The Parallelism of Greater Precision:
Notes from Isaiah 40 for a Theory of Hebrew Poetry
Published in On the Way to the Postmodern: Old Testament Essays 1967-1998, Volume 1 (JSOTSup, 292; Sheffield: Sheffield Academic Press, 1998), pp. 314-36The purpose of this paper/1/ is to draw attention to a hitherto generally unobserved feature/2/ of the parallelistic couplet in Hebrew poetry, a feature which may offer a point de départ for a revised perception of the nature of the parallelistic couplet as such.
The feature in question is that the second half of a parallelistic couplet (line B) is often more precise or speciÞc than the Þrst half (line A). In the Þrst section of this paper (I) I will present some examples from Isaiah 40, which will enable me to construct a tentative theoretical statement about it (section II). I will then examine some further couplets from Isaiah 40, with a view to judging the usefulness of the concept, 'the parallelism of greater precision', for exegetical work (section III). In section IV some similar features of the parallelistic couplet, previously recognized and named, will be examined by way of contrast and comparison. The Þnal section (V) will raise the question whether any new insights into the nature of Hebrew poetry may be developed from the identiÞcation of the parallelism of greater precision.
1. Examples from Isaiah 40
a. Isa. 40.16
r[b yd ya wnblw And Lebanon is not enough to burn
hlw[ yd ya wtyjw and its animals not enough for a burnt offering.Line A taken by itself raises the question, Why should anyone want to burn Lebanon? Lebanon is not elsewhere in the Hebrew Bible usually connected with burning/3/ and while it is natural enough-even while we are still in line A-to suppose that it is the trees of Lebanon that are for burning, even though they are typically used for building, it is impossible to discern from line A what purpose is in view in burning Lebanon's trees, and thus why Lebanon's trees are insufÞcient for burning. Not to know that is to be ignorant of the whole point of the afÞrmation. Only with line B, and with its last word, is it made clear that the image of the whole couplet is of sacriÞce; the burning of line A must be of wood upon the sacriÞcial altar. We are conÞrmed in our impression that line A is not perspicuous by the fact that a modern version like the niv Þnds it necessary to offer an expanded rendering: 'Lebanon is not sufÞcient for altar Þres' (my italics).
We may say, then, that line A is less precise, less speciÞc than line B. Line A is not swallowed up in line B, however; it is not the case that once we have line B we can dispense with line A. Rather, line B provides the clue or the context within which the uncertainty of line A is resolved. Line B drives us back to read line A again in the light of what line B has added to line A.b. Isa. 40.22
µymv qdk hfwnh who stretches out the heavens like a thin thing,
tbvl lhak µjtmyw and spreads them out like a tent to dwell in.
In reading line A we may well ask, What, precisely, are the heavens like that God stretches out? The problem is not primarily that qd is a hapax legomenon, for there can be little uncertainty that it means 'something thin' (from qqd 'crush'). Even if we knew from a multiplicity of attestations that the thinness in question was speciÞcally a thin curtain or veil or gauze,/4/ we would still not know what precisely the image was, for a curtain could be vertical or horizontal, it could be used to divide or screen or cover, it could serve the one who spreads it out, or someone else for whom it is spread out. Line B, however, disambiguates line A. From line B we learn that the 'thin thing' is a 'tent for dwelling in', i.e. a tent from the viewpoint of its occupants, a curtain that is both horizontal and vertical, and spread out not to hide the one who spreads it but to serve as a covering for those under it. The blurred and indeÞnite image of the line A is brought into focus in line B. The 'parallelism' of qd and lha is a parallelism of increasing precision.c. Isa. 40.3
hwhy °rd wnp rbdmb In the wilderness prepare the way of Yahweh,
wnyhlal hlsm hbr[b wrvy make straight in the desert a highway for our God.The imprecision of line A, as compared with line B, lies in the meaning of °rd, 'way', in the genitival relation hwhy °rd, 'the way of Yahweh', and in the connection between 'prepare' (wnp) and 'make straight' (wrvy). °rd, of course, may mean not only 'way, road, path' in a literal sense, but also 'way, manner, habit; way of life, moral behaviour'. In genitival relation with a personal name it can be taken for granted that °rd has a metaphorical meaning, e.g. 'the way of Jeroboam' (1 Kgs 15.34; 16.2), 'the way of David' (2 Kgs 22.2)./5/ When °rd is linked with hwhy, as it is frequently, the usages can be analysed, with BDB, as meaning (i) his creative activity (Job 26.14), (ii) his moral administration (Exod. 33.13; Deut. 32.4), (iii) his commandments (Gen. 18.19, 'to keep the way of the Lord'; and often). Even if this analysis is open to question at some points, it cannot be doubted that some such metaphorical sense is the most natural one (within the same writing as our text, Yahweh's 'ways' are clearly his 'way of life' or 'moral administration' at Isa. 55.8, 9). Natural or not, however, such an understanding of the present text would be wrong; for the parallel term hlsm, 'highway', is certainly used in a literal sense. Indeed it is used, in 26 of its 27 occurrences, literally; in the other case, 'The highway of the upright is to depart from evil' (Prov. 16.17), we have the impression that the metaphor is freshly minted for the occasion, and there can be little question that hlsm is typically a literal road.
