In his article, "The Fourteen Triads of the Sermon on the Mount (Matthew 5:21–7:12)," Glen Stassen makes a fairly standard claim about alms-giving, prayer, and fasting in late-Second Temple era Judaism.1 He writes,
Almsgiving, prayer, and fasting went together as the three traditional Jewish practices of righteousness in the first century. "The three disciplines were almost certainly traditionally associated with one another. . . ." Clearly these are "traditional righteousness, as expected. (284; citing Davies and Allison, Matthew, 1:575; Weder, "Rede der Reden," 158; Davies, Setting, 305–15)
To be clear: I don't have any issue with Stassen's claim; he, like the scholars he cites, are almost certainly right. But notice what Stassen has done (rather that what he has claimed). Stassen has made a claim about how Jews in the ancient world perceived the practices of alms-giving, prayer, and fasting, and then he has cited modern historians in support of that claim. This, I suggest, is the problem.
How could Stassen have been more precise and helpful in his use of documentation? I think either of two options would have been preferable. First, Stassen could have made the claim that his documentation supported. That is, he could have said something along the lines of, "Scholars typically understand Jesus' teaching on these three practices together, which suggests that they see them as largely overlapping expressions of piety." If Stassen had cited Davies and Allison, Weder, and Davies in support of this claim, his citation would have provided examples of the very thing he was discussing.
Alternatively, Stassen could have provided documentation for the ancient phenomenon in question (viz., the association of alms-giving, prayer, and fasting together). That is, if these three practices did go hand-in-hand in Second Temple Judaic piety, then presumably we have texts that exhibit the connection between them. Certainly Matt. 6.1–18 might be one of these texts. But I would want to see other texts, whether from pseudepigraphical and/or apocryphal texts, the DSS, Josephus, Philo, etc., that make this connection and make it more explicitly than Matt. 6.1–18. These texts—and not claims made by Davies and Allison, Weder, and Davies—support the claim Stassen is making. Now, I assume that the secondary literature Stassen does cite provide references to the primary sources, but that actually exacerbates the problem rather than eases it. If other scholars have already gathered the primary evidence, then it should have been very easy for Stassen to provide references to that evidence.
Instead—and this is the point I want to make—the pattern of (a) making a claim about the ancient world and (b) supporting that claim with a reference or references to secondary sources enables us to root our historical claims in scholarly consensus rather than concrete historical data. Even when we're making fairly straightforward claims that probably are rather indicative of the ancient world (as is Stassen in this example), we need to get in the habit of more rigorously distinguishing the historical world(s) we have reconstructed in the critical mass of historical scholarship, on the one hand, from the actually historical data upon which that scholarship depends (and beyond which it extends).
1 Glen H. Stassen, "The Fourteen Triads of the Sermon on the Mount (Matthew 5:21–7:12)," JBL 122/2 (2003): 267–308.