The alleged person or document from which a text derived information. Any legend or theory about the source of an anonymous work is a hypothesis that has to be tested by the contents of the work itself.

For example, Origen's identification of the author of the gospel according to Matthew with a character in that gospel named Matthew can be validated only if the description of the character fits the characteristics of the gospel. Papias popularized the hypothesis that identifies Peter as the source of material in the gospel of Mark. Whether he was or not can be determined only by comparing the contents of Mark with what is known about Peter. Very little is known about Matthew & Mark & Peter, however, & what is known does not confirm the claim that the material in our current gospels came from them. Thus, most modern scholars put less stock in legends about persons as sources than in tests indicating that one written document was used as a source by the author of another.

The Latin church father, Augustine, was the first to opt for a literary solution to the synoptic problem by proposing the hypothesis that the gospel of Matthew (rather than the preaching of Peter) was Mark's source of information. Fourteen centuries later J. J. Griesbach revised this Augustinian hypothesis by suggesting that Mark edited both Matthew & Luke. 19th c. scholars turned this around by proposing Mark as the basis of the narrative structure of Matthew & Luke. But if Matthew & Luke do not depend on each other, then one needs a hypothetical source like Q to account for shared material in these two works that obviously did not come from Mark. Less than 50 years ago, A. Farrer proposed that Q could be dispensed with by correcting Augustine's hypothesis to recognize the priority of Mark as a source. Which source hypothesis is more valid can be determined only by comparing the contents of all three synoptic gospels.

Source hypotheses are complicated by the fact that many, or even most, early Christian writings have disappeared & are known only through references in later works. Papias, for instance, claimed that Matthew compiled the sayings (logia) of Jesus in Hebrew. The canonical gospel of Matthew, however, is a polished narrative composed in Greek. The hypothesis that identified the first document as the source of the second was repeated by Christian scholars for centuries. Yet neither Papias nor those who echoed him claimed to have actually seen a copy of the hypothetical Hebrew source. To complicate matters further, Papias' work has itself been lost & is known only through quotations (or paraphrases) by Eusebius of Caesarea.

Recently, George Howard published a Hebrew version of Matthew that he reconstructed from passages in a medieval Jewish document. Whether these extracts represent an early Hebrew version or a late medieval translation from Latin (or Greek) is a question for further research & debate. But since there is no surviving independent ms. evidence, a Hebrew version of Matthew remains a hypothetical source. [For further details see G. Howard, The Gospel of Matthew according to a Primitive Hebrew Text (Macon GA: Mercer U. Press, 1987)].

Some modern scholars revised this hypothesis by interpreting Papias' statement as referring to an Aramaic document. There is ms. evidence of a version of the gospel of Matthew in Syriac (a dialect derived from ancient Aramaic) as early as the 4th c. CE. But there are many Greek mss. of Matthew from the previous century. So, the alleged original Semitic work ascribed to Matthew is still a hypothetical source.

The synoptic source that modern scholars call Q is another type of hypothetical document. Like the Hebrew logia collection reported by Papias, no independent copy of this collection of sayings has yet been found. But it can be extracted & reconstructed from Greek mss. of Matthew & Luke. So, if Matthew & Luke are independent compositions, there is ms. evidence for Q in the passages cited by these two early Christian witnesses. This textual base is even more solid than the evidence for the testimony of Papias, Clement and other early Christian authors whose works are known only through brief citations in a single author, like Eusebius, writing centuries later.

Unlike the sayings source mentioned by Papias, however, Q was probably composed in Greek. Although there are still some synoptic scholars who dispute the evidence for Q, it has become the most widely accepted hypothetical source among critically trained NT scholars. Research on the reconstruction of Q is an international project that has involved many scholars for more than a century. The first installment of a critical edition of the Greek text of Q (Documenta Q) appeared in 1996.

Other On-line Resources:

• Source Criticism - 1977 essay by David Wenham explaining the reasons for & methodology of gospel source criticism; chapter 8 in I. Howard Marshall's New Testament Interpretation: Essays on Principles and Methods (posted by Biblical Studies.org.uk).