Robert L. Mowery, «Son of God in Roman Imperial Titles and Matthew», Vol. 83 (2002) 100-110

The christological formula qeou= ui(o/j, which appears in the NT only in three Matthean passages (14,33; 27,43.54), exactly parallels the two-word Roman imperial son of god formula found in the titulature of Augustus, Tiberius, Nero, Titus, and Domitian. This formula occurred more widely in first century imperial titulature than has previously been reported; in addition, various three-word imperial son of god formulas also deserve notice. The Matthean formula qeou= ui(o/j would have evoked Roman imperial usage for at least some members of Matthew’s community.

disciples in the boat worshiped him saying, ‘Truly you are God’s Son (qeou= ui(o/j)’ (14,33). These words have no synoptic parallels; though the final six words of 14,32 repeat words found in the Markan parallel (6,51), the twelve words in Matt 14,33, including this christological formula, occur only in the First Gospel. The second occurrence of this formula is in the crucifixion narrative in the religious leaders’ assertion that Jesus had said, ‘I am God’s Son (qeou= ei)mi ui(o/j)’ (27,43). This statement occurs only in Matthew; it is absent from the parallel synoptic accounts⁷⁰. The third occurrence of qeou= ui(o/j is in the words ‘Truly this man was God’s Son (qeou= ui(o/j)’ spoken by the centurion and those who were with him (27,54). Matthew produced the wording of this christological formula by inverting the word order of the phrase ui(o_j qeou= found in Mark 15,39.

        While all three of the Matthean occurrences of qeou= ui(o/j appear to be redactional, this formula does not possess typical Matthean linguistic features. The three occurrences of this formula represent (1) three of the four Matthean phrases which have only a prepositive anarthrous genitive noun and an anarthrous governing noun⁷¹, (2) three of the four Matthean phrases which place a dependent genitive noun ahead of the governing noun ui(o/j⁷², and (3) all three of the Matthean phrases which have a prepositive dependent genitive qeou=⁷³. How can it be explained that this presumably redactional formula possesses such atypical linguistic features? Matthew must have been mimicking an existing formula, either a formula like the imperial son of god formula or a christological formula that was already in use within his community⁷⁴.

        Scholars who have examined the relationship between early Christian Son of God formulas and Roman imperial usage have often focused on the Markan phrase ui(o_j qeou=, especially its use in the centurion’s ‘confession’ in Mark 15,39⁷⁵. Yarbro Collins, for example, has argued that this Markan phrase would have evoked the imperial cult for at least some members of Mark’s audience⁷⁶. One objection to her conclusion, though not a fatal objection, is that first century imperial son of god formulas normally had the word order qeou= ui(o/j rather than the inverted word order found in Mark 15,39⁷⁷. Unlike this Markan formula, the Matthean qeou= ui(o/j exactly parallels the two-word son of god formulas found in the Greek titulature of Augustus, Tiberius, Nero, Titus, and Domitian. In addition, this Matthean formula has the same opening word (qeou=) and the same closing word (ui(o/j)