Anyone who has investigated a proposed connection between two or more texts has likely encountered familiar objections: coincidence, a shared cultural tradition, or language too common to establish any direct relationship. The best-known response is a set of seven tests used for assessing echoes of the Hebrew Bible and its Greek Septuagint translation in the letters of Paul. These tests have since been applied more broadly across the New Testament, the Hebrew Bible, and ancient Near Eastern literature. The tests pose a series of straightforward questions: Was the proposed source available to the author? How much of its language is repeated? How frequently is the same passage used elsewhere? How well does the echo fit its new context? They also ask whether the author and audience could plausibly have recognized it, whether earlier interpreters identified it, and whether the resulting reading is coherent.[1]
Spelled out like this, these rules can look like a rigid scorecard, and they are often used that way. That was not their intended purpose, however, since certainty in interpretation is practically impossible. Interpretation is as much an art as a science, so these processes are better understood as filters than as a procedure that guarantees a result.[2] Much of the debate over the past three decades has therefore centered on how much weight to give each criterion.
Criteria as Guides to Interpretation
Methods for detecting intertextuality are often compared to musical analysis because they depend as much on judgment and sensibility as on objective method.[3] Even one of the most widely used tests comes with an explicit warning: asking whether earlier interpreters heard the same echo is unreliable by itself and should rarely outweigh other evidence that makes the echo more plausible.
Consider the proposed echo of Job 13:16 in Philippians 1:19. The criteria produce a mixed result: strong on availability and fit within the argument, moderate on repeated language, and weak on recurrence. Paul may plausibly have intended the reference, his Philippian audience may not have recognized it, and no ancient or modern commentator appears to have read it that way. The result can therefore be stated only as uncertain. Applying the criteria to particular texts always leaves unanswered questions, and in some cases the tests fail altogether.[4]
Shared Language as Evidence of Borrowing
Among the criteria, repeated language is comparatively reliable because shared language can be identified and counted. Extensive verbal overlap makes borrowing relatively easy to recognize, as in Jonah's complaint, which recalls the description of God given to Moses at Sinai.
Such clear examples can be misleading, however, because there is no threshold a proposed connection must cross. Jonah's reference would be no more allusive had it also included the expression for faithfulness from Exodus, nor would it be less allusive had it omitted one of the elements it preserves. An allusive marker may consist of a single word or even a name. The theory of allusion underlying the criteria therefore requires neither a particular length nor exact verbal correspondence.[5]
Much of the resistance to treating repeated language as decisive comes from the study of Mesopotamian and Ugaritic literature, where allusions may be quite subtle. Requiring substantial overlap is useful for excluding spurious parallels, but the principle loses value when treated as an inflexible rule. One Neo-Assyrian oracle asks whose wing has not been clipped, apparently alluding to the story in which Adapa broke the wing of the south wind. Beyond that single image, the two texts share little obvious similarity.[6]Want to learn more about this? Allusions
Dating Texts to Assess Possible Borrowing
The first test asks whether the earlier text was available to the later writer. This is straightforward when a New Testament author quotes the Hebrew Bible or Septuagint, but much harder elsewhere. The limitation is built into the criterion, since analysis of an echo is possible only when the relative dating of the texts is reasonably secure.[7] Within the Hebrew Bible, establishing the priority of one passage over another is rarely possible with confidence, partly because texts that appear early may contain later material. Availability is therefore not necessary for identifying a connection; it becomes necessary only when determining which passage refers to which.[8]
That question has criteria of its own, but they suffer from the same limitations. One survey identified eight arguments used to establish the direction of borrowing and found that none applies universally or carries the same weight in every case. Sometimes both passages fit their contexts equally well, leaving the test for contextual awkwardness with nothing to measure. In other cases, the same criterion can support opposite conclusions about the direction of dependence.[9]
How Assumptions Can Exclude Possible Connections
Esther provides an example of a criterion working against the very question under investigation. Little published research examines Esther as a source used in Second Temple Judaism, so the availability test yields little evidence. Yet the literature is thin partly because Esther is often assumed to have been unknown to early Christianity, which in turn reduces the incentive to investigate its reception. Applied rigidly, the first three tests can therefore rule out proposed echoes of Esther by presupposing the conclusion. The recurrence test faces a connected problem, since recurrence requires a body of material large enough to establish a pattern, something much of the New Testament does not provide.[10]
Competing Sets of Intertextual Criteria
Another problem is that there is no single agreed set of tests. One formulation proposes that only two of the seven criteria, availability and volume, are necessary because the remaining five overlap enough to be dispensable.[11] A competing framework for detecting allusions to sayings of Jesus proposes eleven criteria, including formal agreement and dissimilarity to Graeco-Roman and Jewish traditions, neither of which appears in the original set.[12] Another widely cited set of eight guidelines for inner-biblical allusion focuses entirely on shared language, treating it as the most objective and verifiable evidence available.[13]
Shared language, shared content, and formal resemblance are the three most common tests, but interpreters apply them in very different ways. Nearly everyone treats lexical similarity as the strongest evidence, and some regard it as the only reliable measure. A substantial number of interpreters work from shared content alone, whereas formal resemblance is rarely used.[14] The largest objection, however, concerns not any single criterion but whether Paul's original audience could have recognized the echo. Five of the criteria are better suited to showing that Paul intended an allusion than to showing that anyone actually heard it.[15]
The Same Criteria Produce Different Results
A comparison of three studies of allusion, each claiming precise criteria and each applying the same general method, found that they reached substantially different results.[16] Interpretive competence varies from one reader to another, so one may identify an allusion, another only a faint echo, and a third no reference at all.
The weighing of evidence, and therefore the identification of allusions, has consequently been described as an art rather than a science. Mechanical application of any fixed set of criteria is of limited use because the criteria function as guidelines rather than rules and do not hold equally well in every situation.[17] An interpreter may assemble numerous lexical, thematic, and formal links among several passages and make a strong case for authorial intent. Another reader may still regard the same evidence as coincidental. At most, the criteria establish relative probability: whether a proposed allusion lies within the bounds of historical and literary possibility.[18]
Why Criteria Need to Fit the Texts
Methods of this kind are best treated as provisional guides rather than checklists that guarantee a result.[19] Every existing set of criteria emerged from the particular texts its author was studying, and no single model fits all of Scripture. A survey of the available approaches suggests that common sense may be more useful than increasingly complex sets of criteria.[20]
General criteria are therefore most useful as a starting point for narrower investigations. Deutero-Isaiah, for example, reworks Jeremiah through reversal, the application of an earlier prediction to a new situation, and the reshaping of language, patterns that are uncommon outside Isaiah 35 and 40 through 66. These can function as meaningful criteria, but only within the particular body of texts from which they were derived.[21]