described by Origen, with the gospels appearing in the order Matthew-Mark-Luke-John. However, many scholars over the past two centuries have come to doubt this tradition. Today it is widely believed that Mark's gospel was the first to appear, since it is the shortest of the four and does not include some valuable teachings of Jesus that were recorded by Matthew and Luke. Those who favor Markan priority argue that it is difficult to explain why Mark would have omitted such material if it had been available to him.

Historical tradition, then, says that Matthew came first, followed by Mark. But current scholarly opinion, based on critical examination of the gospel texts, argues that it was the other way around. Which, if either, of these two views is correct? The late Jean Carmignac, one of the translators of the Dead Sea Scrolls, came to believe that in some sense both are right. In his book The Birth of the Synoptics (Franciscan Herald Press, Chicago, 1987), Carmignac proposed a model that reconciles this difference between ancient tradition and modern scholarship. My purpose here is to describe his proposal, which deserves to be more widely known.

An Original Hebrew Mark?

Through his work with the documents of Qumran, Jean Carmignac became very familiar with the Hebrew of Jesus' time. He also noticed many connections between the Dead Sea Scrolls and the New Testament. In order to study these connections further, he decided in 1963 to make a Hebrew translation of the Greek text of the gospel of Mark. When he began the project, he was immediately surprised to find how easy it was to carry out the translation:

``I had imagined that this translation would be difficult because of considerable differences between Semitic thought and Greek thought, but I was absolutely dumbfounded to discover that this translation was, on the contrary, extremely easy. Around the middle of April 1963, after only one day of work, I was convinced that the Greek text of Mark could not have been redacted directly in Greek and that it was in reality only the Greek translation of an original Hebrew. The enormous difficulties which I had envisioned for myself had all been resolved by the Hebrew-Greek translator, who had transposed word for word and who had even preserved in Greek the order of the words preferred by Hebrew grammar'' (The Birth of the Synoptics, page 1).

Carmignac later found out that his impressions about the text of Mark were by no means unique to him. Other translators had independently made similar observations. 3 He considered various possible explanations for the Semitic nature of the Greek of the synoptics. Was this Greek produced by writers who were simply trying to imitate the style of the Septuagint, the Greek translation of the Hebrew scriptures? No, it seemed too thoroughly Semitic for that. Or was it the Greek of writers whose native language was Semitic but who were trying clumsily to express themselves in Greek? This hypothesis did not seem to fit the data either:

``The appearance is perfectly Greek, too Greek to have been derived from people who possessed a poor knowledge of this language; but the reality is perfectly Semitic, so Semitic that it could only have come from people expressing themselves very naturally in their mother tongue. To put it another way: the Greek of the Gospels is not a poor Greek, nor a clumsy Greek: It is the good Greek of a translator who had respect for a Semitic original, who conserves the flavor and scent of the original'' (pp. 3-4).

Carmignac finally reached the conclusion that our Greek texts of Matthew and Mark were translated from Semitic (probably Hebrew) originals, and that Luke wrote in Greek but was drawing upon Semitic sources. A significant minority of scholars over the past century and a half has held roughly similar views, as he discovered upon further research. A list of forty-nine such scholars is catalogued in the sixth chapter of The Birth of the Synoptics.

A Taxonomy of Semitisms

Carmignac also found that over the past five hundred years, a number of scholars have observed and studied many Semitisms in the gospels-i.e., features of the Greek text that betray the influence of a Semitic language. He compiled a list of these Semitisms and classified them into nine possible categories:

1. Semitisms of Borrowing: Some words, like amen, abba, alleluia, messiah, and sabbath, are directly carried over or transliterated from a Semitic language.

2. Semitisms of Imitation: Some Hebrew phrases translated into Greek may have entered the gospels through a writer who was very familiar with the Septuagint translation of the Old Testament.

3. Semitisms of Thought: Classical Greek writing often contains many subordinate clauses, while coordinate clauses connected by ``and'' are more typical in both the Hebrew scriptures and the synoptic gospels. The gospels also contain phrases reflecting a fullness of expression typical of Hebrew writing. For example, Matthew 5:2 says that Jesus ``opened his mouth, and taught them'' rather than simply saying that he spoke or that he taught his disciples.

4. Semitisms of Vocabulary: Sometimes a single word in one language can have a range of meaning that is covered by several words in another language. For example, we use the word ``son'' to refer strictly to our male offspring, but ``son'' or ``child'' is used much more broadly in Semitic languages and in the synoptic gospels (e.g. Matt. 8:12; 9:15; 23:15; Mark 3:17; Luke 10:6; 16:8; 20:36).

5. Semitisms of Syntax: Characteristics of Hebrew syntax sometimes appear in the Greek text of the gospels. For example, in a phrase like ``in the house of the king,'' the first ``the'' is suppressed in Hebrew. Articles are often omitted in this way in the New Testament. In Hebrew, the verb ``to say'' or ``to speak'' is often used with the preposition ``toward,'' a construction that also appears in the gospels.

6. Semitisms of Style: Semitic prose is designed to be read aloud. It often includes repetition of words for emphasis and as an aid to memorization. Such repetition is also common in the gospels, where we see phrases like, ``A sower went out to sow his seed: and as he sowed...'' (Luke 8:5); ``With desire I have desired...'' (Luke 22:15); ``they rejoiced with exceeding great joy'' (Matt. 2:10). Repetition is also important in Hebrew poetry, where the most important feature is parallelism. The poetry in the gospels also exhibits parallelism.