What line B also makes precise is the relation of the 'road' to God. The construct chain in line A is open to many interpretations, but the phrase 'a highway for our God' in line B is unambiguous: it must mean a road for Yahweh to walk along-just as a 'highway for the remnant of his people' (wm[ ravl hlsm) in Isa. 11.16, and a 'way for the redeemed to pass over' (µylwag rb[l °rd) in 51.10, are paths for Israel to walk on. Here, the 'way of Yahweh' (line A) has meaning in the sense of 'a highway for our God' (line B); line B then speciÞes line A.
Furthermore, we may perhaps see in 'make smooth, straight' (wrvy) a greater precision than in 'prepare' (wnp). wnp in line A is properly 'clear away', i.e. 'remove any obstacles', which could apply to an already existing road-which might be what is meant by 'the way of Yahweh'. Line B, however, in specifying that the 'clearing' is a matter of 'making smooth, or straight' a highway for God, more evidently envisages the building of a new road. wnp 'clear' is to be understood in the sense of wrvy 'make smooth, straight'. The obstacles to be cleared away are not just boulders or other debris lying on the road (which °rd wnp could refer to) but are the natural features (v. 4 will mention them as mountains and valleys) that are to be overcome and eliminated in the construction of an entirely new road.d. Isa. 40.6
ryxj rcbhAlk All þesh is grass
hdch yxk wdsjAlkw and all its loyalty is like the þower of the Þeld.The metaphor in line A is open to various understandings. Grass is indeed a symbol of what is short-lived, and is sometimes used of humankind's transitoriness. So we Þnd, for example, that mortals are like grass renewed in the morning but withered by the time of evening (Ps. 90.5); their days are like grass which is gone when the wind (jwr, as here) passes over it (103.15-16); the 'mortal who is made like grass' (tny ryxj µdaAb) is parallel to the 'human who dies' (Isa. 51.12). It is in a rather different sense that it is said of the wicked that they will soon fade like the grass (Ps. 37.2), for there is not an inbuilt weakness of the human constitution that accounts for the imminent death of the wicked, but a fate peculiar to wrongdoers. In other passages, however, the symbolism of grass is completely different: in Isa. 44.4, though the text is open to doubt,/6/ the redeemed are to spring up like grass (avd), and in Job 5.25 Eliphaz promises Job that his descendants will be many and his offspring like (i.e. as plentiful as) the grass (bv[) of the earth.
The image of grass is therefore ambivalent, and nothing in the lines preceding our present text has conditioned its readers to interpret the image of line A in the way that line B requires. Indeed, the previous reference to 'all þesh' in v. 5 as being about to see the glory of Yahweh could have created an expectation that the image in v. 6 will be positive rather than negative. At best, any reading of line A is bound to be provisional; the precise sense of the 'grass' image-that it is grass in its aspect of impermanence (absence of dsj)-awaits a reading of line B.e. Isa. 40.17
wdgn yak µywgh lk All nations are as nothing beside him;
wlAwbvjn whtw spam less than nothing and emptiness they are reckoned
by him.What line A leaves open and what is further speciÞed by line B is the question: in whose estimation are the nations as nothing beside him? Without line B, we might reasonably conclude that this evaluation is the poet's judgment, just as we would suppose for the preceding verses, 15-16 (though the presence of wbvjn in v. 15 may give us second thoughts). What line B speciÞes is that the reckoning is God's: the wl indicates 'the efÞcient cause (or personal agent)' with the passive (GKC, §121f). Line B makes clear that the perspective upon the nations and their signiÞcance in comparison with God (wdgn, line A) is God's own.
There may be further precision in line B's phrase whtw spam by comparison with line A's yak. Being literal-minded about it, we might say that yak means only 'about nothing', 'roughly nothing', whereas spam means, to be precise, less than nothing'. But it is not at all certain that the m signiÞes 'less than'/7/ rather than 'consisting of, made from',/8/ and there is also the possibility that the text should read spak (so 1QIsa)./9/
No doubt whtw 'and emptiness' adds a further precision or perhaps elaboration to the ya of line A; that is to say, to the idea of ignorable non-existence is added that of chaotic absence of form. But the most signiÞcant precision in line B is that which I have mentioned Þrst: the speculation of the standard or judgment that gives value and signiÞcance to the comparison (k) contained within line A.f. Isa. 40.21
w[mvt awlh w[dt awlh Have you not known? Have you not heard?
