Back to List of Chapters



Here we will look at the linguistic evidence for Matthew’s Gospel being written first in Hebrew.

Eusebius quoted Papias as stating that Mark was very accurate when recording the speech of Peter, but: “Matthew collected/composed the oracles [logia or sayings] in the Hebrew language/style; but each recorded them as he was able.” ((EH 3: 39.16 and RO 166r)).

Modern research shows that at that time Greek shorthand was in use, but the Jews had to write fast, and use abbreviations, when recording Hebrew speech. The words of Papias imply that when Matthew’s Gospel was publicly read, listeners had difficulty making notes because it was in Hebrew.

Quoting Irenaeus Eusebius wrote: “Now Matthew published among the Hebrews a written gospel also in their own tongue/ dialect”. ((EH. 5: 8, 2)).  Kirsopp Lake in a footnote to his translation explains that the word ‘also’ indicates that Matthew’s Gospel had an earlier spoken form. The only reason Markans have for rejecting the evidence of Eusebius is because it challenges their theory. Without evidence, most Markans have asserted that it is not possible to back-translate Matthew’s Greek Gospel into Hebrew. The Pontifical Biblical Commission in 1912 stated that this assertion did not undermine the words of Eusebius.

In recent years two Frenchmen have thrown more light on this question. Jean Carmignac, a specialist in Greek and Hebrew, spent eleven years from 1943 translating ‘The Book of Chronicles’. Then from 1954-63 he was engaged in translating ‘The Dead Sea Scrolls’ from Qumran. Although his work was mainly concerned with the Old Testament, he noticed connections with the New.

In 1963 he attempted to translate one of the Greek Gospels back into the form of Hebrew used at Qumran. Carmignac was ‘absolutely dumbfounded’ to find how easy it proved to be. Realising his translation would meet with ferocious criticism, he searched in the old monastic libraries of Europe to see if the translation had been attempted previously. At the same time he hoped to correct and improve his own work. Although lacking the time to make a thorough search, he soon found sixty translations of Gospels or portions of them. Some were by Rabbis who had become Christians and others by those wishing to dispute with Christians.

Carmignac was aging and had not published his scholarly findings. So his friends persuaded him to write a small book for the general public. This appeared in 1987. He not only asserted that Matthew’s Gospel was the first to be written, but that it was in Hebrew.

According to him, the order of composition was a Hebrew Matthew followed by a Greek Luke, and then by a Greek Mark. In order to conform to Markan priority, he presumed there must have been an earlier Hebrew version of Mark, which Mark later translated. If we ignore this presumption, his work dovetails with the Clementine tradition of Matthew-Luke-Mark. He dated Greek Mark as before 70 AD, so Matthew and Luke at an earlier date.

In 1953 Claude Tresmontant, a French Hebrew scholar, published a book dedicated to Hebrew thought. He then worked for years to produce a Hebrew-Greek dictionary. As he did this he was overwhelmed by how easy it was to back-translate the Gospels into Hebrew.

In his 1980 ‘Le Christ Hebreu’ [In English -1989], he explained in detail the Hebrew basis of the Gospels. He formed the opinion that Matthew in both Hebrew and Greek could be dated as being written soon after the Resurrection. Luke between 40-60 AD, and Mark 50-60 AD.             ((CTH 324)).While not confirming the Clementine order, this pointed to its likely possibility.

Tresmontant detected a Hebrew substructure to Mark’s Gospel. At first sight this could conflict with the normally accepted view that Mark recorded Peter speaking in Greek. But, if over 90% of the talks by Peter consisted of quotes from Matthew and Luke, the Hebrew substructure of these quotations would automatically show in Mark’s Gospel. The additions made by Peter would also have a strong Semitic tone because Peter was a native Aramaic speaker.

J. Kurzinger has shown that the description given by Papias, of Matthew’s Gospel, would be best rendered as ‘style’. ((RO 128-9)). This would be a fair description of a Greek document written with a Hebrew substructure. When discussing Semitisms in the Gospels many could be explained as a Greek author borrowing familiar Hebrew words. It would have been easy for the translator to carry over some Hebrew words. But the Hebrew sub-structure pointed out by Carmagnac and Tresmontant is of a different kind.

Hebrew likes wordplay and takes great pleasure in using similar sounds to assist memorisation. In Matthew 3: 9 we read of ‘stones’ and ‘children’. In Greek and English there is no linguistic connection to assist memorisation. But in Hebrew it reads as ‘abanim’, and ‘banim’ ((CTH 64)). In Matthew 9:16 the tear (qera) becomes worse (ra).  In Matthew 13: 6 shemesh (sun) is linked phonetically to the word for root (shoresh). In Matthew 21:12, shulehanot (tables) is similar to shulehanim (changers). ((JC 29)).