7. Semitisms of Composition: In some gospel passages, meanings that would exist in a Hebrew version are lost or obscured in Greek. One striking example occurs in the Benedictus (Luke 1:68-79), the words Zacharias was inspired to speak after his son John the Baptist was born. In verses 72-73, Zacharias praises God for his promise

``To perform the mercy promised to our fathers, and to remember his holy covenant; the oath which he sware to our father Abraham.''

In Hebrew, the word for ``perform the mercy'' is hanan, the root of the name Yochanan (John); the word for ``remember'' is zakar, the root of the name Zacharias; and the word for ``sware'' is shaba, the root of the name Elishaba (Elizabeth). Here an allusion to Zacharias and his wife and son is present in Hebrew but lost in translation to Greek or English.

There are also a number of places in the gospels where wordplay or assonance would be present in a Hebrew version. Carmignac compiled a list of over twenty examples from the synoptics. One is from Mark 3:14-15, where Jesus chooses twelve disciples

``that they should be with him, and that he might send them forth to preach, and to have power to heal sicknesses, and to cast out devils.''

In a Hebrew version of this passage, the word for ``send'' would be shalah, the word for ``have power'' would be shalat, and the word for ``cast out'' would be shalak. The fact that these Hebrew verbs are so similar may be a coincidence. It is also possible that in an early Hebrew version of Mark, the verbs were chosen for their assonance.

8. Semitisms of Transmission: In some parallel passages, the synoptics use different Greek words whose Hebrew counterparts are very similar. In cases where one of the Greek words seems more likely than the other, it is tempting to speculate that the discrepancy may have resulted from a slight transcription error in an original Hebrew source. For example, Matt 3:11 has John Baptist saying that he is not worthy to carry the shoes of the one who will come after him, while in Mark 1:7 and Luke 3:16 he states that he is not worthy to unfasten them. It turns out that the Hebrew words for ``carry'' and ``unfasten'' are very similar.

9. Semitisms of Translation: Sometimes two permissible Greek translations of the same Hebrew word appear in parallel passages in the synoptics. For example, Mark 5:29 speaks of a ``source'' of blood, while Luke 8:44 mentions a ``flow'' of blood. It turns out that the Greek words used in these two verses are also both used in the Septuagint to translate two appearances of the same Hebrew word in Lev. 20:18.

Another example involves Mark 4:19, which mentions ``lusts of other things,'' and Luke 8:14, which speaks of ``pleasures of this life.'' Carmignac points out that the Hebrew words for ``other things'' and ``the flesh'' have identical consonants, so that different vocalizations could lead to either ``lusts of other things'' or the more likely ``lusts of the flesh.''

Carmignac also notes (The Birth of the Synoptics, p. 34) that the documents at Qumran may shed light on a puzzling phrase found in Mark 9:49: ``For every one shall be salted with fire...'' Research on the Dead Sea Scrolls has uncovered the fact that the Hebrew word for ``salted'' can also mean ``vaporized.'' It is possible that the latter was intended in Mark 9:49.

Carmignac observed that the existence of the first six categories of Semitisms does not prove that the Greek synoptic gospels include material translated from Semitic sources. Semitisms of these types could also appear if native speakers of a Semitic language had originally composed the gospels in Greek. However, examples of categories seven through nine must either be coincidences or evidence for the presence of an underlying Semitic (probably Hebrew) text. Carmignac believed that there were too many possible instances of the last three types for all of them to be brushed aside as coincidences.

Carmignac's Model

Twenty years of research on the synoptics led Jean Carmignac to the view that someone, probably the apostle Peter, wrote an early Hebrew gospel, perhaps by 40 A.D. Matthew would have had this text available among his sources when he composed his own Hebrew gospel. Later, in the early 60s A.D., Mark translated Peter's gospel into Greek and made some brief additions (e.g., Mark 1:1-15) to produce the gospel of Mark.

By placing Peter's Hebrew ''protogospel'' earlier than the other gospel accounts, this model is consistent with modern scholarship that asserts Markan priority. On the other hand, it also agrees with Christian tradition by interpreting patristic statements about Matthew preceding Mark as references to the final Greek text of Mark. Such an interpretation is very reasonable, since early writings by Papias and Irenaeus describe Mark as Peter's ``interpreter'' or ``translator.''

This model differs somewhat from that of Robert Lindsey, who believed that Luke's gospel came first. However, in many ways Carmignac and Lindsey were kindred spirits. In particular, both strongly believed in the existence and importance of Hebrew substrata of the Greek synoptic gospels.

Carmignac had planned an extensive research program to further test his hypotheses. Unfortunately, his untimely death prevented him from carrying it out. It is hoped that a new generation of researchers will continue to explore the trail blazed by this pioneering scholar.


1See the article ``Who Wrote the Four Canonical Gospels? The Earliest Christian Traditions'' in Issue 7 of Grace & Knowledge.

2This is a possible reference to 2 Cor. 8:18.

3Compare Carmignac's impressions with those of Dr. Robert L. Lindsey described in the article, ``Which Language did Jesus Speak-Aramaic or Hebrew?'' in Issue 7 of Grace & Knowledge.

Issue 14


File translated from TEX by TTH, version 3.01.On 15 Jan 2003, 17:07.