µkl varm dgh awlh Has it not been told to you 'from beginning'?
rah tdsym µtnybh awlh Have you not understood from the foundations of the earth?This tricolon displays a double speciÞcity. Line B adds precision to line A, and line C to line B. What is not speciÞed in line A is when Israel's coming to know and hearing is supposed to have taken place. In line B it is more speciÞcally said to be 'from beginning' (varm), but that also raises a question: the beginning of what? Line C explicates B by its phrase 'from the foundation of the earth' (rah tdsym, emended text);/10/ the 'beginning' in question is the beginning of the earth. There is certainly a difÞculty in understanding how Israel can be expected to have known anything from primaeval times,/11/ when it did not exist, but there is no doubt that we have here a parallelism of increasing precision.
2. A Theorem
In the parallelism of greater precision, line B speciÞes line A or some element of line A. There are different functions which precision or speciÞcation may serve:
1. B may disambiguate A (as in examples a, b, c, and d above). In these cases A is to a greater or lesser extent unclear, ambiguous. It is not incomplete, but it is vague or question-provoking, especially when compared with B or with the total effect of A plus B.
2. B may explicate A (as in examples e and f above). In these cases there is no ambiguity or uncertainty about A, but it is patient of further elaboration, in directions that it does not perhaps explicitly state, but which can be seen-especially on reþection after reading B-to be latent in it.Further Comments
1. The movement towards greater precision is characteristically from A to B. This does not mean that A cannot contain more than B, but that when there is a relation of greater precision between A and B, it is B that typically exhibits that greater precision. Some examples of where A contains more that B but is not more precise may be noted.a. Isa. 40.27
bq[y rmat hml Why do you say, Jacob,
larcy rbdtw and speak, Israel?A contains hml, 'Why', a surplus to B. But just because B follows A, hml is implied in B, and B is therefore no less precise than A.
b. Isa. 40.18la wymdt ymAlaw And to whom will you liken to God,
wl wkr[t twmdAhmw and what image/likeness will you compare to him?Let us suppose for the sake of the discussion that 'liken God' and 'compare a likeness to him' are strictly synonymous, or else strictly separate (i.e. that line A refers to a deity and line B to an image of that deity). Let us suppose, that is to say, that B is not more precise than A. Then, we may ask, is A more precise than B? It contains 'God' (la) whereas B contains only '(to) him' (wl). A is indeed more precise, but only in a trivial sense, for wl in B does present us with the same referent as la in A. If, on the other hand, the lines changed places, so that we read:
wl wkr[t twmdAhmw And what image/likeness will you compare to him,
la wymdt ymAlaw and to whom will you liken to God?or even if just the noun and pronoun changed places:
wta wymdt ymAlaw And to whom will you liken to him,
lal wkr[t twmdAhmw and to what image/likeness will you compare God?B would be more precise than A, for in A we should not know who was being spoken of ('him') but in B should Þnd it stated ('God').
2. Not all parallelistic couplets, of course, exhibit the parallelism of greater precision. And because the comparison is between sense-units and not just words, there is often enough room for debate over whether a particular B-line is more precise than its A-line. But I would argue that in Isaiah 40 slightly more than half the parallelistic couplets exhibit it. In Psalm 21, almost every verse is an example of this parallelism. In short, the parallelism of greater precision is a common enough feature of Hebrew poetry to make it worth asking of every parallelistic couplet whether any gain in understanding may result from applying the present concept to it.
3. Exegetical Applications
Under this heading I am examining some texts in Isaiah 40 for which the possibility of an analysis such as the parallelism of greater precision may have some exegetical value.
