Tae Hun Kim, «The Anarthrous ui(o\j qeou= in Mark 15,39 and the Roman Imperial Cult», Vol. 79 (1998) 221-241
This article points up evidence by which the language of the Roman imperial cult might help make clearer what a reader of Mark's Gospel might understand when the centurion (Mark 15,39) refers to Jesus as ui(o\j qeou=. Knowing how an audience familiar with this cult language would react, Mark intentionally speaks of Jesus as ui(o\j qeou= at 1,1, as well as at 15,39.
authenticity of the centurions confession, because not only does he deny its validity as a full Christian confession but also because he questions the necessity of having such a confession at the conclusion of the Gospel. Johnson argues that, "a Roman soldier of a centurions rank and experience would be too sophisticated and would have been exposed to too many gods to make that kind of quick judgment at an execution, and Marks readers would have known it"10.
Although Johnsons argument on the cultural and historical background of the Roman centurions in general appears to be thorough, how each individual centurion would have reacted to Jesus crucifixion cannot be generalized by what the Roman centurions as a group would have done. The Gospels and Acts list a few Roman centurions who were receptive to the Gospel and the Jewish faith (Matt 8,10; Luke 7,9) and even one who himself became a Christian (Acts 10,1-48). Further, the image of a Roman centurion that Marks readers were likely to have should not compromise its claim to historical authenticity, because it is improbable that Mark would invent a saying that he knew none of his readers would find believable. It is also possible that the miracles and signs that occurred at the moment of Jesus death convinced the generally superstitious Romans and prompted the centurion to acknowledge Jesus divinity.
It also seems highly unlikely that the definite article in 15,39 is lost due to authorial negligence or scribal error. Other Markan references to "the Son of God" (1,11; 3,11; Gods "My Son" and "the Son of the Blessed" 1,11; 5,7; 9,7; 14,62) all have the definite article except 1,111. The context of the confession would have been clearer had the Roman centurion spoken in Latin instead of Greek and the Gospel had preserved the original Latin. If this was the case, the association of the confession with the Roman imperial cult, or the lack thereof, would have been obvious, because the Latin ephitet divi filius would immediately direct ones attention to the Roman imperial cult. Latin, of course, does not have a the definite article.