Sunday, December 17, 2006

The new magical imperial toolkit: part 2

In part one I summarised the proposals of Freedman and his colleagues, which present a count of what they called "prose particles":

the sign of definiteness (or "article") ה
the direct object marker את
and the relative אשׁר

as a simple way to distinguish biblical Hebrew prose and poetry. By Andersen and Freedman's 1989 Amos commentary (in the Anchor Bible series) this new toolkit was seen as almost magical, they applied their counts to units as small as a single verse!

Amos 1:1 was prose, with a count of 8.7%, while 1:2 was poetry with a neat 0.0% (somehow in during the 1980s the article ה had ceased to count, or the scores would have been 16% and 15% respectively (or both "clearly prose") on the old scheme!

In this part I'll present my first attempt to test their hypothesis.

Three Chapters of Ezekiel as a Test Case

In 1987 Freedman had proposed:

We can test the system in a provisional way against the book of Ezekiel... First, it is clear from every point of view that much of Ezekiel is straight prose. There are fourteen chapters over 15%, while another eighteen are in the range between 10% and 15%. The remaining sixteen chapters are under 10%; of these, twelve are in the range 5% and 10%, while four are under 5% (chs. 19, 21, 27, 28).¹

So, let's look at these chapters.

Ezekiel 21 is treated the same by both BHS and BHK with 109 words printed as poetry and 420 as prose. The 109 words of poetry with 6 of the particles have a score of 5.5%, which is only very slightly higher than Freedman's 5% threshhold, however the 420 words of prose have 19 particles and so score only 4.5% or more poetic than the poetry!²

Ezekiel 27 is printed somewhat differently by BHS and BHK. BHS has 181 words in prose and 226 as poetry, while BHK had 145 prose and 262 poetry. The scores are:

BHS prose 2.2% poetry 4.0%
BHK prose 1.4% poetry 4.2%

Either the editors of neither BHS nor BHK can accurately distinguish prose and poetry or the method is flawed. Every test so far has given the reverse of the predicted results!

At first sight Ezekiel 28 gives some comfort to Freedman's proposed toolkit. BHS prints it all as prose, but BHK gives 122 words as prose with a score of 6.6% and 230 as poetry with only 2.2%. The sort of result Freedman et al. would predict.

However, things get more complex and more interesting when we look at the location of the "prose particles". All of the particles in the prose sections occur in just three verses, vv.24-26. Many commentators see these three verses (or perhaps only two of them) as later additions to the text. If even the two verses are left out of the calculation the score drops to 1.3% (it would be 0% of course if all three were omitted).

These particles are commonly seen as typical of later biblical Hebrew. (Kaiser in his introduction to exegesis regards this as a fact not needing footnote support, Rooker and others include them in their characteristics of Late Biblical Hebrew.)

So, my provisional evaluation of the new magical imperial toolkit is that it does not function as neatly or well as advertised. But:

there is considerable evidence (not least from Andersen and Forbes³) that these particles are more common in biblical Hebrew prose than poetry
there is some evidence (in a notoriously difficult area) that they may be more frequent in later texts.

1. D.N. Freedman, "Another look at Biblical Hebrew Poetry" in E.R. Follis (ed.) Directions in Biblical Poetry (JSOTS 40; Sheffield, 1987), p.17. [return]
2. My counts differ from Andersen and Forbes only by one word (or less than 0.2%), therefore the differences in scores are negligible. [return]
3. F.I. Andersen & A.D. Forbes, "'Prose Particle Counts of the Hebrew bible", in C.L. Meyers & M. O'Connor (eds.) The Word of the Lord Shall Go Forth: Essays in Honor of David Noel Freedman in Celebration of his Sixtieth Birthday(Philadelphia, 1983), pp.165‑183.[return]

Labels: hebrew bible

¶ 12:38 PM [to Del.icio.us]

If I understand you correctly, you conclude that high counts of the two or three function words singled out by Freedman, Andersen, and Forbes distinguish prose from poetry at the highest level of abstraction only (all the prose vs. all the poetry of the Bible), but not on a pericope per pericope or verse by verse basis.

That rings true to me. For example, the relative pronoun, the object marker, and the article occur often in Zephaniah 1, but the contents of the chapter are marked by a high density of semantic, syntactic, and morphological parallelisms, and scan as verse just fine, as a look at the JPS Hebrew-English Tanakh will show.

What distinguishes poetry from prose in ancient Hebrew is that the former is characterized by a system of versification that the latter lacks. Students of literature have long recognized this to be a language universal. As Albert Willem de Groot put it, “Continuous correspondence of successive segments, called ‘lines,’ is the only constant feature which distinguishes verse from prose.”

By John, at 9:40 PM

Pretty much I agree, though I'd say that if the frequency of these particles changes dramatically it is usually a clue that something is "going on". Sometimes this is a change of genre between poetry and prose, sometimes it may be a change of writer (or editor?), BUT these changes at the level of individual pericopes are not reliable as a measure.

By Tim, at 7:35 AM