a. Isa. 40.26
µabx rpsmb ayxwmh who brings out by number their host,
arqy µvb µlkl to all of them by name he calls.This is a textbook example of a parallelistic couplet, with a mirror chiasmus./12/ ayxwm is parallel to arqy, and µabx to µlkl. Those parallelisms are syntactic rather than morphological, whereas with rpsmb // µvb 'by number' // 'by name' we have a strict morphological parallelism as well (preposition plus noun)./13/
Such an analysis is accurate as far as it goes, but when we consider the sense of the lines it is quite misleading./14/ For 'by number' is not a method of summoning the stars that is parallel to 'by name'. Each star has a name, but each does not have a number. rpsmb in fact does not mean 'by number' at all but 'in number' (cf. 1 Chr. 9.25). As far the sense goes, we would be better to say that line B speciÞes the way in which God brings the army of the stars out in full number: it is by summoning each one of them by name. Line A acquires its speciÞc meaning in the light of line B. It is not so rewarding to ask which terms of A correspond to which terms of B, as to ask whether, and in what way, line B speciÞes line A.b. Isa. 40.31
jk wpyljy hwhy ywq Those who wait for Yahweh will gain new
strength;
µyrvnk rba wl[y they will put forth (?) plumage like eagles.Here the parallelism of greater precision may help to solve a long-standing exegetical problem. Is wl[y qal or hiphil? Do they 'rise up (on) wings like eagles'/15/ or 'put forth,/16/ or grow, wings (or, plumage) like eagles'?/17/ In favour of the former is the use of hl[ qal with µyrvn in Jer. 49.22, and 'soaring' is undoubtedly an obvious thing for eagles to do. But in favour of the latter is the awkwardness of the absence of a preposition before rba, the lxx rendering pterofuhvsousin, 'they will grow wings', and the Targum's wytdrwn I'wlymwthwn kßymw dslyq 'l gdpy n¡ryn 'and they will be renewed in their youth like the sprouting (of plumage) that rises up upon the wings of eagles'.
If line B may be more speciÞc than line A, we can ask, In what way can the somewhat vague language of A, jk wpyljy, 'they will change, substitute, renew (their) strength', be speciÞed further by B? Going up on wings in no way speciÞes 'changing strength' but putting forth new wings or new plumage is indeed a sign of new strength, whether or not there is an allusion to a belief that in old age eagles grow new wings./18/ The exiles would then substitute new strength for old strength, and 'change' their strength in the sense that old eagles 'change' theirs.c. Isa. 40.27
hwhym ykrd hrtsn My way is hidden from Yahweh,
rwb[y yfpvm yhlamw and from my God my right is disregarded.A further example of the parallelism of greater precision may be found here, though commentators do not generally see the connection of thought between the two lines. The A-line taken by itself is open to more than one interpretation. For Israel's way to be hidden m Yahweh could mean that God cannot see Israel's way, or will not see Israel's way, or has caused Israel not to be able to see its own way. 'Way' itself can mean here 'course of life' (BDB, s.v., §5), perhaps with the sense of 'destiny', 'fate'/19/ (neb 'plight') or its 'direction' in some metaphorical sense.
The B-line however makes everything plain, not without some help from the A-line. fpvm is itself not very precise (rv 'judgement', rsv 'right', neb 'cause'). But for a fpvm to rb[ can only mean that a 'right' (which perhaps already became a legal 'cause' or 'case') has 'vanished', sc. from the notice of the one to whom it ought to be a matter of concern. In such a light, the A-line most probably means that Israel's 'course of life' or 'state' (or 'plight', since it is a bad state) is hidden from God (A) in the sense that its claim to restitution has failed to attract his attention (B). And Israel only knows that its 'way is hidden from Yahweh' (A) in that its cause is ignored, it does not receive its rights (B). The resulting interpretation of the couplet is no spectacular advance on the appropriate sense many readers attain very quickly; but our investigation has built a surer foundation for the validity of the interpretation.
We should also note that the principle of the parallelism of greater precision concerns primarily the relation between the two lines of a couplet; it does not focus on the relationship between the members of the two lines, either grammatically or semantically. In the present case, for example, we have an example of a complete or 'mirror' chiasmus (as in 40.26, discussed above), but that is of no signiÞcance for the relation of the lines. It is likewise unimportant whether °rd is a good parallel to fpvm or not, both because the relation of the lines does not depend upon the relation between the terms that are 'in parallel' (as we say), and because the notion that an ideal parallelism is a truly synonymous parallelism has been exploded by the principle of the parallelism of greater precision. Duhm found the parallelism here 'rather feeble' (etwas ärmliche)-which is indeed the case if the ideal is synonymity, but not at all the case if other relationships between the lines of a couplet can exist.
The foregoing examples yield evidence of the value of having the model of the parallelism of greater precision in mind when approaching the exegesis of an individual text. I conclude this section with an exegetical example of where the parallelism of greater precision is a 'false friend' and potentially misleading.d. Isa. 40.28
hwhy µlw[ yhla God of eternity/the world is Yahweh,
rah twxq arwb Creator of the ends of the earth.To anyone approaching this verse with a background in rabbinic or mediaeval Hebrew, the A-line is quite ambiguous: is µlw[ 'eternity' or 'the world, universe'? The parallelism of greater precision would seem to set the matter beyond doubt: 'the ends of the earth' must surely function as speciÞcation of the ambiguous µlw[. It is generally argued, however, that the meaning 'world' of µlw[ is a postbiblical development,/20/ and a mere possibility in the relationship of parallelistic lines ('greater precision') cannot be set against a linguistic certainty.
