Tyson draws on the criteria devised by Andrew Clark in his Parallel Lives to further his discussion of Peter’s and Paul’s characterization in Acts. (Have been discussing Tyson’s argument that our canonical Luke-Acts was largely a second century response to Marcionism.) Before continuing with notes from Tyson, am listing here Clark’s criteria.
Literary criteria are internal controls for assessing parallels, and Clark establishes these by comparisons with Plutarch’s works comparing famous Greek and Roman figures of the past.
Tyson comments if Acts was composed after 115 c.e. there would be added to the internal controls a powerful external control: Plutarch’s Lives were published after that time, and presumably known to the author of Acts. (Marcion and Luke-Acts, p.63)
The usual caveats apply to Clark’s literary criteria listed here: these cannot be applied mathematically — judgment is always a factor; one cannot expect to find all in any parallel; sometimes they will be weakly present even when passages are obviously meant to be read together (e.g. Paul’s 3 different accounts of his conversion); etc.
1. Similarity in content
Too vague to stand on its own as a criterion of authorial intention for passages to be read in parallel. May complement other similarities.
2. Similarity in language
Lexical repetitions or synonyms. Rare words are more likely to be significant. Consider synonyms, too. Are compound forms forms apparently used as intentional parallels to their original forms?
3. Literary form
May not stand on its own but can complement other similarities. Healings of paralytics by Peter (Acts 3:1-10) and Paul (14:8-10) share a common literary form — both contain information about the place, action of the man, word of healing, gesture of healing, immediate occurrence of healing, demonstration of healing, and effect on the crowd (from Lüdeman, Early Christianity, 53).
Sometimes better to speak of distinct literary motifs in common: example, the double visions in each of the conversions of Saul (9:1-19) and Peter (10:1-48).
The more extensive a sequence is the stronger it is as an indicator of intentional parallelism. Sequences may not always be in the same strict order, however.
Larger parallel structures, even though not always perfectly matched, are another strong indication of an intent to create a double pattern. Examples: Talbert’s 32 parallels of content and sequence between the Gospel of Luke and Acts; between Acts 1-12 and Acts 13-28. The parallel structures suggest an intention to highlight a theme of continuity between Jesus and his disciples, and between the apostles and Paul.
Another complementary criterion that carries weight when in conjunction with other criteria. Perhaps also an essential criterion.
Also note: Disruption of the text
If the flow of the text is disrupted, or if a pericope is awkward internally, where a parallel appears, this is a strong indicator that the parallel was an important feature in the author’s mind.
From Clark’s Parallel Lives, pp.73-80.