Hebrew Matthew in 26: 38 probably used the words ‘imdu’ (stay) and ‘immadi’ (with me). The word ‘immadi’ was translated into Greek, and therefore into English versions. This is acceptable in English but superfluous in the Greek language ((JC 30)). Peter, reading a Greek copy of Matthew and speaking in Greek, omits it (Mark 14: 34). We may ask Markans why Matthew, if writing in Greek at the end of the first century, adds a superfluous Greek word.

Translators and copyists are liable to make small errors, and these show when we compare the Gospels. It is necessary to point out here that the use of small dots or dashes in Hebrew are not aids to pronunciation, as they are in some languages.  In Hebrew they signified completely different letters. Many apparent discrepancies between the Gospels may be explained if we accept that Matthew’s Gospel was originally in Hebrew. In Matthew 22: 19 we read ‘show me a coin’, while Mark 12:15 has ‘bring me a denarius’. In Hebrew a coin is HBW and a denarius is HRW.

In Hebrew a small difference distinguishes B and R from one another, so could easily have been misread when being translated into Greek ((JC 32)). Matthew 13:17 has, ‘the just’ or ‘righteous men’, but Luke 10: 24 has ‘kings’. As ‘the just’ corresponds to WYSRYM and ‘kings’ to WSRYM, and as the symbols for W and Y are very close, we are able to see how easy it would have been to read W in place of WY.  Misreading in Greek the word ‘king’ for ‘just’ would be unimaginable ((JC 33)).

Some Hebrew theological concepts present in Matthew’s Gospel would be alien to Greek thought. Here we may note some of the more common Hebrew words and expressions used by Matthew but not by Greeks: Beelzebub, Woe, Flesh and Blood, deliver into the hands, hardness of heart, to set one’s face. Why would a Greek use these if not translating from Hebrew? They are not used by Luke or Mark ((CTH 67-71, 90-92 and 112)).

Tresmontant pointed out that a Hebrew-Greek Lexicon had to be produced when the Old Testament was translated into Greek (the Septuagint). He held the opinion that the Christians used the same lexicon in New Testament times. On both occasions, the translators tended to transcribe word for word, even though this could produce a Greek which did not flow well.

So the evidence produced by Carmagnac, Tresmontant, and over sixty Rabbis, is that Matthew’s Gospel was first written in Hebrew. The examples above are a few of those provided by the two French translators. This is in accord with the historical records of Papias and Irenaeus. They say a Hebrew version for the Jews of Palestine, was composed first. A Greek version appeared when converts were made amongst those who spoke Greek only.

These Frenchmen agreed the Synoptic Gospels were written pre-70 AD, but not take a stand on the sequence in which they were composed. But it is interesting that Tresmontant refers to them in the order of Matthew-Luke-Mark four times on one page of his writings ((CTG 14)). As they showed that Papias and Ireneaus were correct to report that Matthew’s Gospel was first written in Hebrew, confidence in the other information these historians provide, is greatly increased. In 47 AD Western sailors learnt how to utilise the monsoons to sail from Egypt to southern India in less than 100 days. Indian records show that Thomas the Apostle landed at Malankara, Kerala in 52 AD ((SGP 5)).After establishing seven churches he moved to Coramandel on the east coast where he was martyred in 68 AD.

Irenaeus records that Pantaenus, director of the Alexandrian School of Sacred Learning from 180-192, paid a visit to India in 190 and was shown a copy of the Gospel of Matthew in Hebrew characters. Pantaenus reported that it was the Apostle Bartholomew who had taken it to India.            ((EH 5: 10)). Thomas was called ‘Mar Toma’ [Lord or Bishop Thomas] so Pantaenus, who knew Hebrew but not the Indian languages, may have misheard this as ‘Bar Tolmai’, the Hebrew name for Bartholomew ((SGP 19)). Whether or not this copy of the Matthew’s Gospel had belonged to Thomas or Bartholomew, Panteanus had seen it in Hebrew. Yet Markans say it was written for a Greek community a hundred or more years after Thomas and Bartholomew were both dead.

In his ‘De Viris Illustribus’, Jerome states that Matthew wrote in Hebrew letters and words for the sake of the Jews and it was translated into Greek. He writes that: “the Hebrew itself is preserved even now in the library at Caesarea …” Jerome also says he: “was given the opportunity of transcribing this volume by the Nazarenes who use it in Beroea, a city of Syria”. He adds that Matthew, when quoting from the Old Testament, had used the Hebrew Scriptures not the Greek Septuagint ((RO 203 & DVI, ch 3&7)).

Doubt, regarding the use of Hebrew, rests on an assertion in 1555 by  J. A. von Widmanstadt that Hebrew was a dead language  at the time of Jesus  ((JC 83)).  Yet Luke, in  Acts 21: 40 and 22: 2 reports Paul using it. Its use has since been confirmed by ‘The Dead Sea Scrolls’ ((JC 65)).


V: 13/2/13

Back to List of Chapters