The couplet exhibits the parallelism of greater precision nevertheless-in another respect. For if we ask, In what respect is the 'eternity' of Yahweh signiÞcant in the present context?, it is in the sense of line B, that he was in distant times (µlw[) creator of the earth; and it is his creative power that lies behind the subsequent lines (vv. 28c-31).
4. Some Similar Aspects of the Parallelistic Couplet
It is perhaps unlikely that at this stage in the history of research into Hebrew poetry any completely new observations can be made. There are indeed several points at which the parallelism of greater precision corresponds to features that have previously been noted. But no one, I think, has focused upon the matter of 'greater precision' as an important deÞnition of the relation of the lines of the parallelistic couplet.
To take the last point Þrst, it is true that several scholars have observed that 'speciÞcation' is a function of the B-line of a couplet. James Kugel, for example, has recently remarked thatthere are quite a few lines in which B is clearly a continuation of A, or a going beyond A in force or speciÞcity./21/
But his principal concern is to afÞrm that the relationship of A to B is only in a minority of cases an exact repetition, or a saying of the same thing in different words. Kugel's emphasis is on the additive or emphatic aspects of the B-line: 'A is so, and what's more, B is so'. He portrays the B-line as a continuing, seconding, emphatic carrying further, echoing, reiterative statement. All of that may be true of various couplets of Hebrew poetry, but it is quite other than what is being urged in this paper.
To similar effect James Muilenburg remarked thatparallelism is in reality very seldom precisely synonymous. The parallel line does not simply repeat what has been said, but enriches it, deepens it, transforms it by adding fresh nuances and bringing in new elements, rendering it more concrete and vivid and telling./22/
More 'concrete', yes; but he said nothing about more speciÞc, more precise.
Other aspects of parallelism fall now to be compared to and contrasted with the parallelism of greater precision.a. Staircase parallelism
This feature of the parallelistic couplet, otherwise known as 'climactic' or 'repetitive' parallelism, or the 'expanded colon', has long been recognized./23/ Wilfred Watson notes some forty examples,/24/ including Jer. 31.21:larcy tlwtb ybwv Return, O virgin Israel,
hla °yr[ la ybv Return to these your cities.From a formal perspective, the description 'staircase' is apt, but from a more semantic perspective it is easily seen that such a form is an instance of the parallelism of greater precision. In the example quoted, the Þrst line leaves unstated the place to which Israel is to return; the B-line speciÞes the full signiÞcance of 'return' in A. It is true of course that it is not only that the B-line is more speciÞc than the A-line but also that the A is more speciÞc than the B. Such cases do not negate the parallelism of greater precision; they are a subset of the examples of our feature.
b. Number parallelism
Examples of number parallelism can be divided into two categories: those with itemization and those without.
i. Itemized number parallelism. The best known example is Prov. 30.18-20:ynmm walpn hmh hvlv There are three things too wonderful for me,
µyt[dy al [braw four I do not understand:
µymvb rvnh °rd (1) the way of a vulture in the sky;
rwx yl[ vjn °rd (2) the way of a serpent on a rock;
µy blb hyna °rd (3) the way of a ship in the middle of the sea;
hml[b rbg °rdw (4) and the way of a man with a maiden.The itemization makes it plain that it is precisely four things that the speaker does not understand. Line A contains the approximate number ('three'), line B the precise number. In all cases of itemization the precise number is the second one mentioned./25/ It is true that 'three' is in itself just as precise as 'four', but in this context, where 'four' is the precise number intended, 'three' is imprecise. M. Haran has listed twelve instances of itemized number parallelism, six from the OT,/26/ Þve from Ben Sira, and one from the Babylonian Talmud citing Ben Sira. All of these, we may now say, exhibit parallelism of greater precision.
The recurring couplet in Amos, through it is not usually followed by an itemization, Þts here best:y[vp hvlv l[ For three transgressions of X,
wnbyva al h[bra l[w And for four, I will not revoke it. (Amos 1.3,6,
9, 11, 13; 2.1, 4 6)In these cases (except perhaps in 2.6-8) it is only the fourth transgression that is speciÞed, almost certainly because the fourth is so climactic that the others may be left out of consideration./27/ It is because of the fact that they total to four, and so include the fourth, climactic one, that Yahweh's punishment falls. So again the B-line of the couplet is the more precise or speciÞc.
