In previous articles, I have shown that the three columns present independent versions (see Three (in)dependent versions of one text). However, at several places we can observe harmonizations between the three versions. That was the case in Gal 1:21 where mention of Cilicia is absent in the three columns (see Where did Cilicia go? About Gal 1:21).
The case of Rm 4:25 (folio 133r) seems to present a similar phenomenon.
In Greek, we read (normalized version):
ὃς καὶ παρεδόθη διὰ τὴν δικαίωσιν ἡμῶν
In regard to the known Greek text, we would have expected:
ὃς παρεδόθη διὰ τὰ παραπτώματα ἡμῶν καὶ ἠγέρθη διὰ τὴν δικαίωσιν ἡμῶν
In the Latin column, there is:
qui traditus est propter iustificationem nostram
Same here, in the Vulgate text, we have:
Finally, in the Arabic column, we read:
الذي تُل به لاجل تُبرّرنا
We are not aware (unlike for Greek and Latin) of the text that the scribe of Marciana Gr. Z. 11 (379) used as model for his copy but with the help of Madrid BN 4971, we can suggest:
الذِي تل بِهِ لاجل ذنوبِنا وقام حيا لاجل تُبرّرنا
It is clear here that we face a case of haplography (or homoioartcon) where the scribe copying the text goes from one word in the text to a similar word, skipping what is in between. This is due here to the prepositions διὰ, propter or لاجل.
However, it is very unlikely that the same error occurs three times at the same place! More plausible is here that the error made in one of the column led the scribe to adapt the text in the two other columns. Now in which language did the error occur first? Probably in Greek as we can assume that the Greek column was the first to be copied. In fact, the Greek column is always very regular till the end of the page while the Latin and the Arabic columns are uneven. See the example in folio 128r:
Folio 128r: disparate line widths
This one case of scribal intervention in the three columns allows us to postulate elements about the copy of the manuscript. To our opinion, this shows that one scribe copied the three columns: it is difficult to imagine one scribe asking another to omit part of the Scriptures to cover up his mistake. Consequently, it can be assumed that not only the object manuscript was trilingual but also the persons who made it.
Feel free to comment this article if you have any remark!
 In Madrid BN 4971, we have الَّذِي تُلَّ بِهِ لأجْلِ ذُنُوبِنَا وقَامَ حياً لِي يُعِيدُنَا صِالحِينَ. It is not clear why we have لِي here, more intelligible would be لِ.