I can probably make a series out of "Not Always a Citation," and I may or may not do so. Today, I'm somewhat pressed for time, so I want to cite one example.
In Biblical Interpretation in Ancient Israel (New York: Oxford, 1988), Michael Fishbane says that a biblical author referring to a tradition may not be a direct citation of a specific chapter. He states: "Even in the best of circumstances, as the apparent reuse by Amos of the patriarchal tradition regarding the destruction of Sodom, one may validly claim that Amos drew independently from a fluid of traditions quite distinct from that found in Genesis 19" (8).
Fishbane is saying that Amos may not have had Genesis 19 in mind when he referred to the destruction of Sodom, for such a motif was probably floating around in his culture. And he may have a point there. The Hebrew Bible contains a variety of traditions about Sodom and Gomorrah. Deuteronomy 29:23 refers to the destruction of two more cities, Admah and Zeboiim, which also crops up in Hosea 8:11.
What was the sin for which Sodom and Gomorrah were destroyed? Genesis 19 depicts an attempt at homosexual gang rape, whereas Ezekiel 16:49 states that Sodom "had pride, excess of food, and prosperous ease, but did not aid the poor and needy" (NRSV). One cannot claim with certainty that every biblical reference to Sodom has Genesis 19 in mind, since there are traditions that go beyond Genesis 19. Genesis 19 was probably one expression among many Sodom and Gomorrah traditions.
One critique of Fishbane that I've heard is that he doesn't take into account cultural repertory. Rather, he assumes that biblical authors are interpreting specific biblical passages, when they may have been responding to traditions floating around in their culture. This critique relates to the influence vs. intertextuality debate. At one point (and even today in many circles), scholars analyzing an author tried to discover the sources that influenced him or her. Today, many prefer the intertextuality model, which holds that there are ideas that loosely circulate in culture, which means that an author may not have a specific author in mind when he says something that resembles another work. The intertextuality model is not as linear as the influence one. I may have to dig up an old paper to find citations about this debate!
But we see that Fishbane does not buy into the "influence" model in every case. So how does he identify a citation? I'm not in the mood to look up every reference on this right now, but two relevant factors are similarities in vocabulary between two passages, and whether one passage attempts to clarify another in some way.