ii. Unitemized number parallelism. Most such instances display ordinary 'synonymous' parallelism, though of course no number is ever strictly synonymous with another./28/ For example, Mic. 5.4:µy[r h[bv wyl[ wnmqhw Then we will raise against him seven shepherds,
µda ykysn hnmvw eight chiefs of men.We cannot say that 'eight' is more precise, in the context, than 'seven' (nor vice versa). Haran has suggested however that in a few cases 'the intended number is the Þrst of the two';/29/ but in Job 33.14 and Ps. 62.12, which he cites, it can be maintained much more convincingly that it is the second number that is 'intended' or the more precise. This category of graded numerical sayings thus yields two more examples of the parallelism of greater precision, but on the whole it does not display the feature under consideration here.
c. Automatism
The phenomenon of 'automatism' as set forth by Haran/30/ involves the use of one element of a word-pair solely for balance between the lines, and not at all for its semantic signiÞcance. An example that is adduced is Prov. 24.30:ytrb[ lx[ vya hdcAl[ I passed by the Þeld of a sluggard,
blArsj µda µrkAl[w and by the vineyard of a man without sense.The argument is that hdc and µrk form a Þxed pair, corresponding to other pairs like bread and wine, threshing-þoor (rg) and wine-vat (bqy), corn (gd) and new wine (vwryt), farmers (µyrka) and vintners (µymrk). But in the context only the vineyard is really meant, for the following verse refers to 'it' as having a stone fence-which points to a single parcel of ground, and that a vineyard, not a Þeld. 'Field' is not 'intended' at all.
If such is the case, the line that contains the 'intended' word is inevitably more precise than the one with the 'automatic' variant. So for those cases where the 'automatic' variant occurs in line A,/31/ we would have the parallelism of greater precision./32/ However, the existence of such automatism is, in my opinion, open to question.
In the example above, it is entirely probable the hdc, 'Þeld', is a more general term that includes 'vineyard'-as is amply attested by Judg. 9.27, 'they went out into the Þeld (hdc) and harvested their vineyards (µrk)'. So it is not true that the place the wisdom poet saw 'can be either a Þeld or a vineyard, but not both'./33/ The relationship between A and B is not one of 'automatic' word-pairing but of what is here being referred to as greater precision: the hdc in question is a µrk.
As for cases where it is claimed that the A line contains the more speciÞc term, here also the existence of 'automatism' may be doubted. Haran cites Prov. 4.3 as an example:ybal ytyyh b yk For I was a son to my father,
yma ynpl dytyw °r tender and alone before my mother.He comments, 'Only the father is actually kept in the poet's mind, while the mother is mentioned because of a mere automatic adherence to the verbal pattern'./34/ Indeed, the succeeding lines do refer exclusively to the father: 'and he taught me, and he said to me, "Let your heart hold fast my words " ' So it is quite true that it is the father that is principally in focus; but that does not necessarily mean that 'mother' is only there by accident. For if we suppose, quite reasonably, that the sense of the couplet is distributed across the two lines,/35/ we have a perfectly satisfactory meaning, viz. 'When I was a young only child with my parents . . . '/36/ This is certainly not an example of automatic writing.
My conclusion is, though I have not reviewed here all the examples adduced for the phenomenon of 'automatism', that the phenomenon does not exist in the Hebrew Bible,/37/ and that some supposed examples of it really exibit the parallelism of greater precision, while other examples contain other types of parallelistic relationships.d. Ballast variants
A ballast variant is deÞned by W.G.E. Watson as 'simply a Þller, its function being to Þll out a line of poetry that would otherwise be too short'./38/ As an example he cites Judg. 5.28:awbl wbkr vvb [wdm Why is his chariot so slow in coming?
wytwbkrm ym[p wrja [wdm Why does the clatter of his war-wagons tarry so?On this he comments: 'Since låbø (or its equivalent in meaning) does not reappear in the second colon, the longer expression pa'ÿmê markëbôtâyw is used instead of simply one word (e.g. markëbôtâyw // rikbô)'. No doubt a major question is begged here, viz. which comes Þrst, the over-shortness of the line or its Þlling? Since it is unanswerable, the terms of the discussion are probably inappropriate. And the premise of the whole concept of 'ballast' or 'Þlling' depends on the 'isocolic principle' that the two lines of a parallelistic couplet should balance-which is open to question./39/
The language of 'compensation' long ago introduced by G.B. Gray, falls under the same criticism,/40/ and so perhaps the use of 'expletive' by R. Austerlitz./41/ But there can be no doubt that some such terminology is needed for this feature of the parallelistic line.
Another conceptualization of the same phenomenon uses the language of 'gapping', which refers to the absence in the B-line of the verb or some other term of the A-line. This conceptualization, no different intrinsically from that of 'Þlling', has been used with discretion in the systematic analysis of M. O'Connor./42/ His primary example is Num. 23.7:qlb ynjny µraAm Balak brought me from Aram,
µdqAyrrhm bawmA°lm the king of Moab from the Eastern Hills.'Gapping' refers to the absence of anything equivalent to jny in line B. It is not clear what corresponding term O'Connor would use to describe the relationship of 'Aram' to 'the Eastern Hills' and of 'Balak' to 'the king of Moab'. He does indeed regard the latter case as a splitting of the name 'Balak, king of Moab' across two lines/43/ (he calls it 'binominated'), but that cannot be said of the former case.
