The term ‘Midrash’ is used both for the process of exegesis itself and for the various collections of rabbinic materials. It is conventional to divide rabbinic Midrash into two types: midrash halakhah (legal midrash)—for example, Mekhilta, Sifra, and Sifre—and midrash aggadah (non-legal midrash)—for example, Genesis or Exodus Rabbah. However, all these collections contain materials of both sorts. Some collections offer a running commentary on the biblical text, or sections of it, whereas others, described as ‘homiletic’, are built around a variety of texts and are related to festival days, possibly representing the material used in sermons.
The approach of Midrash is illustrated by Jay M. Harris, focusing on Deut. 14: 16: ‘Fathers shall not be put to death for sons, nor sons for fathers; a man shall be put to death for his own sin.’ He writes:
A modem reader…will have no difficulty understanding the simple meaning (peshat) of this verse. It means that each individual will be held accountable for his or her own actions, and that only he or she may suffer the ultimate punishment, if appropriate. Guilt is not hereditary, nor is it visited upon ancestors.…None of these readers is likely to be overly troubled by the verse's verbosity and redundancy.…
Those with some legal training, aware of the importance of statutory construction, may, however, be troubled by these features. They may seek to tease more meaning from the verse or they may simply attribute the verbosity to different legal standards, or perhaps, to sloppy construction.…But what if the option of sloppy construction were not available? What if one took for granted that the author was incapable of sloppy construction, indeed, incapable of less than perfect construction? One would then be forced to find in each of the clauses a distinct statement that eliminates the redundancy, and, indeed, that is precisely what the darshan (the author of a piece of Midrash)…does with this verse. (Harris 1995: 8)
Harris's approach focuses on the triggering factor within a scriptural verse that leads to the need to interpret. Jacob Neusner, examining the broader purpose, notes that Midrash is
how ‘our Sages of blessed memory’ wrote with Scripture.…For the Holy Scriptures were transformed by the Judaic sages. Thus Midrash works in three dimensions: first, as explanation of meaning imputed to particular verses of Scripture; second, as a mode of stating important propositions, syllogisms of thought, in conversation with verses or sustained passages of Scripture; and, third, as a way of retelling scriptural stories in such a way as to impart to those stories new immediacy. (Neusner 1995: 94)
Because of the variety of ways in which the term ‘Midrash’ is applied, including to virtually any rewriting of biblical material today, a helpful definition is furnished by Gary Porton: ‘Midrash is a type of literature, oral or written, which has its starting point in a fixed canonical text, considered the revealed word of God by the midrashist and his audience, and in which this original verse is explicitly cited or clearly alluded to’ (Porton 1992: 819). Neusner notes that what are important in Porton's definition are three elements: (1) exegesis (2) starting with Scripture and (3) ending in community (Neusner 1995: 96).
As indicated above, scholars are divided on a number of issues regarding how to read Midrash. Daniel Boyarin points to one of the problems:
I wish to discredit the opposition between reading which is value-free and concerned with the difficulties of the biblical text and that which is unconcerned with those difficulties and speaks to the needs of the moment. It is clear then that I am not denying the reality of ideological concerns on the part of the rabbis nor that these ideological concerns may have often had an effect on the interpretive choices they made. I am asserting that we will not read midrash well and richly unless we understand it first and foremost as reading, as hermeneutic, as generated by the interaction of rabbinic readers with a heterogeneous and difficult text, which was for them both normative and divine in origin. (Boyarin 1990: 5)
Boyarin bases his approach on inter-textuality, which for him has three important dimensions:
[T]he first is that the text is always made up of a mosaic of conscious and unconscious citation of earlier discourse. The second is that texts may be dialogical in nature—contesting their own assertions as an essential part of the structure of their discourse—and that the Bible is a pre-eminent example of such a text. The third is that there are cultural codes, again either conscious or unconscious, which both constrain and allow the production (not creation) of new texts within the culture. (Boyarin 1990: 12)
It is helpful to cite Boyarin further as the argument regarding the nature of midrashic reading offers an approach to the interpretative methods and schools of later periods.
Rather than seeing midrashic departures from what appears to the ‘simple’ meaning of the local text as being determined by the needs of rhetoric and propaganda and rooted in the extratextual reality of the rabbinic period, or as being the product of the creative genius of individual rabbis wholly above time and social circumstance, I suggest that the intertextual reading practice of the midrash is a development (sometimes, to be sure, a baroque development) of the intratextual interpretive strategies which the Bible itself manifests. Moreover, the very fractured and unsystematic surface of the biblical text is an encoding of its own intertextuality, and it is precisely this which the midrash interprets. The dialogue and dialectic of the midrashic rabbis will be understood as readings of the dialogue and dialectic of the biblical text. (Boyarin 1990: 15)