St. Matthew used about 1475 words, 137 of which are apax legomena (words used by him alone of all the New Testament writers). Of these latter 76 are classical; 21 are found in the Septuagint; 15 (battologein biastes, eunouchizein etc.) were introduced for the first time by Matthew, or at least he was the first writer in whom they were discovered; 8 words (aphedon, gamizein, etc.) were employed for the first time by Matthew and Mark, and 15 others (ekchunesthai, epiousios, etc.) by Matthew and another New Testament writer. It is probable that, at the time of the Evangelist, all these words were in current use. Matthew’s Gospel contains many peculiar expressions which help to give decided colour to his style. Thus, he employs thirty-four times the expression basileia ton ouranon; this is never found in Mark and Luke, who, in parallel passages, replace it by basileia tou theou, which also occurs four times in Matthew. We must likewise note the expressions: ho pater ho epouranions, ho en tois ouranois, sunteleia tou alonos, sunairein logon, eipein ti kata tinos, mechri tes semeron, poiesai os, osper, en ekeino to kairo, egeiresthai apo, etc. The same terms often recur: tote (90 times), apo tote, kai idou etc. He adopts the Greek form Ierisiluma for Jerusalem, and not Ierousaleu, which he uses but once. He has a predilection for the preposition apo, using it even when Mark and Luke use ek, and for the expression uios David. Moreover, Matthew is fond of repeating a phrase or a special construction several times within quite a short interval (cf. ii, 1, 13, and 19; iv, 12, 18, and v, 2; viii, 2-3 and 28; ix, 26 and 31; xiii, 44, 4.5, and 47, etc.). Quotations from the Old Testament are variously introduced, as: outos, kathos gegraptai, ina, or opos, plerothe to rethen uto Kuriou dia tou prophetou, etc. These peculiarities of language, especially the repetition of the same words and expressions, would indicate that the Greek Gospel was an original rather than a translation, and this is confirmed by the paronomasiæ (battologein, polulogia; kophontai kai ophontai, etc.), which ought not to have been found in the Aramaic, by the employment of the genitive absolute, and, above all, by the linking of clauses through the use of men . . . oe, a construction that is peculiarly Greek. However, let us observe that these various characteristics prove merely that the writer was thoroughly conversant with his language, and that he translated his text rather freely. Besides, these same characteristics are noticeable in Christ’s sayings, as well as in the narratives, and, as these utterances were made in Aramaic, they were consequently translated; thus, the construction men . . . de (except in one instance) and all the examples of paronomasia occur in discourses of Christ. The fact that the genitive absolute is used mainly in the narrative portions, only denotes that the latter were more freely translated; besides, Hebrew possesses an analogous grammatical construction. On the other hand, a fair number of Hebraisms are noticed in Matthew’s Gospel (ouk eginosken auten, omologesei en emoi, el exestin, ti emin kai soi, etc.), which favour the belief that the original was Aramaic. Still, it remains to be proved that these Hebraisms are not colloquial Greek expressions.
newadvent.org/cathen/10057a.htm