My present description of the parallelism of greater precision of course prescinds from questions of causes or poetic psychology and even from issues of grammatical relationship, and it cuts across distinctions that have previously been made. That is to say, in some cases, perhaps many, of ballasting, Þlling, expletives or gapping, the parallelism of greater precision can be seen; but in other cases, different parallelistic relationships occur.
A consideration of Num. 23.7, just quoted, which certainly contains parallelism of greater precision, leads to a closer deÞnition of 'greater precision'. It could well be argued that both 'Balak' and 'Aram' in line A are more speciÞc or precise than their counterparts in line B, 'the king of Moab' and 'the Eastern Hills', because they use proper names for identiÞcation. Yet, on the other hand, it could just as well be argued that while many people may be called Balak, it is 'king of Moab' that speciÞes which one is meant. In reality, such a debate misses the point of the parallelism: the question is not whether, taken in isolation, B is more precise than A, but whether 'king of Moab' in B adds any precisions to what we already have in A. Unquestionably it does; however, we must admit that in the case of 'Aram' // 'the Eastern Hills' we cannot say that B adds precision to A (unless perhaps the poet knows something that we do not).
In brief, so-called 'ballast variants' are prima facie candidates for the parallelism of greater precision, but a roughly synonymous parallelistic relationship is also quite possible.e. Word-pairs
This very frequently discussed feature of Hebrew poetry needs to be mentioned here primarily in order to distinguish it from the subject of this paper. At its most conventional, the use of word-pairs is a substitute for creative poetic activity, whereas the parallelism of greater precision is a subtle relationship between or among the lines of poetry that can only be designed in by a relatively sophisticated artist.
In a recent analysis building upon psycholinguistic studies, Adele Berlin has argued persuasively that pairing of words is a manifestation of a common linguistic phenomenon of word association./44/ Two principal types of word associates can be denominated paradigmatic (e.g. good-bad, father-mother, descend-ascend) and syntagmatic (Zion-Jerusalem, mercy-truth, heavens-earth [as a merismus]). The phenomenon of Þxed word-pairs, which has been a primary focus of attention in Hebrew poetry especially since the discovery of Ugaritic poetry, is only a subset of the broader category of word-pairing. It may indeed be possible to regard the present subject of study as yet another type of word-pairing-except of course that the 'greater precision' may be a function of the whole line in relation to the previous line, and not just one word in relation to another word.
My principal concern here, however, is to observe that the phenomenon of the Þxed word-pair is a contra-indication of the parallelism of greater precision-at least as far as those words are concerned. It is indeed true that in word-pairs the B-word is often less frequently attested, more poetic, more esoteric than the A word./45/ But it is not typically more speciÞc than the A-word; indeed, whether the pair is classiÞed as paradigmatic or syntagmatic,/46/ or as 'synonymous', 'antonymous' or 'correlative',/47/ there is usually a parallelistic relationship of balance, in which there is no question of progress from word A to word B.
Things may be different, however, if we adopt a more þexible deÞnition of 'word-pair', as is advocated by several recent authors. Adele Berlin, for example, afÞrms that 'there is no qualitative difference between the so-called "Þxed" pairs and pairs that have not so been labelled. The only difference is that Þxed pairs are attested more often than non-Þxed pairs.'/48/ M. O'Connor writes that 'any single word in a language can be paired with another', a process he terms 'dyading'./49/ And William R. Watters uses throughout his study, Formula Criticism and the Poetry of the Old Testament, a concept of word-pairs that can incorporate almost any kind of associative relationship./50/ In such cases the 'word-pair' may exhibit a parallelism of greater precision-or any of the other kinds of parallelistic relationship.
It may Þnally noted that the presence of a 'synonymous' word-pair within a parallelistic couplet may not necessarily form an obstacle to the parallelism of greater speciÞcity. For example, Ps. 7.17:wvarb wlm[ bwvy May his sin redound upon his head,
dry wsmj wdqdq l[w and upon his pate may his violence descend.Here 'head' (var) and 'pate' (dqdq) are the pair, and are perhaps completely synonymous (though the latter could reasonably be thought more speciÞc or 'concrete' than the former). But there is nevertheless a greater precision in line B, since wsmj 'his violence' speciÞes the kind of 'sin' or 'trouble' (wlm[) that the A-line speaks of.
