There is a discussion of the translation of soul at various blogs, started by Joel Hoffman here and continued on the BBB here where Kurk Gale of BLT and other blogs drew my attention to comments by Robert Alter in his translation of the psalms. Along with the translation problem of the multiplicity of possible referents for a word, there is a metaphor in the discussion related to hardware-software.
Re Alter: his statement, "as I observed above" (page xxxii) I presume refers to his comment on page xxvii where he says 'nefesh means "life breath" and by extension "life" or "essential being"'.
I have to ask: whose extension? It is the extension by time and the dispersion by tongue in the usage, but particularly through translation into foreign tongues - psyche in Greek and anima in Latin. I can barely spell Greek and Latin, so I have to talk about them as languages that I have scanned, not languages that I read fluently. They have a positive estimation in my mind in spite of my limited apprehension of them. Greek is famed for its precision and Latin for its theological discourse. We have extended the meaning of soul in English and by implication nefesh in Hebrew. But I don't find the conclusion that nephesh may be immaterial believable. I don't see an immaterial usage anywhere in the words I have read from Scripture. In fact, Alter's objection to anima as its translation in the Vulgate seems to me to work against his argument.
One thing I have learned from software and hardware over the years is that symbols have meaning only from their usage. Polyvalent symbols will be used in areas of independent scope. Polyvalent symbols within the same scope are very bad practice in programming. They create confusion and frequently make code unmaintainable. Hardware-software as a metaphor cannot get much further in understanding life than the mechanical metaphor of the 19th century, but it does tell us something about tokens and human idiosyncrasy in their invention.
If nephesh is polyvalent then what hope have we of deciding what it 'means' based on its usage? We have to separate those places where it means throat, neck from those where it means breath, life, or just me, you, us, them etc. If we decide sometimes, depending on our whim, to render nephesh as 'life', this will confuse us when life is rendered for xayah (after the style of Miss Piggy). This sort of translation practice really messes up the poetry. I have worked through every instance of every word in the psalms and eventually I refused to let two completely different Hebrew words be rendered as the same word in English in this case or any other (with a few insurmountable exceptions). My refusal was not subjective. If it is true that the Hebrew poets used word-recurrence as a significant structuring ploy in their poems, then it is important to follow a rule that prevents confusion.
There are significant difficulties here especially with English helping verbs and certainly with nephesh. Nonetheless, when nfshi, means me rather than my being, the translation is a compromise. Psalm 69 is a good example. Nephesh occurs 3 times, verses 2, 10, 19 (Hebrew numbering). Though in English the metaphor of water seems broken in verse 2, the unity of verses 1-19 is compromised when a different word is used. But maybe this particular repetition is not significant. That is a subjective decision. In this case, it is possible that the water metaphor of verse 2 is separate from the use of reproach as an organizing principle for the poem. In this psalm, Alter uses three different approaches:
- verse 1 is neck (this creates a different problem for him when the neck as body part turns up, but he can't use throat here since he uses it in verse 4 - not nephesh!)
- verse 10 is being.
- verse 19 is me.
In this usage he misses the sound of one of the frames of the section of the poem from verses 8-22. And he confuses these frames as well. There is an eight-fold concentric circle here - see for yourself in the text and the table at the link. It's deep indeed.
My contention, following the software metaphor, is that we need to see the pattern of usage and not cross-match words if possible. Then we will see different usage patterns and will be able to 'distinguish those things that are different'.