Wednesday, September 8, 2010

Debugging Mounce: "Contraction" (?!) in the Neuter Plural

We are all using the very popular Mounce textbook for the introductory Greek course in the Duke Divinity School. On the Duke Newt blog, we are planning on blogging some of our experiences with this textbook.

Occasionally, Mounce's explanations don't seem quite right. One example occurs on p. 32 of the 3d edition, in section 6.12 on case endings. As a bit of background, Mounce takes the unusual approach of giving the first and second endings apart from the stem-vowel (alpha/eta or omicron, respectively), rather than, as more common, expect students to learn the endings with the stem vowel.

Mounce's approach, however, runs into a wrinkle for the second declension, nom./acc. neuter plural ending, as in ἔργα (from singular ἔργον), because the omicron stem vowel is no longer there. Here's how Mounce attempts to explain it:

6.12 ... The underline (α means that the case ending joins with the final stem vowel.5 ...

5 This is called "contraction," and I will discuss it in detail later. For example, the stem of the noun ἔργον is ἔργο. When it is in the neuter plural its form is ἔργα. The omicron and alpha have "contracted" to alpha. ἔργο + α > ἔργα.

This explanation is simply wrong. In particular, omicron and alpha do not "contract" to a (short!) alpha. Rather, omicron + alpha contracts to a (long) omega (as Mounce later recognizes in the chart on p. 343).

The explanation is also wrong as a historical matter. The alpha of the neuter plural is not the result of an o-vowel plus an alpha ending; rather, it is the ordinary reflex of the Proto-Indo-European ending of *-eH2, which the vowel *e plus the a-coloring laryngeal H2. In this case, the stem vowel *o changes to *e by ablaut, which then is colored to α by the following H2.

Now, I don't expect introductory grammars of Biblical Greek to be giving the details of Proto-Indo-European morphology, but I do expect that their explanations of the Greek be correct. In this case, a more accurate explanation is that the neuter plural nom./acc. ending -α simply replaces the second declension stem vowel. There is no need to appeal to "contraction," which never occurred historically here and which apparently behaves differently from how omicron and alpha normally contract.


  1. Ooh...good catch. Curious if Dr. M will respond!

  2. Got referred to this blog by one of your students. This reminds me why I hate language textbooks. But, I suppose, if you're OK with telling a pedagogical lie in the first place, this one isn't totally misleading. Certainly, the "o" always colors the vowel in o+a and often in a+o. But long alpha is the result of an not uncommon Attic crasis or "grammatical" contraction of o+a (cf. haploa => hapla or ho anêr => hanêr -- I'm sure Sihler has something on this....) And the PIE *-e-H2 of o-stems *does* develop originally to a long alpha.

    So, I suppose Mounce could mitigate the deception if he mentioned that a short alpha comes to replace the long, being generalized from other n. nom/acc pl. endings -- which is actually what does happen with the development from *-e-H2: Latin generalizes the o-stem long a and Greek generalizes the short alpha from the consonantal n. nom/acc pl.