5. Towards a (Somewhat) New Theory of Hebrew Poetry
The new feature that has emerged from this study of the parallelistic couplet is not as much the identiÞcation of a particular relationship of the lines of the couplet (greater precision) as a movement towards a statement of relationships within the poetic couplet. Within the couplets that we have examined here, we can afÞrm, the relationship of the two lines is unpredictable. What is predictable about Hebrew poetry generally is its structure as couplets (or triplets, i.e. extended couplets). What is unpredictable is how the lines of that couplet (or triplet) will turn out to relate to one another. Will they be synonymously parallel, will they exhibit the parallelism of greater precision, or staircase parallelism or synonymously-sequential parallelism,/51/ or some other parallelistic relation-or no parallelism at all?/52/
This unpredictability is encountered by the reader at the beginning, middle and end of the poetic couplet. At the beginning, before we start to read a couplet, we are aware that it is a couplet (whether through modern conventions of typography, or through our familiarity with the poetic convention itself); we can see that it will end after a snatch of words of between four and about ten, and we can expect that the couplet will constitute a complete sense-unit. But we do not know how that self-contained unity will present itself to us. We have some patterns of expectation available to us, indeed, but we are at the mercy of the poet for which of those patterns may lie ahead of us. At the end of the Þrst line (colon) when we pause momentarily for processing the sense so far, we make a provisional judgment on the completeness or otherwise, the self-sufÞciency or open-endedness, of the sense-unit thus far read or heard. But here also we cannot predict how the couplet will proceed; not only do we not yet know the grammatical or syntactical pattern of the next line, or its lexical contents, but, more importantly, we do not know what kind of a relationship what we have read will bear to what we have yet to read. (Many popular expositions of parallelism, formulaic poetry, word-pairs and the like implicitly encourage us to believe that we do.) The unpredictability remains even when we have read to the end of the couplet. For even when we have processed both lines of the couplet, our understanding will not be complete until we have gone back over the lines from the viewpoint of their relationship. And what that relationship will be is not securely indicated by any surface clues (like grammatical morphology or the presence of a Þxed word-pair). It is often not difÞcult to determine that relationship, but the point about it is that it is not given but must be Þgured out by every reader./53/ James Kugel was entirely right in asserting, as against a crude popularization of Lowthian parallelism, that 'Biblical parallelism is of one sort . . . or a hundred sorts; but it is not three'./54/ But I believe he is wrong to describe the 'one sort' as a matter of 'A, and what's more, B', since that restricts the relationship of the lines to those of emphasis, repetition, seconding, and so on. The relationships of A and B are so diverse that only some statement such as 'A is related to B' will serve as a valid statement about all parallelistic couplets. And such a statement is equally valid for non-parallelistic lines. Biblical poetry in general is overwhelmingly composed of couplets (or triplets, extended couplets), and of such couplets we could state that they are of one sort (A is related to B) or of a hundred, but not of three or four or Þve.
Our study of the parallelism of greater precision has alerted us to something that is true of Hebrew poetry generally. The meaning of the couplet does not reside in A nor in B; nor is it in A+B (if they are regarded as capable of being added like 2+2 or 3+2). It is in the whole couplet of A and B in which A is affected by its juxtaposition with B, and B by its juxtaposition with A. The whole is different from the sum of its parts because the parts inþuence or contaminate each other./55/ A has its meaning within the couplet only in the light, or sense, of B, and B in the light, or sense, of A. In the case of Isa. 40.3, for instance, the couplet does not mean B, even if B is more precise than A. It means (i) prepare Yahweh's way in the sense of making straight a highway, and it means (ii) make straight the highway as an act of preparing a way for Yahweh, and it means both of these things concurrently.
Because the relationship of the two lines within the couplets is not predetermined, the reader is more fully engaged in the process of interpretation, a more active participant in the construction of meaning, than when a text presents itself in more straightforward linear fashion. Edward Greenstein has reminded us of Marshall McLuhan's distinction between 'hot' and 'cool' media, 'hot' media presenting a complete pattern of stimuli, 'cool' presenting an incomplete pattern and therefore requiring greater processing and hence a higher level of engagement on the reader's part./56/ Such engagement, I would suggest, is systematically demanded by the nature of the Hebrew poetic couplet. The reader is constantly involved in the delicate and tantalizing/57/ question of the relation of the parts and the product of their interrelationship. That relation, as we have seen in the study of the parallelism of greater precision, is a dynamic one which cannot be mechanically delineated, but which often yields itself only to patient exegetical probing, each couplet in its own right. Any future theoretical study of such phenomena as that discussed in the present paper will have the converging resources of philosophical hermeneutics, current literary theory (especially reader-response criticism and reception theory), and psycholinguistics to draw upon./