QUESTION
Can you give an example how language pattern evidence (i.e., construction) helps decide exegesis and translation?
RESPONSE
Romans 9:5 gives a reason for studying language patterns (constructions) as given in our forthcoming project, Greek New Testament - Grammatical Commentary (GNT-GC).
Is the Justification for the interpretation of Rom 9:5 based on imagination or construction precedence?
Romans 9:5 (ESV)
To them belong the patriarchs, and from their race, according to the flesh, is the Christ, who is God over all, blessed forever. Amen.
ὧν οἱ πατέρες, καὶ ἐξ ὧν ὁ Χριστὸς τὸ κατὰ σάρκα, ὁ ὢν ἐπὶ πάντων Θεὸς εὐλογητὸς εἰς τοὺς αἰῶνας, ἀμήν.
Attributive means an adjectival modifying a substantival head. Most modern translations like the ESV take ὁ ὢν to be an attributive participle modifying Christ, thus giving Rom 9:5 a trinitarian interpretation. It seems pretty straightforward to them. The articular participle is regularly attributive and attributive participles are regularly translated as WHO with a finite verb.
Most modern translates take Θεὸς to be the substantival predicate complement of ὁ ὢν. This also seems pretty straightforward to them. The verb to be regularly takes a substantival predicate complement.
The translators are quite comfortable with this translation: It is orthodox; they have grammatical precedents; they do not stray outside the consensus.
GNT-GC constructions, however, introduce a new, previously unavailable tool to evaluate interpretations. The simple question is: does the proposed construction have precedents or is it unprecedented. An unprecedented interpretation is not necessarily wrong, but if a highly precedented interpretation is available the unprecedented interpretation is very suspect.
The construction proposed by the translators is the substantival predicate complement of the link participle ὢν. There are 23 such participial constructions coded in the quarter of the GNT coded by GNT-GC. They are mostly circumstantial participles, and none is attributive. These constructions can be listed and examined.
Likewise the construction proposed by the translators is the attributive participle ὢν. There are 20 such participial constructions coded in the quarter of the GNT coded by GNT-GC. None of them takes a substantival predicate complement.
Likewise, the coded GNT contains 34 attributive relative clauses with the verb to be that take substantival predicate complements. These constructions can be listed and examined.
So it is clear enough that the attributive participles and participles taking predicate complements are common enough individually, it is also clear that the combination of the two required by the translation/exegesis is unprecedented.
Why would Greek speakers avoid combining two common grammatical constructions into the same grammatical construction? The answer is probably complexity. Although the attributive participle is often translated as a relative clause, both in English and modern Greek, it is only equivalent in simple constructions, not difficult ones, the way the translators are combining them. Interpreters are forced to use their imaginations to translate. However, constructions provide a tool to check imagination against an objective reality that was previously unavailable to translators.
The Greek is actually "Christ being over all God blessed”, not “Christ who is over all God blessed”. English and Greek allow the second phrase. English does not allow the first one because it overburdens the attributive participle; apparently Greek makes the same distinction.
The main alternative to the trinitarian interpretation of Rom 9:5 is elliptical, creating a separate main clause “Blessed be God forever” along the lines of “Blessed is the coming Kingdom of our father David”. These constructions are nominal with an elliptical verb, requiring the verb is to be supplied. This is the non-trinitarian interpretation chosen by GNT-GC because it is better established by precedence.
Romans 9:5 (KJV)
Whose are the fathers, and of whom as concerning the flesh Christ came, who is over all, God blessed for ever. Amen.
The King James translators hesitated to give the modern trinitarian interpretation of Rom 9:5, possibly because they did not find a precedent or possibly because the general lack of explicit trinitarian passages in Paul. However the standard English meaning of God blessed is “blessed by God” which is not what the Greek says.
One of the 20 attributive participial constructions of to be, does take an adjectival predicate complement (not a substantival predicate complement): 2 Cor 11:31, and it is quite similar to Rom 9:5 in some ways.
2 Corinthians 11:31
The God and Father of our Lord Jesus, who is blessed for ever, knows that I do not lie.
It may be studied to see if it might form a looser precedence for the modern trinitarian interpretation of Rom 9:5. GNT-GC does not claim that precedence studies creates definitive answers to cruxes, only that it is objective evidence supporting options that ranks higher than grammatical imagination concerning possibilities.
GNT-GC contains verse notes for some cruxes. The GNT-GC notes use language pattern precedence as partial criteria for selecting preferred interpretation options. Head-depedent and grammatical relations diagramming forces choice and eliminates the possibility of ambiguity inherent in translations and commentaries in general. Note that constructions depend on precise grammatical descriptions of the constructions that the translators select to support the translations. Commentaries seldom supply such exact grammatical descriptions. The GNT-GC objective constraints tightens up such arguments.