Identifying the Reference Sources of αὐτὸν ἦν in Acts 22:29

Acts 22:29 Alexandrian text

εὐθέως οὖν ἀπέστησαν ἀπ’ αὐτοῦ

οἱ μέλλοντες αὐτὸν ἀνετάζειν·

καὶ ὁ χιλίαρχος δὲ ἐφοβήθη

ἐπιγνοὺς

ὅτι Ῥωμαῖός ἐστιν

καὶ ὅτι αὐτὸν ἦν δεδεκώς.

Therefore those who were about to interrogate him

stepped back from him right away,

and the commander was afraid

when he found out

that he (Paul) was a Roman citizen

and that he had bound him.

Part of the Greek New Testament (GNT) grammarian’s job is to identify the reference sources of finite verbs and infinitives (known as the subject) and of substantival pronouns, like direct object τοῦτο or genitive ἐκείνου. About 8% of GNT word occurrences are substantival pronouns. (Adjectival pronouns do not need references.) A little more are finite or infinite verbs with no explicit subjects. Transparency requires the syntax diagrammer to identify the rules by which he makes interpretative decisions.

The first step for the exegete or diagrammer in examining a GNT sentence is to look for textual variants. It turns out that there is a transposition. The Byzantine text is ἦν αὐτὸν instead of αὐτὸν ἦν (still with unchanged meaning, he had bound him). Normally a transposition of parsed words is not critical because the connections are determined mainly by declension, case, conjugation and other lexical signals. So the impact of the transposition is unremarkable in itself here because the meaning of the 3 Koine words he had had him bound is evident from the narrative.

ἦν δεδεκώς he had bound is a perfect active unitary paraphrastic verb pair. Since the two words αὐτὸν ἦν both need to have the reference source identified, the first (general) rule is that the first word in the ἦν δεδεκώς clause copies its reference from the immediately preceding coordinate ἐστιν clause, namely Paul. Since the first word is αὐτὸν in the Alexandrian text, the default assumption is that αὐτὸν refers to Paul because it carries over from the pervious coordinated subordinator clause. Since there are sometimes exceptions to the first, general, rule, we look at the next 2 words and see that they naturally refer to the commander, not to Paul. This confirms that the default rule 1 was accurate and αὐτὸν does indeed refer to Paul.

However, in the Byzantine text the first word of the ἦν δεδεκώς clause is ἦν, meaning that the default rule for coordinate clauses points to Paul as the default subject, not the object. However, when we look at the following word αὐτὸν, we see that that cannot very well refer to the commander because the commander is the subject of binding, not the object. So we get a semantic correction of the original, default hypothesis. So in the Byzantine, we are forced to override the default assumption with a semantic exception to the general rule.

What are the lessons of these observations? First, we see that word order may have an impact on how syntax, diagramming and antecedent identification is conducted, contrary to what the GNT students are sometimes taught about GNT word order. Second, the Byzantine reading is the lexio difficilior, the more difficult reading. It is the Alexandrian scribes, not the Byzantine scribes, who have simplified the original, contrary to the claims of the critical editors.

Apparently the real selection rule for the critical text is that we choose the lexio difficilior if it is Alexandrian, but otherwise we choose the lexio simplicior. In other words, the UBS and NA critical text is an Alexandrian text based largely on Vaticanus and Sinaiticus much as Westcott and Hort published it in 1881, in the same way that the Majority Text of Robinson is a Byzantine text. Critical selection does not prove closer to the autograph because the selection criteria are not axiomatically true and are not rigorously followed anyway.

—Dennis Kenaga, October 2026